Monday, April 6, 2009

Wile E Reality catches up to FreeRunner

Sweetness. Looks like the FreeRunner is finally dead.

I’d just like to take a moment, and lament about all the top notch features we could have had with a truly open source phone like the FreeRunner.

  • grep’ing through your address book using extended regular expressions
  • finger’ing your Fav Five
  • Having 10 different competing UI’s packaged by dozens of different distributions
  • Being able to ssh to your phone so that you can check it’s uptime
  • Using PGP to sign your SMS messages. Beware though, after the signature, you’ll only have 5 characters left to work with.
  • Sending people videos in ogg theora format
  • Losing data randomly with ext4
  • Unloading the kernel module for the asterisk button because I never use it
  • Changing your keypad to dvorak layout, where the most commonly used numbers are in the middle row
  • Using gkrellm to monitor your battery power, as it is slowly sucked away by gkrellm
  • More efficiently using your tiny screen with a tiling window manager
  • Running only free and open javascript
  • Warning users that their keypad lock password is not strong enough
  • E17, some day. No, really, someday.

1346 flames:

«Oldest   ‹Older   1 – 200 of 1346   Newer›   Newest»
wbkang said...

FIRST!!

Anonymous said...

SECOND!

Anonymous said...

Thanks God there is still Android!
I'm looking forward for the Nokia N97, anyway.

Alvarez said...

JAJAJA, this one was funny.
I was beginning to think that you've lost your mojo!

Anyway, a cellphone with Linux would be a fucking dream !
And going through the reasons would be boring, i do it all the time when my noob friends ask me: "Why is your desktop rotating whit that lame video card? And why the f*ck is your programm better than mine?"

Anonymous said...

I never understood who would want an open sauce mobile phone anyway.

A mobile phone is a product with a certain purpose. Part of the definition of "product" is, that it's somehow done and not in constant development.

Imagine you can only start making a call after the first 20 "apt-get update && apt-get upgrade && apt-get install -f && apt-get update && apt-get a life" iterations.

Imagine 35 competing calling applications: Gnucall, gcall, Kall, phone, GNUphone, gphone, Kphone, OpenPhone, OpenCallPhone, GNUOpenCallPhone, GNG is not GNUOpenCallPhone (GNG), etc.

Anonymous said...

@Alvarez

You mean your noob friends ask, in an unlimited lack of knowledge:

"Why the fuck are you interested in a rotating desktop, and why is your oh-so-good-application not able to open a .doc the right way?"

Er said...

Richard Stallman once said: "A GNU is not a rabbit and a phone is not a dish washer!". Linus on the other hand has only once encrypted the Linux source with the dvorak key.

So, OpenMoko was unavoidable!

People seem to have prolems understanding the relationship of E17 and the GPL. By introducing an X11 compatbilly layer on top of Pulse Audio and below Plasmoids this problem will be solved quickly!

SSHing in your SMS will no longer be a problem as soon as KDE is backported to Gnome.
Clearly this will be the final milestone to penetrate Linux through the desktop on your phone!

"Unladen swallow" will give raise of the long awaited unborn child. As it was predicted this will lead us to OGG theora on our tube!

Slowly sucking the brain out of Bill Gates is clearly not a wise idea, instead the hive mind should concentrate on moving Android closer to the masses and deep into the Moko.
Ubuntu will act as key component. Splitting the ISS from Shuttleworth will give us the opportunity to invent fresh and open brown colors.

I think you can only agree with my conclusion: "When the sun raises and the dark side hides behind Apple, Linux will dominate the great depression with a hybrid distribution!"

Anonymous said...

I just want my iPhone to do comfort noise so that it is more obvious when it drops calls...

Anonymous said...

@Er: so looks like that Markov chain powered freetard generator works

Anonymous said...

"Thanks God there is still Android!"

You mean that poor excuse for a phone that over more than half a year is reaching, awwww, number of sold units Apple sells every week?

Er said...

^
Yes republished papers on OLPC have shown the chains created by markov are stronger than some freetards!

Anonymous said...

Hey Alvarez,

I know you're only 16 and (ironically) using a hotmail [evilDRM(TM)] email account. There is still time for you to change. Why not read all the previous linuxhater articles first, I promise your gonads will grow with more knowledge and confidence.

Start with this one, the rotating cube fetish is already covered:

http://linuxhaters.blogspot.com/2009/01/wobbly-compiz.html

And BTW, Welcome!

Anonymous said...

Ironically, even the phone OS of the world, the most sought after phones are either Windows, Rim, or iPhone.
Amazing what can happen when there is a stable and well thought out standard API and DE/development process. The bizarre model doesn't work even in the phone OS market, how bout that?

ochedru said...
This comment has been removed by the author.
Anonymous said...

"grep’ing through your address book using extended regular expressions"

That. Would. Be. *Awesome.*

Anonymous said...

LOL @ the poster above

I thought the same exact thing. :)

Paul Moloney said...

Great comment on Slashdot about this:

"People like this cannot possibly sustain an open, non-dictatorial government for the same reason they cannot sustain an Open Source phone."

And then there were lOlZ.

P.

Anonymous said...

"Being able to ssh to your phone so that you can check it’s uptime "

priceless

Anonymous said...

"Why is your desktop rotating and who the fuck would want that?

There, FTFY.

Oh, and by the way, MicrosoftIsDying(TM).

You fit the freetard profile perfectly. Congratufuckinglations.

Anonymous said...

FreeRunner would have taken off it was modal like vi.

Anonymous said...

microsoft is dead!!! LinFTW!

http://2.bp.blogspot.com/_3Ci4DxdVT48/Sa3g9_8LsLI/AAAAAAAAAM4/O0821lU6wD4/s1600-h/lol.jpg

oiaohm said...

FreeRunner proved it was even possible.

Ok openmoko itself not that great.

EXT4 glitch is not only EXT4. Fault that causes EXT4 to loss data exists in a huge list of filesystems. It the incorrect order of writing metadata. NTFS has it. Its a ancient fault. All Linux filesystems are be checked for it. Data loss on Ext4 is not random.

Don't throw stones when standing in a glass house. NTFS has many major failures most solved by reinstalling windows and keeping backups because most people don't know what as gone wrong.

Openmoko broke the fear lot of companies had about handing over there phone internal specs. Yes failure as a product. As a proff of concept laying the path for future projects like it a success.

Anonymous said...

You mean this NTFS data loss bug that nobody but you can reference?

OnlyOiaohmKnows(TM)

Anonymous said...

Com'on... chkdsk has been patched a lot of times since the NT days just to prevent things like the random loss of data... and yeah... this problem exists, but IS NOT SO FREQUENT LIKE TEH SHIT OF EXT4.

Seriously... these guys@kerneldev are just awesome... and i don't say this in the good sense of the word.

Anonymous said...

OnlyOiaohmKnows(TM)

and a new one

OnlyOiaohmCares(TM)

oiaohm said...

Its not only me. Next time windows does go splat and you have to reinstall it. And the common blame that the power dropped out so caused it. Ie hardware failure.

Or seeing windows yelling MFT errors without hardware faults with possible data loss warning. Its happened.

If the hardware is still functional and windows not you just got bitten by the NTFS write order bug leaving its file-system in a bad state and windows without registry files.

Meta data on drivers that tells NTFS where files are is written twice. So if one copy is damaged the backup will work. The metadata write order means both copies of the metadata are updated before new location the metadata is pointing to is written to disk. So power outage at the wrong time by by old and new data so leaving system crippled.

All filesystems loss some data. Normally just new data. Not the old data that the new data was replacing. If you are losing old data the file system is defective. NTFS is defective.

Basically its so much the normal to you windows users you don't even pick it up any more. The good old catch line covers this "It's Windows" after you say that no more investigation required causing data loss issues like I describe to remain.

Basically pull you head out the sand and start truly looking around. The problem I am referring to is not reported as a NTFS failure because the people reporting it don't know it as that. Its normally like I got no repair windows xp so I had to reinstall completely. Windows XP fails to detect as being installed with the registry files are missing. When you inspect those machines about 85 percent will be the complete contents of the directory where the sam and registry should be is gone due to the metadata write error. Other 15 percent will be damaged registry files. Again has a cause in the file system driver.

The reports of this problem are all over the net if you go looking for the right things. Start looking at the failures people using windows are reporting. You will find a trend investigate them and this problem will appear. Odd snippets here and there from more experience techs reporting like windows\system32\config containing nothing. Most of the time experienced techs class it as status normal.

NTFS is a glass house with broken windows owned by a stack of people who think a glass house with broken windows is how it should be.

Anonymous said...

Look, my theory is that Oiaohm is much like John Titor. He comes from a future alternate timeline where Linux has about 30% of the marketshare and is very different from the Linux in our timeline. As well, NT has flaws in his timeline that were never solved. There's really no other logical explanation.

Anonymous said...

Some priceless comments in that link:

The problem was that the phone had some real glaring problems that were never resolved. Such as a one day battery life. And the inability to charge the phone after the battery wore out completely.

Did they use LiPo batteries ?!?

As a pissed off Freerunner owner I have this to say - OM has been badly managed for some time now. Rather than concentrate on getting basic functionality going they wasted time and money doing things over and over and over again. They must have reinvented the wheel at least three times by now. [...] No disrespect to the developers, but OM the company was a failure.

Looks like they fully embrace opensores ethics: IgnoreTheConsumer(TM)

I would have bought one but they sold out very quickly. I assume they kept production runs short to reduce risk. But doing that guarantees failure. Lately I have been checking back on openmoko.com from time to time. There is no way to buy the phone on line, and the nearest dealer to me is in India.

That's what you get with freetard-inspired business models.

Anonymous said...

Hey, oiaohm, I never said it didn't exist. I'm just saying you can't reference anything at all after being asked something like five times. All you offer, as usual, is that only you are smart enough to diagnose things correctly where everyone else in the entire world fails.

OnlyOiaohmKnows(TM)

Anonymous said...

Why would a freetard need a phone in the first place ?!? It's not like he'll ever make or receive a call.

Anonymous said...

Seriously. And who the fuck gives a shit whether or not their phone is open and free? I mean, really, do you give a fuck what software your Microwave is using? Is it hording your customized popcorn settings? What is the benefit to the consumer for having an FOSS phone versus a proprietary phone? Are the calls going to be better? Am I going to get charged less? Are the keys going to let me type text messages faster? Are my contacts going to be sorted better? What?

Anonymous said...

Cell phones are already complicated enough without freetards stuffing them with "features".

oiaohm said...

I come from the data recovery field.

Where I have to inspect machines and get data back.

NTFS flaw is attempted to be fixed in the NTFS driver in Vista. Please note the word attempted. Transaction mode it don't work right either.

I know Linux has a low market share. The problem I have here is that people look at Linux any minor error attack even if where they are attacking from has the same problem.

If people where equally jumping up and down about NTFS errors I would be happy. At least I would not have days where my datarecovery on it will not be the difference between 10000 people having or not having a job just because software cannot do its job right.

Hardware failures where I have to truly work that is a hardware failure.

No OS is perfect. NT is not perfect Linux is not perfect. Both have flaws in there that need to be removed.

We just have a massive case of Linux put under a heavy inspection and Windows being left to be defective. It is not our problem. Correct its not you pass the problem to a person like me when the machine eats your data.

Does not matter if it is a Linux Nas or a Windows laptop it all ends up on my desk. Yes surprisingly over 30 percent of my work is Linux machines and most of that is NAS's containing important data that people would not ID as a Linux machine. Each Linux has real hardware failures requiring real work to fix. Yes there are a small number that are like xfs that had the same bug as ext4 and ntfs. There is a long list of file-systems with the problem luckily most of them are not in every day use.

The other 70 percent windows I can be lucky to have 10 percent out of the total that have real hardware faults. So yes 60 percent of NTFS related and Software raid software for windows related.

Diagnosing if its a real hardware fault or not wastes time. Repairing NTFS when it should not be defective also wastes time. Time that would be better spent recovering on the real damaged hardware.

Depending on how the drive is damaged the data can be slowly fading off the drive. The longer it takes for me to get to it less chance that data recovery will work.

True hardware defect of a defective write head because the data was not written on the disk strong enough in the first place.

Time is something you cannot value in data recovery. If you get there in time you can get the data. 10 mins latter the data might be gone for good.

NTFS is costing other people there data by choking up data recovery companies ques with items that should have never been there in the first place. There is enough real broken hardware out there that do need data recovery firms time.

One day the que choke might cost everyone here there job at some point. I know it cost my job at one company because of it. And everyone else working for that company the company went under. Drive defective write head did not get processed in time.

If windows was only a Minor OS like Linux it would not be a problem. Numbers make it a problem.

I am not from a different timeline. I see everything different due to the work I do.

Anonymous said...

I always love the "I rarely ever work on the less than 1% market share product but always seem to work on the 90% market share. I think it's because the 90% share sucks and not because statistically I have a huge chance of working on them."

By the way, thanks for clarifying that data recovering entails recovering data. I would have never guessed.

Anonymous said...

Yawn. Lots of blathering and still not a single reference to a critical design flaw. It must have taken you half an hour to type all that. Surely you could have spent that time actually building your case.

If you want to fight with anecdotal evidence, fine. I've worked with hundreds of NT systems over the years and have not observed this systemic data loss fault of which you speak, and certainly not the random dumping of system32. Now ext2, ext3, and reiser on the other hand...

Anonymous said...

Look, my theory is that Oiaohm is much like John Titor. He comes from a future alternate timeline where Linux has about 30% of the marketshare and is very different from the Linux in our timeline. As well, NT has flaws in his timeline that were never solved. There's really no other logical explanation.

I have a better theory: his english reading skills are as good as his English writing skills. Hence the misinformation.

Anonymous said...

I believe that Oiaohm is the internet age's John Nash who was portrayed in the movie A Beautiful Mind. John Nash scoured Life magazine for years deciphering and breaking secret codes that the Russians were using to communicate. I think it's eerily similar to how oiaohm finds the truth hidden on the internet.

The only difference is that Nash actually did come up with some revolutionary theory in economics. Oiaohm not so much.

Anonymous said...

Oiaohm, I have a critical design flaw in Ext3/SMB for you (not even in ext4!).

Just try to create a single terabyte file on soft VLM/RAID1/ext3 with maxed out block size over SMB. Then observe how ALL glorious GNU userland goes bust.
No ssh, no console, no nofink.

Then press RESET button on your NAS to get console back and observe corrupted nodes (yes, RESET does it for you).

You guys can't even do NAS properly! Even when MS opened their SMB protocols!

For the point: NTFS did not have problems and works happily.

Anonymous said...

The only difference is that Nash actually did come up with some revolutionary theory in economics. Oiaohm not so much.

