[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Gnu-arch-users] on RH, part 1
From: |
Tom Lord |
Subject: |
[Gnu-arch-users] on RH, part 1 |
Date: |
Thu, 26 Feb 2004 13:39:11 -0800 (PST) |
I seem to have (understandably) given some people that I only like to
rail indignently against RH -- that I don't recognize the value of
their contributions, the challenges of their business model, the
realities of capitalism, or the quality of their hackers. There is a
sense that I am trying to apply an ethical analysis in a context where
it is inappropriate to do so.
I'd like to put that perception to rest and I've tried to do so by
collecting my thoughts in two parts. Part 1 is a review of what I
think are the most serious ethical issues with RH's business model --
presented from what I hope is a balanced perspective. Part 2 are
some thoughts _towards_ a better business model for RH et al.
-t
Some Problems with RH's Current Business Model
----------------------------------------------
Let's start by looking at RHN and how it relates to the RH distro
products (WS, ES, AS).
For background, RHN (the "Red Hat Network") is:
[....] an easy-to-use systems management platform for your Linux
infrastructure. [....]
* What does it do?
Red Hat Network provides administrators with the tools to
efficiently manage the systems on their network. [....]
administrators can add enhanced capabilities for system updates,
management, and provisioning of their entire infrastructure.
-- http://www.redhat.com/software/rhn/
The various flavors of enterprise linux (WS, ES, AS) are, of course,
distributions. Red Hat distinguishes them this way
[....] It's sold by annual subscription, runs on seven system
architectures, and is certified by top enterprise software and
hardware vendors. It's backed by a Red Hat Network subscription
[....]
-- http://www.redhat.com/software/rhel/
So we have something here that's pretty simple, conceptually:
~ a release-managed distro, tested on various platforms, and
officially supported by some ISVs of non-free software
~ for which the preferred form of management is to use
RHN
What's the catch? There's a two-fold catch:
1) RHN is comprised of non-free software.
The RHN software is protected by trade-secret laws. Customers are
provided with a right to install the RHN software on a limited
number of machines; strictly limited in the number of copies they
may make; are not permitted to make modifications; are not permitted
to "reverse engineer" the software; are required to destroy copies
in their possession at the end of the agreement. Even when this
software is installed on customer machines, RH asserts that those
copies are the property of RH -- that, all appearences to the
contrary, the software has not (for legal purposes) been distributed
to the users.
User's are restricted in their _use_ of RHN software. In
particular, they can not configure it to control more systems than
they have paid RH for the right to manage. RH reserves the right
to inspect a customers site and records to ensure compliance.
(See http://www.redhat.com/licenses/rhel_us_3.html?country=United+States&)
There's a coherent business model reason for these legal
restrictions, of course. Absent them, it would be easy for
customers to buy a single system subscription to Enterprise, obtain
a copy of the RHN software, and control many machines from that
single subscription. (Not that it's all _that_ hard for a customer
to do that anyway -- which is one of the problems I'll be exploring
later.) So long as such a customer limited his support requests to
a scale appropriate to his single-subscription entitlements, he'd
obtain a great deal of the value of a many-system subscription for
the price of a single-system subscription. With the savings, such a
customer could likely buy all the support he needs from a third
party and still come out ahead.
At the same time, those of us who like the ideal of free software
ought to notice that RHN is very far from free. The aim of RHN is
to have a distribution which, though it is comprised (apart from RHN
software) of free software, is packaged in a form that makes users
dependent on an extremely non-free tool. While users are
theoretically free to modify, copy, and redistribute the platform
(apart from RHN itself), the form and legal structure of RHN
strongly discourages them from doing so: it asks users to trade
benefits they might otherwise receive from free software for access
to the Red Hat Network.
This divergence between the core ethical principles of free software
and the RHN business model should raise a red flag for people,
whether they giving prioritiy to ethics ("business is no excuse for
non-free software") or giving priority to business pragmatics
("businesses based on non-free software are distinctly vulnerable to
competitive attack").
(There are people working on working around this problem. One
example can be found at http://www.nrh-up2date.org/ although I
wonder if its creation didn't involve a violation of the law:
http://www.nrh-up2date.org/nrh-up2date.html)
2) The distributions, hence their certifications by ISVs, are not
shared by the community.
