Sunday, August 17, 2008

Flash me a clue

Idiotry at Slashdot continues.

No matter how stably, smoothly, efficiently, and correctly Linux runs on a machine, the public will continue to view it as second-rate if Flash keeps crashing. This is the worst example of being tied down and bound by a crappy 3rd-party product over which no Linux distribution has any control.


This is so typical. Without having any evidence whatsoever, lets blame the closed-sourceness of Flash for all of desktop Linux's problems.

When will you lusers fucking learn? In case you haven't noticed, users want Flash. Say it again. Users want Flash. Anything you do to make Flash not work is your fucking fault. Not Flash's. Yours. At some point in time in the past, Flash worked on Linux. In some configuration. Otherwise Adobe wouldn't have released it. Whatever you did to deviate from that configuration is your responsibility. Not Adobe's. The fact there are a bazillion Linux configurations that Adobe couldn't possibly test: OSS sound, or Alsa sound, or alsa-oss sound, or pa-alsa emulation, or pa-alsa-oss-jack-esound-vagina emulation, or all of the above inside nspluginwrapper, or whatever the fuck... it's all your fault. Not Adobe's.

I don't give a shit if Flash is the worst piece of code since Microsoft Bob. Your users want it. That's all that matters. If you ever cared about your users, you'd give them what they want. Users don't give a shit whether your browser is open source, or whether you have the latest greatest audio system with shitty emulation of the 20 audio systems that came before it.

When you guys have actually bent over backwards to make Flash work for your users, only then can you even beging to complain to Adobe about what they do to make it difficult. Right now, Linux is the making it difficult for Flash, NOT the other way around.

As if it weren't enough, the Slashdot entry finishes with:

I really do have to suspect Adobe's motivation for keeping Flash on Linux in such a deplorable state.


Are you shitting me? Yes, Adobe's spending precious developer resources to release a half-assed version of Flash just to subvert the Linux desktop. If you maybe had one more brain cell (bringing you to a grand total of one), you'd realize that it's actually beneficial to Adobe if the Linux desktop takes off. They have the only credible cross-platform rich application delivery mechanism in Flash and Flex. When Linux spreads, MS-specific technologies like Silverlight look worse and worse to developers. As diversity increases, cross-platform Flash looks increasingly appealing.

But no. You guys not only make it hard for Adobe to ship anything that works, you have to swiftboat Adobe and accuse them of subverting the cause. I hope Adobe just stops supporting Linux altogether for a year or two. Just to teach you lusers a lesson. Then you can concoct some crazy wine-based browser plugin emulation scheme to make the windows Flash run on Linux, and then write another article complaining that Adobe not open sourcing their Flash plugin makes it difficult for lusers to port it to FreeBSD or some shit.

288 flames:

«Oldest   ‹Older   1 – 200 of 288   Newer›   Newest»
Mehrdad said...

Three words: Works For Me!!!!
The next three words: Can't agree more.

OliWarner said...

>>At some point in time in the past, Flash worked on Linux. In some configuration. Otherwise Adobe wouldn't have released it.

Yeah... Because any software released that isn't BETA or ALPHA contains 0 bugs. Adobe are the first to state that they're still chasing thousands of bugs out of Flash player.

Anonymous said...

The wonder is that Adobe bothers with Linux at all. They make the effort to develop their plugin for a system no one in his sane mind would develop for and all they get in exchange is bitching, whining, and bad publicity.

Anonymous said...

Flash on Linux has never worked for me. It always crashes the browser I am using, be it Firefox, Opera or Epiphany (and Konqueror, I don't even bother using). That said, it's downright idiotic to blame Linux developers for the problem. They simply have no control over the code or its integration in internet browsers on the Linux platform. Adobe officially supports two Linux platforms with its provision of Flash packages in the deb and RPM format (which covers all the major distros and then some) and they are simply garbage on any given Linux distro. Given that Adobe officially supports three of the major distros, there is no reason to believe that it's Linux at fault, when Adobe's own products don't work as expected and marketed. A poorly written program is a poorly written program, period. It's the same as Adobe's Acrobat Reader on Windows. It's buggy and bloated and programming at its worst. Would you blame Microsoft for that?

Anonymous said...

Adobe offers docs and sources of it's flash to every developer that is willing to sign NDA with them. So if someone is interested in "porting" flash to FreeBSD he have to sign a NDA with Adobe and do it. But nobody with skills is interested ;) Oposite to linux FreeBSD's main purpose is to run on servers. IT's a single OS not 3000 different distros.
It have stable API and ABI an same for the kernel. It's quite more easy to develop for freebsd compared to any linux distro.

But yes most gnu shit err.. soft works on FreeBSD and have the same problems.

Anyway - overall I fully agree with you. It's not only Adobe's fault that their product is so buggy under Linux.

I bet that even opensourcing flash wont make it any better next 2-3 years.

Anonymous said...

I use Mandriva 2008.0 and it worked fine. Then I upgraded to 2008.1 and the upgrade to 9.0.115.0 broke it. I had to download the previous RPM, install it myself (forced) and now I just know not to download the new version.

When I went to 2008.1 the sound system sounded all screwy, too, and I had to go and disable something in KControl to get it to behave normally.

I couldn't agree more with Linux Hater; there are WAY too many sound systems in Linux.

Homercycles.

Anonymous said...

well adobe shouldn't support open sound system because it's been obsolete for years, so that point is moot.
how is it linux's fault that adobe can't keep their code stable? flash crashes all the time for me too. in fact, it's the ONLY thing on my system that crashes.
can i get something straight here? what happens if an app crashes on mac or windows? is it the app developer's fault or apple/microsoft's fault? every time something breaks on linux, it seems to be the kernel developer's fault and i don't really see any rational argument coming from you here.

Anonymous said...

>> can i get something straight here? what happens if an app crashes on mac or windows? is it the app developer's fault or apple/microsoft's fault? every time something breaks on linux, it seems to be the kernel developer's fault and i don't really see any rational argument coming from you here.

Here's how commercial software works. You pick just enough platforms to cover ~90% of the market. Developers then develop on a supported set of platforms and test on all of them. I.e. you need to test on Windows XP and Windows Vista, 32 bit and 64 bit if that's what you support.

Resellers usually send bug reports back, and unless you fix them you lose them and their customers. At some point, your manager will have to decide whether to spend a few hours of developer time or lose x thousand customers.

Now compare on Linux. There are loads of different distributions, window managers and library versions. You can't possibly test on them all, so you probably pick a couple of common ones and hope the rest work.

But a supported platform is probably Red Hat 5.0. Your Linux setup is very different from this, and it's not surprising there are issues.

No one will pay for anything, so the company has nothing to lose from bug reports. People will scream at you to open source it, which means you basically kiss goodbye to making any money ever.

The fact that there are too many things to fix, too many configurations to test, no money and screaming zealots telling you to give away your intellectual property is the reason that commercial software on Linux is a waste of time.

But you're right, Adobe should fix it. Email them and tell them you'll stop using it unless they do. I'm sure the potential loss of income will make it worthwhile setting up an exact clone of your Linux setup, reproducing the bugs you've found and assigning expensive developer time to them.

Anonymous said...

Because linux is a pile of shit and adobe is letting the plugin rot because of it.

You lusers are SOL.

Anonymous said...

In a way your argument is very pro opensource if flash was opensourse then the first piece of your rand would be fair now it is just stupid.

Stilgar said...

To be honest Falsh is a piece of shit on Windows too. It somehow manages to crash Firefox, IE7 and Opera. Actually I am the first one to support the extinction of Falsh although I am a Windows user. Of course saying that Adobe are sabotaging Linux or that Flash is the reason Linux desktop does not take off is ridiculous.

Anonymous said...

spot on.

penguinrage said...

It's as stable as on windows and mac based on my experience.
So it works for me! (o sht :|)
the ones that are complaining must be using crappy distros lol

Anonymous said...

Spot on indeed!
Very good post, definitely one of the best lately.

Anonymous said...

I'm now officially a `Linux Haters Blog' fan. I tried to introduce my wife to "Desktop Linux" ... for the cause ...

Now, after 1 month, she owns a brand new laptop shipped with M$ Vista and I don't feel like a virgin anymore.

Anonymous said...

Most of this blog is rubbish but I have to agree with you here.

h1d said...

Nice rant.

It's really funny though, even when everything else is open sourced with a Linux desktop, they still don't work? Sorry, I thought you[/.] said, the fault is at the source being closed? no?

hmm...

They just can't get it right and just can't and just need something to blame at... Teens do that... blame parents and such...

Anonymous said...

Linux people have no fucking business sense. Adobe hates Microsoft. Adobe wants to be one of the future leaders of internet technologies hence why they are supporting Linux.

Why is Linux important? Cause Microsoft can't make "cheap" Windows for inexpensive laptops such as the Asus EEE PC, or the newer cheaper tablets that will be coming out. If they did such thing they would end up in Anti Trust court faster than Amy Winehouse does a line of blow.

Linux will continue to have steady growth but not because of the Freetards but because of industry players like IBM, Intel, Canonical and Novell.

Mehrdad said...

"Cause Microsoft can't make "cheap" Windows for inexpensive laptops such as the Asus EEE PC"

Microsoft can and WILL do it as soon as the market is big enough. Look, they're already starting to try (Windows XP OLPC edition, eee PC with Windows XP)

Anonymous said...

"Microsoft can and WILL do it as soon as the market is big enough. Look, they're already starting to try (Windows XP OLPC edition, eee PC with Windows XP)"

Read about how Monopolies are judged, then you will see why that would be a horrible idea for them to keep on with that path. They got away with those two for two separate reasons, first thing the OLPC is a non profit venture, and with the ASUS EEE PC they could say it is a "market test". But charging one price in a particular market then a lower price for the identical product in another product to "drive out competition" is another thing.

Anonymous said...

"Now compare on Linux. There are loads of different distributions, window managers and library versions."

since flash shouldn't directly interact at all with the window manager, that's a moot point.
library versions: ok, you have a slight point there but wouldn't they just ship their own version within the binary to alleviate that issue? not ideal but it works.

"No one will pay for anything, so the company has nothing to lose from bug reports"

you mean you pay for flash? oh dear...

"The fact that there are too many things to fix"

it's their own fault for producing buggy code in the first place

i don't believe there's a conspiracy (for what it's worth). i just think adobe have some shoddy coders working for them. as others have said above me (and on slashdot), the code is quite unstable on windows too.

finally, please explain to me how it's the kernel's fault flash is unstable, as our friendly neighbourhood linux hater implies.

Anonymous said...

Will an RPM package work for every distro that uses RPM (and vice versa for a deb package for deb-using distros)?

Anonymous said...

Great post...couldn't agree more. But you said that things like Silverlight will struggle if Linux market share grows.

I think Silverlight is supported on Linux though. Miguel De Icaza is working on the plugin, calling it Moonlight.

http://en.wikipedia.org/wiki/Moonlight_%28runtime%29

Anonymous said...

You see, Adobe wouldn't have to bend over backwards for 700+ distros with 9000+ versions (with a grand total of 1% of desktop marketshare) if they could just open up their source code in the first place.

amirite?

Anonymous said...

+1

Dear Adobe,

Sorry so many Linux users are assholes. Many of us actually do really appreciate the extreme measures you take to make Flash work with a zillion different browsers, kernels, sound systems, webcams, libcs, X libs, etc.

Love,
A Debian Freetard

PS Please don't open source flash. None of us want to deal with migrating code from OSS to ALSA to ESD to PulseAudio either.

konrad_ha said...

Strangely it works really well for me. Never had a problem with it on my Ubuntu-box.

Anonymous said...

I use Pardus 2008 and Flash&Java works fine out of the box. But with Zenwalk for example - wow it was pain in the ass. And after countless hours spent to install it and configure - it never worked good.
So it in fact depends on distribution - LH is right.

aizmov said...

I agree with you.
Adobe is not the one to blame, but the FOSS community are, yet they go and spread bad publicity.
GNU/Linux lacks stable API and ABI. Things change frequently and as a result they break applications and drivers. I experienced the later first hand, a kernel update to Ubuntu and my nVidia driver stopped working, I had the option of either 640x480 or 800x600, thankfully nVidia updated their driver soon enough and everything was fixed.

Anonymous said...

I agree with the author but has anyone noticed that the distro flash crashes the most on is ubuntu?

Philip Jacob Smith said...

Exactly!

The reason Flash works so poorly is because it needs to do two things which are very difficult to do well in Linux: It attempts to create non-GUI graphics (in other words, it doesn't just create buttons and text and a few static images) and it attempts to synchronize audio with them, which is more difficult than simply playing an MP3 which isn't synchronized with anything. Either one of those things alone is difficult enough to do well, and flash tries to do both.

Note that I keep saying "well." Just doing anything in Linux is much easier than trying to do it in a way that will work on every Linux system. It's comical that a community so vocal for open standards creates an OS with so little in the way of standards, and no documentation for even that much.

Flash did work correctly for me once, years ago. I've tried downgrading to that version of flash, but its broken as well. Apparently it was the OS upgrade, not the flash upgrade, which caused it to stop working.

Anonymous said...

I don't know what you're talking about. Flash works fine on my Ubuntu.

julian67 said...

Any idea about Adobe sabotaging free software is almost certainly very misconceived. Truth is more like it just hasn't very important to them for any of their products or strategically.

But lately they've had this spiffy idea about publishing specs for their protocols without requiring NDAs (google openscreen project).

Why the change of heart?

Two things: those cheap luser freetards have managed to make a damn good replacement for Flash 7 in the form of Gnash and very soon it might be a drop in replacement for Flash 9. All done from RMS's mom's basement using a broken PII, fifteen tubes of Pringles, a crystal radio set, a soldering iron, an underwear catalogue and a big box of tissues.

And the really big one is of course Silverlight. Suddenly Adobe's exclusivity in this area has disappeared with some big players preferring Silverlight (for example NBC in the US, ITV in the UK). Adobe just found out that MS still has very powerful allies amongst the media companies and doesn't like finding out they entirely missed the boat in crucial areas. If Silverlight can be made indispensible to the average desktop user than that's a very serious weapon for defending Windows market share as Linux based systems threaten to become endemic in netbook/mobile phone type devices and to a smaller extent on traditional desktop/laptop machines. MS are even co-operating to allow Moonlight to become a silverlight equivalent in Gnome.

Oh shit. Suddenly it's time for Adobe to start playing nice with lusers as part of a strategy to retain their grip on streaming video and audio, because if they're not careful their ownership of this valuable market is about to disappear entirely. Cue panic and publishing of specs.

Here's my freetard luser prediction (direct from the basement): the next release of Flash will work perfectly reliably on any current distro (or at least as reliably as on other OS). Soon afterwards Gnash will gain all the functionality of Flash 9.

Only 3 people will ever install moonlight on their free desktop. Everyone using Windows or Mac will use Silverlight as a matter of course.

In 3 years we'll be having a similar conversation about moonlight/silverlight as we can now about gnash/flash because Flash will have been strangled, chopped up and buried in the woods by MS.

MS will be feeling pretty pleased with themselves, having retained at least a little (though declining) market share in the netbook/mobile phone market, 80% market share in traditional desktop/laptop market (again declining).

Business as usual.

Anonymous said...

A Linux Loon says elsewhere on the web:

"Since I run Linux/Firefox X64, installing Adobe's offering brings back bad memories of configuring printers in Windows 3.1 (get this addon, download that wrapper, rename something else, copy files here, there and everywhere - then when all's done it plays a few websites until it locks/crashes/dumps). So if it spurs Adobe to produce a player for more platforms I'll be happy. Alternatively if MS produces a silverlight thingy for my platform, I'll drop Flash faster than an oozing baby."

Even if Adobe wants to help the Loons, it can't. When nobody is around to f*ck the Loons, they insist on f*cking themselves (by gratuitously changing something or the other). A bigger bunch of incompetent programmers never got together before to produce any software, let alone a joke of an operating system that failed to garner even 1% of usage share after 17 years of dumping below cost.

julian67 said...

forgot a couple of things:

1: WorksForMe(™) (except when it doesn't)

2: The point about MS co-operating to alllow moonlight to be a free equivalent of silverlight might seem contradictory to the rest of my above argument to some. But please remember that the patent agreements surrounding moonlight are so problematic that it's extremely unlikely that anyone other than Novell will ever be able to dsitribute it and almost nobody wants to use it anyway. Also it doesn't work yet :-)

julian67 said...

toodai mie spellling is fukked baad. sxuse mee. kafeene kafeene twoo march kafeeyne.

Anonymous said...

I completely agree -- Slashdot is definitely not helping the situation with conspiracy theory articles.

paul said...

Here I thought there was hope for you, julian, and then you pull a lusing freetard argument.

those cheap luser freetards have managed to make a damn good replacement for Flash 7 in the form of Gnash and very soon it might be a drop in replacement for Flash 9.

First, Gnash is notoriously unstable and unused; even the developers admit it by leaving it at 0.8.X for a FRICKIN YEAR NOW. To say it's damn fine and usable is a really, really silly argument for alpha software, and one we see through: you may be able to sneak one over on the yokels, but we've all been through this before.

Second, SoftwareFromTheFuture™ arguments are incredibly lame. Saying the next version will solve all problems is the biggest threat to desktop Linux ever being taken seriously, because most people get tired of promises never met.

Anonymous said...

been a ubuntu user for a year now

my conclusion is that windows, linux and macosx all suck. its all slow, crappy, hangs every day, annoying

Anonymous said...

I wonder if organizers hand out tinfoil hats at FOSS conventions.

Microsoft is scanning my brains man!

Anonymous said...

Windows is great for my grandma (on prob is i have to keep cleaning the viruses) but for me Linux works and it makes my life exciting. Windows just does not have that feel..something (a lot of things) missing. It's like the difference between living in a communist country and in a fully free democratic country. "Freedom" I enjoy that. As for rest; windows is ok; stick with it.. no issues.

Anonymous said...

As long as UAC is on in Vista windows is as safe as linux, nothing can harm the user without entering of a password just like in linux so stop it with the viruses, I don't know one vista user who has ever got a virus unless they disabled UAC.

Anonymous said...

But yes most gnu shit err.. soft works on FreeBSD and have the same problems.

You forgot to mention that FreeBSD runs Linux better than Linux does. Seriously, loading the linux flash plugin into a native browser via nsplugin wrapeer through linux binary compatability, runs flash better than the linux binaries native on linux. Even running the whole stack (browser + nsplugin + flash) completely from binary compatability, still performs better than the same binary does natively on Linux.

I think that says something about Linux, specifically about the 2.6 branch, since the Linuxulator in FBSD is 2.4 based.

It's funny how FreeBSD doesn't even have "native" flash, but its users run it happily, all the same, on their workstations, and make little, if any noise about it relative to Linux. Rarely have I heard someone who uses FreeBSD cry about how Adobe should open source Flash.

FreeBSD is about building a stable, reliable, high-performance operating system, Linux is purely about politics.

paul said...

It's like the difference between living in a communist country and in a fully free democratic country.

That is the stupidest goddamn thing I have EVER heard. It also minimizes the fate of those living in a truly repressive atmosphere: they need to worry about, oh feeding themselves and not being eliminated from the face of the earth because of their dissenting views. There are real-work consequences to being a dissident in a repressive regime (like prison or death); there is nothing similar if you're using Windows.

It is the height of a selfish argument. If you really want to see what life is like under a communist country, then get your ass ourt of your mother's basement and some time in Georgia fighting the Soviets. Or hear to Iran and work to topple the government. Or go to Libya and work against Qadaffi. Or go to North Korea and oppose the govenment.

But do not equate what you do with the real dangers faced every day by real dissidents. Worm.

julian67 said...

Gnash is fine in terms of replacing Flash 7 (viewing, not creating) and less so for Flash 8. The big deficiency and reason it is little used is not any stability problem but because to enjoy sites such as youtube, BBC iPlayer etc you really need Flash 9 or equivalent. I have tried Gnash and it was fine in as much as it does as it claims to do, no stability problems afaik, but being in UK I very much want to access BBC's streaming video which requires Flash 9 so I use Adobe's Flash 9 in Epiphany (gecko backend) and it has its problems for sure. I don't do enough web browsing in Windows to make a comparison.

The last release of gnash was 0.8.3 in June this year. It's hardly stagnating, there are regular releases. If you get hung up about version numbering what to do? Try KDE 4? ha ha.

""SoftwareFromTheFuture™ arguments are incredibly lame" Yes they are. It's merely a specualtion and a luser's prediction, I don't pretend anything else. Anything might happen. Who knows, maybe soon we'll all be using Longhorn with WinFS and our problems will be over ;-) But it's reasonable to expect that Adobe releasing their specs unconditionally (as opposed to being conditional on NDAs as in the past) will greatly facilitate the development of Gnash. And also reasonable to expect that now Adobe does see some value in keeping happy any group which might prefer its product to silverlight. Suddenly it's be nice to lusers time. I expect future versions of Flash to be comparable across Win, Mac and Linux based OS. We'll see.

Anonymous said...

Flash needs refinement regardless of OS. The fact that it works for me (it does, perfectly) using Linux on FF3 doesn't mean that it works for everyone.

http://forums.techarena.in/vista-help/762746.htm
http://www.omgili.com/omgili/Internet+Explorer+7+flash+crash+vista
http://www.eggheadcafe.com/software/aspnet/30034147/anyone-else-have-ie7-unde.aspx

So are all of the above, and the 249K other folks with "Vista Flash Crash" in Google Wintards?

Bottom line, Flash as a concept is great, totally market dominant, but needs to be made much more robust for everyone, else Silverlight will eat into their market in a serious way.

Sorry if this is reasonable or balanced, but Flash brings IE down as much as any other browser and any other OS. OK, back to general cursing and luser's suck posts.

paul said...

Gnash is fine in terms of replacing Flash 7 (viewing, not creating) and less so for Flash 8.

Considering how aggressively Adobe is pushing 9, this approach is highly unrealistic. It also shows a great flaw in open-source development: emulating yesterday's outdated software is a formula for failure.

The last release of gnash was 0.8.3 in June this year. It's hardly stagnating, there are regular releases.

Regular releases that never bring it out of alpha!! C'mon. Quit with the BS, man: In one year this group has failed to bring a Flash 7 clone out of ALPHA. Good Lord, Adobe will be Flash 12 by the time Gnash 1.0 is released, still chasing after Flash 7 compatibility.

Again, it's all SoftwareFromTheFuture™ -- and given open source's incredibly bad record in actually delivering on promises from the past, I would not hold my breath.

Anonymous said...

Flash Player 10 solves problems with their Linux implementation. The biggest I think is the WMODE issue. PulseAudio appears to be better supported as well.

Anyway, Adobe should make it easy on themselves and target an LSB release.

That would encourage more adoption of LSB and it's accompanying libraries.

Still, I have never experienced the Flash plug-in crash in Fedora. In fact, I've never experienced it once over the last 4 years I've used Linux.

So its probably something Ubuntu developers are screwing with in newer releases. I wouldn't know, since I haven't tried Ubuntu since 6.06.

Anonymous said...

I'm afraid Gnash is crap.

However, its better than nothing for *nix machines with POWER, ARM, SPARC, etc. CPU architectures.

Anonymous said...