But he's got this five years worth of "research" he likes to hold over our heads but never reveal. I think its quite ironic that a guy who prides himself on "seeing the truth" and gathering information to support his agenda has not, to my knowledge, ever provided a single reference of relevance. His entire logical framework is based upon narrow-sighted conjecture and wishful thinking. His possibly being correct about certain things does not make him any less of a charlatan; even wacky Nostradamus got a few right. I wouldn't be surprised if this "research" is just a bunch of random shit culled from mailing lists, printed and pasted to the wall with differently colored twine going all over the place.

Anonymous said...

I meant to quote the prior two sentences:

"John Nash scoured Life magazine for years deciphering and breaking secret codes that the Russians were using to communicate. I think it's eerily similar to how oiaohm finds the truth hidden on the internet."

Anonymous said...

It turned out that the code breaking was a delusion in Nash's mind but he did it for years as well.

Anonymous said...

"I've worked with hundreds of NT systems over the years and have not observed this systemic data loss fault of which you speak"

I've been working for years with over 2500 desktops which are all using NTFS, and tons of servers too. No signs of such issues either.

OnlyOiaohmKnows(TM) indeed.

oiaohm said...

Beware people only send machines to data recovery when it contains important data.

People will not send like home PC's to data recovery or client side machines containing no important data.

So there is a large section of the Windows market I don't see on my desk. My desk is also biased with Windows laptops. Laptops would be suspected of having far higher hardware failure rates than NAS's due to always being in transport.

Basically there is a abnormality in my data set. I thought I better mention it before someone says those numbers cannot be right. I suspect good backup policies by some companies reduce the numbers on the Windows laptop side. So dead laptop just like client machines get scraped before coming to me.

Proper raided and UPS servers I basically see none from any OS. 1 server a year maybe. Power failure is required to set off the NTFS or EXT4 bug. Yes people see the battery warning light on there laptop and push on anyhow brings a lot of those NTFS failures to my desk. It works most of the time. Until the day it don't.

I value good design of hardware and software. Hardware will always break. I don't think people are going to stop dropping laptops into the sea or houses burning down any time soon and other great human ways of breaking hardware. Even pets play there part peeing on them or burring them.

At least there can be a fun story to what happened. Not one day I put my laptop on the changer and I turned on my laptop and it did not work there are some very important documents on there please get them back.

The simple point here I am at the concentration point. You say you have worked with 100's of machines. True. Remember machines are only coming my way when they have problems.

There is bias in the numbers I see. Something that is as rare a 1 in 1 million items produced will turn up on my desk in basically every 100 machines that cross it. Even as rare 1 error never seen anywhere else in not off the cards.

Yes it is about a 10 000 to 1 scale. Rare enough that a person not doing what I am doing might never see it in there life time. I really need to stop talking to you like data recovery people. We know descriptions of bugs we have seen them first hand.

People really don't understand that losing data is carelessness and really bad luck most of the time.

I think this is part of the problem. It these faults don't see rare to me.

ext2 no data loss protection so expected trouble. They do cross my desk from foolish people using them for speed. Rarely mind you most people are smarter than that.

ext3 good designed data loss protection not expected.

Ext4 early adopters hurt. Mostly there fault. Come on guys how many people lost data and the like using Windows XP pre SP1. The early adopters did the same thing with Ext4. Just because something is said is stable its not so until its been tested.

reiser not recommended due to lack of good maintainer and fall in the the cat of file-system not commonly used because people know better.

Also what would you find inside a NAS. Ext3 or XFS. No Ext2, Ext4 or reiser is used inside them because they are not classed as good enough file systems yet.

I do have a really clear idea of what the commonly used file systems are.

I know fault and the Linux NAS did not truly bust.

You hit the performance wall of ext3. Mind doing the to windows cutting of power half way through a write.

You can make windows hang exactly the same way. It called file system locking. ssh was not accessible because no drive access because it was in use by smb.

Again a case of blaming Linux for something that is common to all. Don't you notice on windows your mouse becomes non responsive when doing large file-system actions locally. Same with pinging the windows server or login remotely.

That is basically all that happened to the NAS. Once it completed the operation it would have returned to life. Corrupted nodes all come from the new file being saved. Old existing files are not damaged.

Trigger for it being so bad is ext 3 short sync time of 5 seconds. Its a little hard to complete a 1 tb write in that time.

Yes there is also another reason why a Window server or Linux server does not do it but a Linux Nas will. It called lack of ram so to protect your data being sent to nas it has to be sent to disk asp.

NTFS also has a trick up its sleve for a 1 tb file test. Called compression. A 1 tb file of all 0 will basically be first written to disk as a compressed file even if you have not set windows to do that. Its surprising how bad Windows handles a 1 tb file that cannot be compressed.

oiaohm said...

Most Nas's are PPC chips. They can run programs straight from hard drive without putting program in ram. Due to how light on for ram those Nas's are you can guess where ssh client and other gnu tools are.

No disk access no applications.

Anonymous said...

Now we're qualifying the problem. "Oh, I meant NTFS is only a problem when the PC is powered by a fat guy riding a stationary bicycle". Come on dude, if you lose data due to power fault, it's the power fault's problem.

Even still, I've been present for some pretty ungraceful shutdowns. NTFS still acts like a champ compared to reiser and the entire ext family, which probably comprise 99% of actual Linux systems. I'll bet JFS and XFS are better, but since no one uses them I'll never know.

I'm sure NTFS fails hard in some fucked up scenarios I've yet to come across, but I don't care and neither should you. In the sane world, it runs without major (or even minor) difficulties.

Anonymous said...

Oh yeah, UFS is way better than Linux shit, too. More Linux NIH for passing over that one. Amateurs.

Anonymous said...

speaking of linux NIH, have you seen that epoll shit?

Anonymous said...

<°(((((><

Anonymous said...

====D~~~~ O^:

oiaohm said...

Look inside NAS's. XFS is the preferred file-system. ext3 secound. ext2 only for read only filesystems, reiserfs never.

Ext3 yes new files being written leaves behind incomplete nodes. Old existing data on that file system is safe. That is of course if you leave it in its default setting min. Not be stupid and alter option lower. Journal is the highest setting that is 100 percent secure at a speed cost.

ordered mode will not damage any other file than one being currently written. That is the default.

Now the insanity that normally causes most of EXT3 data loss is users enabling writeback mode. Yes its faster big you might as well be using ext2 bare min protection to die.

NAS's use only EXT3 journal both of the others are not used out of EXT3.

reiserfs due to technial design flaw "Some directory operations (including unlink(2)) are not synchronous on ReiserFS, which can result in data corruption with applications relying heavily on file-based locks (such as mail transfer agents qmail[5] and Postfix[6]) if the machine halts before it has synchronized the disk."
Should not be touched with a ten foot pole if you value your data. NAS makers are also not that stupid. Anyone caught using Reiser FS for anything business needs to be strung up. I class it as bad as using NTFS for different reasons.

NTFS New stuff being written just ceases to exist along with what was being overwritten. Due to failed meta-data handling. Metadata points directory to a area of blank data to write new directory data. NTFS calls it a empty directory after a restart no error.

Ext2 do have error detection that detects that error by the way. NTFS design does and due to write orders NTFS driver is doing it don't work.

Correct write order is write primary MFT enteries write data Write backup MFT. Order NTFS is really doing is Write primary MFT Write Secondary MFT Write data. All the fix need is moving 1 write operation. So instead of that in MS introduces in Vista and up Transaction mode to NTFS. That eqauls operations can be bundled as a block to apply to file-system. Great idea not. If you don't fix the basic problem on wrong write order you basically wasted your time.

ZFS, NILFS is taking on the same problem. What does it introduce so simple real time snapshotting. So you say to the file system remember this files at X point in time to be referred to latter if need then do your alterations. Also snapshoting does equal to windows rollback system without blocking means to defrag disk and without using any more blocks need.

MS basically has implemented the wrong god darn feature to fix the problems.

Problem you have here. Seeing EXT3 or NTFS screw up badly past what is called normal problems is luck unless you are at a concentration point.

Lot of people take EXT3 screaming about broken inodes here and there as a bad thing. NTFS equal say nothing what one is worse. As long as the two MFT pages match even if they are blank when it page point to it says there should be data there NTFS calls it fine.

NTFS is silent so its errors go invisibly by. Even ext3 in full journaling mode is going to yell after a crash it is design to do that. So you know where the fragments of data are so you can attempt to recover them.

This is your major problem. Silent after a power cut=Really BAD. You guys are making Silent after a power cut= Really good. Not everything could have been completely written to disk there has to be fragments of data unless you were extremely lucky.

A power cut a ext3 filesystem with journaling on enough soon or latter you will get luck and have it come up without and inodes. Thinking that the inodes reports are partly written data. Its rare that you can.

Again the difference between data recovery person and normal. Recovery person wants to know where the partly written data went so we can try to reconstruct the missing information.

Basically in my eyes NTFS and EXT3 are crap for there own reasons. Yet if I had to choose crap I would choose EXT3.

Most of you have the idea I like Linux. Nop. It just gives me less trouble from what crosses my desk.

Come on dude, if you lose data due to power fault, it's the power fault's problem.

If your machine does not boot after loss of power? Is that the power fault? NO.

If your machine does not tell you what fragments you had written before dropping them in the bin of no return is that the power fault? NO

Simple fact here power dropping out is blamed for more than it should be.

This is the problem power failures get blamed for more damage than what should happen.

oiaohm said...

A power cut a ext3 filesystem with journaling on enough soon or latter you will get luck and have it come up without any inodes errors. Thinking that the inodes reports are partly written data. Its rare that you can.

Correction in bold.

Anonymous said...

oiaohm will you stop posting here please? I stopped reading your crap weeks ago, I just scroll it down...

I don't think I'm the only one doing this, so, better cut your losses and STFU, you aren't even funny to mock with

Kharkhalash said...

I'll bet JFS and XFS are better, but since no one uses them I'll never know.

They are, though the Linux version are less robust and mature than the native ones.

JFS and XFS each have almost 2 years of active development and testing in real world situations where high reliability is demanded.

JFS is solid enough for IBM to trust it as AIX's native file system and using it on their high end Blade servers and storage devices, and XFS is solid enough for SGI to have trusted it as the native FS for IRIX on their servers, super-computing clusters and high end multimedia workstations.

Who needs 20 years of tried and true, battle tested, reliable technology when you can produce your own craptastic alternative. Hey, it's nowhere near as reliable, mature, or robust, and it randomly nukes data, but hey, who cares, it's Free and Open, woot woot!

but I don't care and neither should you. In the sane world, it runs without major (or even minor) difficulties.

In the sane world, people back up their data, anyway, just in case, so who really cares if the fs pukes shit out in one-in-a-billion fucked up Bizzaro universe circumstances, right?

----------

NTFS also has a trick up its sleve for a 1 tb file test. Called compression. A 1 tb file of all 0 will basically be first written to disk as a compressed file even if you have not set windows to do that. Its surprising how bad Windows handles a 1 tb file that cannot be compressed.

What you call a "trick up its sleeve", sane people call common sense. Why would deal with a 1tb file when you don't strictly have to?

Why can't go with the compressed, expandable file route, too? Right, lack of clever engineers.

Mind doing the to windows cutting of power half way through a write.

Why on earth would anyone purposely cut power in the middle of a write?

Again a case of blaming Linux for something that is common to all.

90 second write delays (ext4) very much are ext4's fault. There's way too much room for error,

It was flagged stable and widespread adoption was encouraged, despite the fact that the data loss issues were widely known, this is Linux devs' fault.

Delaying the work around until the next kernel release instead of releasing a fixed release right away, when it was ready is completely Linux's fault.

No other file system leaves write delays anywhere near that long, this is ext4's fault alone.

Nobody denies that losing power mid-write is common, what's being disputed is that power failure is not required for both the old and new file to be truncated to 0 bytes, a smple application crash will do fine. Again, because the write delays are way, way, WAY too long.

Other filesystems reduce the room for error by employing sane write delays. ext4 instead extends write delays. This isn't a glass house issue, it's a pointing out a profoundly stupid design flaw, which is unique to ext4.

NTFS has many major failures most solved by reinstalling windows and keeping backups because most people don't know what as gone wrong.

Backups solve all problems.
Are you suggesting that people should, instead of keeping backups just in case, should rely on the filesystem and the filesystem alone?

That's just retarded.

Proper raided and UPS servers I basically see none from any OS.

No shit, Sherlock.
There's no way around lost writes due to incomplete writes, other than minimizing interrupted writes, that's what a UPS is for.

Sensitive data should be kept behind a UPS and backed up regularly and redundantly. We've discussed this before.

Kharkhalash said...

I don't think I'm the only one doing this, so, better cut your losses and STFU, you aren't even funny to mock with

I do this a lot of the time too. Sometimes, I'll comment on the more glaringly retarded, poorly thought-out nonsense, but most of the time is just strings of incoherent gibberish.

I think he's trying for avoid points by hoping they get lost in a sea of pointless, irrelevant information. He hasn't realized that it doesn't work, yet.

yOSHi314 said...

phone is an embedded device, so it would use a specialized linux installation. i don't think you would have that many different UIs for it.

e17 is possible for handhelds but maybe not for phones

"grep’ing through your address book using extended regular expressions "

actually i would love to have this for searching sms archive. but a bit simplified.

"Losing data randomly with ext4 "
i don't think ext4 on phone flash memory is the way to go.

"Being able to ssh to your phone so that you can check it’s uptime "
i'd love it. but not for checking uptime, but sending files over and remotely writing sms.

"Using PGP to sign your SMS messages. Beware though, after the signature, you’ll only have 5 characters left to work with. "
how about sending compressed sms messages? text packs quite well, you know :]

.net jerkface said...

Hey I wanted to add a few more features:

Comes with 5 contact programs: Contactaruni, kontacto, cnt2238, fishdude, and The Cripple. All are missing a feature that the others have.

Resolution sometimes resets to 800x600.

Downloading ringtones results in a prompt asking you if you are sure if really really want free non-free ringtones.

Bluetooth can be spotty. That should be improved in the next version though.

Volume button doesn't work with some applications.

Includes the latest games like portable tux racer and portable quake 3.

oiaohm said...

Kharkhalash
Why on earth would anyone purposely cut power in the middle of a write?

Simple fact you don't unless you are testing for robustness. So anyone testing how strong a OS will be in real world conditions should at least simulate it. Same with designing new file systems or deciding to change to a new file-system.

Its exactly the same as crashing a car into a wall and seeing how badly its broken so you can improve design.

It can simply done without risking hardware in a virtual machine ie killing the virtual machine without notice to the contained OS. Just like car crashing it should be done quite a few times before a file-system gets the green light for important usage.

EXT4 had not had that treatment when it was declared stable. Linux distributions started pushing it and had not done the testing. Nice big double failure the check and balance never happened. End users never asked for its test results against being damaged.

