Slashdot Log In
The File-System Fallout of the Reiser Verdict
Posted by
timothy
on Tue Apr 29, 2008 12:17 PM
from the work-from-home dept.
from the work-from-home dept.
perlow writes "Yesterday, the Open Source community took an emotional hit when veteran Linux programmer Hans Reiser was convicted of first degree murder in the suspicious disappearing of his wife, Nina. While I won't go into the details of the case, as this has been covered extensively in the press, I would like to talk a little bit about how this verdict will impact the technology in play for file system dominance in our favorite Open Source operating system, Linux."
Related Stories
[+]
Your Rights Online: Hans Reiser and the "Geek Defense" Strategy 738 comments
lseltzer alerts us to a story in the Washington Post on the defense strategy in the Hans Reiser murder trial. "In the courtroom where Hans Reiser is on trial for murder, [the evidence] might appear to indicate guilty knowledge. But his attorneys cast it as evidence of an innocence peculiar to Hans, a computer programmer so immersed in the folds of his own intellect that he had no idea how complicit he was making himself appear. 'Being too intelligent can be a sort of curse,' defense counsel William Du Bois said. 'All this weird conduct can be explained by him, but he's the only one who can do it. People who are commonly known as computer geeks are so into the field.'"
[+]
Your Rights Online: Hans Reiser Guilty of First Degree Murder 1395 comments
Anonymous Meoward writes "Today Hans Reiser was found guilty of first degree murder in Oakland, California. Quoting Wired: 'In a murder case with no body, no crime scene, no reliable eyewitness and virtually no physical evidence, the prosecution began the trial last November with a daunting task ahead... The turning point in the trial came when Reiser took the stand in his own defense March 3.' Whether he really did it or not, Hans basically just didn't know when to shut up."
This discussion has been archived.
No new comments can be posted.
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Full
Abbreviated
Hidden
Loading... please wait.
Dear Windows Users... (Score:3, Funny)
Please stop using NTFS.
Re: (Score:3, Funny)
Please stop using NTFS.
But at least Bill Gates hasn't killed his wife.
Re: (Score:3, Funny)
Re:Dear Windows Users... (Score:5, Funny)
Parent
Re:Dear Windows Users... (Score:5, Funny)
I'm sure he's not the only one.
Baaazing! Thankyou thankyou I'll be here all night.
Parent
Re:Dear Windows Users... (Score:4, Interesting)
Either way, when MS-DOS first came out, I don't recall anybody claiming that it wasn't an operating system.
Parent
Re: (Score:3, Interesting)
To most people, the "operating system" is the user interface. That's why some vendors get away with calling a collection of AJAX applications a "web OS". But we're not talking about what people perceive an OS to be, we're talking about what an OS does.
The question of whether some software is an OS is not a religious issue. You take the s [wikipedia.org]
Re:Dear Windows Users... (Score:4, Funny)
Parent
Re: (Score:3, Funny)
My life is over.
Re:Dear Windows Users... (Score:5, Funny)
That's because she'd kick his ass if he tried ;) Seriously, have you seen Bill Gates? ;)
Parent
Offtopic? WTF? (Score:5, Insightful)
For the learning-impaired at "news for nerds", the parent's point was that the only thing that should matter to you or me about Hans Reiser is the file system he designed and the programs he wrote, and then only if you use the damned things.
Did you throw away your Naked Gun [wikipedia.org] movies because OJ Simpeon killed those people?
The tabloid mentality that seems to have overtaken the entire world makes me sick. The file system's designer's personal life is none of my (or your) business.
On a lighter note, the parent was wrong. Bill Gates doesn't strangle puppies, he shoots them! [wikia.com] And he doesn't stomp kittens' heads, he huffs them. [uncyclopedia.org]
Parent
Re: (Score:3, Insightful)
The Naked Gun doesn't require maintenance. Code does, especially when it has several outstanding issues (show of hands, who here likes the quality of reiserfsck?) and how fast the kernel iterates.