From my experience, the vast majority of current Flash issues can be fixed by removing PulseAudio.

More people would probably realize this if they didn't instantly blame everything on the evil closed source app while ignoring every other possible variable in the equation.

Anonymous said...

@ julian67: +1 and a desk forward.

Anonymous said...

Oposite to linux FreeBSD's main purpose is to run on servers. IT's a single OS not 3000 different distros.

There are twelve different flavors of FreeBSD floating around.

There are about 335 active Linux distributions on Distrowatch. My guess is 80-90% of people use merely 5% of the distributions in existence.

Commercial software developers usually only target a few distributions. I just don't see how commercial developers are putting forth significant effort into ensuring that their products works across a broad spectrum of distributions.

The reality that commercial Linux software tends to work on more than those few officially supported is attributed by them targeting a least common denominator version of glibc.

The better solution would be to adopt LSB like RealPlayer has.

It have stable API and ABI an same for the kernel. It's quite more easy to develop for freebsd compared to any linux distro.

Perhaps but 7.0 likely made quite a few changes to the kernel interface ABIs. FreeBSD also lacks quiet a few drivers, especially in the Wi-Fi realm. The kernel also doesn't have the correct features and facilities for Nvidia driver development on x64.

I recall just last year having to use the ndiswrapper implementation to get my Broadcom wireless LAN to work. Problem is it wouldn't work due to a bug in the emulator not supporting my the chipset's Windows driver.

So instead of wireless finding the access point it learned how to panic the FreeBSD kernel.

A developer claimed to have fixed the bug with their ndiswrapper about 14-15 months after I reported the bug, long after I returned to Linux.

Now, I'm not saying something like this couldn't occur on Linux to some unlucky individual but rather that the fish does not taste better on the adjacent side of the lake.

But yes most gnu shit err.. soft works on FreeBSD and have the same problems.

Heh... well FreeBSD doesn't have any ease-of-use features that most desktop-oriented distributions have.

All configuration tasks in FreeBSD and BSD in general are performed by editing config files. Linux is similar with the difference that more and more GUI front-ends are appearing on the scene, consequently making some configuration tasks easier for newbies.

Anyway - overall I fully agree with you. It's not only Adobe's fault that their product is so buggy under Linux.

Well no. Adobe only officially supports RHEL and SUSE with their Flash 9 implementation. Adobe has introduced official support for Ubuntu in Flash 10 beta.

The fact that some distributions might have issues with the plugin is largely in part due to Adobe not supporting those distributions and the various libraries versions they use.

I bet that even opensourcing flash wont make it any better next 2-3 years.

Well developing an open source solution from specifications doesn't happen overnight.

The more ideal solution would be fore Adobe to embrace LSB v3.2 or later with their upcoming Flash 10 release. LSB v4.0 will supposedly address the issues surrounding earlier releases but Flash 10 is right around the corner and LSB v4.0 standard specification isn't even available yet.

Anonymous said...

Usually I like your rants, but this is just sad.

How is it not Adobe's fault for making a shitty, crashy plug-in? It's not just sound that sucks: it crashes all over the place. It simply doesn't work.

Speaking of sound, OSS has been marked deprecated since 2003, but as of 2007 I still had to get sound in flash by using ALSA's shitty OSS emulation layer. I've heard that ALSA sucks, but give me a fucking break, dude. How is it not Adobe's fault that they've been sitting on a deprecated sound system for four fucking years?

Good job on completely missing the point of the first quote, by the way. If you read what he said, he didn't blame all of desktop Linux's problems on shitty flash. He said that if Linux can't even get flash not to suck it will continue to be viewed as second-rate.

I also love your rant about how users want flash. Uh, fucking duh, friend. Nobody said users don't want flash. That was why the post on slashdot was written. Adobe doesn't give enough of a shit about Linux to make flash not suck (and yes, part of the problem is attributable to them), hence flash on Linux sucks.

Bic said...

It seems like developing for Linux is a lot like playing Duck Hunt with a half broken Zapper. You try your damnedest to hit the moving targets (in this case Linux subsystems), and when you meet your inevitable failure you get mocked by the dog (Luser community), even though it's their fault the ducks were scared into fkying around (subsystems keep changing).

Personally I've never had a problem with Flash on a main computer, whether it ran Windows, Mac or Linux. (Granted the Linux version was limited to a Fedora 7 class where even the 10-year veteran Luser didn't know what systems did and was oblivious to Active Directory and NTFS permissions, and we used a bone-stock install...) The only time I've ever had Flash kill my browser was when I ran Firefox 2 and Flash 7 on NT 4.0 SP6. (For the most part, Opera worked.)

julian67 said...

"Regular releases that never bring it out of alpha!! C'mon. Quit with the BS, man: In one year this group has failed to bring a Flash 7 clone out of ALPHA. Good Lord, Adobe will be Flash 12 by the time Gnash 1.0 is released, still chasing after Flash 7 compatibility."

Again the fixation on version number/status.

Gnash is tagged as a beta at the moment, not alpha, but let's not allow any irritating facts or the inconvenience of doing 2 minutes of research spoil a satisfying rant ;-)

As the goal of Gnash is to provide all the functionality of Flash movie player and browser plug-ins I'd expect its versions to remain tagged as alpha or beta until it does so for the current version of Flash. I'm not sure I understand the objection to getting the functionality of 7 done before moving on to 8 and 9. I'm sure if and when Gnash absolutely perfectly reproduces the functionality of versions Flash 7 and 8 it will still be tagged as alpha or beta because it doesn't do the same for the current version of Flash. It's not something to get fixated on. We've almost all used software that's tagged beta, even alpha, and works fine and software that's version 1.* that should never have lost its alpha or beta tags.

Gmail is an obvious example of "beta" software that's widely used.

Examples of software with release versions that are obviously alpha or beta are KDE 4.0, Microsoft Windows XP original release, Vista original release, almost anything named Norton * etc.

Version numbers and alpha/beta/release status ought to have a commonly accepted meaning but they don't (I use some alpha software that as far as I can tell is absolutely ready for release but the author wants further testing and feedback before committing to calling it 1.0. Other projects are less conservative).
Sometimes the version numbers and status are meaningful but in those cases they only mean something if you know what are the goals/conditions to be met before a release status is appropiate. I'd say Gnash versions do seem to be an indication of status.

As for emulating proprietary products: sometimes it's necessary. Almost everyone wants to use Flash. The proprietary product has been less than satisfactory in some respects and several distros won't allow applications in their main repositories if they can't fix its bugs. So people do the needful. It isn't representative of all free software but it is representative of some. If you think Gnash is representative of all free software then that's really your problem and nobody can remove those blinkers except you.

Anonymous said...

How is it not Adobe's fault for making a shitty, crashy plug-in?

Take a look at user's experience re flash: they vary from "works perfect" to "all time failure" with every possible variation in between. The only factor which comes into play whether flash works for you or not is your system and distro. This means linux, not flash, is the problem.

Anonymous said...

Gmail is an obvious example of "beta" software that's widely used.

I highly doubt Gmail is still beta. Google just didn't bother to change the beta logo.

One thing I noticed about Gmail is the massive amount of spam I receive in the Spam folder even though I have *never* used it to register to any sites.

Its almost flabbergasting that I get any spam mail when I never used it let alone 500 pieces of it a month.

paul said...

Again the fixation on version number/status.

In the Real World, the status of the software does matter. And by bringing up KDE 4 and its incredibly botched release, you pretty much confirmed anyone in the Linux world is incapable of a smooth product release -- if that is indeed your model.

I'm not sure I understand the objection to getting the functionality of 7 done before moving on to 8 and 9.

Because Adobe seems to be a lot quicker about releasing new versions than gnash is, and that means a new version is very likely before gnash 1.0 -- never mind 0.9 -- ever sees the light of day. It's an issue in the real world, and most of us here have seen enough to call BS on the practice of promising great things InTheNextRelease.

And you seem to be more typical of a freetard than you want to admit: proclaiming how great it is and then admitting you don't use it. If it's that good, there's no reason you should not be using it.

I agree about numbering and versions potentially being misleading -- that's certainly not an issue unique to the open-source world. But you're focusing on something relatively minor to deflect from the main issue: that after years of development gnash is a piece of crap. And you're using a rational argument to defend a very irrational project in gnash: after years of development it's still unusable.

If you think Gnash is representative of all free software then that's really your problem and nobody can remove those blinkers except you.

As someone dealing with open-source projects for more than a decade now, I can assure you it's FAR more representational than you'll ever admit.

Anonymous said...

I really think Adobe should open source *the player*:

- The player is free (as in beer) anyway, their money is in the development IDE's.

- It could be a very big advantage if flash can be directly bundled with any browser. Hard competition with Silverlight ahead.

- Maybe some OS's and some computer architectures would get better support.

And, yes, I use to read Slashdot but the paranoia theory is retarded.

Timothy said...

But please remember that the patent agreements surrounding moonlight are so problematic that it's extremely unlikely that anyone other than Novell will ever be able to dsitribute it

Since when have patents ever stopped the freetards from anything? In a world where something as simple as a linked list can be patented -- you really think that you're free and clear of any patent infringement in something as large as the Linux kernel?

And considering the Samba project's been around for over a decade and hasn't run into any patent scuffles with Microsoft when they're in a position to hurt Microsoft much more; at this point any hand-wringing about patents in Moonlight is pure FUD.

I know, I know -- it must kill you to know that the most functional and most open source option for RIAs on Linux is a technology that originated at Microsoft.

Also it doesn't work yet :-)

Well that about makes it par for the course for open source software, doesn't it?

paul said...

I really think Adobe should open source *the player*:

- It could be a very big advantage if flash can be directly bundled with any browser. Hard competition with Silverlight ahead.


You do not need to open-source the code to accomplish this.

- Maybe some OS's and some computer architectures would get better support.

Maybe I'm naive, but I do believe the people who know the most about Flash already work for Adobe. One thing I hate about the freetards is their assumption they're better coder than anyone else and could in their spare time improve on a complex codebase.

Anonymous said...

@paul

I least you need to make some changes in the licence, I don't see any browser bundling flash now, although it'll be very handy for the general user.

I'm not saying that linux or BDS or Solaris (not even talking specifically about free OS's) developer's are better than Adobe developers. But anything is much better than [0 developers / no support for your platform].

paul said...

I don't see any browser bundling flash now, although it'll be very handy for the general user.

Since when has that been a consideration for the maintainer of a Linux distribution? ;)

julian67 said...

"In the Real World, the status of the software does matter."

What matters to the end user is "does the software work?". Do you honestly believe that all software with version >1.0 or tagged "release" is free from major problems? Or that software tagged rc or beta is always and necessarily not ready? This applies to proprietary as well as free software btw. Hint: sometimes the beta or rc is found to be fine and no changes are made before the exact same code is renamed someapp_1.0 and released. Sometimes companies release some really dreadful unfinished crap as the finished product. Anyone here enjoying the secure, reliable facilities of MobileMe?

KDE 4 is a good example of why getting fixated on version numbers or status is not a good idea. So is looking back at XP's release or Vista's or lately Apple's MobileMe. There isn't, unfortunately, a universal naming convention that everybody adheres to. Why? Because everyone can do whatever they like and they do, usually through desire for self promotion and the compelling power of marketing depts. This is definitely not a problem exclusive to free software or even software in general.

"And by bringing up KDE 4 and its incredibly botched release, you pretty much confirmed anyone in the Linux world is incapable of a smooth product release -- if that is indeed your model."

If you choose to see KDE's release strategy as representative of free software in general, well that's your choice and again the only person who can remove those very effective blinkers is you.

KDE have been strongly criticised for the way they released 4.0 (rightly imo) and I used it as an example in that light.

Anonymous said...

@julian67

Gnash sucks balls, wouldn't render a damn thing on my FC 8 install.

Doesn'tWorkForMe!

... seriously considering abandoning linux on the desktop.

julian67 said...

"@julian67

Gnash sucks balls, wouldn't render a damn thing on my FC 8 install.

Doesn'tWorkForMe!

... seriously considering abandoning linux on the desktop."


And I (or anyone else outside the Gnash or Fedora teams) should care about this?

Maybe try Vista and the first time a browser plug-in craps out you can abandon that too. Have fun.

Anonymous said...

@julian67

//Maybe try Vista and the first time a browser plug-in craps out you can abandon that too. Have fun.//

I do dual-boot vista, now that you bring it up.

Hasn't crapped out from the plugins I've installed.

Why? Maybe because Vista and its plugins aren't a festering pile of shit on the desktop, while Linux is? Have fun.

paul said...

Do you honestly believe that all software with version >1.0 or tagged "release" is free from major problems?

No, but in general it will be a lot closer than a half-assed open source project at 0.8.x for more than a year. Defending the specific with the general may sound good to you, but the smarter amongst can see it's an empty show of rhetoric. Gnash is a big, stinking, festering pile of poo that shows little life and little chance of replacing Flash.

Plus, you ignored the main point: you're a MASSIVE hypocrite for arguing it works well when you yourself don't find it usable.

paul said...

And I (or anyone else outside the Gnash or Fedora teams) should care about this?

If you're going to defend Linux on the desktop, then do it and don't weenie out of an argument you're lusing. You can't have it both ways.

Anonymous said...

Plus, you ignored the main point: you're a MASSIVE hypocrite for arguing it works well when you yourself don't find it usable.

i'm not the guy you're arguing with, but you're clearly not getting a very clear message: gnash does what it does well; emulating flash 7. i'm assuming the OP doesn't want flash 7, ergo he doesn't use it. simple? and before you start on your pathetic rant; have you not considered that emulating flash is a moving target? with new releases coming out, the team has more features to emulate. if gnash isn't what you need, then don't use it. there's no point being a whingy little moron about it.

julian67 said...

I've read your post again and I'm still not sure what you consider the main point to be. If it's that I use the Adobe plug-in and not Gnash, well I already explained that in a previous post but here goes again:

It worked fine for Flash 7 sites but I need/want something that works with Flash 9 sites so I use the Adobe product.

If your main point was "that after years of development gnash is a piece of crap. " then I don't know what you expect. It's been around for about 2 years. Is there some reference point, a way to assess the speed of replacing a proprietary product whose specifications the author declined (until recently) to divulge?
Apparently the Gnash team aren't fast enough for your taste????? Big deal. I guess "Gnash is a big, stinking, festering pile of poo " is a sound critique by the standards here and enough to convince almost anyone. You've already had a good rant about it being in alpha for a year when in fact it's in beta. Do I detect loudly stated opinions and assertions based mostly on not using the product in question?

If someone has a hissy fit after using Fedora and Gnash so what? You continually extrapolate experiences from the single to the universal. It's a flawed method and makes for meaningless assumptions/assertions. Try doing the same thing for other posters in this thread who variously report Flash not working well/not well in Windows or Flash working fine in a distro or well with Opera or not well with Opera. Now you have to believe everything and nothing or both at the same time.

So I'm not interested if someone finds Gnash + Fedora 8 is a shitty combination on their computer. The people who might care are the Gnash and Fedora developers. If he files a bug report with a line like "...seriously considering abandoning linux on the desktop." I'm sure it will get the attention it deserves and we'll all be able to sleep soundly at night.

paul said...

If your main point was "that after years of development gnash is a piece of crap. " then I don't know what you expect.

You know, I do admire your rhetorical skills. But I am a tad more disciplined than you; this thread started when YOU stated:

Two things: those cheap luser freetards have managed to make a damn good replacement for Flash 7 in the form of Gnash and very soon it might be a drop in replacement for Flash 9.

No one except you thinks it's a damn good replacement. And after two years in alpha and beta it still doesn't fully support Flash. If it took two years to "mostly" support Flash 7, it ain't gonna support Flash 9 "very soon."

And with that, everyone can judge the freetard philosophy at work: overpromise on crap and promise tomorrow's solution will be perfect.

Checkmate, bitch. Oh, I'm sure you'll respond with rhetorical flourishes, and that's fine, but the bottom line is that I've show your philosophy to be hollow, your assertions rhetoric not based in fact.

I've read your post again and I'm still not sure what you consider the main point to be.

I trust this is crystal-clear enough, even for a reality-denier like you.

Anonymous said...

There are twelve different flavors of FreeBSD floating around.

There's only one flavour of FreeBSD, and that's FreeBSD. Now you can argue that OpenBSD is a flavour of FreeBSD because it was forked from it, but it'd be a completely asinine argument.

OpenBSD is a fork of FreeBSD. although there is code shared, they are completely different operating systems, with different kernels, different package mechanisms, different userland, different purposes, and even different sets of supported archetectures.

I know it might be difficult for a Luser to understand the concept of an operating system not being an archipelegio of thousands of individual pieces, maintained individually elsewhere and kludged together.

Believe it or not, the BSD teams actually maintain all of their core systems themselves.

Arguing that any of the BSDs are different flavours of FtreeBSD is uneducated and stupid. Would you argue that OS X is a flavour of FreeBSD? Of course not, but here's a news flash, it's a BSD-style Unix nevertheless.

Do you group all of the System V style Unixes as different flavours of a single given System V style Unice? Of course not, that'd be stupid.

Linux distributions are called "flavours" because it's a repackaging of the same kernel with larhely the same (GNU) userland. It makes sense to lump them together in such a manner.

Perhaps but 7.0 likely made quite a few changes to the kernel interface

Work on FreeBSD 7.0 began when FreeBSD 6.0 was released, back in 2005. 7.0 had been in the CURRENT branch for 3 years before being released. Stuff worked on in 7.0 while it was in Current even made it's way into 6.1, 6.2 and 6.3 when it was deemed stable (not to mistake FreeBSD's definition of stable, (wherefore stable actually means stable) for the average definition of "stable" in Linuxland (which tends to mean "not really quite yet stable").

The API and ABI in 7.0 were stabilized and frozen a long, long time ago. Furthermore, FreeBSD, much like Windows has compatability layers in place for legacy code, see the campatXX ports e.g. compat4x, compat5x, compat6x, much like wow16, wow32 and wow64 in Windows. The lecagy API/ABI is still there, it's just not the primary ABI/API.

Again, the concept of introducing changes to an ABI and API without actually breaking the previous ABI and API may seem completely alien to a luser, and the concept of having separate branches for development and releases may be lost (seeing as 2.6 seems to be both the "stable" and development branch for Linux) on the Linux community. It's called common sense and a sane development model.

Further, there the fact that 7 has been availible for THREE YEARS prior to it's release.

FreeBSD also lacks quiet a few drivers, especially in the Wi-Fi realm

That's a moot point. There's no reason a server OS needs wifi, honestly. You're comparing Linux trying to be a general purpose everything to FreeBSD being highly specialized. It's another completely stupid argument. What next, you're going to whine that operating systems like AIX, OS X and Solaris (which are designed to run on a specific subset of hardware) are flawed because they don't support random exotic hardware?

I'm sure you can make case, but I'm also sure it would have no dredibility.

The kernel also doesn't have the correct features and facilities for Nvidia driver development on x64.

Possibly because development isn't done in the RELEASE branch. That's what CURRENT is for.

A developer claimed to have fixed the bug with their ndiswrapper about 14-15 months after I reported the bug, long after I returned to Linux.

So you're moaning because an OS that specializes in being a server OS isn't all that great as the general purpose OS it wasn't designed to be. Brilliant. (Again, there's absolutely no sane reason why a server needs to be on wifi).

Heh... well FreeBSD doesn't have any ease-of-use features that most desktop-oriented distributions have.

That might be because it isn't designed to be desktop oriented? AIX doesn't have ease of use desktop oriented features, either, nor does any other OS that isn't intended to be a desktop OS.

Linux is similar with the difference that more and more GUI front-ends are appearing on the scene, consequently making some configuration tasks easier for newbies.

FreeBSD has the handbook. The handbook is a compilation of absolutely superb documentation that is second to none. But again, it's a server OS. Further, it's an administrator's OS. Administrative tasks are both superbly documented and trivial to perform, given said documentation. It's also not Intended to be newbie friendly. Again, AIX (or Solaris, HP-UX, or Tru64 or any other server Unice) doesn't have gui facilities for configuration, and that's in no way a bad thing.

Though you can probably run Linux's GUI configuration tools under FreeBSD's Linux ABI compatability subsystem, if you really wanted to.

Well developing an open source solution from specifications doesn't happen overnight.

Nobody said it does. I'm tempted to cite HURD, as an example of open source solutions not hapenning overnight (let alone over more than two decades) though, just for shits and giggles.

The more ideal solution would be fore Adobe to embrace LSB v3.2 or later with their upcoming Flash 10 release.

No, an ideal solution would be for Linux to develop a stable ABI and API to work with. Adobe embracing a standard that nobody actually supports isn't going to magically cause Flash to work properly even though nobody actually supports LSB. It's up to Linux's development community to put their egos aside and pull their heads out of their asses, work together, and standardize of a single given API, and adopt a sane development model while they're at it.

All this talk about Flash and FreeBSD. FreeBSD doesn't even have a native implimentation of Flash (Gnash doesn't count), yet runs the Linux binaries with more stability that they do natively on Linux. Find the error.

Anonymous said...

Gnash isn't usable due to the high saturation of Flash 9 content on the web. It wouldn't matter if it perfectly emulated Flash 7 and was stable on every distro with zero bugs; it's emulating an out-dated technology thus its usefulness is ever decreasing.

julian67 said...

Ok so now you explained what your main point was (something that seems to have been obvious to you alone) and it seems to be that I am supposedly the only person who ever found Gnash worked as a replacement for Flash 7.

And that by expecting Adobe's opening of access to Flash specifications to lead to improvements in future versions of Gnash that this is an espousal of a "freetard philosophy...: overpromise on crap and promise tomorrow's solution will be perfect."

This is all very exciting stuff!!! But would it not be better to address what people actually say instead of what you think they say? I said gnash is a good drop in replacement for Flash 7 because it seemed that way when I tried it, and I said it *might* soon become a good drop in replacement for 9. Nobody's making any promises but you are somehow seeing promises everywhere. Perhaps you have a fever?

Anyway I'll carry on using the Adobe product for now and when Gnash offers support for sites using Flash 9 I'll try it again....which seems to be more than you have done, despite having so much to say on the subject.

Anonymous said...

Speaking of sound, OSS has been marked deprecated since 2003

Only on Linux. OSS is still, by far and large, the standard in audio on UNIX. In fact, the only major Unice not to impliment OSS is OS X. (Linux isn't Unix).

but as of 2007 I still had to get sound in flash by using ALSA's shitty OSS emulation layer.

OSS emulation is one of the all too rare occasions where the Linux devs have had a sudden breakout of common sense and actually implimented a _major_ API change in a means that doesn't completely break the previous (and in this case, standardized across all Unixes) API. Don't whine about it. Be grateful the compatability layer exists.

How is it not Adobe's fault that they've been sitting on a deprecated sound system for four fucking years?

Because OSS is only depreciated on Linux. Adobe is building their player around an established, widely implimented, and stable standard. That's what sane people do. I would venture that if there were no OSS emulation in ALSA, there would be no Flash player at all.

Furthermore, developing around the established standard reduces their development overhead for OTHER platforms. Any platform with binary compatanbility with Linux, but no ALSA (such as FreeBSD via the Linuxulator) can run said Linux binaries. Effectively, they're supporting multiple platforms with a single release, rather than porting and maintaining separate releases for each such platform.