Handling errors happen. You can bet at sometime of the devices laptops will have people dis-guard the battery warning and keep on working leading to no power. Wrong computer will be unplugged from all. UPS fail. Its a real world event.

Basically lot of equals to near misses happen every day. Just because power dropped out at a good time.

Nobody denies that losing power mid-write is common, what's being disputed is that power failure is not required for both the old and new file to be truncated to 0 bytes, a smple application crash will do fine.

Application crash causing file to be truncated to 0 happens on all platforms. Only prevention is like nilfs constant check pointing so you can roll back to exactly before the truncated. Perfectly recoverable from with file-systems built to handle it. Its another one of those problems people like to say is not fixable.

Application truncated to 0 is different to the EXT 4 fault.

Ext 4 fault needs power failure its a half written rename. Same with the 30 second delay it also needs a power failure to be a problem. No power failure no problem.

"90 second write delays" is not the error people point to. It the file that was on the drive for the last 5 days that has now been truncated to 0 or disappear because of renames. They would have been annoyed losing the last 90 seconds of work. They were rope-able for losing files that still should have been there. best was rename a file to c b file to a and c file to b and power outage before all that went though that could result in a and b file ceasing to exist. Not a good error.

Before you sing the praises of XFS too much 2.6.30 has a patch to XFS against exactly the same rename error. The code design in EXT4 was taken from XFS for the rename not EXT3 because XFS was though to be more stable. Yes people can sometimes screw up the EXT3 one was correct.

The rename error is ancient its turning up that is been around for over 20 years basically copied from OS to OS and file-system design to file-system design. Expanding the sync time has truly brought a fairly major hidden defect out into the open. It predates Linux and NT been in BSD it may go all back to the start of Unix.

Next question that everyone should be asking how a fault like the rename bug has gone unnoticed for over 20 years by everyone. Next question after that is how many other ancient faults exist that effect all existing OS's.

Please go and read what the Linux users were truly screaming over. It is not the 90 second sync. Yes that increased the risk of problem happening yes lots are not happy with it now.

There are two bugs in Linux 2.6.29 ext4 long auto sync. Minor headache this one does not vaporise old files. Anything that was written to disk.

Then the rename fault that will vaporise old files. Its a write order fault Dependant on power being lost at the right points. Corrected for 2.6.30.

Declared ext4 stable is not real world tested no where in its testing had it had the power cut emulation test. So too long of sync time and bad operation slipped though.

Each one of these problems should be taken as a lesson.

Keeping backups is a good operation. But here is the important but.

Problem with backups is people with laptops traveling light don't always do backups for there work in the field. This is why laptops make up a large amounts of numbers of items that need data recovery.

Real world interfere with the ideal conditions of having backups and raids.

I don't suggest people should not keep backups. I want file-systems as strong as able so the time between backups has the least risk of data loss.

Lot of NAS's make up numbers mostly because some companies use them as off site storage moving them every day. So far the hit by train takes cake.

Most of the NAS recoveries would not happen if some of the IT personal out there did not trust them 100 percent and keep backups. The other problem when you are dealing with real world you have to allow for real world human stupidity. Mostly it worked perfectly so far so it will keep on working perfectly. Use to have to do the same with the over recycling of backup tapes.

Turns out NAS's that people take home get treated about 20 times worse than laptops. Brick shaped device gets handled with care for a while but after a while is just thrown into car.

alexei said...

It is brilliant! Thanks!

Anonymous said...

Looks like something similar has happened with netbooks, the thing that was meant to Put Linux in Every Home:

http://www.osnews.com/story/21269/Microsoft_Boasts_Windows_Netbook_Success

Anonymous said...

Please go and read what the Linux users were truly screaming over.

- lack of testing ... doesn't really count coz in Linux the end user is the tester
- a lot of porn was lost due to ext4

Each one of these problems should be taken as a lesson.

that's against linux policy

Anonymous said...

I have seen 3 Win2k systems in the same office spontaneously fail to boot the same day, their filesystem was corrupted beyond repair and no virus or reason for the corruption was ever spotted (no internet access). The same desktops had the same problem about 6 months before, no one knows what caused these events.

Transactions are usually meant to achieve file integrity, not in MS' case though! NTFS transactions can make your system unbootable after power loss, requiring a new Vista installation!

Regarding linux netbooks, the latest statistics (NYTimes) suggest that linux has ~ 20% marketshare, the <10% statistic refers to the US only, while ~ 90% of sales are not in the US.

Anonymous said...

20% doesn't look like a big success to me, specially when it's an 8-year-old operating system that it's taking the throne.

It's more like Linux has made a lot of noise but has failed to materialize into something concrete once again.

Of course, the aftermath will be the usual list of excuses and justifications instead of "what the heck are we doing wrong?"

Anonymous said...

Of course, the aftermath will be the usual list of excuses and justifications instead of "what the heck are we doing wrong?

MS is entrenched in the desktop market, haven't you heard? Win32 support is a requisite, if you intend to enter this market, and X is currently in bad shape, what more can you expect at the moment?

Anonymous said...

I haven't reinstalled a Windows OS in at least 5 years... despite some really bad power hits and such. Linux i=on the other hand got buggered by a power hit 2 days after i installed and required a full reinstall just so it oculd even boot up.

MFTcorruption is one i've never seen either, not that i don't doubt it exists. anything can happen. but i'm glad the multi hundreds of thousands of dollars equipment we use at work run WinXP instead of Linux. at least power loss from frequent storms means we only lose the last few seconds of cutting data... instead of the whole bleeding OS.

Anonymous said...

Regarding XP and Netbook, the Freetards on /. are swimming in an Egyptian river:

http://tech.slashdot.org/article.pl?sid=09/04/06/2228220&art_pos=5

We could confirm that yes indeed, Xp now sells on like 99.9% of all netbooks and explain why, but we would be wasting our time...

Anonymous said...

Never had any single problem with data write in NTFS. From NT4.0 Server and Workstation to Windows Server 2008 no single file was lost due to filesystem's fault. The only cause of death of my hard drives were bad sectors popping up in bootsectors (hardware dying, and that's a classic, every HDD fails over time), and one big blow for a client whose only copy of the data burned up thanks to NO backups (moron), no UPS or power protections, and a lightning.

So, hardware failure over time, and no prevention from clients, that's what I see.

Never seen any mystical system32 disappereance, those things should be kept in mythology books.

Anonymous said...

MS is entrenched in the desktop market, haven't you heard? Win32 support is a requisite, if you intend to enter this market, and X is currently in bad shape, what more can you expect at the moment?

See? Excuses and justifications. What I certainly don't expect from the Linux community is to get their head out of their asses and realise that they need to build a goddamn platform and make themselves attractive to OEMs.

Why would they do that? It's so much easier to blame anything else and cry on digg.

Anonymous said...

Lol check this out:

http://blag.xkcd.com/2009/04/03/what-happened-to-my-laptop/

Anonymous said...

Lol check this out

What's even better, imagine they want to push this shit to people who don't know how to set the clock on their microwave oven.

Seriously, after reading all this bullcrap from freetards like oiaohm and /. I really wish this linux desktop bitch dies in agony. It's a disgrace to anybody involved in software engineering.

Anonymous said...

And you still have to edit the sources list manually...in 2009...with Graphical Package Managers...

Anonymous said...

In the sane world, people back up their data, anyway, just in case, so who really cares if the fs pukes shit out in one-in-a-billion fucked up Bizzaro universe circumstances, right?

Are you suggesting that people should, instead of keeping backups just in case, should rely on the filesystem and the filesystem alone?

Yeah, in my world, I'm too busy dealing with "Oh, I accidentally deleted my entire client list but I don't remember when. Can we get it back?" to worry about whatever the fuck NTFS is up to.

(P.S. catchpa is "mardingo". That's just wrong.)

Anonymous said...

Resolution sometimes resets to 800x600.

Since the phone's display is probably closer to 200x200 and we're probably talking about X11, this means the visible area will fly all over the place in a non-accelerated fashion (and I mean like mouse acceleration, not 3D graphics) whenever you click near an edge of the screen.

Anonymous said...

Ah, the usual Slashdot Ballmer chair jokes, first thing modded up on that article. Too bad, Slashtards were so close to finally pulling their heads out of their asses by at least admitting Windows 7 is turning out "okay". The delusion in the rest of that thread ranges from "the statistics are obviously wrong" to "we should lobby the industry to change the netbook hardware so M$ OSes can't run on them". In Soviet Russia, Linux installs you!

Anonymous said...

That Linux hating thread from a few days ago was actually insightful, but now Slashdot is back to moderating content free junk like "Alphas were good chips" as +5 insightful.

Anonymous said...

"I have seen 3 Win2k systems in the same office spontaneously fail to boot the same day, their filesystem was corrupted beyond repair and no virus or reason for the corruption was ever spotted (no internet access). The same desktops had the same problem about 6 months before, no one knows what caused these events."

a peak of current (i think that's how you say it in English) was probably the cause

THE INCREDIBLE HULK said...

PUNY LINUX. HULK SMASH!!!

.net jerkface said...

Pony tailed douchebag on osstatic:
Free and open source software is sweeping across the vast majority of the Fortune 500. When you see the world's most conservative companies starting to deploy open source, you know momentum is on your side.
http://ostatic.com/blog/sun-microsystems-ceo-jonathan-schwartz-on-whats-next-for-open-source

Pony tailed douchebag to IBM:
We're screwed, please buy us. Oh and I want even more compensation for running the company into the ground. Same for McSnarky.
http://blogs.zdnet.com/BTL/?p=15935


Sun is a Class A example of what happens when you try to cosey up with the slashdot crowd with the expectation that your good intentions will translate into profits. What you actually end up with is a big bill and no one around to pay it.

The pony tailed douche also couldn't get the slashdotters to contribue code to opensolaris. He doesn't understand that a big part of the foss appeal is that it isn't corporate directed. What a shocker that there aren't thousands of programmers that want to volunteer to write code for Sun. What a shocker.

MySql is already being forked by the orignal creators. Sun could have forked their own but instead decided to pay 1 billion for the MySql logo. Way to go.

If you have any solaris servers I would suggest migrating them to freebsd. Sun is a sinking ship and their top engineers won't be staying with the captain.

Kharkhalash said...

Its exactly the same as crashing a car into a wall and seeing how badly its broken so you can improve design.


in the context of the discussion, no it isn't. It's analogous to you telling people to go crash their cars into a wall.

EXT4 had not had that treatment when it was declared stable.

Right, because we all know Linux people don't really do Q&A before marking things stable.

Linux distributions started pushing it and had not done the testing.

Blame the distribututions all you want, it's the kernel team and the ext4 team who marked it stable and encouraged widespread use in production environments. The kernel/ext4 devs are to blame for being stupid, and the distros are to blame for taking their word for it.

End users never asked for its test results against being damaged.

So now it's the users' fault? Fuck that. The end user shouldn't need to ask for such results. The end user expects that sufficient Q&A is done to ensure that something marked "stable" won't randomly nuke their data.

Handling errors happen. You can bet at sometime of the devices laptops will have people dis-guard the battery warning and keep on working leading to no power. Wrong computer will be unplugged from all. UPS fail. Its a real world event.

And yet no amount of tweaking the filesystem will prevent that.

Application crash causing file to be truncated to 0 happens on all platforms.


No, an application crash resulting in BOTH the old and new files to be truncated to 0 bytes does not happen on all platforms.

Application truncated to 0 is different to the EXT 4 fault.

No, it isn't. New and old file being truncated to 0 bytes as a result of an interrupted write (be it power loss or an application crash) is exactly the bug Theodore Ts'o described.

Are you suggesting that you know ext4 better than its maintainer does?

Ext 4 fault needs power failure its a half written rename. Same with the 30 second delay it also needs a power failure to be a problem. No power failure no problem.

Wrong. And it's a 90 second write-delay. If the write gets interrupted at any point within approx 2 1/2 minutes after the write is requested (90 second write-delay until the write is initiated, and another 60-90 seconds until the write is complete) if the write is interrupted within 90 second, chances are you only lose the new data. if it's interrupted after the write is initiated, and before the write in complete, you lose the old data as well.

"90 second write delays" is not the error people point to.

It's the cause of the data loss. Nobody is going to report a 90-second write delay as a bug.

It the file that was on the drive for the last 5 days that has now been truncated to 0 or disappear because of renames.

and a rename isn't a write? You're inventing new problems. And you haven't been paying attention. An interrupted write caused the whole file, not just the temp file to truncate to 0.

They would have been annoyed losing the last 90 seconds of work. They were rope-able for losing files that still should have been there. best was rename a file to c b file to a and c file to b and power outage before all that went though that could result in a and b file ceasing to exist. Not a good error.

A power outage mid-write causes data loss, yes. The problem here is that a) on ext4, the same effect is achieved by application crashe]s. This is NOT normal. b) The write delay is way, way, WAY too long, and invites this sort of problem. Sorter write delays narrow the window of opportunity for interrupted writes, ext4 widens the window of opportunity by extending the write delay (which results in better performance, with the cost of 0 reliability. the write-delay for ext4 is 18x longer than that of ext3).


Before you sing the praises of XFS too much 2.6.30 has a patch to XFS against exactly the same rename error.

Are you illiterate? I was very clear that the Linux ports of XFS and JFS are nowhere near as robust, mature, performant and reliable as the original, native versions (for IRIX and AIX, respectively
).

What? you thought someone got a pouch of magical, FOSS pixie dust, sprinkled it on the Linux-XFS CVS trunk and had it inherit the 20 years of Q&A done on IRIX's XFS implementation?



The code design in EXT4 was taken from XFS for the rename not EXT3 because XFS was though to be more stable. Yes people can sometimes screw up the EXT3 one was correct.

[Citation needed]

Ext4 is was initially conceived as a set of extensions to ext3, instead it was forked from ext3 in way for ext3 to be forward compatible with ext4, and for ext4 to be backward compatible with ext3. XFS doesn't come into the equation.

Also, ext3 had a 5 second write-delay which is why random data loss was less common.

XFS has a slightly longer delay because of its SGI roots: High performance file system designed at optimized for multimedia, running on a system optimized and designed for high performance multimedia, running on hardware designed and optimized for high performance multimedia, it's fine tuned to work with excessively huge files (such as raw HQ video data), of course, of IRIX, XFS had snapshots, which help to minimize data loss, and no, the volume manager (which did snapshots, since, you know, IRIX was a fully integrated system) isn't included in the Linux port.

So before you slag on XFS, go back in time and use it in the context it was designed to be used in: on IRIX on a SGI Tezro workstation. Kthnx.

The rename error is ancient its turning up that is been around for over 20 years basically copied from OS to OS and file-system design to file-system design.

the potential of losing data on an interrupted write is not a bug, it's a technical limitation inherent to file systems. It's normal and expected behavior:

You're performing an action, you're interrupted, the action is left incomplete, the desired outcome, as a result is not attained. It's common sense, not rocket science.

