Linux vs FreeBSD clusters (was: how are the Redhat product changes affecting existing and future plans?)
Robert G. Brown
rgb at phy.duke.edu
Mon Nov 10 19:45:55 EST 2003
On Mon, 10 Nov 2003, Alvin Oga wrote:
> hi ya art
> On Mon, 10 Nov 2003, Arthur H. Edwards wrote:
> > 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.
> good point ...
> - i think that "volunteer-based distro" will survive all the commercial
> methodologies ...
> - commercial folks are out to make $$$$ to attempt to cover the
> costs of marketing, sales, advertisement and analysts expectations
> - voluteers do what they do, because its what they like doing and will
> probably continue doing so for the next few eons
Alvin and Arthur,
I don't think either one will disappear anytime soon and think that
we've entered an era where the two can exhibit excellent synthesis.
There is nothing wrong with commercial distributions, or commercial
distributions making money, as long as they remember:
a) They don't own their product.
b) They are therefore at best selling added value, such as support.
c) This puts pretty strict limits on what they can sanely charge.
Some of the major distributions may be forgetting c) just a bit, but the
market will correct this soon enough:-) Or maybe this is just wishful
thinking and some marketing hype to give their stocks a bit of a bounce.
Truthfully, the commercial distributions and individual/volunteer
developers have achieved a moderately healthy synergism in all the
various Linuxes. In fact, the commericial folks have provided a variety
of valuable services -- collecting packages, running a moderately
systematic debugging service (which actually has worked adequately for
serious bugs, however poorly/slowly it works for ignorable/annoying
bugs), applying critical patches, contributing whole applications and
much needed support services to the development process.
I certainly don't begrudge them a living and have gone out of my way to
buy something from them periodically in the hope that the golden goose
stays fat (enough) and will continue laying. I just think it is pretty
silly of them to try to charge as much or more than their major
commercial competitors for a product that they don't own and (for the
most part) didn't develop. They won't get it and in the meantime
they'll irritate a lot of people they should be (and have in the past)
been working with, as well as for.
On the other hand, it is perfectly reasonable to try to re-engage the
volunteer community in the development process and to give the notion of
"supported versions" a bit of a goose. In recent years, we all may
literally have become somewhat complacent, trusting the commercial
groups to PROVIDE those valuable services without our strong
participation. If this is all Fedora (for example) is about, I'm all
for it. However, the whole process >>has<< already started spawning new
alternatives (such as caos) and I think "the community" has plenty of
capacity to support plenty of non-commercial or very low margin
alternatives (as one would expect, given the existence of Gnu itself,
freebsd, debian, all of which support themselves by means of low margin
events like T-shirt sales and donations). So the commercials may find
that they've created something of a monster. Or four.
Overall, I'm less cynical about the process than I was a month ago
because I see some benefit that could arise from it. Companies like Red
Hat, Mandrake, SuSE/Novell may learn from all this what the limits are
on what they can charge and what added value they need to provide and
where to earn a fair living (in contrast to a gross Bill Gates
billionaire profit). They also are firmly reminded that they need us
(the "volunteers" who as often as not own the software they sell) more
than we need them. As I've said before and will repeat -- they aren't
going to get anything like the prices they wish to charge businesses for
"rawhide", and they >>must<< have a meaningful rawhide and community
development process in order to sustain Linux's legendary stability and
The "community" may be reminded that although the software is free,
supporting it in a distributional form isn't free, and if they don't
take steps to fairly compensate whoever it is that is providing the
service, they'd better make arrangements to do it themselves
individually or collectively. The stress will also very likely create a
spate of new programs, which is a very good thing. Some of them may
even be revolutionary products, as I think we're about to enter an era
where computers build themselves an operating environment from certified
open sources in real-time and on demand, except where administrators
deliberately do or mirror a prebuilt repository based on the same tools
and sources for efficiency reasons. The GPL sanctifies the source
package, new XML-based packaging schema and the web itself can guarantee
cross-linux, maybe even cross-linux+BSD combined build compatibility.
That is, SOMEWHERE in the very near future I think we are going to see
the emergence of a completely new paradigm, one that finally ends the
era of commercial software as we have known it in the past. The
requisite tool components are all there, the broadband connections to
the home required to sustain it are there, and I think the creative
juices are cooking in developers' minds. The coming revolution will
make even the notions of java and .net look tame and in retrospect a bit
silly, as the entire notion of java and source application delivery will
be just a tiny fraction of a source application delivery system that can
and will deliver the entire operating system and all derivative tools
(including java). C, perl, python, java, html, php -- sources of all
sorts bundled into GPL packages with attached development processes and
delivered directly to your system on demand for prices ranging from
nothing (as most web services are delivered today) to a trivial amount
for a snazzy subscription/security service.
Hmm, sounds like a whole new .com concept, doesn't it? But I really
think that is where we are going, fairly rapidly now.
I do think that consumer Linux, especially, will suffer tremendously
(indeed already is) from the largely unnecessary and unjustifiable price
inflation -- as if Linux is somehow more expensive to support (badly)
than Windows. 2004 should have been the year of consumer linux, and
still would be if Red Hat would get out there with a $35 box set of the
absolute kitchen sink followed by perhaps $25/year full update support
per household. Or even less -- they now risk the community providing
the installation and update support for free so efficiently that they
can't charge even this.
You can get pretty rich selling $15 objects to people, if you sell them
to a LOT of people. Just ask J. Rowling (and her publisher). But who
would buy even the best of Harry Potter books for $150 a copy?
Especially when they could get pretty much the same book for free, or in
an inexpensive $5 paperback version.
Sounds like Econ 141 time -- I vaguely remember some nifty concepts such
as "supply and demand", and "elastic and inelastic markets". We'll see
if RH, Mandrake, SuSE/Novell remember them too.
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
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