Linux vs FreeBSD clusters (was: how are the Redhat product changes affecting existing and future plans?)

Arthur H. Edwards edwardsa at plk.af.mil
Mon Nov 10 11:26:48 EST 2003


I think your point about newer package management tools is
well-taken. I have tried the apt for rpms (when I was running the free
scyld distribution) and it was clealy better. I have not tried yum,
but I have not had enough (any) frustration with apt-get, and now
apt-proxy, to make a move desirable. I also agree that my attachment
is more to open-source than to Debian per-se, although after using RH,
SUSe, and even turbo, Linux, I have stuck with Debian. I actually wish
SUSe (now part of Novell) well, and I am sorry to see the demise of RH
as we know it, because they are where increased user base comes
from. However, I don't know whether SUSe will have better luck at
generating revenue than did RH, and they may well go the same
direction. It is that possibility that makes me think that Debian, or a
similar, volunteer-based distribution may have the greater longevity.

Art Edwards

On Mon, Nov 10, 2003 at 07:30:13AM -0500, Robert G. Brown wrote:
> On Sun, 9 Nov 2003, Arthur H. Edwards wrote:
> 
> > 
> > It is interesting because one of the initial attractions for Debian
> > was its organization of libraries and configuration files. Afer RH, it
> > seemed totally transparent. I guess this is just a matter of personal
> > taste. I would be surprised, though, if after trying apt-get, you
> > could ever go back to the rpm model.
> > 
> > Art
> 
> It isn't "the rpm model" -- in both cases the packaging and metadata are
> adequate. Comparing apt to rpm is apples to oranges -- apt-get is a
> toplevel toolset to extract and resolve dependencies from the debian
> packages and use them to retrieve and install package(s) and their
> entire consistent dependency trees, by revision.
> 
> The problem is that in the past there has been no comparable toolset for
> RPM packages and all the distributions that rely on them.  For the last
> two or three years, there has been (first yup, now yum).  It, too, is
> "totally transparent" and has, arguably features that some
> administrators might prefer (including considerable and increasingly
> fine-grained control over their own, local, repository images).
> 
> Whether or not you've looked at yum and tried yum and compared yum's
> operation and features to apt, the existence of choices appears to be a
> good thing, as does "competition" of sorts (the friendly, slightly
> religious sort that tends to exist in the open source world:-).  I know
> yum's primary developers quite well (since they work about fifty meters
> away from my office in the same building:-) and they are very, very
> dedicated and not at all religiously inclined towards Red Hat per se.
> Yum has been successfully used to make RPM-based repositories for just
> about all the primary RPM linux distributions, and I believe that people
> have even used it to distribute/maintain RPMs on Solaris boxes.  At this
> particular moment, I think that yum makes RH (or if you prefer, Fedora)
> slightly preferrable to Debian in a scaled/automated LAN installation
> because both effectively automaintain after installation, but RH/Fedora
> permits the easy use of PXE/kickstart.  Kickstart, after all, is (IMO)
> the reason RH maintained its dominance in spite of the otherwise pain of
> manipulating RPMs compared to Debian, and the reason it remained
> dominant among RPM-based distros as well.
> 
> In part because of its existence, there is actually some talk of coming
> up with a rational unification of linux packaging schemes, reviewing and
> getting rid of package features that have proven to be more Evil than
> Good over the years, developing an XML schema, and lots of other good
> things that might actually reduce the "us and them" barriers for linux
> in general.  I personally think that this would be a good thing.  As
> Mark has been saying -- most of us are religious about open source,
> stability, functionality, but at best we are "used" to particular
> distributions and could be convinced to change fairly easily if
> advantages associated with the change outweighed the hassle of learning
> something new.
> 
>   rgb
> 
> -- 
> Robert G. Brown	                       http://www.phy.duke.edu/~rgb/
> Duke University Dept. of Physics, Box 90305
> Durham, N.C. 27708-0305
> Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb at phy.duke.edu
> 
> 
> 

-- 
Art Edwards
Senior Research Physicist
Air Force Research Laboratory
Electronics Foundations Branch
KAFB, New Mexico

(505) 853-6042 (v)
(505) 846-2290 (f)
_______________________________________________
Beowulf mailing list, Beowulf at beowulf.org
To change your subscription (digest mode or unsubscribe) visit http://www.beowulf.org/mailman/listinfo/beowulf



More information about the Beowulf mailing list