On 23 Oct 2003, seth vidal wrote:
> > $2500 USD for RH Proxy server (basically an up2date server, bleeccch)
> > $25/installation of Workstation when in quantities over 100
> > $150/installation of RHES
> > $1500/installation of RHAS
>
> eek $25/workstation? What about beowulf nodes - do they count as
> workstations?
>
> > Purchasing the proxy is a requirement to get the .edu (.ca up here)
> > pricing. I asked about paying a large annual fee for the whole
> > university to get unlimited WS installs with a fee for ES/AS installs,
> > but they're standing FIRM.
>
> bizarro.
Bizarro and Evil. This is sheer madness. It is also what I think many
people have feared concerning Red Hat and the commercialization of
Linux. Linux takes the long slow road from "nothing" up to a world
class operating system written by volunteers, and as soon as it is just
barely "arriving" at a state where it might challenge Windows' dominance
on the desktop they dream up these prices that are absurd even relative
to Windows. These prices make me very, very angry.
This isn't the right time or place for a proper rant on this, but one is
boiling up inside of me. It is perfectly fine for Red Hat to want to
make a profit, but this is highway robbery (as are Microsoft's similar
prices, but Microsoft at least has the excuse of having proprietary
software and a huge staff of full time programmers who do nothing else
but develop and maintain it). What possible justification is there for
prices of $25 per workstation seat when those seats are installed
automatically from a single, University local server, and consume no
additional marginal resources from Red Hat if one is installed or one
thousand are installed? What possible justification could there by for
charging $2500 for ANY server EVER? Are they suggesting that they
somehow "earn" hundreds to thousands of dollars per system by including
apache or nfs server rpm's?
This isn't a fair profit -- it is gouging the market, and the prices
above are their LOWEST prices I'm sure offered to a group that has,
collectively, CONTRIBUTED a rather large fraction of their code base.
They're selling our own stuff back to us with a truly absurd markup.
Red Hat adds value, sure, which is why in the past I've gone out of my
way to buy their products from time to time, but the true value they add
is the SUPPORT they sell. We (as a University) consume almost no
support from them other than the rpm images themselves, and providing
those costs them no more for one system than for the entire University
as it is entirely automated.
> > So it's looking like fedora [legacy] or Debian. Bleh, I could go on for
> > hours about this. I needed to rant, sorry for the wasted bandwidth.
> >
> > Seth, are you on the fedora-legacy list?
>
> yep - last time I checked fedora-legacy list was hosted at duke. :)
>
>
> Why not fedora core? Don't like the model?
It is my own hope that the net result of this is that RH's absurd prices
are uniformly and even violently rejected by the community, and that
they come to their senses (or should I say are driven to their senses)
by the process. The shame of it is that I think that Universities and
businesses and even private individuals would be perfectly happy to pay
a fair price for RH. It's just that a fair price model has to reflect
the actual cost scaling (to RH vs the enterprise) of the distribution
model chosen.
If they wanted to charge a University a couple of thousand dollars a
year for a single up to date (I'd vastly prefer rsync'd to up2date)
primary image of their repository, that seems entirely reasonable. If
they want to OFFER a University a support contract (access to
phone/email/web support services) on the distribution for some sort of
fair market price, that too seems reasonable, as does offering the
service on an a la carte basis. These prices reflect the actual
consumption of bandwidth and RH based resources in the former case (plus
an amortized share of the cost of providing the distribution itself,
plus a healthy profit) and is a straightforward cost+profit recovery
model in the latter case. The differentiation (cost or otherwise)
between "advanced servers" and "workstations" is utterly spurious and
silly.
I work with at least one business that just adopted RH widely across
their enterprise, in part because it was so much CHEAPER than Windows.
Now they turn around and make it more >>expensive<< EVEN than Windows,
EVEN when all sorts of core business applications are still missing and
have to be pieced together. I give up. Not since Sun microsystems
missed the boat back in the early 90's, when they had a PC-capable Unix
and chose to market it for absurd prices instead of $20 seat full
retail, has a company done such a boneheaded thing. If Sun had taken
the low road, they would own the Universe now and linux would never have
been born. RH is doing one of the few things that could actually set
the progress of linux back by years. The Microsoft marketing droids
must be laughing hysterically -- for the first time in a decade they'll
be able to sell their product as being the CHEAPEST solution out there.
Aw, hell. I just hope Fedora works out to be minimally as functional
as RH 9 out of the gate. Otherwise it is off to new-distribution-land,
just when I'd really gotten RH, kickstart, yum all worked out...
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@xxxxxxxxxxxx