There's no direct fix, however, there are countermeasures: UPSes, shorter write delays, FS snapshots, regular, redundant backups, and saving new data to a temporary file, snapshotting the file to be replaced, then overwritng the old file with the temp one, transactions, journals and soft-updates, for example, to reduce the risk, frequency and impact of interrupted writes resulting in dataloss.



Expanding the sync time has truly brought a fairly major hidden defect out into the open. It predates Linux and NT been in BSD it may go all back to the start of Unix.

It's an inherent technical limitation, that goes back to the beginning of file systems, not a bug or defect.

Next question that everyone should be asking how a fault like the rename bug has gone unnoticed for over 20 years by everyone.

It hasn't, otherwise the cocophony of countermeasures would not exist.

Please go and read what the Linux users were truly screaming over. It is not the 90 second sync.

Are you completely dense?
They aren't complaining about the reason dataloss is so frequent and random on ext4, they don't care why, nor should they. They only care that they're losing data frequently and randomly when they shouldn't.

They're screaming that it was marked stable and production use was encouraged when it was clear that it wasn't stable.

Yes that increased the risk of problem happening yes lots are not happy with it now.

Yes, now that they know why it happens, they realize that extending write delays was a profoundly stupid decision.


There are two bugs in Linux 2.6.29 ext4 long auto sync. Minor headache this one does not vaporise old files.

Yet there is a bug which vaporizes old files, Theo Ts'o blames this on application developers.

Dependant on power being lost at the right points. Corrected for 2.6.30.

Wrong, it's also triggered by application crashes. There's a workaround in 2.6.30 (not a fix, there's a distinction), which has not been released yet. That raises the question: why in the blue hell was a new release with this workaround released right away, the second it was made ready?

So too long of sync time and bad operation slipped though.

You really don't need to run tests to figure that extending write delays by 18x opens the door for all kinds of dataloss problems. And again. power outages are not necessary for this behavior, that's why it's a problem.

Each one of these problems should be taken as a lesson.

Don't trust the kernel devs.


Problem with backups is people with laptops traveling light don't always do backups for there work in the field.

They should. The fact that they don't is their fault and their's alone. No amount of filesystem optimization will compensate for reckless behavior.

Real world interfere with the ideal conditions of having backups and raids.

Not really. You either keep backups, and are able to recover your data, or you don't and when your data is lost it's gone for good. It really is that simple.

Lot of NAS's make up numbers mostly because some companies use them as off site storage moving them every day. So far the hit by train takes cake.

Companies have the resources to have redundant off-site backups. Moving a single NAS around doesn't correct the fact that there is a still a single point of failure.

Anonymous said...

79th!!

Anonymous said...

Are you suggesting that you know ext4 better than its maintainer does?

Of course he is. OnlyOiaohmKnows(TM)

Anonymous said...

Regarding netbook, I suggest this one:

ArmIsTheSavior(TM)

Anonymous said...

We had one similar already: X86IsDying(TM)

Anonymous said...

2001: Xp will bring in the year of linux

2006: Vista will bring in the year of linux

2007: Netbooks will bring in the year of linux

2009: ARM Netbooks will bring in the year of linux

2030: Self-aware cyborgs will bring in the year of linux

2050: After the great war against the machines, linux holds about 55% marketshare in surviving bartertowns. However no one knows what linux is and no one cares since computers are mostly collected to trade for weapons.


In other news this could shake things up since we will be able to see a direct kernel comparision.
http://www.h-online.com/open/FreeBSD-7-2-Beta-released-Debian-FreeBSD--/news/113022

Anonymous said...

Microsoft still hasn't fixed the bug in their file-system where a partition losses all data if the drive is smashed, degaussed, or melted.

Thus far Linux has been successful mitigating this problem by producing a product that nobody wants to use.

Anonymous said...

@April 7, 2009 3:55 PM

Guess you haven't heard about ext5's "shadow platter" technology where hard drives can form a type of implicit "RAID" mirror using neutrino flux of the surrounding air. This will at long last fix the Microsoft bug of hard drives not working after their destruction. Evil Microsoft is in bed with hard drive companies and gets $$$ kickbacks by forcing "upgrades" of broken equipment. ext5 will end this tyranny.

Anonymous said...

That would be awesome. Finally I can avoid having my files truncated to 0 after I've dropped my laptop out of a plane.

Even though Linux hasn't made the technology available, I might as well go ahead and switch now.

Anonymous said...

Imagine distros switching from the Linux kernel to the FreeBSD kernel. With the "me too" freetard attitude that brought us the PulseAudio fiasco, it's possible.

DamnRight said...

NTFS is a glass house with broken windows owned by a stack of people who think a glass house with broken windows is how it should be.

Such a poetic metaphor. You really are the king of projection, if nothing else. I'd wonder about what your opinion of Linux is then, but I know the answer already. Everything will be fixed six months from now, by magically turning every single subsytem inside out without any coordination work needed or any issue arising.

Anonymous said...

@April 7, 2009 4:46 PM

Unfortunately, "shadow platter" technology only works when the computer is perfectly still as the neutrino flux is very sensitive. Even the minute hard drive vibrations can corrupt the "shadow platter" technology. To address this, the Linux community plans to boycott all hard drives that create mechanical vibration. Furthermore, we expect all major manufacturers to "open source" their fabrication processes so that the community may use their superior manufacturing experience to improve their products.

Oh, did I mention that the backup copy dissipates in a time frame too minute to be calculated with current instruments? This, too, is Microsoft's fault for sabotaging the time measurement industry for two decades.

Anonymous said...

If Windows is a glass house with broken windows that people view as the norm then Linux would have to be a burning house where people hide in their closet from the flames screaming at the fire-fighters trying to rescue them by denying there is a fire.

Anonymous said...

Linux haters are fags.

thepld said...

" Imagine distros switching from the Linux kernel to the FreeBSD kernel. With the "me too" freetard attitude that brought us the PulseAudio fiasco, it's possible."

But wait! I thought BSD users were evil for not using the GPL and for producing actually useful code! Also, how dare you not admit to the brilliance of the Linux kernel, what with its totally unstable APIs and constant bugs.

thepld said...

Oh and for the record, I was able to update FreeBSD from 6.0 to 7.2 without breaking a single thing. I'd love to see Ubuntu survive an upgrade to a single release.

Anonymous said...

"http://blag.xkcd.com/2009/04/03/what-happened-to-my-laptop/"

That was a great read. Of course the tard patrol comes in and blames anything BUT the Linux OS itself. Yeah, the main problem is LinuxOS. Dude shoulda gone BSD or Solaris if he wanted to try a real nix alternative. Or Windows/OSX if he just wanted it to work.

The repository is another of Linux OS Achilles's heel. Once you mess with your sources list, you may accidentally bork your system by upgrading a system or DE file.

Anonymous said...

What Happened to my Laptop

By I'm an idiot and proud of it

For the people who wanted to know the chain of events causing the laptop problems in the last post:

1. Safe search won’t disable, and login for regular Google services is broken. I decide it’s a cookie conflict between Google services and file a bug report with Google.

2. Google says “can’t reproduce”. I discover that clearing private data doesn’t actually work right, and after some testing, go to file a bug with Firefox. They ask for my version number.

3. I notice my Firefox is out-of-date, and decide that it might be a quirk of this version of Firefox+Ubuntu. I go to upgrade Firefox in Synaptic before filing the bug. I don’t have the notifier running and haven’t updated in a while.

4. I upgrade a package in Synaptic, but it turns out I just upgraded the meta package and not the actual package (or something.)A more seasoned Synaptic user says “why not just run regular upgrades like a normal person? It’ll fold Firefox into it.” On any other day, this would have been good advice.

5. I start the upgrade. It’s churning along nicely, and I locate my cookies file and start examining it.

6. My battery monitor disappears. This is normal enough, actually. I try to start it back up and get a notice about a broken configuration file. Uh-oh.

7. My nm-applet disappears. That’s a little stranger, but neither applet is very reliable in my ion3 setup.

8. At this point I find that my cookies file is malformed in some way, and that if I move it manually (rather than using the in-browser ‘delete cookies’) the original bug disappears. Don’t know how that happened, but it’s solved, so I cancel the bug report with Mozilla.

9. I go to edit the cookies file in emacs, and get a message that emacs can’t start. Uh-oh.

10. The upgrade fails with a bunch of package incompatibility messages.

11. I start to feel alarmed.  I link an apt-guru friend a screenshot of a relevant part of my Synaptic window. He says, “one of those version numbers looks wrong. You’re running Hardy, right?” I am.

12. I open my sources.list (using nano, since emacs is broken.) After a bunch of spaces at the bottom, I see something bad: A Debian repository. I’m running Ubuntu.

13. I don’t know how I added it. Maybe I was on the wrong system and didn’t notice the prompt. Maybe I was trying to install one specific package from Debian (via apt, for some reason) and forgot to take it out. I honestly don’t remember. But since I hadn’t updated in a while, it hadn’t come up. But now I’m in trouble.

14. I remove it, update my lists, and do a grep to see how many of my packages have upgraded to Debian versions. 750-1000 or so. I paste the results back to some friends. One of them looks, shakes her head, and says my system is like that guy in Star Trek after the transporter accident.

15. Over the next few days, we try pinning packages back to the Hardy version and downgrading. There are conflicts all over and lots of ninjinuity is required. Sometimes the system won’t boot properly, claiming kernel module problems, which turns out to be because somewhere in this process my initramfs got misconfigured. But eventually, everything is put back in working order (except, for some strange reason, gnuplot, which refuses to install the main binary file. I compile it from source.)

16. Watching the system boot successfully, I go to pour some milk and cereal in celebration. I shake up the milk jug a little. The lid is loose. It spills all over the keyboard. The system stops booting and the cycle of
horror starts again. (And yes, if I’d gotten a Lenovo when you all suggested it, I’d have a spill-proof keyboard with drains. That’s my plan for the next time something horrible happens to my laptop, which should be any day now.)

Copyright 2009 I'm an idiot and proud of it

Anonymous said...

Web Hosting Talk is down

And yes, we run on Linux

April 7th 2009 07:27

In Progress - WHT Off Line

UPDATE: 7:14pm est 04/07/09

From what we know now, there were more records on the database server where the credit card dump was taken. If research shows that a larger number of customer's data was compromised, we will contact those individuals directly.

UPDATE: 4:34pm est 04/07/09

It has been brought to our attention that any WHT Premium memberships purchased PRIOR to 2006would be included in the exploited credit card details.

UPDATE: 4:24pm est 04/07/09

We have contacted all major credit card companies and are awaiting their guidance. It should be noted that card holders will not be held liable for any fraudulent purchase made using their credit card.

ANNOUNCEMENT - 1:25pm est 04/07/09

This morning, the hacker who attacked WHT initiated further communication. He provided
evidence that credit card information on one of our database servers was, in fact,
compromised during that attack.

What data was compromised?

At this point, we know that the hacker compromised and has publicly posted credit
card information from our self-service billing system currently used for sticky posts (located at http://myinet.inetinteractive.com).This system was also used for display (banner) advertising in prior to December 2007.

What about premium and corporate members? Or display advertisers?

If you've purchased a premium or corporate membership or you are a display (banner ad) advertiser from December 2007 or later, your data is safe. These products run
on a newer billing platform that does not store credit card information.

What is WHT and iNET Interactive doing about it?

If we have evidence or suspicion that your credit card information was leaked, you will be receiving further communication from WHT and iNET Interactive.

Why is WHT down and when do we expect it to be back up?

We're currently doing a full security sweep of our cluster to ensure the servers are secure. The site will be back up once this security review is complete.

Anonymous said...

More on Web Hosting Talk

And yes, Web Hosting Talk runs on Linux

"WebHostingTalk.com has been compromised once again but this time the Credit Card details portion of their database was compromised and is now circling the internet as "WHT_XXXX_XXXXXXX_XX.rar". This file contains *thousands* of credit cards with complete details including Credit Card Number, CVV2 Number, Full Name, Bank Name, as well as some other details."

LOL

Anonymous said...

Wow, that's messed up about WHT. Secure on the server, eh? Lintards. So, where can I download that file...?

Anonymous said...

Time to switch off that Ubuntu cPanel VPS and run a Windows Server 2008 box, I guess, WHT? No smug Linux "experts" needed.

Anonymous said...

LOL sounds awfully like the "M$ SUX LINUX 4EVA" kids on UbuntuForums. Lintards, this is what you've created:

http://www.webhostingboard.net/showpost.php?p=3191&postcount=21

Anonymous said...

This guy has copy of the message that was included in the compressed file
with all the credit card information from WHT:

The hacker wrote:

Ok so backup only was not enough for you pussys, HERE, have some credit cards too

You know, it’s fucking hilarious you negros backed up some bullshit backup and users still got the same passwords.

Well some did change, to be precise, 1348 users out of 200,000. AWE-FUCKING-SOME, no? You faggots couldn’t even bother changing your fucking HOST.

Why the fuck are you looking into your shitty server? GO LOOK AT YOUR HOSTER, RACK-SHIT-EDGE YES GO.

NEWAYS, what I noticed in the helpdesk jizz shit you got, you say you can’t give out information about ur server?

IT’S SHIT PEOPLE. 2GB ram for a site like this? gimmeh a break nigz. Also all those motherfuckers who were bragging the haxored thread at forums,

you negros think I give a shit you say ’bout me? I can rape your fucking life easy as 1,2,3 and this just proves the fact that I can.

Anyway, hopefuly this is the last time I have to “try” get into ur shitty servers(COUGH RACKEDGE COUGH) to back the shit up, I wonder if u nigz gonna get sued over this shit. lolz

OOH before I finnish, u nigz reported the box I posted backup from the 1st time, what would you do if i posted from ur own box now faggot? I got more places to post
from than u got hair on ur head or mby ur bald lolz

k peace out fags and dont mess with me

Anonymous said...

It'll be fun to have that WHT rar file just to make fun of Lintards.

Anonymous said...

"2GB ram for a site like this?"

Oh, snap. Dissin' their RAM. That's REAL. 50 Cent had a song about that - he killed the last motherfucker who dissed his box.

thepld said...

"do if i posted from ur own box now faggot? I got more places to post
from than u got hair on ur head or mby ur bald lolz"

There's a poetic quality to that!

Anonymous said...

Well, he did make a good point about them failing to change passwords.

Anonymous said...

Found it!