That's what standards are for. That being said, since Linux deviated from the standard, it makes a lot of sense for the new API to have backward compatability with the standard.

It makes absolutely no sense for OSS (which is the standard Unix sound API) to support ALSA (which isn't).

Think of it on a broader scale, other than an insignificant market share, why do you think very few commercial entities (we'll stick with Adobe, in this case) support Linux with their GUI apps? There's no bloody standard there, and the actual Unix standard toolkit (Motif) isn't even present on your average Linux system, what standard are they going to build on? GTK? QT? EFL? wxWidgets? TCL? picking one over any other is just as much of a headache as trying to support all of them. It isn't worth the effort. Even if they build their own, even the system libraries aren't standardized across distributions!

A flash plugin exists simply because there's actually a standard, stable API to build the plugin with, and guess what, regardless of what Linux may want to believe, that standard and stable API is still OSS.

So many calls that corporate entities are out to sabotage Linux, it's all bullshit. The only entity actively sabotoging Linux is Linux itself. Just as the only Entitly working against third party, commercial support for Linux, is Linux itself.

Stephanie said...

Why is Linux important? Cause Microsoft can't make "cheap" Windows for inexpensive laptops such as the Asus EEE PC

They can, and they do.

If they did such thing they would end up in Anti Trust court

And you, my friend, have understanding of antitrust. Antitrust violations accur when an entity leverages an existing monopoly in one market, in order to gain a foothold in another, different (as in not the same market) market.

What you're arguing is what, that Microsoft would have to leverage monopoly in one market, in this case, I assume you mean Operating Systems, to gain a foothold in another, different (again, as in not the same) market, in this case Operating Systems. They're not even entering a different segment of the market! They're not even leveraging Windows!

An antitrust violation would be if Dell, for example, cornered the consumer PC market, effectively giving themselves a monopoly, then tried to leverage their newfound monopoly in the consumer PC market to gain a foothold in the operating system market, buy bundling DellOS.

Anonymous said...

@anonymous

I'm sure the people who have actually lived under communist regimes may disagree with you and wouldn't appreciate your analogy of using an operating system to being the same as their plight.

julian67 said...

A note on the brilliance of proprietary software:

Prompted by today's fascinating discussion I thought I'd refresh my memory on how Flash perfoms in Windows. I run XP SP3 (fully legal OEM, no warez here). I have the same version of Flash 9 on Debian and Windows.

I started IE7 and navigated to bbc.co.uk/iplayer to be greeted with the message "you need to install Flash...." Hmm, that's odd, could have sworn I've watched flash movies on this before....but whatever. Installed Flash, got the "Flash install complete" message. No change.Close browser, restart it. No change. Log out, switch to admin account, do it all again. I get the install completed message.No change. Do it again. This time I get the activeX prompt and allow it. Install completes, cute animated confirmation. Restart browser. No change. Reboot. No change. Try youtube: greeted by "Get the latest Flash Player". err, I just did. Lots of times. Checked all browser security/privacy settings are at default. No change. Restart browser. No change. Fuck this.

I also have Mozilla Firefox 3 installed. I start Firefox, navigate to iplayer site, everything works. Go to youtube. Everything works.

Which group of free software developers should I hysterically blame for my proprietary OS's proprietary browser failing to work with a proprietary plug-in? How can I prove that Firefox's lucky strike is evidence of general malaise throughout free software? How can I extrapolate my individual experience to demonstrate that Microsoft is shit, everything they ever made is shit, Adobe is shit, eveything they ever made is shit, they are all lusers? Should I start a blog?

Anonymous said...

There's only one flavour of FreeBSD, and that's FreeBSD. Now you can argue that OpenBSD is a flavour of FreeBSD because it was forked from it, but it'd be a completely asinine argument.

There are multiple distributions based on each. Not everything is the same either. PC-BSD, for example, has its own package format that is not compatible with ports.

I know it might be difficult for a Luser to understand the concept of an operating system not being an archipelegio of thousands of individual pieces, maintained individually elsewhere and kludged together.

Its not difficult to grasp. Virtually all other operating systems are developed in that single-entity fashion.

Beyond a very bare-bones system, FreeBSD needs external software (GNU tools) to render it useful.

Arguing that any of the BSDs are different flavours of FtreeBSD is uneducated and stupid. Would you argue that OS X is a flavour of FreeBSD? Of course not, but here's a news flash, it's a BSD-style Unix nevertheless.

There are multiple distributions of FreeBSD, NetBSD, and OpenBSD that have their own unique features. Linux is mostly the same way in that most distributions are based off another.

FreeBSD has the handbook. The handbook is a compilation of absolutely superb documentation that is second to none. But again, it's a server OS. Further, it's an administrator's OS. Administrative tasks are both superbly documented and trivial to perform, given said documentation. It's also not Intended to be newbie friendly. Again, AIX (or Solaris, HP-UX, or Tru64 or any other server Unice) doesn't have gui facilities for configuration, and that's in no way a bad thing

http://tldp.org/

A very large hub for general linux documentation, books, and HOW-TOs.

In addition, every distribution has their own set of documentation wikis and HOW-TO articles.

Yes, FreeBSD is a server OS by default but someone decided they wanted to favorably compare it with Linux.

The distinction between server and workstation/desktop is only a set packages like Apache, Perl, PHP, MySQL, and OpenSSH. Many desktop distributions bundle these anyway but simply disabled their service daemons.

My father's Fedora 9 installation is strictly desktop-oriented, however, because I uninstalled all server tools.

Nevertheless there are desktop and workstation variations of FreeBSD that include X server and a desktop environment by default.

That's a moot point. There's no reason a server OS needs wifi, honestly. You're comparing Linux trying to be a general purpose everything to FreeBSD being highly specialized. It's another completely stupid argument. What next, you're going to whine that operating systems like AIX, OS X and Solaris (which are designed to run on a specific subset of hardware) are flawed because they don't support random exotic hardware?

OpenBSD generally has superior wireless support and at one point was even better than Linux in that respect; OpenBSD is certainly no desktop OS either--its known for its security as a server.

Besides, Broadcom wireless controllers are very common in laptop equipment.

Nobody said it does. I'm tempted to cite HURD, as an example of open source solutions not hapenning overnight (let alone over more than two decades) though, just for shits and giggles.

Hurd is an example of a theoretical project. The FSF even admitted that if they were aware of the Linux kernel at the time that the Hurd project would have unlikely ever commenced.

Possibly because development isn't done in the RELEASE branch. That's what CURRENT is for

No, there has been desire for drivers on x64 FreeBSD for years and the developers have been glacier slow at implementing the necessary API hooks in the kernel.

Isn't it ironic that users of a server OS would want performance graphics drivers?

No, an ideal solution would be for Linux to develop a stable ABI and API to work with. Adobe embracing a standard that nobody actually supports isn't going to magically cause Flash to work properly even though nobody actually supports LSB. It's up to Linux's development community to put their egos aside and pull their heads out of their asses, work together, and standardize of a single given API, and adopt a sane development model while they're at it.

https://www.linuxfoundation.org/lsb-cert/productdir.php?by_lsb

Anyway, if Flash were to target the LSB libraries and specification then the users would relentlessly encourage their distributors into supporting it until they do.

Flash player can act as tremendous leverage at pushing standardization across the distribution spectrum.

This would be a good thing.

All this talk about Flash and FreeBSD. FreeBSD doesn't even have a native implementation of Flash (Gnash doesn't count), yet runs the Linux binaries with more stability that they do natively on Linux. Find the error.

Does the Linuxulator even support 2.6.16 yet? I remember there was a Google SOC project to enhance it for 2.6 support.

Many modern commercial Linux apps do not work correctly in Linuxulator because they utilize functionality of newer kernels.

Cedega was an infamous example of this. However, Cedega is dog shit these days so that probably doesn't matter much. Cedega 6.1 or 7.0 may redeem itself from the shit pile but the bar currently isn't set very high.

Anyway, back on topic... Gnash currently implements a large subset of Flash 7 functionality. However, Flash 10 is around the corner and further pushes Gnash behind.

The general assessment by Linux users is that Gnash is turds on x86 and x64 (with ndispluginwrapper) but exquisite diamonds on exotic hardware, such as SPARC, ARM, POWER and PowerPC--for which there is no official Flash implementation.

Proccessor emulation via Qemu, while fast for emulation is snail slow compared to any virtualization solution.

Moonlight is in much better shape because Microsoft has agreed to share confidential Silverlight specifications and binary codecs with Novell to aid in Moonlight development.

Anonymous said...

@julian67
//How can I extrapolate my individual experience to demonstrate that Microsoft is shit, everything they ever made is shit, Adobe is shit, eveything they ever made is shit, they are all lusers? Should I start a blog?//

Simple. You have shit hardware. :-p

paul said...

I started IE7 and navigated to bbc.co.uk/iplayer to be greeted with the message "you need to install Flash...."

Works for me! Must be some sort of user error! Thanks for playing!

Anonymous said...

Works for me! Must be some sort of user error! Thanks for playing!

Heh. A Windows user placing the blame on the another Windows user for their problems?

I thought only Linux users were allowed to do this?

Oh the irony is delicious.

Anonymous said...

@anonymous: Read about how Monopolies are judged, then you will see why that would be a horrible idea for them to keep on with that path. They got away with those two for two separate reasons, first thing the OLPC is a non profit venture, and with the ASUS EEE PC they could say it is a "market test". But charging one price in a particular market then a lower price for the identical product in another product to "drive out competition" is another thing.

Not true. The game-changer is the fact that the product they're competing against is ostensibly FREE; therefore, they can't undercut Linux on price.

julian67 said...

Exactly.

The worksforme/sucksforme response so berated here as a luser freetard stock response is precisely the extrapolation that you do Paul, and quite a few others. Instead of saying "works for me/sucks for me" you dress it up a little but essentially many of the respondents on this blog are offering nothing except a verbose version of worksforme. I offered my own small (real) example from the Windows side of things to show how stupid and pointless it is, and you've had the good grace to demonstrate how stupid and pointless you are.

The anonymous poster who thinks a problem with flash/IE combo can be attributed to hardware is only very slightly more stupid than you.

Anonymous said...

@julian67: "Which group of free software developers should I hysterically blame for my proprietary OS's proprietary browser failing to work with a proprietary plug-in?"

None. Blame the luser in the mirror for contriving a false example in a desperate attempt to tar Windows with your FOSS problems. Look, pal, zillions of Windows users use Flash every day on YouTube and countless other sites, they do it without any help and without your blatantly anti-commercial attitude. If you can't get Flash running, you better figure out a different line of work, you fucking moron.

Anonymous said...

"I offered my own small (real) example from the Windows side of things to show how stupid and pointless it is, and you've had the good grace to demonstrate how stupid and pointless you are."

No, I think you proved exactly how humorless you are. Guess freetards really _do_ need the smiley faces to tell when something is funny!

Anonymous said...

Price isn't an issue.

Microsoft is more than willing to eat any losses associated with pricing Windows licenses near free on low-power, budget laptops if it puts the breaks on Linux progress in a new market.

If Linux were in a position to threaten Microsoft's near-monopoly then rest assure that Microsoft would reduce the price of their products down to nothing.

Furthermore, secret monetary incentives would likely be awarded to OEMs who shun the new contender in a discrete way.

paul said...

If Linux were in a position to threaten Microsoft's near-monopoly then rest assure that Microsoft would reduce the price of their products down to nothing.

Furthermore, secret monetary incentives would likely be awarded to OEMs who shun the new contender in a discrete way.


Reducing the price to next to nothing and then pissing away the remaining revenue is a Linux business strategy, not a Microsoft business strategy.

Anonymous said...

@julian67: "The worksforme/sucksforme response so berated here as a luser freetard stock response is precisely the extrapolation that you do Paul, and quite a few others. Instead of saying "works for me/sucks for me" you dress it up a little but essentially many of the respondents on this blog are offering nothing except a verbose version of worksforme. I offered my own small (real) example from the Windows side of things to show how stupid and pointless it is, and you've had the good grace to demonstrate how stupid and pointless you are."

Tell you what. Prove it. Post some screenshots of these problems, and we may take you a little more seriously. As it stands, you sound more like a Windows hater with an ax to grind.

paul said...

I offered my own small (real) example from the Windows side of things to show how stupid and pointless it is, and you've had the good grace to demonstrate how stupid and pointless you are.

Boy, you have no sense of humor and/or irony, do you? I used YOUR exact same argument as a response, and that's your response? Lighten up, son!

Anonymous said...

Reducing the price to next to nothing and then pissing away the remaining revenue is a Linux business strategy, not a Microsoft business strategy.

Right...

Just check out the USA Acer Aspire One netbook configurations:

Linux, 3-cell battery, 512MB RAM, 8GB SSD: $379 USD

WinXP, 3-cell battery, 1GB RAM, 120GB HDD: $399 USD

$20 more for the Windows version with twice the memory and 15x the storage capacity.

paul said...

$20 more for the Windows version with twice the memory and 15x the storage capacity.

Without knowing individual component prices, you can't argue exactly how much Microsoft is charging.

Besides, you can't bitch about the Microsoft tax and then simultaneously bitch when you think Microsoft has lowered its pricing.

julian67 said...

"None. Blame the luser in the mirror for contriving a false example in a desperate attempt to tar Windows with your FOSS problems. Look, pal, zillions of Windows users use Flash every day on YouTube and countless other sites, they do it without any help and without your blatantly anti-commercial attitude. If you can't get Flash running, you better figure out a different line of work, you fucking moron."

That's the worksforme argument again. Idiot.

"Tell you what. Prove it. Post some screenshots of these problems, and we may take you a little more seriously. As it stands, you sound more like a Windows hater with an ax to grind."

OK when I see posted here a verifiable article of proof of every complaint about free software then we can all play on the same level field. But what does a screenshot prove? You could choose to disbelieve that too. Actually the example I used is real. I'll see if I can find a decent free screenshot utility for Windows before I get to diagnosing and fixing the problem.

Works for me! Must be some sort of user error! Thanks for playing!

Boy, you have no sense of humor and/or irony, do you? I used YOUR exact same argument as a response, and that's your response? Lighten up, son!

Not true. That's your argument not mine. The whole worksforme thing is lame. Which is why it's as popular here with Win fanboys as it is at ubuntuforums with ubuntu fanboys.

Paul you're an asshole.

RuralRob said...

Funny thing is, open source is also the reason that there is yet no 64-bit version of Flash (for any OS). Flash contains a just-in-time compiler for its scripting language that compiles to native x86 code as the script runs. This compiler is actually part of the Mozilla project, and as yet nobody has gotten around to making a version that compiles to x86-64. Thus, no 64-bit Flash yet.

Anonymous said...

Without knowing individual component prices, you can't argue exactly how much Microsoft is charging.

Besides, you can't bitch about the Microsoft tax and then simultaneously bitch when you think Microsoft has lowered its pricing.

BS. Judging by this, not only is Windows be free but that Microsoft is lending over some cash as well.

Another point of interest is that they chose a SSD for Linux rather than HDDs for both. SSD Flash memory technology is far more expensive per GB than an HDD but the advantages in I/O performance don't live up to it's hype, especially on real cheap/shitty SSDs.

So in reality the value looks exactly as it appears... the Windows option has 15x more storage capacity and no shitty 8 GB SSD will ever counter that.

I'm not bitching about the MS tax. This is simply predatory pricing, something Walmart has been accused of in the past.

It also confirms that Microsoft will do anything to ensure that nobody can take their bread and butter away.

ButtFuzz said...

Probably already mentioned somewhere else on this blog, but here's the Canonical list of operating systems that suck:

http://linuxmafia.com/cabal/os-suck.html

Enjoy!

Anonymous said...

@Paul: "Besides, you can't bitch about the Microsoft tax and then simultaneously bitch when you think Microsoft has lowered its pricing."

Of course they CAN bitch about it. They DO all the time. Let's face it. The only thing that Microsoft could possibly do that these Freetard cretins would approve is if the company shuttered its offices and stopped competing against Linux completely. Somehow, they seem to have this fucked-up delusion that only Linux has a right to compete -- and anybody who tries to compete must automatically be using UNFAIR PRACTICES because, clearly, Linux couldn't possibly lose on quality, right? [Cough]. They set the competitive price at ZERO (the value of their time), and then they bitch when Microsoft lowers its prices to compete. Un-fucking-believable. What a bunch of whining PUSSIES...

Freetards don't want to CO-EXIST with Microsoft. NO ... they want nothing less than utter DESTRUCTION of Microsoft and domination of desktop and Web spaces. Fucking luser Jihadists. May they burn in Hell...

paul said...

Heh. Julian, you are a trip. Earlier today you wrote:

I have tried Gnash and it was fine in as much as it does as it claims to do, no stability problems afaik,

This is pure WorksForMe! No one else thinks gnash is anything but a pile of crap, but it's OK, you say, because it WorksForMe!

And then you have the cojones to write this:

Not true. That's your argument not mine.

Dude, that's PRECISELY your argument! To a frickin' T!

Paul you're an asshole.

Yeah, being accountable for your words is a bitch, isn't it?

paul said...

SSD Flash memory technology is far more expensive per GB

But not per unit. Try pricing in real-world terms. Quit with the abstract and look at component costs, the way laptop builders do.

And then realize anyone looking to buy a Linux system is going to be a cheap bastard, so the accompanying components need to be cheaper as well. If these are older components subject to price drops, the margins may actually be better on the Linux machine.

But that doesn't make for a very good conspiracy theory, does it?

Anonymous said...

@anonymous: BS. Judging by this, not only is Windows be free but that Microsoft is lending over some cash as well.

There's a name for this psychological condition: PARANOID DELUSION.

Another point of interest is that they chose a SSD for Linux rather than HDDs for both. SSD Flash memory technology is far more expensive per GB than an HDD but the advantages in I/O performance don't live up to it's hype, especially on real cheap/shitty SSDs.

Who are you blaming for this choice? Acer? Microsoft? Or both? Make sure you have your tinfoil hat on. You don't want Microsoft to lock-in on your position...

So in reality the value looks exactly as it appears... the Windows option has 15x more storage capacity and no shitty 8 GB SSD will ever counter that.

So, WTF. Complain to Acer. They're the ones marketing this package, not Microsoft.

I'm not bitching about the MS tax. This is simply predatory pricing, something Walmart has been accused of in the past.

LMFAO! Dude, Linux is FREE. The only way that Microsoft could undercut Linux is if they PAID ACER to use Windows. You don't have any fucking proof that that's happened here, so just SHUT THE FUCK UP, you fucking retard. Besides, who's been harmed here? It's not as if ACER would have to pay anyone for Linux, so no one can complain about being the victim of predatory pricing. Make sense, dimwit?

It also confirms that Microsoft will do anything to ensure that nobody can take their bread and butter away.

No, what it confirms is that you are one paranoid muthafucker with zero evidence to prove your case.

Anonymous said...

BS again.

What would be fair is if Windows cost $20 more or less with the exact same hardware.

Based on this if the two options were identical then the Windows option would be even cheaper.

Microsoft knows their position in the market and Linux's too.

Microsoft is using their dominance as a monster competitor to kill off the little guy seeking market entry.

Consumers aren't stupid. Even if Linux were the superior OS the Windows option would still be the better option, if only because of hardware specs alone. For a Linux user its easy enough to simply buy the Windows machine and install Linux but that still credits as a sale for Microsoft.

paul said...

Microsoft is using their dominance as a monster competitor to kill off the little guy seeking market entry.

And exactly what did Microsoft do to kill off Apple, a much more serious threat?

THEY CAME OUT WITH A BETTER VERSION OF OFFICE THAN WHAT WAS AVAILABLE ON WINDOWS.

A move that most certainly helped Apple move some Macs and establish a presence in the corporate market.

Yup, Microsoft is really nasty toward those competitors. Kill them with kindness.

Your delusional rantings just don't match with reality.

julian67 said...

MS offers a steep discount to OEMs on XP for netbooks and restrictions apply to the hardware allowed, though this was relaxed only recently.

http://www.networkworld.com/news/2008/070208-microsoft-eases-hardware-terms-for.html?hpg1=bn


Comparing products from a single manufacturer is pretty pointless, it only tells you about that manufacturer, nothing else. You could go to Dell and find a $350 saving on the same hardware by choosing Ubuntu instead of Vista and make an entirely different argument. You could look at different versions of the Eee PC and find it works both ways according to which model you choose and who you buy it from. Waste of time.


Paul you fucking cretin, there's a difference in finding something works/doesn't work and saying that because it works/or not for me it works/or not for everyone.

You claim that

"Gnash is a big, stinking, festering pile of poo"

"Gnash is notoriously unstable and unused"

"..gnash: after years of development it's still unusable."

But can offer no evidence, example or justification.

Not one example.

Your comments suggest you haven't even used a recent version. Your comment "Because Adobe seems to be a lot quicker about releasing new versions than gnash is" even suggests a basic lack of understanding of what Gnash is.

But somehow you believe you're in a position to assess it and that you opinion is worth something. All I can do is try it for myself, which I did in the past and again this evening. I also spend time on several Linux and Debian forums and I know that gnash is not a disaster, people are fairly surprised at how well it works. It works well with Flash 7 sites, as claimed. As someone else wrote here:

"gnash does what it does well; emulating flash 7....if gnash isn't what you need, then don't use it. there's no point being a whingy little moron about it."

That's it. That's what people usually find (that gnash does Flash 7 and that you're a whingy little moron). If you tried gnash 6 months ago or 3 months ago well guess what? It's under constant development, it's changed a lot. There's no point judging a developing project on how it performed pre-alpha or alpha when it's currently in beta. It actually works well enough for some Flash 8 sites but lacks features on them like fullscreen toggling. For most of us that isn't enough so we use Adobe's product.

Anonymous said...

And exactly what did Microsoft do to kill off Apple, a much more serious threat?

THEY CAME OUT WITH A BETTER VERSION OF OFFICE THAN WHAT WAS AVAILABLE ON WINDOWS.


If memory serves me correct, Office for Mac lacks VBScript support, which is a major feature that even enhanced versions of Open Office have.

Regardless, Microsoft felt the demand for Office was high enough to warrant a Mac port. It was done as an additional revenue source not to aid their competitor.

If anything Apple aided Microsoft by switching to x86 Intel hardware and implementing BIOS and MBR compatibility within their EFI-based hardware. The Bootcamp bootloader and setup tool was even bigger for Microsoft.

Now its simple as pie to install Windows natively on a Mac.

In the past, when Apple merely sold computer systems and had no music monopoly, they weren't doing so well.

In fact Apple almost went bankrupt in the the mid to late 90's. Now with additional sources of revenue, Apple can afford to sale pricey machines and still own a sub-10% of the market.

Yup, Microsoft is really nasty toward those competitors. Kill them with kindness.

Indeed they are unless you happen to become a partner in business.

Anonymous said...

@julian67: "That's the worksforme argument again."

OK when I see posted here a verifiable article of proof of every complaint about free software then we can all play on the same level field.

You have proof. Read the links in the original blog post. They refer to admissions from verified lusers that FOSS software is utter crap.


But what does a screenshot prove? You could choose to disbelieve that too.

