Public SourceMage Discussion List

["Wyatt Draggoo" <wyatt AT draggoo.com>]

> I agree with you on the point that I never put LILO outside of the MBR
> either.  Maybe we should take a poll somewhere, as I'd personally like to
> change it to disc too.
> 
> Regarding jfs - before we even consider the alterations that would need to
> be done, why the jfs kick from 'ya all of a sudden?  We already have
several
> journalling FS options built in, with the supporting software. 
Personally,
> I would consider getting Raid support going during install a higher
priority
> than adding in jfs, and even that is fairly low on the list, as I'd like
to
> get a 'net install going very soon.  I think a lot of benefits can come by
> the work needed to get an okay net install done.
> 
> But anyhow, let me know your thoughts... I like having pretty good reasons
> behind feature bloat.  =)

I think the biggest reason for me to use jfs is because it doesn't require
additional patches to the kernel.  And since I also like to ues the PREEMPT
patch and the LOW LATENCY patch, the three don't jive together (read: using
PREEMPT, LOW LATENCY and XFS together will completely hose your
filesystems, and I've done this about 6 times in a row just to prove it was
the XFS patch---even without any XFS filesystems...).  I've never really
liked reiserfs (not sure why, or if there is any rationale behind this),
and the small bit I've read on ext3 is that it's not much more than
journaling slapped on top of ext2.  JFS, on the other hand, has been around
for quite a while (I used it in the latter days of OS/2), and has always
been rock-solid for me.  It's also in the standard, unpatched kernel, which
makes it very easy to add to the ISO (as opposed to RAID, which will
probably involve either a *lot* of install changes, or be a PITA to set
up).  Just seems to me like it's something that will only take a few
minutes to add, but add a lot of functionality.

Wyatt

--
Wyatt Draggoo