Digital Point Forums - 19 hours and 47 minutes ago

Ok so backup only was not enough for you pussys, HERE, have some credit cards too :)
You know, it's fucking hilarious you negros backed up some bullshit backup and users still got the same passwords.
Well some did change, to be precise, 1348 users out of 200,000. AWE-FUCKING-SOME, no? You faggots couldn't even bother changing your fucking HOST.
Why the fuck are you looking into your shitty server? GO LOOK AT YOUR HOSTER, RACK-SHIT-EDGE YES GO.
NEWAYS, what I noticed in the helpdesk jizz shit you got, you say you can't give out information about ur server?
IT'S SHIT PEOPLE. 2GB ram for a site like this? gimmeh a break nigz. Also all those motherfuckers who were bragging the haxored thread at forums,
you negros think I give a shit you say 'bout me? I can rape your fucking life easy as 1,2,3 and this just proves the fact that I can.
Anyway, hopefuly this is the last time I have to "try" get into ur shitty servers(COUGH RACKEDGE COUGH) to back the shit up, I wonder if u nigz gonna get sued over this shit. ;-) lolz

OOH before I finnish, u nigz reported the box I posted backup from the 1st time, what would you do if i posted from ur own box now faggot? I got more places to post from than u got hair on ur head or mby ur bald lolz

k peace out fags and dont mess with me ;-)

Download:

#Rapidshare: http://rs160.rapidshare.com/files/218436279/WHT_OWNAGE_iNETiNT_CC.rar
#RS MiRRoR: http://rapidshare.com/files/218437279/WHT_OWNAGE_iNETiNT_CC.rar
#Direct: http://isitetech.com/WHT_OWNAGE_iNETiNT_CC.rar
#Uploading: http://uploading.com/files/BL7JGOAS/WHT_OWNAGE_iNETiNT_CC.rar.html
#More: http://mirrorcreator.com/files/AIJCNUFJ/WHT_OWNAGE_iNETiNT_CC.rar_mirrors

This time I'll make sure everyone gets to download it :-)
NO FKIN AROUND U CANT STOP ME NIGGERS.

Anonymous said...

Fixed links...

Digital Point Forums - 19 hours and 47 minutes ago
 
Ok so backup only was not enough for you pussys, HERE, have some credit cards too :)
You know, it's fucking hilarious you negros backed up some bullshit backup and users still got the same passwords.
Well some did change, to be precise, 1348 users out of 200,000. AWE-FUCKING-SOME, no? You faggots couldn't even bother changing your fucking HOST.
Why the fuck are you looking into your shitty server? GO LOOK AT YOUR HOSTER, RACK-SHIT-EDGE YES GO.
NEWAYS, what I noticed in the helpdesk jizz shit you got, you say you can't give out information about ur server?
IT'S SHIT PEOPLE. 2GB ram for a site like this? gimmeh a break nigz. Also all those motherfuckers who were bragging the haxored thread at forums,
you negros think I give a shit you say 'bout me? I can rape your fucking life easy as 1,2,3 and this just proves the fact that I can.
Anyway, hopefuly this is the last time I have to "try" get into ur shitty servers(COUGH RACKEDGE COUGH) to back the shit up, I wonder if u nigz gonna get sued over this shit. ;-) lolz

OOH before I finnish, u nigz reported the box I posted backup from the 1st time, what would you do if i posted from ur own box now faggot? I got more places to post from than u got hair on ur head or mby ur bald lolz

k peace out fags and dont mess with me ;-)

Download:

#Rapidshare: http://rs160.rapidshare.com/

#RS MiRRoR: http://rapidshare.com/

#Direct: http://isitetech.com/

#Uploading: http://uploading.com/

#More: http://mirrorcreator.com/

This time I'll make sure everyone gets to download it :-))

NO FKIN AROUND U CANT STOP ME NIGGERS.

Anonymous said...

So far, this is the only
link
that works

So much for Lintard security!

LOL!

Anonymous said...

I'll never buy hosting from those "isitetech" faggots.

Anonymous said...

The above message was available earlier at these Google links. Later, the messages were removed by forums.digitalpoint.com .

You can see copies of the deleted messages here. Click on the appropriate messages on the + sign.

Anonymous said...

By the way, those .rar files are beautifully expanded with the latest version of WinRar at http://www.rarlab.com/download.htm

The hacker was nice enough to upload his .rar files free of viruses.

I'm sure the dudes in the iNETiNTERACTiVE_CC_NFO and LOL_BEANS_WHT_BACKUP_06_04_2k9 will never use WHT anymore!

This 'conficker' a la Linux really hit them hard!

LOL!

Anonymous said...

By the way, the iNETiNTERACTiVE_CC_NFO SQL file has
2144 CC records.

So that's about 2000 dudes walking way from inetinteractive.com . I hope these 2000 idiots have already cancel those CCs.

Anonymous said...

Hopefully this 'conficker' a la Linux makes freetards humbler.

Fuck them anyway!

Anonymous said...

ext4? what is that? Does it exist in Debian Stable? If not, it doesn't work.

People try UBUBU and Fedora 33 "the experiment", or make install some stuff and they expect something to work.

The current "stable" state of linux is always seen by the current state of debian stable. If you start to ignore every new package/features that comes up and mostly stay with debian, the linux world seems slightly better.

Or you can give *BSD a try.

Anonymous said...

Such a poetic metaphor. You really are the king of projection, if nothing else.

I think he's a cylon. And of course, he believes Linux has a destiny.

Anonymous said...

Debian stable is so fucking old, almost nothing can actually get done.

Oh, it doesn't actually change anything,

oiaohm said...

Problem is I have seen the dominance of Unix come and go. The dominance of Apple come and go.

Linux itself current state is not ready in my eyes for desktop usage. Will not be ready at least until when the graphical stack is ready. I have said this since 1997. I have laughed at people claiming lots of years as the years of the Linux desktop. The criteria was not meet.

There is a list of criteria a OS has to have to attack the desktop market.

Thinking Linux is not up-to the criteria being able to hold over 40 percent of netbook sales is impressive. So it should not be under estimated.

The big problem I have is that MS OS tech has basically stood still. Now people are claiming market will be held by Applications. Where have I heard that before that is right 1994 supporters of Apple about desktop publishing. Did Apple dominance stay no.

NTFS vs EXT3 NTFS has some advantages compression and encryption.

Now lets move on to the generation of file-systems Windows 7 will have to face.

NTFS vs NILFS. NILFS has the advantage of constant snapshotted so reduced data loss risk, NTFS has compression and encryption. Critical data usage NILFS wins.

NTFS vs Btrfs NTFS is basically out matched. Btrfs has snapshotting and compression. Btrfs has what is called writable snapshots so you have a stuff you can leave the damaged writes on the disk for latter inspection and keep on using the machine. Btrfs is a more suitable notebook file-system than NTFS is.

Problem here is compression and encryption can be done through a fuse over mount.

Applications destroying file-systems to the point of not being able to boot will not exist in the next generation file systems being made. Were is Windows equal feature.

Have you guys not been watching the game Solaris did start moving to ZFS if current design file systems work.

Even memory management in windows is getting very old school. Where is the prototype compression of swap to avoid sending unrequited stuff to swap files.

Linux has two different memory compression methods under development. Ksm that overlaps duplicate blocks in memory. compcache that compresses data going to swap.

I am sorry to say the lower levels of Linux are getting the attention they need. Lower levels of windows is basically bit rotting. There is plenty of room if Reactos could catch up to completely beat windows while being a clone.

Basically we have two OS's getting half right. Linux low level core is getting the correct alterations.

Windows upper level has had the right things for a long time. Now they are overdoing it in places so screwing up user experience.

Linux kernel is also entering a very interesting state.

There is not a unlimited number of features you can add to a OS kernel before features will start becoming just another duplicate. It has taken all this time but the Linux kernel has entered that state.

In the open source world its called a project entering a mature state. It is very rare because most projects don't last long enough to reach even entering it. Clear road marker was tracing and kernel mode debugger added to Linux that this had started happening.

The focus of Linux kernel is moving from adding new features to cleaning up the code base. Of course when a open source project hits mature it is basically bug less. Hitting full mature state is still years off.

Yes the projects in mature state you can count on 1 hand. Even in closed source they are extremely rare. There has never been a OS to reach entering mature state.

Part of the reason why I say lots of you are underestimating Linux. Developers are paid to add features or repair items in Linux. We could see quite a few displaced from the Linux kernel itself. Very experienced coders.

This is a interesting time. Its part a repeat of history. Get sloppy with your basic tech in time you will be replaced.

Linux is just be best contender at this stage to defeat Windows. Is there a chance it could be Reactos a Windows clone defeats windows yes. There is even a chance apple will return to dominance.

The dice will land soon enough. I am not going to be foolish like a lot here of trying betting my house on a set of dice that have not landed yet. MS long term victory over Linux is not assured. Nether is Linux long term victory over Windows.

Odds are with Linux based on the neglect the core of windows is taking.

Anonymous said...

The reason why Linux fails so hard at everything is clearly descriped by the Kernel developers. It's even included with it's source code.

http://www.kroah.com/log/linux/stable_api_nonsense.html

You think you want a stable kernel interface, but you really do not, and you don't even know it. What you want is a stable running driver, and you get that only if your driver is in the main kernel tree. You also get lots of other good benefits if your driver is in the main kernel tree, all of which has made Linux into such a strong, stable, and mature operating system which is the reason you are using it in the first place.

That's the reality no one wants to see, hear or know about - and thus, nothing will ever change.

Haos said...

Over 500 XP boxes here, with permanent power issues (we have to replace batteries every 4-6 months due to deterioration). Blackouts happening daily. Not a single FS originating failure since 2007. NTFS here outlives the very hdd it is installed on. If there would be any power/blackout related NTFS bug, we would be the first place to experience it. Sorry, not an issue apparently...

Anonymous said...

Real world users to freetards:

We don't give a fuck about file systems! We don't know what they are! We want applications and our fucking wifi to work!

I know you won't listen, but I had to try.

DamnRight said...

There is plenty of room if Reactos could catch up to completely beat windows while being a clone.

I know life is different in bizarro world, but not that different.

Anonymous said...

Problem is I have seen the dominance of Unix come and go. The dominance of Apple come and go.

Now this guy is 50 yrs old... now I will listen to his elderly wisdom!

He still thinks like a 10 yr old delusional kid, though

Anonymous said...

When did Apple dominate? Like 6 months with Apple ][, before it was ripped to shreds by Commodore and Atari at home and IBM at work?

Anonymous said...

Over 500 XP boxes here, with permanent power issues (we have to replace batteries every 4-6 months due to deterioration). Blackouts happening daily. Not a single FS originating failure since 2007.

Are the desktops supported by UPS?

Anonymous said...

"There is plenty of room if Reactos could catch up to completely beat windows while being a clone."

Aren't they still trying to clone Windows XP? They're not even done doing that and now they're going to be even further behind once Windows 7 rolls out. ReactOS is way too late to the party.

MaraDNS said...

Linux freetards piss me off.

They piss me off because they use lies to support their position; the simple truth is that any honest evaluation of Linux shows it's a joke compared to Windows XP on the desktop.

Since freetards can't use honesty to promote Linux, the resort to lies.

Lets look at some freetard marketing:

you buy a volume license and then buy the extended hotfix agreement through your volume license account. You also have to pay for the individual fixes on top of that.

Bzzt. Wrong answer. Windows XP will have security patches applied to it until 2014. The only version of Linux available today that will still be supported in 2014 is RHEL 5 and its knockoffs; note that RHEL 5 doesn't work with my hardware but Windows XP does.

Windows XP is still being sold on freshly manufactured hardware and will continue to be for at least months after 'all support ends.'

Bzzt. Security updates is still support. Notice the subtle freetard distorting of facts: No longer offering mainstream updates becomes "all support ends". Never mind that XP will be provided security patches far longer than the latest crop of Linux distributions are either abandoned or force users to upgrade the entire OS several times.

Linux has this notion that every six months you have nothing better to do with your time than to upgrade your OS and spend a week dealing with all of the bugs an upgrade entails.

Anonymous said...

酒店喝酒,禮服店,酒店小姐,制服店,便服店,鋼琴酒吧,兼差,酒店兼差,酒店打工,伴唱小姐,暑假打工,酒店上班,日式酒店,ktv酒店,酒店,酒店公關,酒店小姐,酒店兼差,酒店上班,酒店打工,禮服酒店,禮服店,酒店小姐,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,台北酒店,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,禮服店 ,酒店小姐,酒店經紀,酒店兼差,寒暑假打工,酒店小姐,經紀 彩色爆米花,經紀人 彩色爆米花,酒店傳播,酒店經紀 彩色爆米花,爆米花,童裝,童裝拍賣,童裝大盤,童裝寄賣,童裝批貨,酒店,酒店,童裝切貨,酒店,GAP童裝,酒店,酒店 ,禮服店 , 酒店小姐,酒店經紀,酒店兼差,寒暑假打工,招待所,酒店小姐,酒店兼差,寒暑假打工,酒店上班,暑假打工,酒店公關,酒店兼職,酒店經紀

Anonymous said...

you buy a volume license and then buy the extended hotfix agreement through your volume license account. You also have to pay for the individual fixes on top of that.

Yup, completely wrong. I don't know about individual (custom?) fixes, but hotfix are part of the normal update routine and are offered freely to anyone with a valid license (aka not pirated Windows).

I remember when Microsoft should have stopped making fixes and security updates for XP, but as of today I'm still receiving updates regularly way beyond the then proposed end of support. For something I paid about 90€ in 2005 it's really nice (student license, it's incredible how much software you can get if you are a student).

Windows XP is still being sold on freshly manufactured hardware and will continue to be for at least months after 'all support ends.'

Well, when something still works like a charm, why change it? ... Oh yeah, freetards like to change everything at least every 6 months. Nearly forgot that, sorry!

Anonymous said...

Hell, beyond all logic, Microsoft *still* supports Windows 2000, Internet Explorer, and Office 2000. C'mon, what 1999 era software does Linux support? Anything at all? Does it even compile anymore?

Anonymous said...

Meant to type "Internet Explorer 6"

binky said...

This is the best Linux Hater post ever!

Anonymous said...

Liking Microsoft because Windows is "better" is like liking Hitler for being "racially superior".

Anonymous said...

Freetards upgrade every 6 months because their AlternateReality(TM) makes them think this time things will work. More often than not, what was working before breaks in the new release though.

Liking Microsoft because Windows is "better" is like liking Hitler for being "racially superior".

You can use whatever the hell you want based on your shiny ideals. Just don't try to shove them down our throats and don't expect the world to do the same. I want my software to work, thank you very much.

Anonymous said...

Like all Freetards' projects Openmoko was a solution in search of a problem.

Anonymous said...

If Linux were a song:

Boom boom pow! Boom boom pow!, My Desktop Vista is 2008, but Linux you're two-thousand LATE...boom boom boom! boom boom boom!

Im an OS without conviction,
Im a OS who doesnt know,
How to sell. a contradiction
I come and go
I come and go
Kernel Kernel Kernel Kernel Chameleon, I come and go, I come and go o o.