It would go a lot further toward proving your case than your wild-ass claims. I simply don't believe you.

Actually the example I used is real. I'll see if I can find a decent free screenshot utility for Windows before I get to diagnosing and fixing the problem.

Just press Print-Screen, you dumbfuck, and paste into MSPaint, for chrissakes. Sheezus. No wonder you can't get your fucking platform running. You can't even copy-paste a goddamned screenshot ... that's it, pack everything in your desk, and turn in your cardkey to security immediately...

julian67 said...

I don't want to print the whole screen, just the active windows, and I wanted the ability to add a one or two second delay, which is why PrtScrn isn't ideal. Terrribly complex I know and I hope this isn't too much to take in all at once.

Anyway I found a reasonable free screenshot tool. I'm not sure what it will prove (nothing imo but wtf). Watch this space.

Anonymous said...

@julian67: "Comparing products from a single manufacturer is pretty pointless, it only tells you about that manufacturer, nothing else. You could go to Dell and find a $350 saving on the same hardware by choosing Ubuntu instead of Vista and make an entirely different argument. You could look at different versions of the Eee PC and find it works both ways according to which model you choose and who you buy it from. Waste of time."

Given that the Linux package is a bag of failure, I'm not surprised that you would feel that way. Nonetheless, it can't hurt to take a wider look at all the options available for cruddy, low-powered notebooks ...

Except, the problem is ... when you start to look at scaled-down versions of Windows on low-powered notebooks, they start looking better than Linux ... For example, here's what Michael Gartenberg of Jupiter Research had to say about XP on the XO (Linux fanboys have been saying for months that running XP on the XO is "impossible", "won't work", "too bloated", "please, Microsoft, don't shit in our hats", etc):

http://weblogs.jupiterresearch.com/analysts/gartenberg/archives/2008/08/laptop_mags_xp.html

"[Windows XP] is much better than the stock OLPC OS IMHO and I imagine much more useful as well."

But can offer no evidence, example or justification. Not one example.

Pot, meet kettle. Look, if you want some respect, provide evidence, yourself. Otherwise, you're just another fanboy with an ax to grind and lies to spin...

Anonymous said...

@anonymous: If memory serves me correct, Office for Mac lacks VBScript support, which is a major feature that even enhanced versions of Open Office have.

Apparently, Mac users don't care.

Mac Office sales soar on Apple's gains
http://news.cnet.com/8301-13860_3-9942251-56.html?tag=st.top

"Microsoft's Mac unit is set to disclose on Tuesday that copies of the new Office for Mac 2008 are flying off the shelves at three times the rate of its predecessor. The company wouldn't disclose sales numbers, but said the sales are the highest in the 19-year history of the unit. That continues a trend that has been going on for some time...Finally, the company announced that, for the next full release of Office for Mac, it is bringing back support for Visual Basic scripts."

Anonymous said...

@julian67: "I don't want to print the whole screen, just the active windows, and I wanted the ability to add a one or two second delay, which is why PrtScrn isn't ideal. Terrribly complex I know and I hope this isn't too much to take in all at once. "

Don't be so fucking gay, and just take a screenshot, Francis.

Anonymous said...

[Windows XP] is much better than the stock OLPC OS IMHO and I imagine much more useful as well.

Well of course he would say that. The Sugar UI is worse than diarrhea and its the only thing these poor kids see.

This comes from a Fedora user and supporter, which the Sugar OS is based on--far far underneath.

These kids would have no idea how to enter the terminal and manage to install a real desktop environment like Xfce, Gnome or KDE.

julian67 said...

Linux or XP the better deal on a netbook? I don't know or care, was only pointing out that looking at the recommended prices of a single manufacturer doesn't say a lot, and everyone arguing the point is a waste of time. If it isn't the OS you want on the hardware you want the price doesn't matter too much anyway.

Here are my exciting screenshots

http://img122.imageshack.us/my.php?image=sysinfocs3.jpg
http://img165.imageshack.us/my.php?image=completead0.jpg
http://img241.imageshack.us/my.php?image=iegh7.jpg
http://img164.imageshack.us/my.php?image=ffnv5.jpg

Like I said they don't prove anything but if it makes someone feel happier.....

It's pretty interesting to be called a liar on the basis of describing a bug in XP/IE/Flash but then we all know that all these products are perfect and the fact that lots of people use them without issue proves they have no bugs...right?

Worksformeandmeandmeandmeandme is so much more convincing than silly old worksforme.

cue accusations of manipulated screenshots....

Anonymous said...

Well of course he would say that. The Sugar UI is worse than diarrhea and its the only thing these poor kids see.

Apparently, diarrhea is a delicacy among FOSS users.

This comes from a Fedora user and supporter, which the Sugar OS is based on--far far underneath.

I applaud your frankness.

These kids would have no idea how to enter the terminal and manage to install a real desktop environment like Xfce, Gnome or KDE.

Right. But, somehow, this was supposed to be about "edj00kAshun". Problem is, these kids don't even know where to start. God help them. I wouldn't wish FOSS software on any of these kids. They have enough problems.

Anonymous said...

Here are my exciting screenshots

Why are you running IE7 under Wine?

julian67 said...

I don't know why you think it's running under wine. Care to elaborate?


I don't have wine installed on Debian and I definitely don't have wine installed on XP!

It's running on XP SP3.

julian67 said...

btw I'm running blackbox in place of the explorer shell (this is not exactly the same thing as explorer file manager, I still use that) because it's the best way i found to have multiple desktops and the window management features I like (sticky windows, rollup/shade etc), the bar you see at the bottom is the bblean bar.

There are other ways to do it such as dexpot, emerge desktop, virtualwin etc but tbhis worked best for me.

Anonymous said...

He's joking.

IE 7 doesn't work in Wine yet.

IE 6 with IE 7's trident engine hacked in works (which is pretty simple to do) but not the IE 7 interface.

I've managed to use IE 7 and IE 8 via my STS on Linux but its all pretty pointless because I don't even use IE on Windows.

Anonymous said...

I don't know why you think it's running under wine. Care to elaborate?

For example, the window title bar in completead0.jpg is non-standard. It SHOULD have a little IE icon in the top lefthand corner, not a little dot. That looks way too Motif-like to be Windows. Nice try, though. No wonder you didn't want to show your entire desktop.

Anonymous said...

IE 7 doesn't work in Wine yet.

Um, yeah, it does.

http://www.tatanka.com.br/ies4linux/page/Main_Page
"See our latest beta with exclusive IE7 support!"

Anonymous said...

Better read closer dude...

Currently IEs 4 Linux (due to Wine limitations) can't install ie7 full package. What I have done so far is to install ie7 render engine into an ie6 user interface. So, right now, you will not get all that new interface and new features. But, what's important for web development, you already can get its engine to test your sites.

julian67 said...

Two more screens so you can satisfy yourself that this is indeed native XP SP3

http://img126.imageshack.us/my.php?image=winsg1.jpg

http://img148.imageshack.us/my.php?image=win2vg0.jpg

This is the admin account desktop with a horrible wallpaper to remind me I'm using admin in case I forget.

I've no idea if IE7 works in wine, never tried it.

Anonymous said...

There are multiple distributions based on each. Not everything is the same either.

I think almost none of them are the same, which is exactly the point. The different flavors of FreeBSD serves completely different purposes (i.e. as NAS, firewall, embedded, etc.) They're like the 'specialist' distros for Linux, not like what Fedora is to Ubuntu.

Beyond a very bare-bones system, FreeBSD needs external software (GNU tools) to render it useful.

Actually FreeBSD doesn't use GNU tools, they ship their own ar and ls and mv and cp and such. They do use GCC but I think it's been modified before being packaged with the operating system. Any other external software needed are already outside of what GNU provides for linux.

Anonymous said...

I can buy that Adobe is spending precious developer resources to release a half-assed version of Flash; they've been doing it for years longer on Mac OS X. But not as part of a conspiracy to promote Windows, but just because they allocate 80% of their developer resources to the Windows platform.

Cause like 90% of their customers are on it ya know?

Anonymous said...

There are multiple distributions based on each.

Forks, yes. All BSDs are technically forks of BSD4.4. More accurately, SunOS was a fork of an older BSD release, as was Nextstep. And OpenBSD forked from FreeBSD, and NetBSD from OpenBSD (or vise versa), and Dragonfly from FreeBSD 4. Still, they're distinct separate operating systems, not flavours, not distributions.

PC-BSD, for example, has its own package format that is not compatible with ports.

Brilliant how that's your answer to "with different kernels, different package mechanisms, different userland, different purposes, and even different sets of supported archetectures.".

No shit, Sherlock.

Its not difficult to grasp. Virtually all other operating systems are developed in that single-entity fashion.

Again, you're just taking what I say and throwing it back at me, and trying to look clever. The point is FreeBSD does this, this is how sane people develop an operating system.

Linux is not developed in this fashion. It's an anchepelagio of hundreds of individual, externally writen and maintained applications kludged together into a distribution.

Beyond a very bare-bones system, FreeBSD needs external software (GNU tools) to render it useful.

Cute, but no, that's not how it works.

First, the userland is part of the base system, that is to say, maintained by the FreeBSD team.

Second, Aside from GCC (which is to be phased out once there's a BSD lisenced compiler suite availible), there is no GNU software in the FreeBSD base system. The tools you think are GNU tools (ls, mv, ln, grep, fetch, more, and all that fun stuff) are either written from the ground up by the FBSD team, or based on the Berkeley's own code.

The userland is maintaimned by the FreeBSD team. You have to keep in mind that the GNU userland is in fact a reimplimentation of these same tools. it's the basic Unix userland.

Fail.

There are multiple distributions of FreeBSD, NetBSD, and OpenBSD that have their own unique features. Linux is mostly the same way in that most distributions are based off another.

Do you even understand what the difference between a distribution and an operating system is?

I've had to reiterate this, what, three times now? Here we go again...

Linux = Distributions = mishmash of individual, externally maintained and developed pieces kludged together into a distribution.

BSDs = complete operating system developed in a manner much closer to the cathederal method, where all the pieces are maintained at the same place, by the same team, and are built as a whole system, not individual pieces kludged together.

Furthermore, the BSDs are distinct operating systems (not distributions, or flavours of each other) in that they each have their own, distinct kernel and userland.

Linux distributions are distributions (and not distinct operating systems) as they ship the SAME kernel, and largely the same userland.

Really, how is this distinction so difficult to understand?

A very large hub for general linux documentation, books, and HOW-TOs.

I'm well aquainted with TLDP, thank you. In addition, every distribution has their own set of documentation wikis and HOW-TO articles.

I'll give credit where credit is due, Gentoo's documentation is top notch, as fare as Linux goes. However, compare your built in manpages on Linux to the manpages on FreeBSD.

And still, the initial point remains, compare the quality of the average Linux distribution's wiki's docs to the handbook. The FreeBSD handbook wins hands down.

Further yet, a how-to is not the same as documentation. A how-to, as the term suggersts, shows you how to attain a given task, documentation on the other hand, also as the term suggests, documents how a system or given application works.

Those are two very different things, and citing howtos in a discussion about quality documentation really just misses the point.

As I said, the docs in the handbook are second to none. The administrative tasks being trivial and painless to perform in FreeBSD is a direct consiquence of having top notch documentation.

The lack of FreeBSD how-tos and the general lack of how-to sites (and can think of only bsdguides.org off the top of my head) is, again a result of top notch documentation. Step-by-step how-tos aren't generally needed if you understand how the system works.

Yes, FreeBSD is a server OS by default but someone decided they wanted to favorably compare it with Linux.

It's designed to be a server OS first and formost, and a workstation secondly (not how there's a diastinction between a desktop and a workstation). It's nothing "by default" it is what it is designed to be: highly specialized. At no point does linux influence that in any way.

The distinction between server and workstation/desktop is only a set packages like Apache, Perl, PHP, MySQL, and OpenSSH. Many desktop distributions bundle these anyway but simply disabled their service daemons.

Again, you're looking at it from a Linux-centric point of view. The whole system is designed, from the ground up, to be specialized as server OS.

My father's Fedora 9 installation is strictly desktop-oriented, however, because I uninstalled all server tools.

Because in Linuxcountry, all distributions have the same kernel and base system. That being said, the distinction between a "Desktop" Linux distribution and a
"server" Linux distribution is purely a matter of which packages are installed and which aren't.

Again, you fail to grasp the distinction between a Linux distribution and a BSD operating system. FreeBSD is a server OS not because the server tools are there by default (they aren't. In fact no services are activated by default, and there is no http daemon or relational detabase server as part of the base system), but because the entire system, kernel and all, was designed, specialized and optimized to be a server OS.

A perplexing amount of lusers just can't fathom the concept of highly specialized OSes. It's the Unix philosophy, "do one thing, and do it well". They can't seem to comprehend the concept that a system specialized for the server and a system specialized for the desktop are designed differently.

Seeing as you seem convinced that the only differentation between server Linux and desktop Linux is the packages, all that says is that Linux isn't specialized for either. It tries to be a general purpose OS, but we all knew that already, right?

Nevertheless there are desktop and workstation variations of FreeBSD that include X server and a desktop environment by default.

For the last time. There are NO variants of FreeBSD. Not in the way you seem to think. BSD isn't Linux. FreeBSD is FreeBSD. OpenBSD is OpenBSD. PC-BSD is PC-BSD, Dragonfly is dragonfly, MirOS is MirOS. They aren't variants of FreeBSD, they're distinct operating systems.

Besides, Broadcom wireless controllers are very common in laptop equipment.

I still never understood why non-developers feel the need to run server OSes on their laptops. Just because people do it, doesn't change that it was designed for the server.

Hurd is an example of a theoretical project. The FSF even admitted that if they were aware of the Linux kernel at the time that the Hurd project would have unlikely ever commenced.

Way to miss the "for shits and giggles" part.

No, there has been desire for drivers on x64 FreeBSD for years and the developers have been glacier slow at implementing the necessary API hooks in the kernel.

That possibly has to do with the developpers being busy developing a specialized server OS.

Isn't it ironic that users of a server OS would want performance graphics drivers?

No, it really isn't. It doesn't change that FreeBSD continues to be a server OS, devoting developer time to high performance graphics drivers detracts from devs working on developing a.... server OS.

https://www.linuxfoundation.org/lsb-cert/productdir.php?by_lsb

Jebus. Seriously, do you have any reading comprehension skills at all? I cited LSB, obviously I'm aware that it exists.

You obviously missed the part about how it's a standard that nobody uses.

Anyway, if Flash were to target the LSB libraries and specification then the users would relentlessly encourage their distributors into supporting it until they do.

So what you're saying is essentially, instead of the Linux community getting their heads out of their asses, putting their egos aside, and agreeing on a single, stable, standardized API, Adobe should target a standard that nobody uses in the hope that the users will then pester the Linux community into getting their heads out of their asses, putting their egos aside, get together and embrace a standard API?

Jebus. Do you realize how absurd of a counter-argument that is? Why not Save a bloody step?

Flash player can act as tremendous leverage at pushing standardization across the distribution spectrum.

Just as a standard, stable API can act as a tremendous push for not only working flash support, but support for other third party software. Chicken and the egg.

Does the Linuxulator even support 2.6.16 yet? I remember there was a Google SOC project to enhance it for 2.6 support.

The API changes with every other point release, causing system-wide breakage on Linux. Do you really want to try building a binary compatability targeting that?

Nevertheless. upport for 2.6.16 syscalls is enabled by default in 8.0 CURRENT, 7.0 defaults to 2.4 sycalls with support for 2.6 syscalls availible but disabled by default.

Part of the reason they stuck with support for 2.4 support by default for so long is because 2.4 was stable. 2.4 was from the days when the Linux kernel had a much more sane development model, where there was the stable branch (2.4), and the development branch (2.5, where new APIs and changes to existing APis were introduced and brought to a stable state, resulting in the 2.6 branch).

The 2.6 branch has a completely demented development model. There is no longer separate stable and development branches. the stable branch IS the development branch. New APIs and drastic changes are introduced within the stable branch. Think about that for a moment!

Many modern commercial Linux apps do not work correctly in Linuxulator because they utilize functionality of newer kernels.

Many commercial Linux apps that worked correctly in 2.6.n don't work correctly in 2.6.n +1. That's been whole of the main points all along, the need for stable APIs.

Cedega was an infamous example of this. However, Cedega is dog shit these days so that probably doesn't matter much.

Then why bother bringing it up?

Anyway, back on topic... Gnash currently implements a large subset of Flash 7 functionality. However, Flash 10 is around the corner and further pushes Gnash behind.

Precisely the reason for the parenthetical "Gnash doesn't count" comment. I don't think that anyone can sensibly argue that Gnash is all that terrible of a drop in replacement for Flash 7. The question with Gnash is simply what's the point? How useful is a drop-in replacement for an obsolete platform? The vast majority of flash content online is in Flash 9 format, and as you said, Flash 10 is right around the corner.

Tim said...

Alright, +1 beer for you since you resurrected 'bob' for argument sake.

But its SO much easier to pick on edlin (or ed, or both).

Anonymous said...

Linux on the desktop is just a fancy typewriter.

It's not free, it's worthless.

Anonymous said...

Brilliant how that's your answer to "with different kernels, different package mechanisms, different userland, different purposes, and even different sets of supported architectures.".

Yes it is. PC-BSD is an example of a FreeBSD derivative. It takes FreeBSD then adds the PBI package format and other desktop tools.

Its the same thing as say Ubuntu basing itself off Debian then adding additional features on top.

Again, you're just taking what I say and throwing it back at me, and trying to look clever. The point is FreeBSD does this, this is how sane people develop an operating system.

Linux is not developed in this fashion. It's an archipelago of hundreds of individual, externally writen and maintained applications kludged together into a distribution.


That just your subjective opinion. How effectively these pieces are assembled together depends on the distributor.

Cute, but no, that's not how it works.

First, the userland is part of the base system, that is to say, maintained by the FreeBSD team.

Second, Aside from GCC (which is to be phased out once there's a BSD lisenced compiler suite availible), there is no GNU software in the FreeBSD base system. The tools you think are GNU tools (ls, mv, ln, grep, fetch, more, and all that fun stuff) are either written from the ground up by the FBSD team, or based on the Berkeley's own code.

The userland is maintaimned by the FreeBSD team. You have to keep in mind that the GNU userland is in fact a reimplimentation of these same tools. it's the basic Unix userland.


Still very basic user-land tools.

Are you telling me you don't use the ports tree, which is comprised largely of GPL-licensed software?

What about Bash? You use the shitty default CSH shell instead?

Do you even understand what the difference between a distribution and an operating system is?

FreeBSD, OpenBSD, NetBSD, and DragonFly BSD are seperate operating systems but DesktopBSD, PC-BSD, FreeSBIE, Frenzy, OliveBSD, MirOS, mOnOwall, etc. are all derivatives of those core OSes.

A derivative is equivalent to a distribution in BSD land.

I'm well aquainted with TLDP, thank you. In addition, every distribution has their own set of documentation wikis and HOW-TO articles.

snip.


Then that is all you need to know. TLDP is the hub for general Linux documentation. Distributions also have their own documentation that might provides the specifics.

Again, you fail to grasp the distinction between a Linux distribution and a BSD operating system. FreeBSD is a server OS not because the server tools are there by default (they aren't. In fact no services are activated by default, and there is no http daemon or relational detabase server as part of the base system), but because the entire system, kernel and all, was designed, specialized and optimized to be a server OS.

Desktop-oriented FreeBSD derivative distributions typically recompile the kernel for more optimized desktop app usage. For example, DesktopBSD and PC-BSD may compile in the new ULE process scheduler.

Seeing as you seem convinced that the only differentiation between server Linux and desktop Linux is the packages, all that says is that Linux isn't specialized for either. It tries to be a general purpose OS, but we all knew that already, right?

Yet still does just as good of a job as FreeBSD as a server and much better than any *BSD (besides OS X) as a desktop OS with supported hardware; we'll leave out the Windows comparison, though.

For the last time. There are NO variants of FreeBSD. Not in the way you seem to think. BSD isn't Linux. FreeBSD is FreeBSD. OpenBSD is OpenBSD. PC-BSD is PC-BSD, Dragonfly is dragonfly, MirOS is MirOS. They aren't variants of FreeBSD, they're distinct operating systems.

FreeBSD, OpenBSD, and NetBSD, yes but the others are derivatives. This isn't exactly the same thing as a distribution in that they don't first begin life as a collection of individual components. However, there are still some truth to that with BSD derivatives. For example, PC-BSD and/or DesktopBSD add Xorg and drivers, CompizFusion, and their own desktop tools. These additions are individual components not part of the base FreeBSD.

A Linux distribution is an OS and there can be derivatives of the assembled OS as well.

I still never understood why non-developers feel the need to run server OSes on their laptops. Just because people do it, doesn't change that it was designed for the server.

According to you, it's secondary goal is to be a workstation OS and laptops are very popular in that usage scenario.

No, it really isn't. It doesn't change that FreeBSD continues to be a server OS, devoting developer time to high performance graphics drivers detracts from devs working on developing a.... server OS.

Well the users want it so you're essentially saying that FreeBSD shuns it's users?

NVNews has a huge thread about x64 drivers on FreeBSD. Over 89,000 views and 344 replies on that thread alone.

Jebus. Seriously, do you have any reading comprehension skills at all? I cited LSB, obviously I'm aware that it exists.

Obviously you don't because if you did you would have noticed that there are a handful of distro that support LSB along with a few notable commercial products. There are likely many more developers utilizing LSB libraries but simply haven't bothered to officially certify their products.

So what you're saying is essentially, instead of the Linux community getting their heads out of their asses, putting their egos aside, and agreeing on a single, stable, standardized API, Adobe should target a standard that nobody uses in the hope that the users will then pester the Linux community into getting their heads out of their asses, putting their egos aside, get together and embrace a standard API?

That is exactly what I'm saying. LSB is the only real standardized solution that the Linux world can hope to achieve in the near future.

If Flash utilized LSB libraries then the distributions in question would also need to support LSB and their libraries for the Flash plugin to function.

The high demand for Flash on Linux therefore translates to more LSB adoption.

Nevertheless. upport for 2.6.16 syscalls is enabled by default in 8.0 CURRENT, 7.0 defaults to 2.4 sycalls with support for 2.6 syscalls available but disabled by default.

The 2.6 branch has a completely demented development model. There is no longer separate stable and development branches. the stable branch IS the development branch. New APIs and drastic changes are introduced within the stable branch. Think about that for a moment!

Many commercial Linux apps that worked correctly in 2.6.n don't work correctly in 2.6.n +1. That's been whole of the main points all along, the need for stable APIs.


I agree but such standarization will never happen within the kernel development process. An unstable ABI is a considered a feature (for better or worse) and strongly endorsed by the Linux developers. Linux developers have zero interest in supporting backwards compatibility at the ABI level because eventually baggage builds up and cripples further development.

An unstable ABI seems to really only effect proprietary driver developers anyway.

Most Linux developers oppose developers linking proprietary driver developers with the GPL Linux kernel.

In fact, at one point developers contemplated blocking proprietary drivers from hooking with kernel interfaces altogether.

At the very best the legality of linking proprietary code with the kernel is questionable. Linux is licensed as GPL not LGPL so the linking issue exists.