Namesys is dead in the water, and reiserfs will go with it.
Re:Offtopic? WTF? (Score:5, Insightful)
Reiserfs4 is pretty much dead... Unless somebody picks up the code and runs with it which could happen but I really wouldn't bet the server farm on that.
The simple truth is "Big Deal"
You currently have a choice between.
EXT3 which isn't the fastest but it does work.
JFS which is pretty feature rich and stable.
and XFS which is also feature rich and stable.
EXT 4 is coming soon and ZFS may be GPLd at some point.
The lose of ReiserFS to bit rot isn't the end of the world or really that big of a deal.
Now the lose of Hans Reiser. Well if he didn't do it then it is very sad. If he did do it then it is also very sad. That is just on the human level and nothing to do with FOSS.
Parent
Re: (Score:3, Insightful)
What are you saying, there? The courts got it wrong?
A different kind of file system? (Score:5, Funny)
Re:A different kind of file system? (Score:5, Funny)
mkdir("tmp");
chroot("tmp");
chdir("..");
Parent
But... (Score:5, Funny)
Parent
Re: (Score:3, Insightful)
I'm hoping... (Score:5, Interesting)
Re: (Score:3, Interesting)
Given that the code is Free, it's always been under the control of anyone who wanted it. Since no one had stepped up to the plate before now, I'm betting against anyone doing so in the future.
Re:I'm hoping... (Score:5, Insightful)
This doesn't make sense. There's a clear reason why nobody stepped up before: Hans Reiser. The guy was clearly a bit of a nutbag, however competent he may be at designing filesystems, and I surely wouldn't have wanted to wage a very public and protracted battle with him over his pet project.
There are lots of projects that can use programmers and leadership. One that has nobody running it is a lot more attractive to take over than one where the original creator is going to be lurking hatefully in the background, looking for an opportunity to stab you in the back. (Figuratively; but perhaps in Reiser's case, literally.)
Parent
I'm hoping we'll forget this now (Score:5, Interesting)
Why wasn't it taken seriously by everyone?
- When it encountered problems, it crashed spectacularly. You didn't just lose a file or a block of a file, you lost entire trees and could get metadata instead of file data and vice versa.
- It lacked basic tools, like "dump"/"restore" or "freeze".
- The fsck rebuild process was a security nightmare. A user could craft a file with data that would appear to the fsck process to be metadata, and take over the system after a reboot. In typical arrogance, the solution was "don't reboot, then".
- It didn't support streams or compatible metadata, thus no Mac sharing or SELinux. SELinux in particular is a requirement for many big corporations right now, and not supporting it means no buy.
And, yes, the arrogance of the maintainer played a part too. But even with a much nicer guy running the show, it would not have been a serious contender for the throne.
It's time we forget ReiserFS and move on.
Parent
Re:I'm hoping we'll forget this now (Score:5, Funny)
Parent
Re: (Score:3, Funny)
Re: (Score:3, Interesting)
Re: (Score:3, Interesting)
that maybe control of ReiserFS will now be in the hands of someone who is not a total ****
Well, I've never been exposed to his inter-personal issues, but in reading the article, the journalist involved seems to think that ReiserFS is dead. I'd be rather surprised if that were the case. It's an amazing filesystem, and as far as I knew a few years ago, the only one capable of scaling performance reasonably for extremely large directories.
Hopefully, it will simply be forked (if needed) and continue to be developed now that the original author is clearly not going to be around to maintain it... wel
Re: (Score:3, Insightful)
It's an amazing filesystem, and as far as I knew a few years ago, the only one capable of scaling performance reasonably for extremely large directories.
ext3 supports huge directories. Implementing a hash table to store directories entries instead of a linked list isn't exactly rocket science.
He can appeal, of course, and as I understand it, it's a technical case that they have against him, so there's a stronger chance that an appeal could work than if there were direct evidence of his actions.
