Taras Glek wrote:
>
> http://devlinux.com/projects/reiserfs/ says "3.5.18[ReiserFS] is out, and it
> is stable".
> SuSE is going to support ReiserFS in the 6.4 release, I would imagine that
> they have throughly tested the filesystem. If you need to throughly test the
> fs, why dodn't you intergrate it into the Cooker ASAP(you would have to do
> that at some point). I would be among the first to try it out.
have you tested it ?
the format of the fs change on 2 months, the smp support is not as
stable as ext2 one.
As for 2.4 support, it trashes when you stress it (sometimes just by
untarring a kernel).
And it lost all its speedups on ext2 while porting from 2.2.x to
2.[34].x.
As for suze, they are annoyed as the future 2.4.x reiserfs has another
format and (at this moment) don't support the 2.2.x format.
As for fsck, it's broken and even dangerous on 2.[34].x (when you patch
it in order to get it to compile - i wonder why they do printk in a
userspace program ?).
And don't speak of other arch such as alpha support: the fs bitmaps are
not endian safe ...
Yes, it's a promising fs, but not as stable as ext2 for the moment.
We'll consider it support when its stability, it portability among
ix32/alpha/usparc/ppc will be sufficient.
> jfs and ext3 have not reached version 1.0+ yet and Reiser FS seems to be much
> more machure than xfs(just an impression from the version numbers).
>
> Vandoorselaere Yoann <[EMAIL PROTECTED]> wrote:
> > Taras Glek <[EMAIL PROTECTED]> writes:
> >
> > > Hi,
> > > I have a 27G hard drive and fscking it can take quite a while with
> Mandrake
> > > 7.02 using ext2 filesystem. I would love to see reiser fs supported by
> > > Mandrake as soon as possible so I wouldn't have to wait as long dueing
> boot
> > > up. SuSE and a number of other distros plan to have Reiser FS support in
> their
> > > upcoming releases.
> >
> > It's in our plan too.
> > But don't forget there is some other interesting filesystem to come :
> > xfs, jfs, ext3...
> >
> > Also,
> > Changing the filesystem will require *heavy* stress test...
> > We can't afford a critical bug to be discovered in reiserfs after releasing