Anyway, you'll never ever witness commercial applications targeting rolling release/bleeding-edge distributions like Frugalware and Arch.

Most applications are not dependent on the kernel features, especially if they have no accompanying kernel module that needs compilation.

I think LSB targets applications that aren't dependent on specific kernel features.

Then why bother bringing it up?

Its an example nonetheless. But again most applications aren't specific on kernel functionality, which probably explains why a good majority of Linux applications worked via Linuxaltor.

Precisely the reason for the parenthetical "Gnash doesn't count" comment. I don't think that anyone can sensibly argue that Gnash is all that terrible of a drop in replacement for Flash 7. The question with Gnash is simply what's the point? How useful is a drop-in replacement for an obsolete platform? The vast majority of flash content online is in Flash 9 format, and as you said, Flash 10 is right around the corner.

Nothing we don't disagree about here.

Except that Gnash's Flash 7 support isn't complete either. Gnash implements a mere subset of Flash 7's SWF features and is buggy at that.

I'll admit however that its sad that the Internet has become as proprietary as it has.

julian67 said...

I think we can all see and agree that the BSDs have some awesome powerful word processors.

MommaSaid said...

@julian

It looks like you enabled and disabled your XP theme in between

http://img165.imageshack.us/my.php?image=completead0.jpg

http://img241.imageshack.us/my.php?image=iegh7.jpg

The scrollbars look different.. Any particular reason why?

BTW I've used bb4win, but found that it didnt integrate well. Can you use shell extentions?


On the topic of flash in the following use case:

1> Install XP/Vista
2> Update to latest servicepack, etc
3> Install flash

vs.

1> Install any linux distro
2> Update to latest patch
3> Install flash

Discounting hardware and driver issues the windows case in my experience hits closer to 100%.

I dont think you can say that same for linux.

Anonymous said...

Will the Micro$oft fanbois please stop spreading FUD about Flash in Linux?????? I've been using Ubuntu for 2 years now and I've NEVER had any problems with Flash!!!!!!!

Anonymous said...

So it works for you :)
As a llinux user i must it doesn't work on all distros - and in must cases its very unstable.

George said...

As a Flash dev. I can vouch for Julian's issue. It's not that the player itself is broken, it's that IE7 messes up some value that reports the Flash plugin version. When a site just tries to throw the content out there everything is fine, problems only arise when the page is using javascript flash detection methods (like SWFObject) to detect the plugin version before embeding the content via actionscript.

http://blog.deconcept.com/2006/12/08/corrupt-flash-player-install-after-ie-7-upgrade/

Anonymous said...

Will the Micro$oft fanbois please stop spreading FUD about Flash in Linux?????? I've been using Ubuntu for 2 years now and I've NEVER had any problems with Flash!!!!!!!

WorksForMe™™™™™

julian67 said...

No I haven't changed the theme but I have an admin account for admin and a restricted account for everyday use and what you're seeing is a screenshot from each. I've spent a little time to have the restricted account set up nicely because it's the one I see every day. With the admin account appearnce doesn't matter at all and I haven't paid any attention to it beyond adding an odd wallpaper just so I don't forget I'm running with full admin rights if I log in as admin with brain in slow gear.

My Windows install is XP SP3, fully up to date and no unsigned drivers except daemon tools, no warez. Clearly Flash works because Firefox runs it perfectly. A driver issue is unlikely. As I'm almost always in the restricted account I use Run As where possible if I need to perform an admin task. I've noticed this can be unsatisfactory in several ways and quite a lot of application installers are poorly written (i.e. with the expectation that the admin account is the default and no expectation of multiple or restricted accounts existing), for example installing some applications places menu entries, desktop shortcuts and quicklaunch buttons only for the admin account. Bizarrely some installers add links for the restricted account from which i'm running run as and ignore the admin account! This is the joy of running a semi-multi-user OS with software written by people who can't conceive of more than one account at all. Better installers get it right but they are the minority. I often have to do these things myself if I want them done right. On top of this is the issue of programs that simply fail to even run or often run but can't save settings unless run as Admin. this is usually because the doofuses who write them think C:\Programs\someapp is the right place to store user configs instead of docs&settings\username\application settings. I try to avoid software like this but everyone is at it, even MS are guilty sometimes.

BBlean seems really good to me. Explorer context menus all function as expected. The only issue is that I had to manually add synaptics touchpad helper and power4gear utility launchers to my startup folder. I guess a lot of this stuff is written with the sensible expectation that explorer will be the default shell, but it's just a few minutes to straighten out stuff like this. The benefit I get is a noticable chunk of memory being freed and sensible window management with multiple desktops and a full menu available with a right click anywhere on the desktop. I don't care that much about the look, if explorer had better window management and multiple workspaces that would be fine.

In Debian installing flash is actually trouble free. It's in the unofficial but widely known and used debianmultimedia repository and it's a no brainer either via a gui package manager or a simple command. You can just as easily install it via the in browser prompt the first time you visit a website that requires it, same as in Windows. And the one flash plug-in works for all my gui browsers.

It seems someone else has replied on this subject while I've been writing this and offered some useful information. I had started to think it was a problem with the installer or a permissions/path problem resulting from using run as but I'll have a close read and see what I can find out.

Anonymous said...

@julian67

//The anonymous poster who thinks a problem with flash/IE combo can be attributed to hardware is only very slightly more stupid than you.//

Or slightly more humorous, trying to alleviate the tension. I'm fully aware that the hardware couldn't be a problem, I was just throwin' fuel on the fire of typical rants.

Please get a sense of humor and/or learn to detect sarcasm ... especially when it ends with a :-) ... could've sworn that would have tipped you off, but whatever.

DamnRight said...

Usually I like your rants, but this is just sad.

How is it not Adobe's fault for making a shitty, crashy plug-in? It's not just sound that sucks: it crashes all over the place. It simply doesn't work.


Congrats! You completely missed the rant's point. Which is ...

It doesn't matter how totally broken or completely awesome the plugin is, or whose fault it is. What matters is that a sane dev community would make every effort and devise fixes, workarounds, hacks, patches and kludges to get it to work, right now, with an emphasis on providing a smooth, painless, transparent experience for a Flash user.

What do you get instead? Whining about evil conspiracies (the close sourced Man keepin' ya down) and a demented project to recreate Flash from scratch, which will never catch up and is doomed to suck for eternity.

You've seen it with the Foxconn thing, and it's always the same pattern. Change one line of code, or create a tinfoil shitstorm?

Bodhibuilder said...

As some freetard once wrote on some Ubuntu forum: flash is only for those annoying ads, so who needs it anyway. Using linux is an art of renunciation. Hence this cult scent around linux users. It's just a geek version of the Krishnas, only more annoying.

Anonymous said...

That just your subjective opinion. How effectively these pieces are assembled together depends on the distributor.

The point wasn't how effectively they're pieced together, it's that they're maintained externally, and pieced together, not built as one.

Are you telling me you don't use the ports tree, which is comprised largely of GPL-licensed software?

I use the ports tree, quite obviously, but all I have installed from it is Apache, PostgrSQL, PHP, Python, daemontools, vim and perl. None of which are GPL'ed as far as I know.

Of course, the argument was about GNU tools, being the GNU userland. Something lisenced under the GPL doesn't make it GNU, and certainly not part of the GNU core userland.

What about Bash? You use the shitty default CSH shell instead?

I actually like csh.

are all derivatives of those core OSes.

Obviously. Though to be more precise they're all descendants of BSD4.4. They're derivatives, yes, but they aren't distributions, and OpenBSD being forked from FreeBSD isn't the same as say, Ubuntu being derived from Debian. Ubuntu is a repackaging of Debian, OpenBSD is a full-fledged fork of FreeBSD, with, at this point, a much different codebase.

BSD isn't linux. There is no "BSD kernel" they're all distinct different kernels.

A derivative is equivalent to a distribution in BSD land.

Not quite, The BSDs are forks, there's a big difference between a fork and a distribution.

Then that is all you need to know. TLDP is the hub for general Linux documentation. Distributions also have their own documentation that might provides the specifics.

Excewpt the argument was never about the existence of dicumentation, but about the general quality of the existing documentation.

Desktop-oriented FreeBSD derivative distributions typically recompile the kernel for more optimized desktop app usage.

There's more to it than just a basic kernel recompile, generally speaking. DesktopBSD and PC-BSD are still very young mind you, and as such their codebases haven't diverged much from FreeBSD.

Conversely, look at Dragonfly and OpenBSD, their codebases have evolved into vastly different kernels and systems.

the new ULE process scheduler.

Not to be pedantic, but ULE isn't new, it was introduced in FreeBSD 5.x.

Yet still does just as good of a job as FreeBSD as a server

That's your opinion, frankly my experience differs wildly.

and much better than any *BSD (besides OS X) as a desktop OS with supported hardware

Possibly because OS X is the only one actually designed to be a desktop OS. PC-BSD might target the desktop, but building ontop of a system highly optimized for a server to reach that end is destined for failure.

we'll leave out the Windows comparison, though.

A wise thing to do, since you keep bringing up the desktop.

According to you, it's secondary goal is to be a workstation OS and laptops are very popular in that usage scenario.

Secondary tends to mean "not as high priority as primary" to most sane people.

Well the users want it so you're essentially saying that FreeBSD shuns it's users?

I'm saying that that particular request isn't a priority. Shunning users is more akin to KDE devs announcing that they want or need users, nor do they much care for what their users think.

Obviously you don't because if you did you would have noticed that there are a handful of distro that support LSB along with a few notable commercial products.

Obviously you like to take everything literally. A standard supported by only a handful amoungs the sea of linux distributions is, in practical terms an unused standard. Just like Mozilla is the only browser to actually support SVG, it remains an unused standard, in practical terms.

That is exactly what I'm saying. LSB is the only real standardized solution that the Linux world can hope to achieve in the near future.

And what I'm saying is that if anybody wants third parties to actually bother developing for it, the community is going to have to get together and adopt it first and formost, rather than relying on third parties to essentially target a minuscule niche with a minuscule niche, in the hope that it compells further adoption.

I don't disagree that flash would deffinately strengthen LSB adoption, I'm merely pointing out the catch-22 that strengbthened adoption of LSB as a standard would strength third party development for it.

An unstable ABI is a considered a feature (for better or worse) and strongly endorsed by the Linux developers

That's what I find particularily disturbing.

At the very best the legality of linking proprietary code with the kernel is questionable.

It's perfectly legal if you do it right. The purists are just pissy because there are loopholes to keeping the actual drivers closed.

The whole argument is a purely semantic one of what constitutes derivative work. Some like to believe that hooking into the kernel makes the driver a derived work. So driver makers avoud the headache and hook proprietary drivers into lgpl wrappers, which in turn hook into the kernel. The purists complain that it's still a derivative work because it needs the kernel. Driver manufacturers decide that supporting Linux at all is far too much of a headache, given it's whopping 0.8% of the market, and move on. Honestly I'm impressed that Nvidia even bothers anymore.

Anyway, you'll never ever witness commercial applications targeting rolling release/bleeding-edge distributions like Frugalware and Arch.

That's to be expected. It's the users who go after the unsupported bleeding edge and cry bloody murder when third party applications don't work properly.

julian67 said...

George was right, thanks George!

It seems that upgrading IE6 to IE7 broke flash in IE (but not in Mozilla).

So here's the new headline story:

"Flash me a clue"

No matter how stably, smoothly, efficiently, and correctly Flash runs on a machine, the public will continue to view it as second-rate if Windows keeps breaking. This is the worst example of an honest to goodness 3rd party developer just trying to put food on the table being tied down and bound by a crappy all-american-monopolist-pig-fucker-luser-wintard-corporation OS over which nobody sane or competent has any control.

Aaaah, that's better :-)

Windows update breaks Flash! Flash breaks Windows! Wintards use a broken system! Apt-get devs buy an extra twinkie each! Linspire Torqwrench when interviewed stated "Micro who?" and barfed. R-emenem-S reportedly said "Flash? that's the crappy broken version of Gnash isn't it? Do they have a "Guilty Conscience" or "Just Don't Give a Fuck"? and then fell asleep after ingesting 25 zopiclones and snorting a bunch of crushed zolofts.

shevegen said...

Hi there,

often I agree with you, but this time I do not agree. You stated:

"Users want flash."

No, I don't. I hate flash.

I use it, indeed - but I hate it.
It is a shame that flash became so popular... firefox crashes every so often because of flash. At least on Linux.

It sucks. I hate flash. :-)

julian67 said...

I love Flash but will they ever get it working right with Windows' crappy web browser and breakalot updater? If only it worked right like it does in Debian.

ho hum

Debian is a beauuuutifulll thaaaaaannnng.

Debian, the OS of the Gods.

zippedydoodahzippedydaaay

Choose Debian, you know it makes sense.


Hi I'm a Mac. Hi I'm a Windows PC. Hi I'm Debian, get on your knees bitches.


lalalalalalalala


Flash! aah aah! Flash I love you but we've only got 24 hours to save the wintards! Aah aah!

<-tribute to Freddie Mercury (famous Debian dev and major pervert).

"It was all over in a flash", said one time tycoon Bill G, now a sad wreck of a man living on charity and memories...."One day it was looking so good and all of a sudden windows updated and it all just got fucked up. My computer caught fire, the building was invaded by hairy hackers, my life flashed before my eyes and I woke up with my dick in my hand, my money gone and Melinda had got shacked up with ESR. I don't want to talk about it....can I use your address? I want to get one of those Ubuntu shipit CDs.."

Anonymous said...

@julian67I love Flash but will they ever get it working right with Windows' crappy web browser and breakalot updater? If only it worked right like it does in Debian.


Wow. Maybe we need to get lusers free membership of the National Down Syndrome Congress.


Hi I'm a Debian distro, I own 0.1123% of the market.. Wait!!!! an luser downloaded another ISO!!! We now own 0.112300001% of the market making my opinion 0.000000001% more significant. Please make me feel wanted. Please !!

Cue in luser circle jerk.

Anonymous said...

@julian67: "I love Flash but will they ever get it working right with Windows' crappy web browser and breakalot updater? If only it worked right like it does in Debian."

Hey, douchebag. It isn't Windows or Windows update that's failing. The problem here (if you bothered to actually read about it) is that older installs of Flash aren't handling tightened restrictions on cross-site scripting in IE7. Recommended fix is to uninstall older version of Flash and reinstall latest version. Security's a bitch. But it's a necessary evil.

thecodewitch said...

What the fuck is all this noise? Where the fuck do these linux fundies spawn from??

"I can't get flash working on my Vista box after I raped it with some half assed fluxbox thing. I tried everything!"
Maybe the universe just plain hates you, because fucked if it doesn't work for everybody else.

Try going here. For each version of flash up to version 9, the zip files provide 3 separate ways of installing and using flash - a standalone player, the ActiveX control, and what I assume is the IE plugin. If you can't get it going after this, install ubuntu and get it going in linux. Here is my howto on doing this.

1) Open up a terminal.
2) Open up fugsy.dat in a text editor. If you aren't 1337 enough to know where this is, its in the /etc/gnash.d/RMS/googoo/apt/ directory, n00b.
3) Find the line:

crawf -o -yy -fu:rms:338 -yomama

and change it to:

crawf -o -yy -fu:rms:33xxv -barrelroll -yomama -eats -chickensauce

This puts the two spurving bearings in a direct line with the pentametric fan and prevents side fumbling in the dingle arm.

4) Go to the site with Gnash in it. You know what you doing. Use Konqueror.

5) Download it and install it. Use some combination of rpm, sudo apt-fuckallyoustupidfucks, synaptic, and xEyes. There are some extra files in /etc as well. Ask Freddy, he might know.

Me: Hey Freddy!
Freddy: What?
Me: Wtf Linux wtf??
Freddy: Fuck off!

6) autoconf gnash -eats -shit
7) configure, make
8) After the compile error, go to RMSIsYourGod.cpp
9) Fuck around in there for a bit
10) Go back to the makefile
11) Fuck around there too
12) Do this for 7 hours
13) 2 weeks later, discover some linux fucknut writing in some forum somewhere that you need gcc version 2.448 RC12 - fuck you edition, along with 17 libraries, especially SideFumbler-1.3.55.2.552.5.2.7.so, but be careful not to use it with BarescentSkor-4.2.45.1111.2.so, or you'll get cancer.
14) After you finish paying off your mortgage, there is a good chance you'll get Gnash running. But the only flash files you'll be able to run are some 35 year old Flash 5 and 6 games that feature a square being followed by another square.

Enjoy your life, linux uberdorks!

Anonymous said...

@julian67 loves cock.

Anonymous said...

julian67: "Hi I'm a Mac. Hi I'm a Windows PC. Hi I'm Debian, get on your knees bitches."

What a fucking homo.

Anonymous said...

The point wasn't how effectively they're pieced together, it's that they're maintained externally, and pieced together, not built as one.

Well it isn't necessarily a bad thing when done right. Ubuntu seems to be having problems lately with their use of Debian Unstable upstream packages.

I use the ports tree, quite obviously, but all I have installed from it is Apache, PostgrSQL, PHP, Python, daemontools, vim and perl. None of which are GPL'ed as far as I know.

Of course, the argument was about GNU tools, being the GNU userland. Something lisenced under the GPL doesn't make it GNU, and certainly not part of the GNU core userland.


Wait... there is Daemon Tools for *nix?

I wasn't refering strictly to GNU userland tools but GPL software in general. I already knew that BSD had its own basic userland commands.

I actually like csh.

To each their own.

Obviously. Though to be more precise they're all descendants of BSD4.4. They're derivatives, yes, but they aren't distributions, and OpenBSD being forked from FreeBSD isn't the same as say, Ubuntu being derived from Debian. Ubuntu is a repackaging of Debian, OpenBSD is a full-fledged fork of FreeBSD, with, at this point, a much different codebase.

Yes, I'm aware that OpenBSD and FreeBSD aren't were forked.

However, PC-BSD and FreeBSD is akin to Ubuntu and a minimal CLI install Debian.

DesktopBSD and FreeBSD aren't destined to become forks, judging by the projects' goals. They are simply customized FreeBSD images--a derivative distribution.

Not quite, The BSDs are forks, there's a big difference between a fork and a distribution.

The four stated BSDs are forks but the strict derivatives of those forks are essentially distributions IMO.

Except the argument was never about the existence of documentation, but about the general quality of the existing documentation.

The quality of documentation is sufficient. TLDP's purpose is to compile a collection of quality documentation resources for Linux users.

There's more to it than just a basic kernel recompile, generally speaking. DesktopBSD and PC-BSD are still very young mind you, and as such their codebases haven't diverged much from FreeBSD.

Those two aren't meant to diverge, at least for now. Diverging from the FreeBSD code base would result in having to maintain the entire OS independently.

That doesn't leave much time for their small developer teams to focus on desktop usability.

Not to be pedantic, but ULE isn't new, it was introduced in FreeBSD 5.x.

Yes, however the scheduler wasn't ready as a drop-in replacement for production systems until now. This is why ULE will become the default scheduler in FBSD 7.1

That's your opinion, frankly my experience differs wildly.

They are about the same and achieve similar uptimes. A Linux server with Xorg can potentially be more compelling with GUI front-end solutions.

Granted, X Server can crash occasionally (especially with proprietary drivers) and reduce uptime but things seem to be becoming more stable lately.

Proprietary Xorg video drivers from Nvidia and AMD are slowly improving in quality and performance.

Possibly because OS X is the only one actually designed to be a desktop OS. PC-BSD might target the desktop, but building ontop of a system highly optimized for a server to reach that end is destined for failure.

True, though PC-BSD is probably the most successful BSD desktop effort to date.

Possibly because OS X is the only one actually designed to be a desktop OS. PC-BSD might target the desktop, but building ontop of a system highly optimized for a server to reach that end is destined for failure.

Satisfying the need of their users should be the number one priority.

And yes... Linux distribution developers often fail at this too.

The whole argument is a purely semantic one of what constitutes derivative work. Some like to believe that hooking into the kernel makes the driver a derived work. So driver makers avoid the headache and hook proprietary drivers into lgpl wrappers, which in turn hook into the kernel. The purists complain that it's still a derivative work because it needs the kernel. Driver manufacturers decide that supporting Linux at all is far too much of a headache, given it's whopping 0.8% of the market, and move on. Honestly I'm impressed that Nvidia even bothers anymore.

Perhaps but LGPL wrappers to load kernel binary blobs is commonly considered a circumvention of the GPLv2 license.

Developing kernel-mode drivers for Linux is unintuitive but once they are made its typically then only a matter of a few modifications to the driver code to "fix" whatever ABI changes where made in the new kernel point release.

Over time, I imagine that quite a few #ifdefs pile up in the driver source code in order to accommodate older kernels and their slightly different driver interfaces.

It would appear to me that to maintain a reasonably clean code base would be to periodically remove support for aging kernels and the associated code.

That's to be expected. It's the users who go after the unsupported bleeding edge and cry bloody murder when third party applications don't work properly.

Yes. Distributions with six month release cycles or longer are the way to go because packages have more time to bake and stabilize. The kernel also becomes a more stable target for that specific distribution release.

Anonymous said...

And what I'm saying is that if anybody wants third parties to actually bother developing for it, the community is going to have to get together and adopt it first and formost, rather than relying on third parties to essentially target a minuscule niche with a minuscule niche, in the hope that it compels further adoption.

I don't disagree that flash would deffinately strengthen LSB adoption, I'm merely pointing out the catch-22 that strengthened adoption of LSB as a standard would strength third party development for it.


Well as far as Flash is concerned Adobe only officially supports RHEL and SUSE, both of which are LSB certified.

Anonymous said...

OMFG what a bunch of lusers...

A fish bowl for a desktop? WTF?

http://www.youtube.com/watch?v=ib0V2ND3gd8

tchill said...

I use gentoo and it works for me in 32- and 64-bit.

Anonymous said...

WTF? This blog is for a bunch o M$ b1tch3s... isn't it?

julian67 said...

"Anonymous said...

@julian67I love Flash but will they ever get it working right with Windows' crappy web browser and breakalot updater? If only it worked right like it does in Debian.


Wow. Maybe we need to get lusers free membership of the National Down Syndrome Congress.


Hi I'm a Debian distro, I own 0.1123% of the market.. Wait!!!! an luser downloaded another ISO!!! We now own 0.112300001% of the market making my opinion 0.000000001% more significant. Please make me feel wanted. Please !!

Cue in luser circle jerk.
August 19, 2008 4:13 PM
Anonymous said...

@julian67: "I love Flash but will they ever get it working right with Windows' crappy web browser and breakalot updater? If only it worked right like it does in Debian."

Hey, douchebag. It isn't Windows or Windows update that's failing. The problem here (if you bothered to actually read about it) is that older installs of Flash aren't handling tightened restrictions on cross-site scripting in IE7. Recommended fix is to uninstall older version of Flash and reinstall latest version. Security's a bitch. But it's a necessary evil.
August 19, 2008 4:29 PM
thecodewitch said...

What the fuck is all this noise? Where the fuck do these linux fundies spawn from??

"I can't get flash working on my Vista box after I raped it with some half assed fluxbox thing. I tried everything!"
Maybe the universe just plain hates you, because fucked if it doesn't work for everybody else.