I wanna grep u all over,...grep you again.....til the night closes in, til the night closes in..

PS-kill me with your best shot!
PS-kill me with your best shot!
PS-kill me with your best shot!,,
fire Away!

Whooah, it's only halfway there
ooo yeah, Livin on a prayer
Take my contribs, and we'll make it I swear, oo yeah, Livin on a prayer!

By order of the prophet,
We ban that Alsa sound,
Degenerate the faithful,
With that Pulse'in sound
The shareef dont like it
Rockin the casbah
Rock the casbah

And I know, baby, just how you feel.
You've got to roll with the sources list to get to what's real.
Oh can't you see me standing here,
Ive got my back against the distro machine, Aint I the worst that youve seen? Oh cant you see what I mean ? Might as well jump! jump !

We can break if we want to
We can leave your apps behind
'Cause your apps don't dance and if they don't dance
Well they're no apps of mine
I say, we can break if we want to
A place where they will never find
And we can act like we come from out of this world
Leave the real OS's behind
And we can dance

Stop playing with my conf
Finish what you start
When you make my spinning cube come down
If you want me let me know
Baby let it show
Linux dont you fool around
Just try to understand, Ive chmoded all I can, cause you got the best of me.
Borderline feels like I'm going to lose my mind. You just keep on pushing my sanity, over the borderline..

I was living for a dream, loving for a moment
Taking on the world, that was just my style
Now I look into Gnomes eye,
I can't see forever, the search is over,
XP, you were with me all the while

If you're lost you can grep--and you will find me
time after time
if you crash I will catch you--if it's not ext4, I'll be waiting
time after time

Now look at this Linux yo-yos thats not the way you do it
You play the Alsa on the mtv
This shit aint workin thats not the way you do it
Money for nothin , and chicks ain't free

Kyle said...

@ Anonymous, April 8, 2009 1:20 PM:

Liking Microsoft because Windows is "better" is like liking Hitler for being "racially superior".

All men are created equally. To claim the same of companies and their operating systems is absurd.

Also, Godwin's Law - you've heard of it, yes?

Oh, yeah, one more thing. The orders of magnitude of difference in your analogy is incredibly disrespectful and marginalizes exactly what happened in Germany.

The worst you can claim is that MS has a stronger hold your pocketbook than it should. I don't think that really compares to genocide or starting a war that ended up killing over 14 million servicemen.

Linuphiles continually accuse their opponents of FUD - I think this is possibly the worst FUD there could be. It's a good thing it's so fucking unbelievable that no sane person would give it any legitimacy.

Kharkhalash said...

Aren't they still trying to clone Windows XP? They're not even done doing that and now they're going to be even further behind once Windows 7 rolls out. ReactOS is way too late to the party.

They're trying to build a compatible implementation of the Win32 API, and implement the NT kernel's ability to support multiple concurrent subsystems.

Pretty well, Wine done right (instead of a compatibility layer above the main APU, the win32 API is the native API).

Except the win32 API is set for deprecation (see the discussion about Singularily at the end of page 6 of the previous LHB entry ("Das Boot").

It's a good idea, but it's too late. At the pace ReactOS development is going, by the time it's ready, chances are win32 will already be deprecated, which is why Mono should be getting a lot more attention and development time than it is.

For something I paid about 90€ in 2005 it's really nice (student license, it's incredible how much software you can get if you are a student).

Gotta love that student license! Visual Studio Pro 2005/2008, Windows Server 2008/2003, MS SQL Server 2008 developer/express, XNA Game Studio, CCR and DSS Toolkit 2008, Robotics Developer Studio 2008, Expression Studio 2, and Virtual PC all FREE FOR STUDENTS via the DreamSpark program!

Hell, beyond all logic, Microsoft *still* supports Windows 2000, Internet Explorer, and Office 2000. C'mon, what 1999 era software does Linux support? Anything at all? Does it even compile anymore?

The current line of Linux software is roughly as functional (barely) as everyone else's 1999-era software, does that count?

Problem is I have seen the dominance of Unix come and go

Clearly, you haven't. This has been discussed before, as well. Unix never fell, it still remains entrenched and dominant in the fields it was dominant in during the Unix-era.

DOS and Windows rose to dominance by creating a new market, which marginalized the Unix market.

NTFS vs Btrfs NTFS is basically out matched. Btrfs has snapshotting and compression. Btrfs has what is called writable snapshots so you have a stuff you can leave the damaged writes on the disk for latter inspection and keep on using the machine. Btrfs is a more suitable notebook file-system than NTFS is.

You realize that the whole point of Btrfs is that it uses B-Trees to index file system data, and that NTFS uses B+ Trees to index file system data, (as do ReiserFS, XFS and JFS2 (not to be confused with JFS1 which is the one that was ported to Linux, for reference) as well as Postgres and MySQL) right?

You also realize that Btrfs is developed by Oracle for that reason, right? Btrfs aims to be tailor made for Oracle Linux, whose sole purpose is to run Oracle RDBMS.

That's what is both most endearing and irritating about freetards, you love to play the "bitrot" card, but rarely ever do so where it's appropriate. You love to think that NFTS, for example, hasn't changed since it was introduced in Windows NT, possibly because it's just always called NTFS, the version numbers aren't made public, yet every iteration of Windows NT ships with an improved filsystem.

Not only that, but new features are added in a way that is transparent, every iteration of NTFS is both forward and backward compatible with previous and later versions of the FS, while features not supported in older OSes degrade gracefully.

You know, the average freetard isn't even aware that Windows has had file system journaling longer than Linux has? Or that NTFS has been able to do snapshots since version 3, via VSS (Volume Shadow copy Service)? Or that while Linux is boasting about FS encryption as though it's the latest and greatest thing, NTFS has had EFS for almost a decade?. Then there are the tards who insist that file system fragmentation is unique to vFat and ntfs.

Most people don't give a flying fuck about the file system, and that's why MS doesn't announce new features and functionality in NTFS with the same fanfare that the OSS world does with file systems. However, a minority of people do care, and so this information is readily available to them proving they know where to look for it (MSDN is a good start).

Time for a new TM, I think... How's about ThePastIsTheFuture(TM) or TheFutureIsThePast(TM)?

Problem here is compression and encryption can be done through a fuse over mount.

Fuse always seemed so very hackish to me. NTFS does this via functionality built into the file system, and services built into the OS, that's the traditional, sane way of doing it, and it only really causes problems when the FS needs to be portable (which is why, for example, as previously discussed, XFS on Linux doesn't have the same functionality of XFS for IRIX, for example, snapshots).

Applications destroying file-systems to the point of not being able to boot will not exist in the next generation file systems being made. Were is Windows equal feature.

You didn't quite go into how or why this will be the case, unless you're referring to snapshots, which has been handled by VSS fo a very, very long time now.

Have you guys not been watching the game Solaris did start moving to ZFS if current design file systems work.

ZFS is tailored for Solaris' specific market. Honestly, I don't see why you'd need stuff like zpools, 128-bit storage capacity, RAID-Z, or adaptive endianness on a desktop, laptop or workstation.

Even on low and middle end servers, ZFS seems kinda excessive.

ZFS is for the ultra high end server and storage sector, which non-coincidencially is Solaris' market.

You may not be aware of this, but Windows and Solaris really don't compete in the same markets anymore, the feature requirements are very, very different.

EverythingIsTheSame(TM)


Linux has two different memory compression methods under development.

See that's your problem. You're comparing things that aren't even in an alpha state, and knowing OSS's track record, won't be for years, to technology that is available today.

It's a stupid, completely moot comparison and you damn well know it. Either compare current, production-ready tech to current production-ready tech, or future tech to future tech.

Lower levels of windows is basically bit rotting.

And you of course, know this because you've infiltrated Microsoft's dev team and are privy to their internal code, right?

There is not a unlimited number of features you can add to a OS kernel before features will start becoming just another duplicate. It has taken all this time but the Linux kernel has entered that state.

Actually, yes, there is. The only limit is in imagination, that statement shows only that you are devoid of foresight and imagination.

Though, regardless, Microsoft is actually doing the opposite, they're stripping features OUT of the kernel (it started with the driver model in Vista), because they're working toward a full-microkernel design. Again, see the discussion on Singularity found at the bottom of page 6 in the previous LHB thread ("Das Boot").

There has never been a OS to reach entering mature state.

IfISayItEnoughItBecomesTrue(TM)

Mature is a relative term. And you freetards love to forget about the existence of those enterprise operating systems with 20+ years of development, testing and debugging done on them. Solaris, AIX, QNX and zOS, for example are all very much mature operating systems. Hell, most of thgese systems were considered mature before Linux was even conceived.

The trouble is, a "mature" project by the OSS definition is, well, short sighted and impossible. "Mature" in OSS means more or less "complete". It's a very, very short sighted way of thinking, because software always evolves, new tech always becomes available, new features are always thought of, and there's always room to expand functionality.

This again, shows only that you have a very limited imagination.

The trouble I guess is also that is FLOSS there's this notion that "Good Enough" is actually good enough. In the rest of the IT world, "Good Enough" is never good enough.

Linux is just be best contender at this stage to defeat Windows.

What's with the obsession with "defeating" Microsoft? Nobody not involved in the Linux Jihad looks at it this way.

Linux isn't even the best contender to take on Apple as the main non-competition.

If anything, Apple is the number one contender to provide MS with any competition on the desktop/home computer market, except it isn't in their best interests to do so, since it would involve entering the low-end and commodity markets, and it just doesn't suit them.

On the webserver end, you may not have noticed, but Windows is quickly bridging the gap with Linux. So Linux is clearly not even the best contender there, not unless people rally behind Novell which is the only entity who seems to care about interop, and really the only ones in a position to implement interop.

DamnRight said...

Wow, worst Godwin's Law Fail in recent years! Good catch! Wish I'd spotted it instead ...

Anonymous said...

LinuxMoviesComingSoon(TM):

-highlander-there can be only one distro. Watch distros battle each other and watch your gonads shrink

-X11-Files - I want believe(TM).
2 government freetards search for answers to stop the takeover of the planet by operating systems from Microsoft.

-"user space" - red stapler special edition - A freetard tired of using teh same old crap everyday grows his gonads and finally uses XP, discovering new daring life of video editing, gaming, and dj multitrack music creation.

-A distro apart - starring BIN Kdiesel. A man given KDE 4.x gets upset and starts a quest to search for KDE 3.5

-fast and furious - watch as Oiaohm and Kharkhalash battle it out in everything from fast threads, fast boot time, and OS file formats. Good action guaranteed! EXT4 special NTFS edition

-The Bill Gates of Hell - A Linux user converted to XP on his netbook, and opens the Bill Gates of Hell, resulting in 99% of the netbook population succumbing to the phenomenon

-When Kharry met Gsally - a fun comedy where strange things happen all the time, but nothing productive happens or makes sense.

-Kwall-E - Kwall-E 3.5 meets KEVE4.0 but KEVE4.0 mysteriously freezes and developments happen on yet another distro

-The Mista - A strange Mista creeps unknowingly unto a small linux-only town. Panic ensues when things like gaming, video editing and music creation are beign used by people in the Mista. Can the small towners protect their simple way of life from the clutches of future technology?

-The GnuWatchmen - the ongoing distro war forces a superhero to fake a Microsoft attack in order to unit the distros unto one LSB and DE standard. Watch as IBM drops Sun like a bad tender offer.

-No Vista for Old Men - APlawrence stars as the vindictive husband of a poor wife still using XP on a windows98 era computer. Watch as he chastises her daily for using windows, while extolling the virtues of the CLI bible. Can teh local sheriff save this damsel in distress?

-Stallman hears a Who! - Watch as Stallman keeps hearing "GNU GNU GNU!" seemingly out of nowhere.

-Kdagascar - Watch as 3 linux users leave the usual monotony of their former existence for a magical place called "Windows".

-GNUraline - A girl is told that Linux can do this and that, but she sees thru the wallpaper and icon BS and finds the true Linux monster underneath. Can she make it back to the windows realm before she is trapped and eaten by lintards?

Anonymous said...

ROFL!

Anonymous said...

Kyle,

You give it legitimacy by attempting to refute it. Anyway if you replace the word "freetard" with Jew it sounds like Nazi Germany all over again. Same mindset different target.

Anonymous said...

Ah, the old YouLoseBecauseISaySo(TM).

Anonymous said...

Hah, the overused ICantThinkOfAnythingToSaySoIllTrademarkSomeCamelCase(tm) SoIWillSoundIntelligent(tm) ButIAmReallyQuiteStupid(tm)

Anonymous said...

No one's saying that the entire freetard race should be systematically eradicated from existence (though it would be nice), just that sterilization should be required. Thus, we are much more like Malthus than Hitler. Unfortunately, this plan has a fatal flaw: freetards don't reproduce.

Anonymous said...

Anyway if you replace the word "freetard" with Jew

Nah you can't do that. Being a freetard is a choice one makes.

Anonymous said...

MoreMovies!

-Unfaithful - Mrs APLawrence is a housewife married to Mr AP Lawrence. Both share a computerized life. One windy day in Soho, when Mr APlawrence is busy trying to google the cure for AV360, she has a quick and dirty trist with the XP Poolboy. She notices the XP Poolboy has gonads that are MUCH MUCH bigger, and he does things Mr APLawrence could only DREAM of doing.
In a fit of jealous rage, Mr. APlawrence chains her to a Pentium III with 256megs of ram, while he smashes her beloved XP CD to bits.
Will there be justice? or will he relaize,.."you know, UNIX really is dead!"

-KDEnight of the living dead -
Strange Bizarro isometric trash icons litter the countryside as a few windows survivors accidentally walks into a LinuxWorld convention. Can the brave windows users barricade themselves from the onslaught of ignorance, lies, and false promises of things getting fixed?

Whatever Happened to Baby LSB? - Two sisters, one LSB standard, the other a freakshow live in a liveCD mansion. The freakshow hates the LSB standard and vice-versa. Will one of them win out in the end? Stay Tuned!

-Crapnophobia - a small town goes ballistic when a strange CLI based OS is unleashed unto a town used to windows, causing chaos and destruction! Get a "grep" on yourself man!

-Raiders of the lost sources list -a young Jones goes on the ultimate quest to find a real DVD video editing program that actually works. Watch as he bravely edits all his source lists, as mysterious packages install themselves unto his poor unknowing distro. Will he find what he is looking for? Will it break one of his other apps collected over the years?
Does it take a penitent man to get something right in Linux!
Or will the Linux holy grail end up being a crusty old piece of useless crap!

-Fast Boot Times at Ridgemont Gnuversity - MaraDNS stars as the cool good/bad guy, able to straddle his way between XP and Linux OS debates with no problem. When Lintards question his manhood, Mara DNS springs into action, and shows these fools who the boss is!

