OT: redhat

[Bill Vermillion]

When asked his whereabouts on Fri, Nov 05 22:22 , Fairlight took the 
fifth, drank it, and then slurred: 

> >From inside the gravity well of a singularity, Bill Campbell shouted:

> > When I installed my FreeBSD system here, I looked at the
> > default sizes that came up with an automatic allocation, and
> > doubled them all.

> > Every couple of years it seems that I have to double the
> > amount of space I allocate for the ``/'' file system on Linux
> > systems.

> Yeah, I know what you mean.  When I spoke of my X11R6 stuff...actually, at
> the time, it was R5.  :)

And Xorg seems bigger than XFree86.

> > As a rule, if a file system gets nuked, it's ``/'', and I
> > really like to have all my critical data in another file
> > system. I've also found that when a journaling file system
> > goes bad, it goes *REALLY* bad.

> A case for good backups. :) I suppose it's good to know the
> facts, even if they're not particularly reassuring. Thanks
> for the heads-up. I can't wait for ReiserFS to dive. *tries
> -really- hard to look enthusiastic*

> My biggest disappointment so far is Reiser's lack of
> implementation of the immutable bit. I'd pay for that feature
> right about now. Unfortunately, only ext2 and ext3 actually
> support it, AFAICT. If you use chattr to set it, it will show
> up as set, but you can still do whatever you like.

Clarification please.  Do you mean reiserfs will accept but no use
it, or that ext2 and ext3 will set it but no use it.

How do you see if the immuteable flag is set.  In BSD it's the -o
along with the l flag.  ls -lo  at the minimum.  The chflags in BSD
also can do more than set immuteable.  You can use it to make
files append only - so that if a hacker makes it in they can not
change the log file [as an example], or you can set other things
such as nodump.

If you properly combine the file flags with the security level
that prevents a time change greater than one second you close up
a few of the favorite hacker tricks for deleting trails pointing to
their efforts.

> The apparent claim for not supporting it is that it's not a
> POSIX standard. Like that's ever stopped someone. You'd think
> they'd second-guess this design decision, given FBSD's support
> for such a mechanism at the very least.

> But hey, if that's the most ReiserFS disappoints me, I'll count
> myself fortunate. I dunno--people swore they had nightmares
> with ext2, and in all the years I used it, on all the systems
> in which I saw it deployed, I only saw a few screenfuls of
> inodes lost, and it was 99.9% indicative of incipient hardware
> failure. As someone said to me when the journalling filesystems
> (ext3 and ReiserFS) were really on the rise, "Yes, it performs
> really admirably for a 1994 vintage filesystem."

That's why I think the UFS2 has such great potential. It [along
with things like SGIx XFS] are really major changes to things
that are long overdue for change in this world of rapidly
increasing drive sizes.  And prices.  Today's ad for Staples
has 250GB drives at $130 and recordable DVD blanks at $0.40 [ TDK ].

Bill
-- 
Bill Vermillion - bv @ wjv . com