Try going here. For each version of flash up to version 9, the zip files provide 3 separate ways of installing and using flash - a standalone player, the ActiveX control, and what I assume is the IE plugin. If you can't get it going after this, install ubuntu and get it going in linux. Here is my howto on doing this.

1) Open up a terminal.
2) Open up fugsy.dat in a text editor. If you aren't 1337 enough to know where this is, its in the /etc/gnash.d/RMS/googoo/apt/ directory, n00b.
3) Find the line:

crawf -o -yy -fu:rms:338 -yomama

and change it to:

crawf -o -yy -fu:rms:33xxv -barrelroll -yomama -eats -chickensauce

This puts the two spurving bearings in a direct line with the pentametric fan and prevents side fumbling in the dingle arm.

4) Go to the site with Gnash in it. You know what you doing. Use Konqueror.

5) Download it and install it. Use some combination of rpm, sudo apt-fuckallyoustupidfucks, synaptic, and xEyes. There are some extra files in /etc as well. Ask Freddy, he might know.

Me: Hey Freddy!
Freddy: What?
Me: Wtf Linux wtf??
Freddy: Fuck off!

6) autoconf gnash -eats -shit
7) configure, make
8) After the compile error, go to RMSIsYourGod.cpp
9) Fuck around in there for a bit
10) Go back to the makefile
11) Fuck around there too
12) Do this for 7 hours
13) 2 weeks later, discover some linux fucknut writing in some forum somewhere that you need gcc version 2.448 RC12 - fuck you edition, along with 17 libraries, especially SideFumbler-1.3.55.2.552.5.2.7.so, but be careful not to use it with BarescentSkor-4.2.45.1111.2.so, or you'll get cancer.
14) After you finish paying off your mortgage, there is a good chance you'll get Gnash running. But the only flash files you'll be able to run are some 35 year old Flash 5 and 6 games that feature a square being followed by another square.

Enjoy your life, linux uberdorks!
August 19, 2008 4:44 PM
Anonymous said...

@julian67 loves cock.
August 19, 2008 6:10 PM
Anonymous said...

julian67: "Hi I'm a Mac. Hi I'm a Windows PC. Hi I'm Debian, get on your knees bitches."

What a fucking homo.
"


Was it something I said? Gosh!

Hi wintards!

Anonymous said...

what the fuck is adobe flash for?

bitching crashing all browsers may it work for?

does it work, anyway??

Anonymous said...

@julian67

Hi to you too, freetard.

Anonymous said...

Well it isn't necessarily a bad thing when done right. Ubuntu seems to be having problems lately with their use of Debian Unstable upstream packages.

That's still not the point :p The point is that they're very different approaches resulting in very differently designed systems (and marking the distinction between a distribution and a whole system). I'd venture to assert that the pieced together aproach , inherently cannot achieve the the level of cohesion, integration and streamlined nature of a system designed to function as a single entity.

Wait... there is Daemon Tools for *nix?

Not that DeamonTools, mounting images is built in with md. daemontools for FreeBSD watches daemons and restarts them if they unexpectedly quit.

I already knew that BSD had its own basic userland commands.

And not so basic ones, either. Don't underestimate Berkeley's contributions to Unix, and especially to open source Unix.

The four stated BSDs are forks but the strict derivatives of those forks are essentially distributions IMO.

Okay, so there are two alternate repackagings of FreeBSD. I'll give you that much.

The rest are forks having diverged into different and distinct codebases (though much sharing of code still occurs, ex. OpenBSD developed pf and OpenSSH, pf has found its way into FreeBSD, and OpenSSH is included, by default even on the average Linux system; NetBSD developed pkgsrc which runs on more platforms that NetBSD runs on archetectures, as mind boggling a prospectt as that is, etc, etc)

Satisfying the need of their users should be the number one priority.

I'll agree in most cases. However, one must keep in mind that FreeBSD was never designed with the goal of being everything to everyone, it was, however designed to be a robust, stable, secure and high performance server operating system. If users want FreeBSD to diverge from its intended pupose, they're going to have to understand that such requests won't be given high priority, since they're in direct conflict with the project's intended goals.

That's not not say that such feature requests will never be implimented, mind you. However, feature requests which are in line with the primary objective that is making FreeBSD a robust, stable, secure and high performance server
OS are a completely different matter.

They are about the same and achieve similar uptimes.

Uptime alone is a horrible basis on which to make a comparison. Lots of people use it as a means to gauge stability, for example, and cite uptimes in comparisons with Windows, where windows tends to have lower uptimes because updates and installs tend to require reboots, which in no way is a result of poor stability.

Overall performance, reliability and scalability need to factor in as well. In my experience, FreeBSD is a system I can set up, forget about and expect to work properly, even after a buildworld. My experience has shown me to not place that kind of trust in Linux/

I've also observed that FreeBSD sails smoothly under loads that would cause Linux on the same system to grind to a halt, if not die an ungly death. Solaris/Sparc beats the pants off of FreeBSD in this department, mind you.

This is of course based on my own experience with generic installs of each, YMMV.

A Linux server with Xorg can potentially be more compelling with GUI front-end solutions.

A server with a GUI is hardly ever more compelling, not in Unix country, anyway. Further, you say that as though Xorg and these same frontend aren't availible on FreeBSD.

Granted, X Server can crash occasionally (especially with proprietary drivers)

Nvidia's drivers are still the best availible. AMD opened up their drivers, btw.

True, though PC-BSD is probably the most successful BSD desktop effort to date.

Not by a long shot. Don't forget OS X.

Perhaps but LGPL wrappers to load kernel binary blobs is commonly considered a circumvention of the GPLv2 license.

It's sheerly on a semantic level, though. It's more about zealotry and politics than about the lisence. It's not a circumvention of the GPL2, it's a circumvention of headaches. The purists and zealots will cry that it violates the "spirit" of the GPL, whatever that's supposed to mean. They're just pissy about not getting the source code.

I'd presume the aggressive, and extremist nature of GPL3 expressly forbids such practices, and frankly Linus refusing to adopt GPL3 for the kernel is quite possibly the best thing he's ever, and will ever do for Linux.

Well as far as Flash is concerned Adobe only officially supports RHEL and SUSE, both of which are LSB certified.

So one could then argue that Adobe already has chosen to support LSB, and as stated earlier it's up to the rest of the Linux community to put their egos aside, take their heads out of their asses and work together to standardise on a single ABI/API, in this case, LSB. Good luck with that, I really don't see it happening in the near future, given that ever changing and ever breaking subsystems and API are considered a feature.

julian67 said...

LSB as it is now is unlikely to ever gain universal adoption, or even majority adoption. It's very nice for Red Hat & Novell (maybe Mandriva too) and derivatives because RPM is part of the LSB standard, but most of the non-commercial and a few commercial distributions are built on .deb and there's also a sizeable group built on Slackware.

As to its relevance to Flash that just diminished considerably because Adobe have started to release details of their Flash protocols without NDAs (from May this year I think). This probably points to the core issue being more to do with publishing protocols than userland binary compatibility, though of course proprietary software companies who don't want to publish without NDAs will see this differently to those who can do it.

As to the Linux kernel needing a stable interface, this looks terribly important to some and an irrelevance to others. Those others include the kernel developers and everyone whose device is supported by the kernel. Considering that the Linux kernel supports more hardware than any other kernel yet developed it might be worth looking at why the Linux kernel was never intended to have a stable interface (lots of people will get excited at this point because they think the words stable and unstable mean something they don't mean).

Good reading:

http://lxr.linux.no/linux/Documentation/stable_api_nonsense.txt

https://www.linuxfoundation.org/en/Linux_Device_Driver_Model

The lack of a stable kernel interface is only an issue for those who desperately want to avoid revealing information about their hardware, or are themselves entangled with NDAs. It's also a handy fallback excuse for those companies which simply aren't interested in their hardware working with the Linux kernel but don't want to say so explicitly. Somehow major manufacturers like Intel manage to offer their hardware specs without the sky falling on their heads and without their profits vanishing. They find that the workload of developing a suitable driver is shared, their costs diminish and the responsibility for maintaining any subsequently developed driver is taken up by the Linux kernel developers. They don't need a '98 version, a 2000 version, an XP version, a Vista version etc. With every update to the kernel their driver is automatically maintained at no cost to them.

This is really imo a long term issue/process whereas as users we look at the immediate and short term and believe all problems should be solved now and that this can be achieved by the kernel allowing closed binary drivers, having a stable ABI etc. But the benefits are already quite clear, the process has been ongoing for about 15 years and more and more hardware vendors (big important ones) have come round to the idea, AMD/ATI being the most notable recent example.

cue shrieking hysterical flames:

julian67 said...

""I can't get flash working on my Vista box after I raped it with some half assed fluxbox thing. I tried everything!"

codewitch as well as being a lame MS apologist/advocate you're brain dead.

Please explain how the default shell has any effect whatsoever on a browser plug-in or installation of activeX controls or binaries?

Clue: it doesn't. You have no fucking idea what you're talking about.

For what it's worth the upgrade from IE6 to IE7 was done a long time before I changed the default shell.

BECKER said...

You are some serious troll.

I'm not falling for your linux hate, it's waaay too ignorant.

Anonymous said...

becker is a pecker!

Anonymous said...

ARE YOU GET PAID? SERIOUSLY, WHY ALL OTHERS DISAGREE WITH YOU. ANYWAY, MY POINT WANT TO SAY THAT I HAVE THIS OLD LAPTOP (I MEAN IT IS 5 YEARS OLD), MY SISTER WANT TO THROW IT BECAUSE IT COULD NOT KEEP RUNNING MORE THAN 10MIN (YEAH IT IS FROM TOSHIBA). ANYWAY, I TOOK IT AND I PUT XUBUNTU (XFCE WAS CRAP) AND I PUT INSIDE KDE. FOR 3 MONTHS I DID NT EVEN CLOSE THE COMPUTER. BUT I WANTED TO WATCH ALSO FLASH. THAT WAS A FUCKING STUPID IDEA, I PUT THE FLASH AND EVERYTHING ON MY SYSTEM WAS TOTALLY SLOW (AND THE LAPTOP CLOSED). DID YOU REALISE MY POINT? NOP BECAUSE YOU ARE STUPID. SO I WILL GIVE YOU THE SOLUTION: flash was already a shit from the start. ALSO THAT CRAP ABOUT AUDIO AND FLASH IS UNTRUE , EVERY COMPANY SHOULD REALISE THERE CODE OPEN SOURCE SO EVERY DISTRO CAN WORK IT. YOU STUPID ASSHOLE... WHY YOU TALK , WHY YOU TALK , TELL ME MORE , WHY YOU TALK, WHY YOU DONT SAY YOUR NAME IN FRONT OF THE PAGE BECAUSE YOU KNOW WINDOWS LOVERS ARE LESS THAN LINUX LOVERS. BUT DONT WORRY NO ONE WILL HARM YOU, I JUST WANT TO POINT WHEN I WILL LAUGH AT YOU.

Silvio said...

Great post!!
Very funny and has a lot of truth to it.
But for me, I still stick with linux for some reason!
Keep up the rant! :D

Anonymous said...

@julian67

It's obvious that LH is a M$ payed shill. Working his fachist drivel against Linux. All his ignorant hate is offensive to us. He should be banned from the Internets.


- Sir Hat de Tinfoil

Anonymous said...

That's still not the point :p The point is that they're very different approaches resulting in very differently designed systems (and marking the distinction between a distribution and a whole system). I'd venture to assert that the pieced together aproach , inherently cannot achieve the the level of cohesion, integration and streamlined nature of a system designed to function as a single entity.

We agree to disagree then.

But I'll add that the fundamental parts of Linux are only the kernel and some GNU userland tools like glibc, gcc, make, grub, bash, etc.

I don't think there are any distributions in existence that fail at combining those basic, essential components.

Distribution quality tends to be judged on higher-level components, such as KDE 4.0 keeps crashing; my printer or scanner doesn't work; wireless doesn't work; I cannot hear sound, etc.

Customized FreeBSD and OpenSolaris distributions can suffer from the same problems.

Okay, so there are two alternate repackagings of FreeBSD. I'll give you that much.

The rest are forks having diverged into different and distinct codebases (though much sharing of code still occurs, ex. OpenBSD developed pf and OpenSSH, pf has found its way into FreeBSD, and OpenSSH is included, by default even on the average Linux system; NetBSD developed pkgsrc which runs on more platforms that NetBSD runs on architectures, as mind boggling a prospect as that is, etc, etc)


OliveBSD is essentially OpenBSD with a X server; FreeSBIE is a LiveCD version of FreeBSD; Frenzy is another liveCD derived from FreeBSD; MirOS is derived from OpenBSD and synchronized with its code base.

Overall performance, reliability and scalability need to factor in as well. In my experience, FreeBSD is a system I can set up, forget about and expect to work properly, even after a buildworld. My experience has shown me to not place that kind of trust in Linux/

I've also observed that FreeBSD sails smoothly under loads that would cause Linux on the same system to grind to a halt, if not die an ungly death. Solaris/Sparc beats the pants off of FreeBSD in this department, mind you.


I've seen Linux uptimes of over 600 days only be shutdown for updates and maintenance.

Linux scales better than FreeBSD on multi-processor and multi-core systems. Linux's SMP implementation is more advanced.

FreeBSD 7 can scale to perhaps 8 to 16 processors; whereas the latest Linux 2.6 kernel can scale to hundreds or even thousands of processors. This is partially why Linux has great adoption on Supercomputers.

A server with a GUI is hardly ever more compelling, not in Unix country, anyway. Further, you say that as though Xorg and these same frontend aren't available on FreeBSD.

There are commercial server products for Linux and some distributions include GUI methods for configuring Apache.

Whether these work on FreeBSD would depend on the Linuxaltor.

Not by a long shot. Don't forget OS X.

OK... lets just dispel the OS X comparison by saying that its certified UNIX and that it's GUI which renders it user-friendly is proprietary.

PC-BSD is the most successful BSD desktop effort when compared to the other open-source BSD forks and derivatives.


It's sheerly on a semantic level, though. It's more about zealotry and politics than about the lisence. It's not a circumvention of the GPL2, it's a circumvention of headaches. The purists and zealots will cry that it violates the "spirit" of the GPL, whatever that's supposed to mean. They're just pissy about not getting the source code.

I'd presume the aggressive, and extremist nature of GPL3 expressly forbids such practices, and frankly Linus refusing to adopt GPL3 for the kernel is quite possibly the best thing he's ever, and will ever do for Linux.


I agree but the GPLv2 didn't intend for these loopholes exist. Its just that FSF didn't account for these issues at the time--1991.

The GPLv3 very likely patches the loopholes by clarifying the derived works section. I haven't bothered to read the license yet.

I suppose it was ideal for Linus to stick with the GPLv2 license, since many hardware manufacturers still are not willing to hand over their intellectual property (in the form of specifications) to the FOSS community.

So one could then argue that Adobe already has chosen to support LSB, and as stated earlier it's up to the rest of the Linux community to put their egos aside, take their heads out of their asses and work together to standardise on a single ABI/API, in this case, LSB. Good luck with that, I really don't see it happening in the near future, given that ever changing and ever breaking subsystems and API are considered a feature.

Naw, I think LSB-certified software are dependent on LSB libraries. All distributions seeking to support Flash would need adopt the LSB libraries and pass a majority of compatibility tests for the software to function correctly.

Anonymous said...

It's obvious that LH is a M$ payed shill. Working his fachist drivel against Linux. All his ignorant hate is offensive to us. He should be banned from the Internets.

Freedom of speech man.

LH should be allowed to say whatever he wants even if some of us disagree.

Anonymous said...

It's a favorite Loon tactic to deflect and redirect any outrage expressed at the lameass Linux system. Some Loons seem to be using this very tactic on this board now. "Windows has its problems, Linux has its problems - no OS is perfect", they say (or imply).

This is what is known as the Big Lie. Windows problems are in most cases annoyances. Considering that the vast majority of computer users use Windows, any problems will also be fixed relatively quickly. Linux suffers from show-stopping issues. You are expected to be a complete loser, and wait several years for the open sores bums to fix the problems.

No OS is perfect, my ass. Linux is an absolute disaster on the desktop. It can't handle hardware, it can't handle software, and it can't handle people. Years from now, people will still be laughing at Linux's unsteady lurch on the desktop!

.net jerkface said...


The lack of a stable kernel interface is only an issue for those who desperately want to avoid revealing information about their hardware, or are themselves entangled with NDAs.

So what? Maybe they don’t want people to muck around in their hardware with custom drivers. Maybe they want to keep the latest version of the driver on their own website. Maybe they are just a bunch of pricks that don’t like Linux. You know what? All of these reasons are equal in weight because at the end of the day it is their hardware, not yours. Courting potential partners instead of chastising them is the way to gain market share. But then this is common sense to me and I actually feel odd explaining it.

With every update to the kernel their driver is automatically maintained at no cost to them.

You make it sound like once you open your specs the driver will magically appear. It doesn’t work that way and in fact it can easily take months for your driver to get in the kernel. Linux has a poor history of supporting new hardware for exactly this reason.
What happens if a company wants to make a change to the driver after discovering a bug? With the stable abi model they can make the change themselves and have the new driver available for download overnight. In Linuxland they have to contact the driver maintainer and *request* that it be fixed. Hardware companies don’t like being in a position where they have to make *requests* to fix their own product.

But the benefits are already quite clear, the process has been ongoing for about 15 years and more and more hardware vendors (big important ones) have come round to the idea, AMD/ATI being the most notable recent example.

Where are the benefits clear? Linux still has poor driver support for consumer peripherals. Even if your printer works in Linux you will probably not have access to all the printer options available in Windows. Some printers that have a Linux driver still have weird color/contrasting issues.

On the issue of AMD, how long has it taken to get ATI video cards working properly? And you call this a success?

For a couple years OpenBSD actually had better wireless support than Linux and they have probably had less than 1% of the funding that Linux has received.

There are some serious problems with Linux development, beginning with the unstable abi/api. That coupled with the separate kernel/OS model has made a software engineer’s nightmare when it comes to programming productivity and a complete headache for third party companies.
Oh but you say everything is fine in Linuxland? It has been 15 years and Linux still hasn’t hit 1% desktop share. The iphone will probably get 1% before Linux. Win server and IIS are gaining at greater rates than Linux/Apache.

The love affair needs to end. Linux is a trashy whore pretending to be a suitable housewife.
I’m surprised that the big companies haven’t caught on to this. A Red Hat BSD could really help in putting the whore back on the streets.

Anonymous said...

PC-BSD is the most successful BSD desktop effort when compared to the other open-source BSD forks and derivatives.

Way to stack the deck there. OS X a BSD plain and simple, an adapted NeXTSteP, which was derived from BSD4.3. The fact that it has proprietary frameworks and libraries on top of it doesn't change that it's a BSD system, weather a system is BSD-style Unix or not is in no way contingent on the availability of source code.

I could argue that Interix (aka SFU aka SUA aka Window's POSIX subsystem aka a OpenBSD ported to run on the NT kernel) would possibly be the third best attempt at a desktop BSD.

I'd say that Nextstep would come in second, I'd even argue that Nextstep 10 years ago was more desktop suited that Linux is now. But I digress, you're the one who brought up desktops, I've been talking servers the whole way through, go go gadget strawman argument!

Customized FreeBSD and OpenSolaris distributions can suffer from the same problems.

I never claimed otherwise I'm saying that a system designed to be a single integrated, streamlined system tends not to. You're replying that a kludge system derived from one of those systems is subject to the same problems.

Way do ignore the point being made, some people call that a strawman.

I don't think there are any distributions in existence that fail at combining those basic, essential components.

I never suggested otherwise, but why bother with arguments that are actually being made when you can instead address made up ones that are far more suitable and easy to dismiss, right?

This is partially why Linux has great adoption on Supercomputers.

I'm talking about servers, and you're going on about supercomputers. Whereas FreeBSD supports the IntelMP limit of 8 CPUs (with N cores per CPU) on x86, this is a limitation set by IntelMP per board, not by FreeBSD itself. Conversely it supports T1 on Sparc, which itself has 8 cores. But that was way back in 5.0, as far as 7.0 goes, I can't even find a mention of the maximum amount of CPUs it supports.

Nevertheless it's an apples to oranges cpmparison. Supercomputers are large clusters, hell, supercomputers are supercomputers, servers are servers.

You're essentially trying to argue that because Linux is used on supercomputers, it scales better than [other OS] on servers.

I doubt anyone in the FreeBSD team has a fancy 1024 CPU monster from Sun to test this stuff out on, but I also don't think FreeBSD is designed to compete with Solaris on the ultra high end (again, servers, not supercomputers, though there are a number of Solaris-powered supercomputers)

John said...

I think Linux Haters blog is important, but you can't just say that it's not adobe's fault - Most users who have problems with Adobe Flash either blame it on their PC (dumb) and then others just blaim it on Adobe. So now we're on Linux, and you bitch about something that's actually not important, nor can the Linux community fix it.

Sometimes I question whether you're really out to help Linux or just to have an excuse to hate it.

julian67 said...

"The lack of a stable kernel interface is only an issue for those who desperately want to avoid revealing information about their hardware, or are themselves entangled with NDAs.

So what? Maybe they don’t want people to muck around in their hardware with custom drivers. Maybe they want to keep the latest version of the driver on their own website. Maybe they are just a bunch of pricks that don’t like Linux. You know what? All of these reasons are equal in weight because at the end of the day it is their hardware, not yours. Courting potential partners instead of chastising them is the way to gain market share. But then this is common sense to me and I actually feel odd explaining it."


Yes, I agree it's the hardware manufacturers' choice, I have no problem with that and didn't express one. Their reasons are their own business and they don't owe anyone an explanation.

With every update to the kernel their driver is automatically maintained at no cost to them.

You make it sound like once you open your specs the driver will magically appear. It doesn’t work that way and in fact it can easily take months for your driver to get in the kernel. Linux has a poor history of supporting new hardware for exactly this reason.
What happens if a company wants to make a change to the driver after discovering a bug? With the stable abi model they can make the change themselves and have the new driver available for download overnight. In Linuxland they have to contact the driver maintainer and *request* that it be fixed. Hardware companies don’t like being in a position where they have to make *requests* to fix their own product.


No I don't make any claim that a driver will magically appear, but once the driver is written it's a good situation for everyone. As to bug fixing a pre-existing driver this isn't the issue you state. Some manufacturers share that responsibility with kernel developers and some just happily let someone else do it all. An important point is that once a manufacturer supported driver is accepted into the kernel then by definition there is some openness and co-operation between all parties. Nobody has to go begging to anyone else. And many drivers depend on other drivers such as usb/firewire/sound subsystems and if those drivers receive improvements then the drivers depending on them benefit too, with no extra effort on the part of the original manufacturer.

But the benefits are already quite clear, the process has been ongoing for about 15 years and more and more hardware vendors (big important ones) have come round to the idea, AMD/ATI being the most notable recent example.

Where are the benefits clear? Linux still has poor driver support for consumer peripherals. Even if your printer works in Linux you will probably not have access to all the printer options available in Windows. Some printers that have a Linux driver still have weird color/contrasting issues.