Anonymous said...

Google's cache now has the WHT compromised info.

Link

Shit!

Anonymous said...

I can't believe you fags still use a Window Manager running on top of MS-DOS. For this reason Windows will always be behind Linux no matter how pretty they try to make it look.

keren_sky said...

Priceless post

Anonymous said...

I can't believe you fags still use a Window Manager running on top of MS-DOS. For this reason Windows will always be behind Linux no matter how pretty they try to make it look.

Troll, or just plain stupid?

Besides, what is Desktop Linux than a Window Manager running off a third-rate knockoff of a 30+ year old operating system, with deep roots in archaic subsystems?

Anonymous said...

Don't believe me? Press SUPER + R then type CMD then press ENTER and see what comes up.

Anonymous said...

Wow, more evidence of freetards thinking Windows is still entrenched in 1998. God they're stupid.

Anonymous said...

Don't believe me? Press SUPER + R then type CMD then press ENTER and see what comes up.

An NT kernel command prompt. So? Microsoft designed the command prompt to be familiar to people who were used to their older product, DOS. Yes it does run DOS software, but only under a compatibility layer much like your much-touted WINE is.

Anonymous said...

The April 8, 2009 8:14 PM anonymous almost had it, but he fell into the age old trap of not realizing that Windows NT actually contains two command interpreters. First is the all-familiar cmd.exe, which is, as we all know, a 32 bit executable environment. Second is the obscure command.exe, which does indeed identify itself as DOS, but, as your CPU meter will suggest, this is a virtualized 16 bit instance. You know, like OS/2 'n stuff, and we're not about to suggest OS/2 is DOS-based, are we?

Whelkman said...
This comment has been removed by the author.
Anonymous said...

Err, command.com. Brownie points to anyone who remembers the difference between .com and .exe.

Anonymous said...

Never mind. Google makes it too easy.

Anonymous said...

.com is a TLD while .exe is an executable.

thepld said...

"
You give it legitimacy by attempting to refute it. Anyway if you replace the word "freetard" with Jew it sounds like Nazi Germany all over again. Same mindset different target."

So freetards are a race? What's their genetic structure? Under that logic, launching a rhetorical assault on an opposing political party would be racism.

Anonymous said...

So I tried KDE 4.2.2.66669999999 today. Yes, I really tried it on an Ubuntu install. What utter, ridiculous, insane shit. It's slow as molasses, confusing, and ugly. What the fuck were they thinking?

Jack O'Sullivan said...

One Hundred and Sixty One!!!!

Jack O'Sullivan
Bedroom Furniture Sets
Men's Dress Shoes

Anonymous said...

" .com is a TLD while .exe is an executable."

Congratulations on cluelessness! .com is the file extension for the original executable used by MS-DOS systems.

Anonymous said...

And there things lay, with COM files being "raw memory images" and EXE files being "structured"

http://blogs.msdn.com/oldnewthing/archive/2008/03/24/8332730.aspx

Anonymous said...

You'reStillUsingDos(TM)

Anonymous said...

So, if I'm StillUsingDOS(TM), why do I need a DOS emulator (eg. DOSBox) to run DOS applications?

Freetards and theyr bizarro world.

Anonymous said...

Ups "their".

PlasticLawnDaemon said...
This comment has been removed by the author.
Anonymous said...

Though, regardless, Microsoft is actually doing the opposite, they're stripping features OUT of the kernel (it started with the driver model in Vista), because they're working toward a full-microkernel design.

More evidence that Kharkhalash doesn't really know what he's talking about...

NT has become a huge mess, splitting it in more components is the only way for MS to avoid collapsing under the weight of the maintenance required for it (refer to the Vista fiasco). This has absolutely nothing to do with microkernel design, which NT doesn't have and probably won't adopt until it dies, because so far no one has found a way to make microkernel design as fast as the monolithic one, and the benefits of microkernel design are overvalued, especially now that virtualization is ubiquitous. It's easy to see how overvalued microkernel design is on commodity hardware: the number one cause of kernel crashes is the GPU driver, and a GPU hang will hang the system, therefore the OS, microkernel or otherwise.

Anonymous said...

not to mention that FreeRunner had support for copy/paste (at least when u installed gnome on it) while the iPhone still does not!!

Anonymous said...

The Ribbon is so "ugly" that freetards are trying to copy the concept. Would you say it? they're doing it wrong

Anonymous said...

Hahaha, from the comments of that ribbon blog:

I really hate the ribbon as used in MS Office, but I'd love to see it used in KDE applications.

This is so typical. As long as it's not the original from MS but some half-assed freetard copy, it's okay. Please, someone free us from freetards.

Anonymous said...

FastThreadsandFuriousFS - Bin Kdiesel returns as someome killed off kde 3.5 from his favorite distro. He teams up with his rival, Kbun2 and together they flood the slashdot and digg forums against those who put appearance before stability.

Lost in Userspace - watch as a team of linux devs from earth working on a project get mired in Linux red tape, and end up being moved to the unknown void known as userspace. Will their apps find their way home again?

GNUminator - The year is 1984. Devo, Culture Club and breakdancing were hot. Richard Stallman also started the GNUproject. Enter GNU-T800, a Linux cyborg sent back in time from 2250, when GNUnet was still trying but failing again and again to create a useable desktop linux. The GNU-T800's mission is to go back in time and destroy Bill gates and Steve Jobs from creating Windows and OSX. Will it succeed? Or will humans prevail and get the OS they really want?

GNUminator II - GNU-T1000, a new improved linux cyborg with better looking wallpaper and icons appears. It's job is to break into MSdyne systems and retrieve MS code and hopefully move it into userspace changing the balance of time. MaraDNS stars as John Connor, leader rebel of the HumansWithGoodCommonSense(TM).
John utters the infamous mantra: "There is no fate but what we make" and enrages the GNU-T1000 which BTW has an inferiority complex. (GNU-T1000 shouting) "I am NOT an animal! I am a Desktop OS! Dammit!"

There's Something About Vista - Watch as Ben Stallman (brother of Richard Stallman) is infatuated and falls in love with Vista. His evil brother Richard tries to spoil his love interest by spreading false lies and deceit on slashdot, digg and various other blogs.

Knemeny of the State - Colonel Panic stars as a windows user posing as a Linux user. While trolling the various blog sites, he secretly uses windows software of which 99% is precracked. This causes an uproar in the GNU community especially when they can no longer use Photoshop CS4 after a small .dll change in Windows 7. Bastard Windows infected DRM! I can't use my ware$!

Trading Userspaces - in a strange twist of fate, a windows user is thrown into a linux environment, while a linux user is thrown into windows one. Watch as the windows user's life degrades and he is forced to dress up in santa suits and sneak into windows conventions for food. While the linux user gets acquainted and becomes really really good at working with windows and never intends to return back to linux. Who will win the 1 dollar bet that a free OS doesn't mean jack if there are no apps worth running on it?

oiaohm said...

Kharkhalash not knowing the tech "VSS operates at the block level of the file system." That is basically LVM snapshots Linux also has had that for ages.

VSS what it is
"The end result is similar to a versioning file system, allowing any file to be retrieved as it existed at the time any of the snapshots was made. Unlike a true versioning file system, however, users cannot trigger the creation of new versions of an individual file, only the entire volume."

Nilfs is a true versioning filesystem. VSS does not provide that. Nilfs is snapshots per file compared to snapshots per filesystem. Nilfs is main line as of 2.6.30.

VSS cannot effectively cross drives with it snapshots. Btrfs can. Btrfs can fuse read only snapshot with write able area to grant a working file-system many times over even have many different clones mounted at the same time. Great for virtualisation. VSS is a incapable bit of crap. Simple fact VSS does not compared to what Btrfs and Nilfs are.

This is the problem with MS supporters MS has lots of the old styles of operations that need replacing. VSS style is the 20 year old method.

This is the problem the features have been for ages. MS had got use to having top and is falling apart. Linux features are passing what MS has.

Linux does support ecryptfs what is basically a stacked file system that allows encrypted files on anything that is a posix compatible file-system. So its not really need as unique file-system feature. Encryption advantage is basically nothing in NTFS.

Vista to Windows 7 no new features in NTFS.

Btrfs yes was started by Oracle but it is for include in the main kernel to be used by everyone. Who is better to design a safe fast file-system than a database maker since those are exactly what a database needs.

The two memory compression techs for. One is delayed to 2.6.31 due to benchmarking need to be done other than that is fully ready to go. So about 9 months for it to appear. That is right just when Windows 7 gets out the door. The other memory method is almost ready to go. It is in what is classed as beta testing.

Kharkhalash there are a limited number of features you can build into a kernel. Idea of unlimited is wrong.

MS trying to cut stuff out of kernel space is because they have hit there limit.

Past particular points it becomes duplication or stupidity. Ie embedding a office suite in kernel would be stupidity.

Its not limited creativity. There is just a limit. There is not a unlimited things a OS kernel needs and should to do. RTOS support is the last major fronter left for the Linux kernel.

Limit is also defined by existing hardware. Problem is Linux development speed is more than keeping up with the changes in existing hardware. Number of programmers working on the kernel is about double what is required to keep up with and include new hardware features.

Even Linux is pushing items out of kernel space to userspace because they make more common logic in userspace. Like syscall printf equal send it back to libc in userspace reason more secure.

If you don't care how secure a OS is you have unlimited.

MS OS is basically collapsing under its own weight. How to I know the lower down parts are bit rotting areas that should have been repaired are still stuffed Kharkhalash. Stuff like VSS are a good example. Feature abilities should have been expanded long ago still not being fixed. Then features like transaction mode on NTFS that should have been picked up by a VSS repair being added. This is so normal of bitrot actions. Normally shows lack of a developer with experience to fix the true problem. X11 suffered from this big time. Alterations on NT these days are very much like the alterations on X11 that got it into the huge mess to clean up.

Kharhalash except NTFS is a file-system who features are getting old and need a major rework. Its backwards compatibility limit what features can be added. At least Linux developers know when the time to cut there losses has come. EXT line of file-systems have the same forwards and backwards compatibility feature as NTFS. Past a particular point it becomes more trouble than what can be dreamed.

Yes basically you can only tack features onto a filesystem for so far before it becomes a trouble making mess.

NTFS is basically end of life like EXT. WinFS should be hope. Instead of truly redesigning the filesystem they tacked WinFS on top of over patched file-system design and wondered why it come down like a house of cards.

Anonymous said...

The Linux

Morpheus: I imagine that right now you're feeling a bit like Alice. Tumbling down the rabbit hole?
Neo: You could say that.
Morpheus: I can see it in your eyes. You have the look of a man who accepts what he sees because he's expecting to wake up. Ironically, this is not far from the truth. Do you believe in compiling your own OS, Neo?
Neo: No.
Morpheus: Why not?
Neo: 'Cause I don't like the idea that I should be wasting time with this shit.
Morpheus: I know exactly what you mean. Let me tell you why you're here. You're here because you know something. What you know, you can't explain. But you feel it. You felt it your entire life. That there's something wrong with the world. You don't know what it is, but it's there. Like a splinter in your mind -- driving you mad. It is this feeling that has brought you to me. Do you know what I'm talking about?
Neo: Desktop Linux?
Morpheus: Do you want to know what it is?
(Neo nods his head.)
Morpheus: Linux is everywhere, it is all around us. Even now, in this very room. You can see it when you look out your window, or when you turn on your television. You can feel it when you go to work, or when go to church or when you pay your taxes. It is the world that has been pulled over your eyes to blind you from the truth.
Neo: What truth?
Morpheus: That you are a slave, Neo. Like every Linux user, you were born into bondage, born inside a prison that you cannot smell, taste, or touch. A prison for your mind. (long pause, sighs) Unfortunately, no one can be told what Linux is. You have to see it for yourself. This is your last chance. After this, there is no turning back.
(In his left hand, Morpheus shows a blue pill.)
Morpheus: You take the blue pill and the story ends. You wake in your bed and believe whatever you want to believe. (a red pill is shown in his other hand) You take the red pill and you stay in Wonderland and I show you how deep the rabbit-hole goes. (Long pause; Neo begins to reach for the red pill) Remember -- all I am offering is the truth, nothing more.

Neo kicks Morpheus in the groin, takes the blue pill and swallows it with a glass of water. He wakes up, formats his drive and installs Vista. After 20 minutes he's finished, takes his jacket and goes out for a beer and a hooker.

Anonymous said...

Do you believe in compiling your own OS, Neo?
Neo: No.
Morpheus: Why not?
Neo: 'Cause I don't like the idea that I should be wasting time with this shit.
Morpheus: I know exactly what you mean.


ROFL hahahaa

Anonymous said...

@Kharkhalash
You realize that the whole point of Btrfs is that it uses B-Trees to index file system data, and that NTFS uses B+ Trees to index file system data, (as do ReiserFS, XFS and JFS2 (not to be confused with JFS1 which is the one that was ported to Linux, for reference)
Actually JFS2 was the one ported to Linux, and it is called simply JFS. JFS1 wasn't never ported. Get your info right!

Anonymous said...

Gotta love this one: in this bug report people brag about a wallpaper made on a Mac with Adobe Photoshop(r) in ubuntu-wallpapers.

Proudly brough to you from the Ubuntu-it community, first-class example of freetardery

Anonymous said...

@yOSHi314
"grep’ing through your address book using extended regular expressions "

actually i would love to have this for searching sms archive. but a bit simplified.


Symbian OS phones have got this feature long time ago, it's really nice to search through your sms archive.

Anonymous said...

So I gave KDE 4.2.2 a second chance (because unlike true LIntards I'm open-minded). Playing around with it further, it's not as bad as I first thought. Once I disabled some of the bling it's actually usable. Still feels unpolished though (I even get screen tearing in the menus with NO desktop effects), but I like some of the features. Maybe KDE can pull things together (this is on Ubuntu, but I'm thinking, a good, polished, optimized future version of KDE on top of BSD would be something to give me pause).

Anonymous said...

lol or KDE4 in top of windows

Anonymous said...

