Linked by Thom Holwerda on Thu 28th Jun 2007 18:50 UTC, submitted by Punktyras
Linux "Hans Reiser is waiting for me, standing on the other side of an imitation-wood table. The room is small, the concrete walls bare. A guard locks the steel door from the outside. There is no sound. Reiser is wearing the red jumpsuit of a prisoner in solitary confinement, though he has been allowed to meet with me in this chilly visiting room. There was a time when he was known as a cantankerous but visionary open source programmer. His work was funded by the government; he was widely credited (and sometimes reviled) for rethinking the structure of the Linux operating system. Now he is known as prisoner BFP563."
Thread beginning with comment 251736
To view parent comment, click here.
To read all comments associated with this story, please click here.
RE: a sad day
by sbergman27 on Fri 29th Jun 2007 22:17 UTC in reply to "a sad day"
sbergman27
Member since:
2005-07-24

"""
i loved reiserfs and am teribly sad to see it go.
"""

A plug for ext3, here. I've been noticing that while reiserfs still has the reputation for being faster for lots of small files in a directory, ext3 has been steadily improving, in the absence of fanfare, and in the last informal (bonnie++) testing I did between reiserfs 3 and ext3 on my desktop box, ext3 absolutely blew reiserfs away for that situation. And generally beat it on block io as well.

I was shocked at the results, as it was not what I was expecting. And this was on Suse where reiserfs is the default, so it is unlikely some problem with the default configuration.

Also, on a performance note, I always like to remind people that the reiser4 vs ext3 benchmarks on the Namesys site had the phases of the mongo benchmark in which Reiser4 was solidly trounced ("embarrased" might be a better word) by ext3 removed before publication. Hans admitted it when confronted on Linux kernel a couple of years ago, mumbled something about fixing it, and then never got around to actually correcting the highly misleading published results.

I can did up a link to the fiasco if required.

Edited 2007-06-29 22:20

Reply Parent Score: 3

RE[2]: a sad day
by SEJeff on Sat 30th Jun 2007 05:45 in reply to "RE: a sad day"
SEJeff Member since:
2005-11-05

No need to dig it up dude, google bookmarks to the rescue. Hans was called out by a former Namesys employee for "doctoring" the benchmarks. He must have "forgot" 90 second stalls in file operations.

http://lkml.org/lkml/2004/8/27/70

Hopefully justice is served. Fry him if guilty and let him get on with his life if innocent.

Oh yeah:
http://linuxgazette.net/122/piszcz.html
Reiser4 actually isn't all that fast. Scroll down and read the conclusion.

Edited 2007-06-30 05:47

Reply Parent Score: 2

RE[3]: a sad day
by sbergman27 on Sat 30th Jun 2007 17:45 in reply to "RE[2]: a sad day"
sbergman27 Member since:
2005-07-24

"""
Oh yeah:
http://linuxgazette.net/122/piszcz.html
Reiser4 actually isn't all that fast. Scroll down and read the conclusion.

"""

Thanks. That's another benchmark that makes me wonder why no one ever seems to get excited about JFS.

Reply Parent Score: 2