The benefit of supporting more hardware than any other kernel is quite a big one ;-) The manufacturers benefit from lower and shared costs of development and maintenance. The consumers benefit from "just works" installation on supported hardware. Generally hardware can be made to work even if not exactly officially supported but it's not always fun and isn't noob friendly. But supported hardware is a different experience. Examples: taking the dreaded Ubuntu/Debian family (because that's what I know best): In Ubuntu how to install the driver for my Safecom USB wireless adapter? Plug it into a USB port, done. Same with my D-Link USB wireless. And built in intel 2200 and 3945 on my laptops. Just hit the on switch, done. Same with my HP printer. Connect it and it's installed (with all features available). Same with my Creative Audigy (whose XP drivers were an abomination), put it in the PCI slot and boot. Done. Out of these the only ones I bought with a view to Linux compatibility were the printer and the safecom wireless, the rest was just what I had from when I used 2000/XP exclusively. I do appreciate that there is plenty of consumer hardware that will demand a lot more effort but most stuff these days is true plug and play.

I have a laptop with a silly sound system (Intel ICH7 controller, stereo pair and subwoofer). It came with XP drivers and I tested it with XP and the official drivers to test the hardware when new. The soundcard driver install bluescreened the machine twice but did actually work on reboot. There is a Vista driver but it's completely useless (hardware detected, all looks good, Vista shows no errors, but there's no actual sound) and this is common, perhaps even universal on this model. In linux last year (when I bought it) I found after some googling that i needed to make a one line change to a config file and got it working. Recently I noticed an update in Lenny included new ALSA configs. I deleted my modified one, installed the new one, crossed my fingers. It works. So Debian Lenny is the only one of XP, Vista and multiple distros that would have sound working out of the box if I installed today. It's a "designed for XP" laptop and to XP's credit despite the bad install it did work but nothing beats just installing or plugging a device and giving it no further thought.


But of course anyone who has a Lexmark printer and Broadcom wifi and switches to linux will endure some serious pain.

Another nice feature of kernel support is motherboard drivers. No working your way through the manufacturer's driver disk. Install the OS and it's all taken care of. And what if your motherboard dies? Put a new one in the case, doesn't have to be same or even similar as long as is same architecture, connect peripherals and HDD and boot. It all works. Something approximate to this used to be possible with XP but that got killed with SP2.

On the issue of AMD, how long has it taken to get ATI video cards working properly? And you call this a success?

For a couple years OpenBSD actually had better wireless support than Linux and they have probably had less than 1% of the funding that Linux has received.


AMD have released hardware specs but their older drivers were bound with NDAs of ATI's partners so couldn't be opened, so the free driver project is one that has started from scratch in effect. I don't have an ATI card so I haven't paid a huge amount of attention but Novell are sponsoring this and apparently things are going well.

OpenBSD wireless: the point isn't as clear as it first appears. OpenBSD supports a more limited range of wireless devices but has the reputation for building extremely good drivers, usually done after gaining the support of the hardware manufacturers. Mr De Raadt is famously keen on smaller Taiwanese vendors who are keen to take advantage of any marketplace advantage available and his strategy is obviously intelligent and fruitful but nevertheless OpenBSD supports many fewer wireless devices than other BSDs and Linux. And if we're talking about consumer/desktop systems the huge drawback in OpenBSD is that the drivers might be spiffy with sanitised bug free code but there is no WPA support (unless this changed in the last 6 months). FreeBSD does support WPA but has no friendly gui tools for them, so it's into the shell to switch access points. There are areas where bash in a terminal is equal to or better than a graphical tool but wireless roaming never seemed one of them to me.

There are some serious problems with Linux development, beginning with the unstable abi/api. That coupled with the separate kernel/OS model has made a software engineer’s nightmare when it comes to programming productivity and a complete headache for third party companies.
Oh but you say everything is fine in Linuxland? It has been 15 years and Linux still hasn’t hit 1% desktop share. The iphone will probably get 1% before Linux. Win server and IIS are gaining at greater rates than Linux/Apache.


I didn't say everything is fine in linuxland, it's pointless to say i did. A lot of companies are getting the benefit of having their hardware supported directly in the kernel (it's their choice, the same as the companies who aren't interested) and obviously they see a benefit and don't see overwhelming problems in the way that people who find linux threatening/annoying. Their hardware keeps working with every kernel release and a stable binary kernel interface is something they have ceased to find relevant because they've discovered they don't need one and stopped worrying about it. I think your comment mixes two issues, kernel interfaces on one hand and application/userland interfaces on the other.

The love affair needs to end. Linux is a trashy whore pretending to be a suitable housewife.
I’m surprised that the big companies haven’t caught on to this. A Red Hat BSD could really help in putting the whore back on the streets.


You were quite interesting until the last paragraph when you let your personality out ;-)

Anonymous said...

You make it sound like once you open your specs the driver will magically appear. It doesn’t work that way and in fact it can easily take months for your driver to get in the kernel. Linux has a poor history of supporting new hardware for exactly this reason.
What happens if a company wants to make a change to the driver after discovering a bug? With the stable abi model they can make the change themselves and have the new driver available for download overnight. In Linuxland they have to contact the driver maintainer and *request* that it be fixed. Hardware companies don’t like being in a position where they have to make *requests* to fix their own product.


Open specifications enables kernel developers write the driver. At this point assistance from the hardware manufacturer is at most appreciated but not required.

They don't have to request anything either. They can download the kernel source code and make the changes then submit the fix.

If this were a problem then they can join the development team.

Anyway, here is a guide posted recently for potential contributors.

http://news.zdnet.co.uk/software/0,1000000121,39461788,00.htm

Where are the benefits clear? Linux still has poor driver support for consumer peripherals. Even if your printer works in Linux you will probably not have access to all the printer options available in Windows. Some printers that have a Linux driver still have weird color/contrasting issues.

What devices? If you were going to say web cams then you would of had a point prior to Linux 2.6.26 but now a shit load of web cams are supported with the new driver.

It depends on the printer and how well the printer manufacturer cooperates with the Linux community. My HP printer is perfectly supported. The HP Device Manager has tools to clean & align print heads, shows ink levels, etc.

On the issue of AMD, how long has it taken to get ATI video cards working properly? And you call this a success?

ATI's drivers sucked because very little effort was put forth into developing a quality driver and the driver wasn't maintained quickly enough to fix breakage by new kernel releases.

My guess was that ATI's Linux driver team comprised of only a few people. It unlikely even remotely compares to the amount of energy spent on the Windows driver.

Following AMD's acquisition of ATI the drivers have improved immensely and have also released hardware specifications for several Radeon GPUs.

There are some serious problems with Linux development, beginning with the unstable abi/api. That coupled with the separate kernel/OS model has made a software engineer’s nightmare when it comes to programming productivity and a complete headache for third party companies.
Oh but you say everything is fine in Linuxland? It has been 15 years and Linux still hasn’t hit 1% desktop share. The iphone will probably get 1% before Linux. Win server and IIS are gaining at greater rates than Linux/Apache.


An unstable kernel doesn't mean that all other programming APIs are unstable.

In fact, unless your application utilizes a kernel module, it is unlikely an important factor to consider.

Which standard C/C++ library (glibc) versions to target is what needs to be considered--assuming the software is developed in C or C++.

Look at GTK+ 2... the toolkit APIs have become so stable that many are claiming stagnation. GTK+ 3 might break backwards compatibility in order to progress forward but that isn't the point.

LSB has the potential of becoming a savior for commercial developers targeting a range of distributions, especially now that LSB 4 will address most of the concerns.

There just has to more incentive and/or incitement for distributions to adopt the standard. I mentioned earlier that Adobe Flashplayer can leverage its dominance to make it happen.

The good news is the three main distributions that really matter (RHEL, SUSE, and Ubuntu) have jumped on board.

thecodewitch said...

""I can't get flash working on my Vista box after I raped it with some half assed fluxbox thing. I tried everything!"

codewitch as well as being a lame MS apologist/advocate you're brain dead.

Please explain how the default shell has any effect whatsoever on a browser plug-in or installation of activeX controls or binaries?

Clue: it doesn't. You have no fucking idea what you're talking about.

For what it's worth the upgrade from IE6 to IE7 was done a long time before I changed the default shell.


Yes...oh god YESSS!!!! Bring it on douche!

The comment about your use of fluxbox as a shell replacement was more a comment on your poor judgment and lack of taste. This would be evident if you didn't quote me out of context, because following this, I posted a link to the Adobe flash installers, which include the flash ActiveX controls.

One of the reasons I post here is I love seeing you morons bounce around desperately trying to convince yourselves that linux isn't a broken, worthless ejaculation of a bunch of basement dwelling rejects. Deep down, you know this is true, but like some broken drug addict or scientologist, you lie through your teeth to promote it, and hopefully suck others in so they can share in the misfortune/assrape. You quoting me out of context is just the latest example of the acts of desperation fundies resort to.

After 15 years, Linux is still a bitterly disappointing failure from the point of view of users, who are ignored and practically spat upon by the "community", and its especially disappointing to developers, who still don't have a stable set of core binaries to develop for.

julian67 said...

The comment about your use of fluxbox as a shell replacement was more a comment on your poor judgment and lack of taste. This would be evident if you didn't quote me out of context, because following this, I posted a link to the Adobe flash installers, which include the flash ActiveX controls.

One of the reasons I post here is I love seeing you morons bounce around desperately trying to convince yourselves that linux isn't a broken, worthless ejaculation of a bunch of basement dwelling rejects. Deep down, you know this is true, but like some broken drug addict or scientologist, you lie through your teeth to promote it, and hopefully suck others in so they can share in the misfortune/assrape. You quoting me out of context is just the latest example of the acts of desperation fundies resort to.

After 15 years, Linux is still a bitterly disappointing failure from the point of view of users, who are ignored and practically spat upon by the "community", and its especially disappointing to developers, who still don't have a stable set of core binaries to develop for.


1st it's not fluxbox, it's blackbox. They are similar but definitely different, with fluxbox in fact being a derivative. I explained the reasons for using it (multiple desktops, proper window management) and also that if native explorer shell offered multiple desktops and configurable window management I would be happy with it. I did explore alternatives, some, like Emerge desktop, are extraordinarily good looking but that wasn't terribly important. Blackbox offered the functionality and the best performance. It's very customizable and can be made to look different. I use it very close to the plain defaults as it only has to work, not be pretty.

It's very commendable that you posted a link (Well done! Good boy! Sit!) but the previous poster, George, a flash dev had already confirmed the bug and linked to some actually useful information. As you can see in this very thread I had it fixed before you even posted, and all within the blackbox shell. The problem does not lie with blackbox. The bug is a pure MS goodness arising from the Windows Update.

So here's an example of a problem in Windows but you still see it as something to do with Linux. You're a clueless cretin.

Your posts contains references to ejaculation, assrape, acts of desperation, basement dwelling and being spat on.

Is there something you want to tell us all about your recreations?

You're a gimp.

Anonymous said...

Way to stack the deck there. OS X a BSD plain and simple, an adapted NeXTSteP, which was derived from BSD4.3. The fact that it has proprietary frameworks and libraries on top of it doesn't change that it's a BSD system, weather a system is BSD-style Unix or not is in no way contingent on the availability of source code.

The comparison is with fully open source operating systems. Darwin by itself is virtually unusable and anyone who tried to download plain Darwin and install Gnome or KDE on it will say the same--like me.

The components that make OS X usable are proprietary and not anyway related to BSD.

OS X is no longer compliant with Nextstep anyway.

I could argue that Interix (aka SFU aka SUA aka Window's POSIX subsystem aka a OpenBSD ported to run on the NT kernel) would possibly be the third best attempt at a desktop BSD.

SFU is only good for a few *nix tools. Windows ports of GNU tools, Cygwin and Cooperative Linux are better options overall.

I never claimed otherwise I'm saying that a system designed to be a single integrated, streamlined system tends not to. You're replying that a kludge system derived from one of those systems is subject to the same problems.

Yes if the integration isn't well tested but that is an exception not a rule. My point is once you start adding high-level FOSS components on top of a base BSD installation then the same theoretical integration problems can occur.

I'm talking about servers, and you're going on about supercomputers. Whereas FreeBSD supports the IntelMP limit of 8 CPUs (with N cores per CPU) on x86, this is a limitation set by IntelMP per board, not by FreeBSD itself. Conversely it supports T1 on Sparc, which itself has 8 cores. But that was way back in 5.0, as far as 7.0 goes, I can't even find a mention of the maximum amount of CPUs it supports.

Nevertheless it's an apples to oranges cpmparison. Supercomputers are large clusters, hell, supercomputers are supercomputers, servers are servers.

You're essentially trying to argue that because Linux is used on supercomputers, it scales better than [other OS] on servers.


I'm arguing that Linux scales better on servers than FreeBSD, especially multi-processor server systems.

At one point FreeBSD 7 may of had a performance edge over earlier 2.6.2x kernels on single or dual processor/core servers. Certain benchmarks appear to back that up but that advantage was mostly eliminated with Linux 2.6.25.

Solaris is a different story and a more creditable contender on HPC equipment. However, I've not actually seen any benchmark comparisons between Solaris and Linux to make any judgment here.

My point wasn't to discredit FreeBSD as a credible server operating system. I just simply disagree that FreeBSD serves that purpose better than Linux. I'd agree that it is similar on typical server hardware.

thecodewitch said...

So here's an example of a problem in Windows but you still see it as something to do with Linux. You're a clueless cretin.

Its not a bug in windows, its a bug in IE7, the web browser. All browsers have wacky edge case bugs, so why don't you just install Firefox or Opera?

When that LH blog page with the 300+ disqus comments caused firefox to sit for 5 minutes scratching itself, I had a look at Opera. But I didn't write a comment bullshitting about how bad linux is because my firefox has a bug because 1) A Firefox bug has nothing to do with linux, 2) There are other, actually legitimate and devastating reasons why linux is a waste of time, as opposed to your nicely written but misguided attempt to equate the braindead jackassery that is flash in linux, with one bug in one browser on Windows.

julian67 said...

Gimpboy can you read? I said in the original description of the problem that Flash worked fine in Firefox on my Windows install. It was only the discussion of flash in this thread that prompted me to boot XP and compare the experience of Flash in Windows to Flash in Debian. I thought using IE would be best to make the comparison useful to me and for it not to depend on a 3rd party browser. In Windows I web browse very little but when I do it's with Firefox which is why the bug had passed unnoticed by me for several months at least.

IE is an integral part of Windows in a way that Firefox isn't in a distro (default browser in Gnome is Epiphany, in KDE is Konqueror. Epiphany in fact isn't integral to Gnome, though Konqueror is integral to KDE in a similar way that IE and explorer are in Windows). IE is in no way an optional extra and is not uninstallable under normal circumstances. Same goes for Windows Update. It's not an optional extra (nor should it be).

It's true that a bug in Firefox doesn't imply a bug in the OS it's running on. It seems we agree on something. Strangely though when MS makes bug counts to try to demonstrate Windows has fewer bugs than a distro the way they measure it is a bare default Windows install vs a fully loaded distro and they count all the distro's applications' bugs as "Linux bugs". That's for information only and so that you might better conform to the will of your gimpmasters.

You mentioned asses again.

Anonymous said...

I haven't had any problems with Firefox and this blog on either OS.

julian67 said...

Loads fast in Epiphany.

Haven't visited it from Windows, that would be heresy.

Perhaps gimpboy has found a bug in Windows?

.net jerkface said...


The benefit of supporting more hardware than any other kernel is quite a big one ;-) The manufacturers benefit from lower and shared costs of development and maintenance.


So exactly where is the benefit for consumers? Any new hardware bought at a box store will come with a driver for windows, and any hardware built after 2000 likely has a driver available online. Are you going to tell me that for consumers the Linux situation is better?

You claim that manufacturers benefit, but how do you know this? You admit that most don’t want to open their hardware specs, yet they all somehow benefit financially? Why are you so sure that they wouldn’t benefit more from a stable abi? In the software world shared development and maintenance between organizations can easily increase overall costs. It’s often cheaper to have a dedicated individual or team work without having to coordinate with a group that is not under the authority of the company. There are also additional costs that come with the Linux model, such as increased support costs from being forced to support multiple distributions.
At the very least an ISV needs to support two distributions which is a lot of testing and support for a platform that has less than 1% of the market.
Furthermore since I have only heard good things about Windows driver development when compared to Linux, I do not see a clear benefit here for manufacturers. I think it is arguable that a free unix like freebsd that has a stable abi would be financially better for hardware companies. Some companies probably benefit from the Linux model but others obviously dislike it to the point of deciding to not make a linux driver even if it means a loss in overall sales. At the very least it is debatable so no clear benefit.

OpenBSD wireless: the point isn't as clear as it first appears.


The point is simple: Linux has had massive financial support yet OpenBSD had better wireless support for a few years, and OpenBSD isn’t even trying to get on the desktop. Pathetic, that’s the point. A small group of geeks in Canada did a better job than the entire Linux army. Totally pathetic.


An unstable kernel doesn't mean that all other programming APIs are unstable.

In fact, unless your application utilizes a kernel module, it is unlikely an important factor to consider.


VMWare was recently broken by a kernel update. Nothing to worry about there at all.
Even with the LSB you still have to test and support for multiple distributions. This doesn't make the situation much more appealing, especially when these other factors are considered:
1. Linux has less than 1% of the market
2. Linux users are cheap and don’t like to pay for software
3. Linux users are often gpl advocates who think that software companies are evil and actively try to convince others to adopt their beliefs.
4. Linus and his pals don't care at all if their kernel update breaks your applications. They don't care about commercial software in the least.

Don’t wait too long for that Linux port of photoshop.

Anonymous said...

@julian67: "I said in the original description of the problem that Flash worked fine in Firefox on my Windows install. It was only the discussion of flash in this thread that prompted me to boot XP and compare the experience of Flash in Windows to Flash in Debian."

You're cherry-picking defects, dickweed. Flash has had a ton of problems in Firefox on both Windows AND Linux.

http://www.youtube.com/watch?v=cY_KrB8Mepg

http://portableapps.com/node/2528

http://www.brainonfire.net/2006/05/25/firefox-crash-on-flash-content-in-ubuntu/

So, big surprise, Flash is crap on BOTH platforms. The larger issue here, though, is that Flash problems on Linux tend to be WORSE because (a) Linux market share is SO LOW that Adobe (obviously) gives those fixes lower priority, (b) Adobe can't possibly test Flash on all flavors and versions of the hundreds of Linux distros available. In other words, Freetards who think that diversity is their friend are fucking kidding themselves when it comes down to developing stable software. Flash on Windows will ALWAYS be better because Adobe only has to test a small number of platforms. You can whine and protest that reality, and try to find as many counter-examples as you like, but it's the truth. The fact that we're even have to have this conversation is proof.

thecodewitch said...

Strangely though when MS makes bug counts to try to demonstrate Windows has fewer bugs than a distro the way they measure it is a bare default Windows install vs a fully loaded distro and they count all the distro's applications' bugs as "Linux bugs".

All this sad whining and nit picking about which metric to use and which bugs to count pales into oblivion when you consider that when you offer the thing to everybody for free, and they tell you they still don't want it, there is probably a deeper problem than bug counts.

Linux the kernel is an interesting thing in itself. If Linus Torvalds didn't release the code, obnoxious, self righteous douches like Stallman and Eric Raymond wouldn't be going around the world giving bullshit interviews and talking shit. Stallman would still be fiddling with his GNU HURD in obscurity, and Eric Raymond might have been shot in the face on a hunting trip by now.

People who persist in using it are just the self righteous douches who can't do shit, and need something trendy to set them apart from people who actually want to use their computer get something done.

The fact that you insist on defending this piece of shit tells me that the sadist here is you.

.net jerkface said...


I'm arguing that Linux scales better on servers than FreeBSD, especially multi-processor server systems.

...

My point wasn't to discredit FreeBSD as a credible server operating system. I just simply disagree that FreeBSD serves that purpose better than Linux. I'd agree that it is similar on typical server hardware.


So Linux has had hundreds of millions invested into it in the form of full-time programmers and yet it can only outperform a free Unix maintained by small group of volunteers in the area of high-end multi-processing. Got it. Go Linux.

.net jerkface said...


They don't have to request anything either. They can download the kernel source code and make the changes then submit the fix.


A submission is a request. To make a change to an existing driver they have to request that the changes be made. With the stable abi model they can simply put the latest version on their website. They don't have to coordinate with an outside party for a minor update.

thecodewitch said...

.net jerkface:
A submission is a request. To make a change to an existing driver they have to request that the changes be made. With the stable abi model they can simply put the latest version on their website. They don't have to coordinate with an outside party for a minor update.

Holy fucking shit! You mean I don't have to sift through 50 freetard dev forums trying to guess what they're going to break next before I start my project? You mean, I don't even have to email these obnoxious hippies if I don't want to? You mean I live in a free country where I can earn my own money, and don't need special permission from some commie politburo to start work and buy a house?

A better analogy than "bazaar vs Cathedral" for comparing the ESR and RMS accredited open source method versus a design and implementation created by a single competent programmer, or a closed team of competent programmers is "slum vs Cathedral"

With open source, you end up with everyone reduced to the level of the lowliest, scraggliest, most incompetent bum. (Stallman comes to mind) Whenever someone tries to fix anything, some fuckhead or group from the "community" comes along and takes a dump in the middle of the street, ensuring eternal incompatibility, and the fuckheads who did this are answerable to nobody, because nobody in the community cares, and those that do eventually either stop caring, getting used to the shit in the street, or they escape.

Anonymous said...

So Linux has had hundreds of millions invested into it in the form of full-time programmers and yet it can only outperform a free Unix maintained by small group of volunteers in the area of high-end multi-processing. Got it. Go Linux.

Eh? Well considering FreeBSD beats all the other open source BSDs in most benchmarks then Linux beats them as well.

Linux also beats Windows at some benchmarks. Windows beats Linux in encoding and file compression performance.

Windows probably beats Linux at 3D rendering performance as well but that is a direct result of less-performant proprietary drivers in Linux.

http://mssaleh.wordpress.com/2008/05/19/ubuntu-804-lts-vs-windows-xp-sp3-application-performance-benchmark/

A submission is a request. To make a change to an existing driver they have to request that the changes be made. With the stable abi model they can simply put the latest version on their website. They don't have to coordinate with an outside party for a minor update.

If they don't want to make requests then they can join the development team to manage their own driver in the source tree.

Linus is aware of the difficulty at becoming an active participant.

Did you see the guide that was linked with the article?

http://ldn.linuxfoundation.org/how-participate-linux-community

VMWare was recently broken by a kernel update. Nothing to worry about there at all.

That is because VMWare has its own kernel module for loading the guest addition drivers. >.>

Even with the LSB you still have to test and support for multiple distributions.

Nope, not really.

http://ldn.linuxfoundation.org/lsb/all-about-linux-application-checker

Technical highlights of this tool:

# Checks for cross-distro portability and gives a good indication of how well the app may perform on various distributions of Linux

# Gives guidance for application developers on how to enhance the portability of their applications

# Gives a good indication of how close the app is to LSB certification

1. Linux has less than 1% of the market

The Wintards can never omit this cliché statistic.