Red Hat courts non-free ISVs and seeks semi-exclusive certifications
for a distribution which they produce privately. While they more or
less conform to the requirements of making source available, and
even make it available to non-customers, they do not do so in an
upstanding manner (see http://whiteboxlinux.org and
http://whiteboxlinux.org/news.html) or in a form that ISVs
recognize (the ISVs don't certify "a system built from such and such
sources", as one might expect of a free software OS -- they certify
whatever is supplied to paying customers of RH.)
Here again, there is a coherent business strategy. Semi-exclusive
ISV certification has value to customers relying on software from
the certifying ISVs. For example, a customer who already relies on
Oracle but who wants now to move to commodity hardware and lower
their platform costs is driven to Red Hat.
But, again, there is a violation of ethical and pragmatic
principles. For many projects (the kernel, apache, GCC, Gnome,
etc.) Red Hat is happy -- eager even -- to cooperate fully with the
community. They author and submit changes in the usual way. In
some cases, they contribute to helping maintain the projects as
public projects to which anyone may contribute. Red Hat doesn't
insist on private control of these projects. They recognize the
gain that comes from running the projects in public. But for the
distributions themselves -- although there _are_ public distribution
projects -- RH turns its back on the community. Rather than sharing
work and obtaining the efficiencies of free software processes --
they make the gambit of doing work privately and in a form where
nobody but them will get the chief benefits.
That's not an ethical violation of the same sort as the non-freeness
of the RHN software. It's a behavior that is entirely permitted --
indeed _protected_ -- by free software licenses. There are even
areas of technology (say, a computer-chess competition) where such
behavior would be expected as "part of the game".
But it is ethically odd, at least, that in this one area of making
and proving the quality of distributions -- an area in which there
is demonstrable gain (Debian) from cooperation -- they turn their
back on the community. They _could_ have invested to help the
larger community become more efficient at this task and put the
matter behind all of us -- gone on to spend their labor resources on
other tasks. But they didn't: they elected instead to reinforce
and exploit an inefficiency in the larger community.
And, once again, even if business pragmatics are your primary
concern rather than theoretical ethics, the split here between what
would be an unproblematic relation with the larger community and
what they actually do should give pause: if there is a more
efficient and more cooperative way for the larger community to
produce the same result that RH produces privately, won't the market
eventually find this optimization?
Projects like http://whiteboxlinux.org, Debian, and UserLinux
suggest that the barrier RH has erected around their distribution
will, inevitably fall. The Fedora project may even indicate
recognition of that by RH.
Examining these two elements of their business, RHN and Enterprise, we
see that the RH strategy towards customers is a pincer strategy:
customers are encouraged to continue to rely on non-free applications
from third-party ISVs. That's one arm of the pincer. In order to run
those applications, they must use a supported RH distribution,
deployed using the RHN software that excludes the customer from many
of the benefits of software freedom. That's the other arm of the
pincer. It's hard to reconcile that, in my opinion, with part of RH's
stated mission:
That's why open source is inevitable. It returns control to the
customer. The code is open and you can see it, change it, learn
from it. Bugs are more quickly found and fixed. And when customers
don't like how one vendor is serving them, they can choose another
without overhauling their infrastructure. That means: No more
arbitrary pricing. No more technology lock-in. No more monopolies.
-- http://www.redhat.com/about/mission/opensource.html
And the RH strategy towards the free software community is
exploitative -- a kind of "Maxwell's Daemon" approach. They are happy
to support community efforts (e.g., Gnome, GCC, the Linux Kernel)
which enhance their pincer strategy while, simultaneously, turning
their back on, competing against, and even rendering illegal free
software projects that undermine their pincer strategy (e.g., Debian
vs. Fedora, RHN vs. NRH.)
The fundamental act of friendship among programmers is the sharing
of programs; marketing arrangements now typically used essentially
forbid programmers to treat others as friends. The purchaser of
software must choose between friendship and obeying the
law. Naturally, many decide that friendship is more important.
-- RMS, The Gnu Manifesto
-t
[Prev in Thread] |
Current Thread |
[Next in Thread] |
- [Gnu-arch-users] on RH, part 1,
Tom Lord <=