Re: (Score:3, Interesting)
Re: (Score:3, Funny)
Re:I'm hoping... (Score:5, Informative)
Sturgeon has a bit of a credibility problem in that none of the eight people he's confessed to killing are, um, dead.
And it's "could have".
Parent
Re: (Score:3, Insightful)
Yeah, it's odd. Lots of things about this case were odd. But that's why we pay people to spend months and months sifting through these things, rather than deciding them via Slashdot.
The same as it affected... (Score:5, Insightful)
It will affect ReiserFS the same as it affected the company that wrote the NTFS file system was convicted of their crimes. (being a monopoly).
Or when the creator of Unix (and the C language) was convicted of their crimes (being a monopoly)
Or the same as it affected Union Carbide when they poisoned a whole town killing everyone. (the former president will be jailed for life if he ever sets foot in India).
The thing is.. Things are decided on their merits, and price, etc. Not on their creators. Otherwise, Walmart would be bankrupt!
So what? (Score:3, Insightful)
Develop from Prison (Score:5, Interesting)
Re:Develop from Prison (Score:5, Funny)
Parent
Re:Develop from Prison (Score:5, Funny)
Parent
Re: (Score:3, Funny)
Good sience and inovations stands apart (Score:3, Interesting)
If this file system is superiour, then it should stand.
However naming the delete function 'Wife' would be lacking in class....
Not a lot, really (Score:4, Insightful)
Re: (Score:3, Funny)
Stupid Article (Score:3, Interesting)
XFS (Score:3, Funny)
XFS (Score:4, Insightful)
ReiserFS was already dead before today (Score:3, Interesting)
I was sad when I found out he was on trial. I had one of his quotes on my cubicle wall from him about development when I found out about the arrest. Hans Reiser is a good developer. Apparently he's not a very good person.
Recent versions of ReiserFS have had stability and performance problems for some time and will die a normal open source death on merits.
Re: (Score:3, Funny)
Re: (Score:3, Insightful)
Re:Article a bit one-sided? (Score:5, Informative)
True, the other developer moved the code and patches to his server months ago when the namesys.com site went down.
Rieser 4 Patches and Programs [zelnet.ru]
According to the article they are still in active development. They got patches for the 2.6.25 kernel so somebody is still working over there.
Source Article [news.com]
Parent
Re: (Score:3, Insightful)
Re:What happened? (Score:5, Informative)
Now, you're right, that list alone creates a plausible but by no means airtight case that he killed her and disposed of the body. If that were all, I probably would have voted to acquit if I were on the jury.
But then Hans took the stand for eleven days, against his attorney's advice, and tried to explain all that. And he did such a massively poor job of it that the jury believed he was lying about why he'd done those things. He said he removed the passenger seat because he was sleeping is his car and wanted extra space; but in a Honda CRX, the passenger seat is probably the most comfortable part of the car to sleep in, and removing it leaves a non-flat surface with bolts and bars across it (Hans said that with a good sleeping bag he didn't notice). There was an inch of water in the interior because he hosed it out; Hans said he doesn't remember it bothering him, sleeping on the floor where the passenger seat used to be.
Since there's little worth lying about that's more important than being falsely convicted of murder, the jury concluded that his lies were covering up a murder. In short, he talked himself into jail. He's not the first defendant to do that.
Parent
Re: (Score:3, Insightful)
If he is a half-assed coder why don't you write a better filesystem.
Why is this form of rebuttal always trotted out? One's inability to do any better doesn't invalidate their original criticism. All those rejects on American Idol auditions are pretty half-assed singers and I can recognize that plainly (as can most people), but I'm certainly not going to claim I can do better because I know I suck as a singer.
Back to filesystems: thought it wasn't terribly common, I did have a ReiserFS filesystem throw up on me once a long time ago. Luckily I was just tinkering with a ne