2. Linux users are cheap and don’t like to pay for software

I dunno... I think we'd pay for quality commercial software if the Linux port were significantly better than what is available for free.

As of now the selection of commercial products for desktop Linux users is extremely limited.

Some of us instead buy the Windows version of a product (say Photoshop) and run that within Wine, which contradicts everything you said.

3. Linux users are often gpl advocates who think that software companies are evil and actively try to convince others to adopt their beliefs.

Disliking Microsoft is very different from disliking software companies in general.

4. Linus and his pals don't care at all if their kernel update breaks your applications. They don't care about commercial software in the least.

This is true but Linus is the more pragmatic of the developers.

Anonymous said...

Linux advocates are racists and rapists.

Stallman and Torvalds are pedophiles.

Anonymous said...

Linux advocates are racists and rapists.

Stallman and Torvalds are pedophiles.


Now thats just sad.

Your trolling attempt is made of fail. You achieved 0 (zero) lulz.

There are some pretty good trolls around these parts, if you know where to look. Lurk moar, and learn to troll properly.

And you epic fail for making me defend the bearded dork.

Anonymous said...

SFU is only good for a few *nix tools. Windows ports of GNU tools, Cygwin and Cooperative Linux are better options overall.

Except SFU runs natively on NT, and interacts with the win32 subsystem just like any other native subsystem would.



Colinux is inefficient in that you're essentially running full fledged Linux on top of the win32 subsystem, and Cygwin is a sort of kludge. There are obvious advantages to taking the native subsystem approach.

As for it only being good for running Windows ports of GNU tools (actually, Interix runs Interix ports of BSD tools, we went over this, there's no GNU code in a BSD base system, and Interix is essentially OpenBSD ported to run on the NT kernel as a subsystem)

Cygwin has the same limitation in that packages need to actually be built for Cygwin to run, just like they need to be built for Interix to run. Of course there are the obvious limitations and designed puproses, Interix isn't designed to replace the win32 base system, rather copliment it, so you're obviously not going to be running KDE or something on Interix. (though there's no reason it can't be ported)

My point is once you start adding high-level FOSS components on top of a base BSD installation then the same theoretical integration problems can occur.

Your point is then both moot and convoluted. I'm talking about the base system, you're talking about adding external packages to the base system. Adding external packages (foss or otherwise) has absolutely no baring on how integrated and streamlined the core system is. It's external, it isn't part of the core system.

The comparison is with fully open source operating systems.

No, it wasn't. It wasn't even about BSD vs. SYSV to begin with. Hell, the original argument didn't even have anything to do with the desktop!

The components that make OS X usable are proprietary and not anyway related to BSD.

Other than the fact that they're built around and on top of a BSD system, and run only on that BSD system, and are part of the core of that BSD system (OS X, and in the whole package, not just Darwin),

The masjor component and concepts behind OS X, such as being a Unix operating system based on the Mach kernel and code from BSD Unix, Display Postscript, the Opjective-C language and runtime, the object oriented application layer, and the accompanying dev tools are all straight from Nextstep. In fact, after NeXT and Apple merged, NeXTAStep was modernized into OS X.

Those frameworks and libraries that makes OS X usuable, chiefly Cocoa, are the APIs OS X inherited from OpenStep. Hell, Cocoa class names are still prefixed with NS!

And all of the fancy stuff that make me love my macs are all written ontop of and with those frameworks and APIs

OS X is every bit as much BSD as FreeBSD, OpenBSD, NetBSD or any other BSD, given that it's what NeXTstep evolved into. Arguing that OS X isn't BSD, implies that Nextstep wasn't (since they're the same codebase) BSD either, and in turn that BSD4.3 (from which NeXTStep was forked) wasn't BSD either.

OS X is no longer compliant with Nextstep anyway.

It's a direct descendant of Nextstep, which is turn was a direct descendant of BSD 4.3.

FreeBSD 7 isn't compatible with BSD4.4 anymore either, doesn't mean that it isn't BSD.

No SYSV Unix is still compatible with System V, either, doesn't change thnat they're System V derived Unix.

Linux 2.6.x isn't compatible with 2.2 or 1.0 (2.6.25 is barely compatible with 2.6.24, for that matter) anymotre, but that doesn't stop it from being Linux.

Anonymous said...

Apple steals your identity with its proprietary software. Since you can't see the code then you must assume the worst. Imagine if people had secrets; how would you ever uncover their plots against you if they kept their secrets?

RockMeLinuxDamus said...

Are both BSD users commenting on this blog? I think it's pretty cool that the entire community came out to play.

Anonymous said...

Decision was to go with pulseaudio and with that you inevitably cause breaking (not because of crappy engineering of PA, but because of technical inability to emulate old systems perfectly).

If that's the price, I want to go with it. I don't want to have half-broken archaic sound system to remain (with only one app playing sound at the time), even if flash worked on it perfectly.

That said, Flash 10 beta works properly with pulse, so I consider the problem resolved (at least when full v10 is released). For other commercial applications, we might wait a bit longer (Skype being a prime problem now).

Anonymous said...

Or as this link points out.. install windows in vmware and run your flash applications..

https://blueprints.launchpad.net/ubuntu/+spec/webcam-chat

julian67 said...

"So exactly where is the benefit for consumers? Any new hardware bought at a box store will come with a driver for windows, and any hardware built after 2000 likely has a driver available online. Are you going to tell me that for consumers the Linux situation is better?"

Can you actually read???? I already outlined some clear benefits.

On supported hardware the Linux model is seamless, install free and just easier. I've used Windows long enough to know that the existence of a driver is no guarantee that the driver is any good, yet that seems to be an assumption made as soon as people compare Linux and Windows drivers. The worst drivers I've encountered in Windows have been Creative's Audigy (original version), Nikon's Super Coolscan 4000 ED driver (professional 35mm negative scanner), the Intel/Lenovo sound driver for Vista and while I didn't remember it earlier the Lenovo/Intel 3945 wireless driver for Vista was absolutely crap as well.

The Creative driver was eventually reworked and satisfactory but it took Creative months to even admit a problem and some more months to produce the fixed driver. Meanwhile I used a no-cost alternative, the kX driver http://kxproject.lugosoft.com/

The Nikon driver was awful. It's a firewire device but Nikon managed make a driver that killed the device driver on hotplugging. If you got the start up or power down sequence wrong or hotplugged at the wrong time your next task was to pick your way through system32 and find the Nikon .inf files, delete them, uninstall the driver and start over. A professional digital lab nearby had the same scanners and the same problems. I used this device for a couple of years. In that time Nikon never even acknowledged the problem let alone released an improved driver despite numerous complaints from customers.

And how to install this troublesome device in a distro? If you already have sane backend (scanner backend) which is installed default in many distros the procedure is this: Plug the fucking firewire cable into the fucking firewire port. Doh! In comparison the Nikon handbook goes through several pages of install instructions for Windows and it's still crap.

The procedure for repairing the Creative driver was the same, find those .inf files and uninstall and reinstall. Regularly. Install procedure in a distro? Insert card into PCI slot. Boot. Done. It never failed ;-)

Another advantage for consumers: You might have noticed that in Windows some driver updates arrive via Windows Update. You might have even realised that this is confined to signed drivers and not even all of those and that many are incredibly out of date. It's basically inevitable that if you run Windows and have several peripheral devices that you are running unsigned drivers. Even OEM suppliers' laptop/motherboard driver disks routinely include unsigned drivers. It's up to you to remember to search for updates. In a distro the drivers are in the kernel, any needed patches are automatically dealt with via the package management.

For manufacturers: the "fact" that you claim you heard something is not an argument. doh! What would a stable kernel binary interface offer manufacturers? It would mean that 100% of the development costs and time fall on them. They would be 100% responsible for handling bug reports and fixing those bugs. If the linux kernel reverted to the old stable/unstable release model (necessary if a stable kernel binary interface was required) it would mean that a manufacturer might have to wait a very very long time to get their hardware support into the distros, i.e. to the people who matter, the customers who might buy the device. At the moment very few manufacturers find this attractive, this situation is much less common than manufacturers who see the benefit of getting their hardware supported in the kernel, available to all in good time and maintained, even improved, for all future versions. The really notable exception is Nvidia who make very well featured drivers but unfortunately those binary only drivers regularly feature in the the top 10 list of kernel oops (kernel crashes), alongside many other binary only drivers.


VMWare. Last weeks big breakage was cross platform. It screwed up on every OS. This is nothing to do with Windows, Linux, Mac, Solaris, BSD or any other OS. If there's been another situation recently where VMware's kernel module crapped out then this is in fact not an example of the need for a stable binary kernel interface but an example of why it's better to get the driver into the kernel and have it tested and maintained properly.

Market share: You cannot possibly have any idea about this, you're blowing gas.


"2. Linux users are cheap and don’t like to pay for software"

OK let's forget for a moment that unlicensed (aka pirated) Windows installs seem to be as common as licensed ones shall we? And that the proportion of people using Photoshop who actually have a legal version is small. Let's just quietly forget the most popular method for obtaining retail software is getting a warez version via bittorrent shall we? Let's play Alice in Wonderland.

And then look at Red Hat and Novell. An awful lot of people seem to be happy to pay for their software. As do Asus Eee PC users who get a version of Xandros a commercial distro whose price inevitably is built into the Eee PC purchase price in the same way the cost of my XP Pro license was built into my Asus M5's purchase price.

Other points: you still are too fucking dumb to spot the difference between proprietary and commercial. they are not the same thing. There is no-cost non commercial software and there is commercial open source software. You're an idiot and apparently determined to remain one.

OpenBSD wireless: seems the point is had, temporarily, according to you. Yawn. It's convenient that you forget that most people don't consider an open or wep encrypted network to be useful at all.

I'm not waiting for a port of Photoshop. Gimp is fine for me as I don't need full CMYK support (I don't prepare artwork for publication) and I'm still curious as to why people with 8 bit and 10 bit and 12 bit capable consumer digicams feel they desperately need the 16 bit support in Photoshop, which can offer some advantages in handling HDR and in some unusual circumstances in normal picture editing but in fact most people will never be able to tell the difference, in the same way they can't tell the difference between 320 kbps and a CD. Add to this the fact that all those features are available in other free tools if I feel I need them. Or if I really really want it...guess what...Google have funded a huge effort to make Photoshop work better on wine and atm apparently CS2 works fine. So all you warez kiddies can now run your crappy torrent-sourced warez and keygenz and crackz in wine.

Anonymous said...

What's the actual benefit of having a whole, holistic operating system such as *BSD where the kernel and userland apps are developed nd synced as a single piece, whereas Linux kernel and userland apps are "kludged" together? I can see that OS purists prefer the former solution but are there any technical or other merits?

Anonymous said...

Obligatory!

http://i170.photobucket.com/albums/u252/calla_teryn/takeittux.jpg

Anonymous said...

@julian67

I normally disagree with your take on Windows vs. Linux ... but nice work shutting down the codewitch.

Anonymous said...

@julian67: "On supported hardware the Linux model is seamless, install free and just easier.

And on the huge pile of hardware that isn't "supported", the experience is an abysmal nightmare.

I've used Windows long enough to know that the existence of a driver is no guarantee that the driver is any good, yet that seems to be an assumption made as soon as people compare Linux and Windows drivers."

The overwhelming majority of the time, the Windows drivers just work. Sure, some of them suck, but the benefit of actually having a profit-driven hardware market is that the market, itself, acts as a pruning mechanism to eliminate products which simply don't work. Not so with Linux. Manufacturers have practically zero incentive to fix problems.

Another advantage for consumers: You might have noticed that in Windows some driver updates arrive via Windows Update. You might have even realised that this is confined to signed drivers and not even all of those and that many are incredibly out of date.

Most OEMs (Dell, HP, Lenovo) provide one-stop update/installation of drivers from their websites using a unique identifier on the machine. So, really, this is a non-issue.

It's basically inevitable that if you run Windows and have several peripheral devices that you are running unsigned drivers.

Bogus assumption.

Even OEM suppliers' laptop/motherboard driver disks routinely include unsigned drivers.

They can. But they don't "routinely" do so.

It's up to you to remember to search for updates. In a distro the drivers are in the kernel, any needed patches are automatically dealt with via the package management.

And when those drivers are not part of the kernel, you're fucked.


And then look at Red Hat and Novell. An awful lot of people seem to be happy to pay for their software. As do Asus Eee PC users who get a version of Xandros a commercial distro whose price inevitably is built into the Eee PC purchase price in the same way the cost of my XP Pro license was built into my Asus M5's purchase price.


The most telling data is the fact that THERE IS NO THIRD PARTY COMMERCIAL SOFTWARE APPLICATION MARKET FOR LINUX. None. Zip. Zilch. Maybe if Linux users cared a little less about free-as-in-beer, ISVs might actually market commercial software to them.

Anonymous said...

@julian67: "On supported hardware the Linux model is seamless, install free and just easier.

And on the huge pile of hardware that isn't "supported", the experience is an abysmal nightmare.

I've used Windows long enough to know that the existence of a driver is no guarantee that the driver is any good, yet that seems to be an assumption made as soon as people compare Linux and Windows drivers."

The overwhelming majority of the time, the Windows drivers just work. Sure, some of them suck, but the benefit of actually having a profit-driven hardware market is that the market, itself, acts as a pruning mechanism to eliminate products which simply don't work. Not so with Linux. Manufacturers have practically zero incentive to fix problems.

Another advantage for consumers: You might have noticed that in Windows some driver updates arrive via Windows Update. You might have even realised that this is confined to signed drivers and not even all of those and that many are incredibly out of date.

Most OEMs (Dell, HP, Lenovo) provide one-stop update/installation of drivers from their websites using a unique identifier on the machine. So, really, this is a non-issue.

It's basically inevitable that if you run Windows and have several peripheral devices that you are running unsigned drivers.

Bogus assumption.

Even OEM suppliers' laptop/motherboard driver disks routinely include unsigned drivers.

They can. But they don't "routinely" do so.

It's up to you to remember to search for updates. In a distro the drivers are in the kernel, any needed patches are automatically dealt with via the package management.

And when those drivers are not part of the kernel, you're fucked.


And then look at Red Hat and Novell. An awful lot of people seem to be happy to pay for their software. As do Asus Eee PC users who get a version of Xandros a commercial distro whose price inevitably is built into the Eee PC purchase price in the same way the cost of my XP Pro license was built into my Asus M5's purchase price.


The most telling data is the fact that THERE IS NO THIRD PARTY COMMERCIAL SOFTWARE APPLICATION MARKET FOR LINUX. None. Zip. Zilch. Maybe if Linux users cared a little less about free-as-in-beer, ISVs might actually market commercial software to them.

Anonymous said...

@julian67: "On supported hardware the Linux model is seamless, install free and just easier.

And on the huge pile of hardware that isn't "supported", the experience is an abysmal nightmare.

I've used Windows long enough to know that the existence of a driver is no guarantee that the driver is any good, yet that seems to be an assumption made as soon as people compare Linux and Windows drivers."

The overwhelming majority of the time, the Windows drivers just work. Sure, some of them suck, but the benefit of actually having a profit-driven hardware market is that the market, itself, acts as a pruning mechanism to eliminate products which simply don't work. Not so with Linux. Manufacturers have practically zero incentive to fix problems.

Another advantage for consumers: You might have noticed that in Windows some driver updates arrive via Windows Update. You might have even realised that this is confined to signed drivers and not even all of those and that many are incredibly out of date.

Most OEMs (Dell, HP, Lenovo) provide one-stop update/installation of drivers from their websites using a unique identifier on the machine. So, really, this is a non-issue.

It's basically inevitable that if you run Windows and have several peripheral devices that you are running unsigned drivers.

Bogus assumption.

Even OEM suppliers' laptop/motherboard driver disks routinely include unsigned drivers.

They can. But they don't "routinely" do so.

It's up to you to remember to search for updates. In a distro the drivers are in the kernel, any needed patches are automatically dealt with via the package management.

And when those drivers are not part of the kernel, you're fucked.


And then look at Red Hat and Novell. An awful lot of people seem to be happy to pay for their software. As do Asus Eee PC users who get a version of Xandros a commercial distro whose price inevitably is built into the Eee PC purchase price in the same way the cost of my XP Pro license was built into my Asus M5's purchase price.


The most telling data is the fact that THERE IS NO THIRD PARTY COMMERCIAL SOFTWARE APPLICATION MARKET FOR LINUX. None. Zip. Zilch. Maybe if Linux users cared a little less about free-as-in-beer, ISVs might actually market commercial software to them.

Bob said...

the 16 bit support in Photoshop, which can offer some advantages in handling HDR and in some unusual circumstances in normal picture editing but in fact most people will never be able to tell the difference,

The difference is sometimes subtle, but people do notice the lack of colour-burn and smoother gradients. One could argue that 8-bit colour isn't true HDR either, since there aren't enough colours for it to actually qualify as high dynamic range.

The thing is HDR has become a buzzword, and people like to think that it's simply a matter of merging differently exposed images and tonemapping the result. Whereas that is indeesd the basic technique to doing HDR, the end result isn't HDR unless it actually covers the high dynamic range. HDR refers to the end result, not the process.

It's not even only useful for HDR, more colour information is again equal to less colour burning and smooth gradients, which may not be particularily useful for ordinary photo editing, unless the photos being edited are already in 16-bit, but opens up a lot more possibilities in terms of creating and post processing artwork in a wholly 16-bit workflow.

As for CMYK, you may not need it, but the entire print industry works in CMYK and requires Pantone spot colours. I've mentioned in previous comments (to other posts) that certain government flags and logos, for example, by legislation, have to be in done with certain Pantone colours.

Further, working with Pantone saves many headaches and time, since although colours display differently on different screens, on different gamma settings and under different lighting, a Pantone colour is ALWAYS the same Pantone colour, it may not look the same on different screens, the colour always comes out the same in print.

But you don't work in print, so none of this is important to you. Others however do, and it's the gamebreaker.

Pantone, by it's very nature is unlikely to ever be supported in free software due to its proprietary nature. That kinda sucks, but it's a requirement for work in the industry.

While CMYK support does exist in some FOSS applications (Hello Krita), the colour managent facilities are far from decent, and again, the lack of Pantone support makes it largely useless. This may change in the future, but that's of no use to anyone in the present.

And while 16-bit support does exist for Cinepaint, it's primarily designed for use on retouching film, (and it's great for that) and itself lacks Pantone support and decent colour management facilities.

While certain parts are implimented in free software, no one package, or group of packages provides the whole deal. Most other high grade commercial and proprietary applications don't even offer the featureset and facilities that Photoshop does (Photopaint comes relatively close, but is mostly developed as an afterthought, Corel's babies are DRAW and Painter, and PSP was repositioned as to not compete directly with PhotoPaint, Fireworks was repositioned to overlap as little as possible with Photoshop, etc).

This isn't a proprietary vs FOSS issue, this isn't even a Windows vs. Mac vs. Linux issue. It's purely a Photoshop vs the rest issue, where Photoshop is the clear leader of the pack, and is the industry standard not because of some conspiracy against Linux or open source, but sheerly because it's the only solution that meets the industry's requirements.

Pretty well the same situation in Vector land, except there are two industry leaders, which see about equal use (many print shops even use both) where they're on relatively equal footing featurewise, and it essentially comes down to Corel DRAW if you prefer to work with a line-based workflow, and Adobe Illustrator, if you prefer a curve-based workflow. Illustrator having a slight advantage due to being able to purchase it at a discount as part of the suite.

The third in line, by a wide margin is Xara, except the component that makes the Windows version a consideration, the lightning fast rendering engine, isn't availible in the Linux/OSS version, since Xara doesn't want to give away their crown jewels, they're rwriting the renderer to be based on cairo, removing the OSS/Linux version as a contender, (they still want you to pay for Xara Xtreme, which isn't ported) feature to feature it doesn't compete in the same league as Draw or Illustrator.

Wine is a kludge solution at best, and I generally wouldn't trust sometghing mission critical to it. And altought I've considered it, and it might be acceptible for a single application, I really didn't see a point in running a Linux syustem where the bulk of the userland was to be run on a compatability layer, especially considering that I do a lot of work on audio composition and production as well, where here, too, Windows and Macintosh are clearly better suited platforms.

It's not that there are no samplers availible for Linux, it's just that none of them are in the same league as Native Instruments' Kontakt. It's not that there aren't any softsynths or instrument/effects building frameworks, it's just that nothing is anywhere near the dsame level as NI's KORE or Reaktor, or Steinberg's Hypersonic. It's not that there aren't guitar amp simulators, just notrhing on the quality of NI's Guitar Rig.

And it's not that Ardour isn't a great DAW, it's just a DAW is fairly worthless if you don't have the facilities to use your indtruments. And as good a DAW as it it, it doesn't possess some of the extra features that make Ableton's Live (second to none in terms of the facilities it offers for live performance) or Nuendo (for film scores and ultra highend post processing) or Cubase (for mastering and ultra high end post processing) so appealing.

It's also not that there isn't any live scratching or mixing or DJ'ing solutions for Linux, just not any particularily good ones.

The only reason to work audio in wine is because politics are more important to you than quality, workflow and convenience, and I'll even go so far as to say stability. The overhead from wine negates any advantage Jackd's low latency provides.

Linux just isn't well suited for some markets, sure it might be some day. But it isn't now, and the promise that something may work well down the line doesn't put food on the table today, nor does it justify trying to get work done with shitty tools today.

Software politics are a waste of time for people with work to do. Using the best suited tool for the job is a far more important and worthwhile consideration.

Anonymous said...

What's the actual benefit of having a whole, holistic operating system such as *BSD where the kernel and userland apps are developed nd synced as a single piece, whereas Linux kernel and userland apps are "kludged" together?

Integration, ease of maintenence, seemlessness, fewer bugs and glitches since the whole package is designed to work as a single system, and allows for better QA.

Sanity, fewer things tend to go wrong, and they're easier to fix (no waiting on upstream, DIY doesn't result in incompatability with upstream).

Stability, no chasing after the upstream, no lack of cohesion, no conflicting setups on different systems.

Less (if any) breakage between updates (because, again, the whole system is synched and developed to function as a whole, and the whole base system is updated at once, instead of each piece being updated individually)

A more sane, effective third party development effort, since, again, you're more than likely to have the same set of base libraries on all systems).

No conflicts where package A depends on version B of a base library, whereas package B depends on version A of the same library.

Things just tend to work a lot more seemlessly, resulting in a much better user experience.

Anonymous said...

I'm still curious as to why people with 8 bit and 10 bit and 12 bit capable consumer digicams feel they desperately need the 16 bit support in Photoshop

I'm curious as to why freetards use this piss poor excuse to accept the Gimp developers moronic choice to not merge Silicon Grail and Rhythm and Hues's patches that provided this basic functionality 10 years ago in favor of the vaporware GEGL. Oh right, freetards don't care about functionality, they care about the damned license...

Perfect example: http://lists.freedesktop.org/archives/xorg/2008-July/037374.html Patch from nvidia to allow xorg's autoconfiguration to recognize the driver being met with open hostility while user's are left to figure out how to create an xorg.conf from scratch so they can have a functioning GUI.

«Oldest ‹Older   1 – 200 of 288   Newer› Newest»