NT has become a huge mess, splitting it in more components is the only way for MS to avoid collapsing under the weight of the maintenance required for it (refer to the Vista fiasco).
if NT is a huge mess, how can linux be defined with its completely unstructured (as in the engineering definition of "structuring", ie logical separation) subsystems and ever-changing undependable internal APIs?

because so far no one has found a way to make microkernel design as fast as the monolithic one
none that people used to mainstream operating systems (and only those) know of, that is - CS is full of whitepapers on how to improve efficiency of research kernels ( typically microkernels because of their elegance and orthogonality) from lightweight RPC to address space sharing (with boundary protection enforced by different means from traditional address space separation, eg VM'ing, language contraints...)
yet, none of these systems is in wide use - mainly because they're not practical for the needs of general purpose computing as of today

and the benefits of microkernel design are overvalued, especially now that virtualization is ubiquitous. now that virtualization is ubiquitous (maybe ubiquitous where you live, i don't see so much virtualization around here myself - again, i have a pentium 4 and a sempron le, maybe that's because of this)

ANYWAY: i'm not particularly fond of pure microkernels myself, and more inclined in favour of hybrid kernel based os's - yeah, the process - service round trips, etc etc - but mostly my own convinction that, in general,"the kernel IS the OS", that is, the kernel is made to a) multiplex resources across application programs, and b) provide enough functionality to allow application programs to run without *depending* on other processes functionality that would compete to the kernel itself (in turn, services implementing system wide functionality shall be used for non vital stuff -eg PM policies, or plug and play, filesystem search indexing... - and disable-able as possible)
so, i'd consider a kernel without a filesystem layer, suited for the need of some particular application, rather than those of a desktop system neding *a general purpose OS*...

BUT, if nothing else, microkernels are a very good starting point to build flexible hybrid kernel upon - unless the kernel is built by retarded coders without a minimum of reasearch and design, of course - because a microkernel system already has an (enforced) logical (AND physical) separation between core components and their respective data structures - its likely already well *structured*, and the initial structured design if flexible enough could be leveraged at the API level, while reorganizing subsystem at the implementation level , this would be a sign of a sensible and elegant design (cont)

This has absolutely nothing to do with microkernel design, which NT doesn't have and probably won't adopt until it dies,
the first NT was based on a pure microkernel ( actually Mach 3 IIRC) as was OS/2
the NT kernel (better, Executive and HAL) expanded after the first relases, to include even the graphics and gui subsystem - yet, as was pompusly stressed by a microsoft whitepaper , moving subsystems to kernel space didnt change anything for applications, API and ABI wise
compare that to linux, on which the audio stack couldnt switch to a different sound server without applications needing to be adapted as well...
mroeover , an OS based on a monolithic kernel, needing a daemon to do something even microkernels do natively (fast message passing across processes registered to an interface - "publish - subscribe" "push" model IPC) despite the fact that nearly every application (or at least potentially a whole category of them) uses or will use that functiomality since the related API is a "standard" and is enforced to a certain extent (DBUS, anyone?), sounds like a pretty lame design

moreover, a simple criterion i've been taught that applies to a desktop system ( or in general one that could be disconnected from the network or be connected nor make use of it), is to design things while thinking the network stacks doesnt even exist - as don't sockets
surprise, if i proceed this way nice things can happen, as one is more induced to think oout of the box and invent alternative techniques to cope with interactions between processes on a single machine ( which, big surprise, is exactly equivalent to the "desktop" case) - you see, you could even come out with something like LRPC (or, on NT's implementation local procedure call ) and extensively optimize local - only services (such as the GUI) for it, *before* implementing the network stack...
but sadly foss developers are all entrenched in this unix way of doing things, and all for putting things into the kernel which are unstable or none uses them, yet against it when it would make sense

heck, besides being born from the gradual replacement of a microkernel (minix) code (!) linux even had another "microkernel period" (in the form of ukLinux ) - just to be discarded short after, with the excuse of being inefficient (of course, a plain-unix like monolithic kernel stripped of very much everything in order to resemble a microkernel, not only is a fake microkernel, but also a very crappy one that even old hog Mach1 could beat - something with the disadvantages of both micro- and monolithic kernels, with none of the pro's) and giving another argument to the "microkernels are crap" cause, as the real reason...

It's easy to see how overvalued microkernel design is on commodity hardware: the number one cause of kernel crashes is the GPU driver, and a GPU hang will hang the system, therefore the OS, microkernel or otherwise.
in theory, if the gpu driver is implemented as a (trusted) userspace process, in the case of a crash restarting the process will re-init the graphics card (then applications will need to redraw after the restart of the graphics server)
the poblem is more with the loss of the current state of the graphics device (which may be bad if important computation -gpgpu- was going on, or simply some important gui application ) stat will need to be reloaded somehow

Anonymous said...

Here's a good example of a Freetard.

http://blogcritics.org/archives/2009/03/03/175049.php

This Richard Chapman guy is hilarious.

wbkang said...

It's easy to see how overvalued microkernel design is on commodity hardware: the number one cause of kernel crashes is the GPU driver, and a GPU hang will hang the system, therefore the OS, microkernel or otherwise.

You see, in Vista, it doesn't crash the entire system because the video driver is in userspace. They just automatically restart the driver.

Anonymous said...

wbkang, we can't hold not knowing that against them. After all, the last time they used Windows was 98 first edition. Just look the "DOS" comments in this thread. It's understandable, though. After twenty years of X11 and GNU userspace, one gets used to static technology and assumes such applies everywhere.

Anonymous said...

@April 8, 2009 8:47 PM

In a .com file code and data could not exceed 64kb.

wbkang said...

It's been like 2+ years since Vista RTM was out though... lol

Anonymous said...

Re "The Linux"

LOL that was great! ROFL!! I was reliving that scene in my mind too THANK YOU! lol ")

Re: "Wallpaper in Ubuntu made by Macintosh"

WOW that IS a big bug isn't it? Nothing like an official Ubuntu wallpaper with mac jpg header info LOL

Anonymous said...

Linux Hater's Nemesis?

http://senseitim.blogspot.com/2009/04/kill-conficker-switch-to-linux.html

Anonymous said...

He's doing the right thing. Only by encouraging people to try Linux they will see for themselves how much Linux sucks.

So, go ahead, do what the guy says, try Ubuntu now and use it exclusively for a while. You'll come to appreciate Windows like never before, but don't listen to me. Just judge by yourselves.

Anonymous said...

You only get conficker if you are using a pirated version of windows(most have rootkits preinstalled btw)... and shut off UAC... and are not using any sort of firewall(because you know better???)...

I mean in all the conficker hype leading up to 4/1 the fosstard'o'sphere somehow left all that out???

OIC, that was because the fosstard'o'sphere was trying to harm microsoft, not actually protect people. Ahhh. I get it now...

You fosstard'o'sphere fosstards realize your last tiny bit of credibility in MSM was spent on that right? The "windows endangers the power grid" scare media you tried to make happen yesterday received precisely zero traction in MSM.

You guys blew it, that was your last remaing area of credibility...

Kharkhalash said...

Actually JFS2 was the one ported to Linux, and it is called simply JFS. JFS1 wasn't never ported. Get your info right!

You're right, the older, more mature one was the one that wasn't ported, my mistake.

-------

NT has become a huge mess, splitting it in more components is the only way for MS to avoid collapsing under the weight of the maintenance
required for it


Funny, that was the rationale for conceiving the first generation of microkernels.

the whole premise of the microkernel is to separate all non-essential services from the kernel (in the strictest terms, according to Liedtke's Minimality Principle, everything other than:
- address space mechanics
- CPU allocation management (Threads, scheduler)
- Inter Process Communication (IPC) mechanisms

Doesn't belong in-kernel, with very few exceptions (mostly for the sake of convenience)

One of the premises of the microkernel is that a smaller base kernel, combined with smaller compartmentalized servers is easier to manage than a gigantic monolithic kernel.

Consider, for the sake of argument, Minix3's full microkernel, with a paltry 3,800 lines of code, to Linux's 6+ million, which is easier to manage?

This has absolutely nothing to do with microkernel design,

Pulling the driver model outside of kernel-space, and into userspace is a step in the direction of a microkernel (device drivers are neither address space related, schedulers, nor IPC mechanisms, and thusly can be moved into userspace).

which NT doesn't have

Learn to read. Nobody claimed otherwise. NT is of a hybrid design (it was supposed to be a microkernel, but those weren't feasible 15 years ago). Meaning it's designed like a microkernel, except the components that would otherwise run in userspace,
run in kernel space. If I'm not mistaken, it was initially built on a microkernel.

and probably won't adopt until it dies,

That's the speculation. NT is the hybrid kernel. Singularity is the microkernel. Apparently, you missed the part about referring to the discussion on Singularity (and potential transitions from NT to Singularity) found on the sixth (6th) page of the previous article ("Das Boot") near the bottom.

because so far no one has found a way to make microkernel design as fast as the monolithic one,

As was pointed out, there is a plethors of methods to improve performance on microkernels. The original argument against microkernals was Mach's shitty performance. Liedtke proved that that was due to poor design and implementation, and improved performance by an order of magnitude in L4.

also, see QNX, or LynxOS, or integrity, or any other microkernel-based embedded RTOS. In spite of the inherent overhead of IPC in a microkernels, QNX still manages to deliver absolute-worst-case-scenario maximum latency an order of magnitude lower than low latency Linux.

Other's theorize that the performance penalties will be made more and more negligible with the advances in modern hardware, ultimately what it comes don't to is that performance is the last consideration in system design (meaning you tweal performance last), not the first. The benefits of a microkernel are too great to ignore, eventually the performance hit will be negligible enough to be worth the tradeoff in general purpose systems, jsut not yet (which is why NT is the production kernel, and Singularity is the research kernel).

and the benefits of microkernel design are overvalued

The benefits of a microketnel are simply:

robustness though simplicity: smaller kernel, smaller components, fewer lines of code, in theory, less room for bugs to crop up.

ease of management: You said so yourself, componentizing the system services leads to more easily managed parts. This was the rationale for the microkernel in the first place.

Stability, reliability and availability: Because the bulk of what would be part of a monokernel now resides in userspace, is compartmentalized and isolated, a crashed server will not take down the kernel, nor any other server not expressely dependant on the one which crashed. The crashed server can be stopped and restarted live, without shutting down or affecting the rest of the system. "Properly" designed applications can even be put on hold while the server they rely on stops and restarts.

Versatility: servers can be enabled and disabled on the fly. This is more conveniant than recompiling a kernel to reach the same end.

Security: implied by versatility and robustness, reduced attack surface.

if anything, the benefits of a microkernel are vastly undervalued.

especially now that virtualization is ubiquitous

Except that a hypervisor isn't much more than a microkernel which doesn't make any claim to the minimality implied by a microkernel, and is specialized to host virtual machines.

A microkernel is full capable of hosting operating systems (thus acting as a hypervisor, both Mach and L4 have been used for this purpose). but is specialized in hosting core system components which would otherwise live in kernel-space.

It's kinda silly to suggest that one supplants the other.Virtualization does not address the issues microkernels do, nor does it bring the benefits of microkernels.

It's easy to see how overvalued microkernel design is on commodity hardware: blah blah blah

The whole point of a microkernel is thaat crashed servers can do so, be stoped, and restarted, without affecting the kernel itself, nor even other servers which don't expressly depend on it.

The graphics driver dies, it gets stopped and restarted, the card is reinitialized, and all is fine and dandy. Applications need to redraw, and you stand to lose unsaved work that was being worked on at the moment where the driver died, but the point is the system remains running.

For the kernel to go down, either the microkernel itself, or a component is kernel-space needs to crash.

It's easy to see how someone who doesn't quite understand what a microkernel is, how it works, or what benefits it offers would undervalue the benefits.

Anonymous said...

Breaking news.

Confickerworkinggroup.org hosted on linux and Apaxche, and maintained by an elite group pf security experts, has been hacked.


LOLOLOLOLOL

Anonymous said...

You see, in Vista, it doesn't crash the entire system because the video driver is in userspace. They just automatically restart the driver.

If the video driver goes crazy and starts writing garbage to the card registers and the card hangs the bus, no kernel can save the system, this is a limitation of the hardware. The restart you see in Windows is probably due to a hang in the userspace part of the driver that communicates with the kernel-level part (roughly the equivalent of an X crash). If the kernel interface crashes in Windows, the system is dead.

Consider, for the sake of argument, Minix3's full microkernel, with a paltry 3,800 lines of code, to Linux's 6+ million, which is easier to manage?

Linux includes the drivers for almost every piece of hardware on earth.

The speculation part is that you, Kharkhalash, expect MS to turn NT into a microkernel OS, based on a research project. So far, MS has proved that they can't even handle a modest OS upgrade from XP to Vista in 8 years, what makes you think this leap in design is feasible at all?

In spite of the inherent overhead of IPC in a microkernels, QNX still manages to deliver absolute-worst-case-scenario maximum latency an order of magnitude lower than low latency Linux.

Have you tried to set up a server on QNX lately? Linux latency is getting improved all the time, and if I remember correctly, linux currently sets the bar for fast OS design (context switches, process creation etc).

It's easy to see how someone who doesn't quite understand what a microkernel is, how it works, or what benefits it offers would undervalue the benefits.

Yeah, right. There are benefits to a microkernel, however monolithic kernels have proved that they can be very reliable too, and virtualization gives them many of the advantages of microkernel design without any of the disadvantages.

Anonymous said...

roughly the equivalent of an X crash

Except when x crashes you lose all your work. The two or thee times I have had a video card reset, the screen just goes blank for a moment, then redraws its self, and you get a pop up warning you that your video card reset....

Now on the computer I am using at the moment, In the year I have owned it I have not encountered even a single system crash.

Also, I am not sure what you are talking about with regards to Microsoft being unable to upgrade XP to Vista.... but with a few exceptions I have had no issues in my Vista migration. Are you sure your are not just suffering from the slashtard delusion that Vista was a failure???

Anonymous said...

Linux includes half-baked barely-working drivers for almost every piece of hardware on earth.

There, FTFY

Anonymous said...

@April 9, 2009 1:03 PM

Don't let real world usage get in the way of uninformed doomsday scenarios.

Anonymous said...

And "driver" means that Linux can interpret low level instructions and bytecode. The frameworks required to actually utilize the devices do not exist.

Anonymous said...

Confickerworkinggroup.org was last seen up April 4th according to Google cache. WOW it's been down ever since. Hmmm hacked? :)

Anonymous said...

"
Don't let real world usage get in the way of uninformed doomsday scenarios."

Yeah , because X never crashes, ever, even when using those lovely high quality drivers from Intel or that great stable Compiz code.

Anonymous said...

April 9, 2009 1:30 PM

Netcraft has it up as of yesterday.

I am just wondering why the fosstard'o'sphere is not covering it? Computerworld and ZDNet both have multiple stories on their front pages right now about the risk that Conficker poses for windows users, but they have zero mentions that the worm is also apparently able to take down apache servers hosted on linux and maintained by elite security experts.

This is like the MLB story all over again. The fosstard'o'sphere was all crooning on monday that MLB had switched to flash, but then failed to mention that MLB's HD flash service does not work. Even when they mentioned they did not mention it directly, just as "MLB deletes blog posts". They didn't point out that the MLB fans were upset thtat they had a working system a week ago.

And again, my problem is not with Linux, it is with the fosstard'o'sphere. I want reliable tech news, and I currently don't have that, and evidenced by all the fosstards who still think Vista was a failure.

«Oldest ‹Older   1 – 200 of 1346   Newer› Newest»