The GNU/Linux Desktop Alternative
THE SITUATION
I am advising a friend, who is planning a major upgrade of thousands of
desktops, in a major department of Minnesota state government, that he
ought to insist on a cost-benefit analysis that includes GNU/Linux and the
GNU Office Suite (I need the list of these products). Windows NT and
Microsoft Office seem to be a shoe in choice but are very expensive because
of the software licensing cost and the cost of re-training support staff.
I am doing this as volunteer work. All of my writings are licensed under
the GNU General Public License (GPL) <http://www.gnu.org/copyleft/gpl.html>
and may be freely copied and modified provided that any derived works are
under the GPL.
THE HELP I NEED FROM YOU
Please recommend URL's that show cost-benefit analyses of the GNU/Linux
desktop alternative. Also, it would be helpful to have URL's for major
installations that currently use GNU/Linux. Finally, if you were trying to
persuade my friend of the efficacy of this choice, what would you say?
I intend to collect up the feedback I receive and forward it to my friend,
post it on the web and, perhaps, inject it into an upcoming conference that
looks at the technology futures for the department.
BACKGROUND
The major thesis of my argument is that, since the department must re-train
all their support staff anyway, that cost could be considered a wash.
Therefore the decision should be made on other factors. I use the "good
money after bad" argument to help diminish any residual value of previous
training in "non-free software" <http://www.gnu.org/philosophy/free-sw.html >.
I contend that the real needs of the government workers, in terms of where
they spend their real computer usage time throughout the day, are primarily:
TN3270 mainframe connectivity
email
occassional intranet browsing
occasional, simple word processing outside email and document creation
for some people, intensive web document preparation
I argue that the specialized workers doing desktop publishing will be
converting to an HTML solution (I actually encourage SGML but that is
another story). That is but one example of a specialized need. As we look
at each kind of specialized worker's needs, I hope to identify usable, if
not excellent free software alternatives. Where a non-free solution simply
does not exist, I support using existing non-free products under the
GNU/Linux DOS and Windows emulators.
I contend that the addiction to non-free software is costing state
government, in his agency alone, hundreds of thousands, if not millions of
taxpayer dollars per year. The cost for all of local, state and federal
government is staggering but I am not making that legislative point at this
time. I hope, instead, to persuade one major agency of government to make
the transition and become a showcase that proves to the rest of government
that becoming free (from their non-free software addiction) is possible,
attainable and has wonderful side-effects. There is benefit in investing
in human resources rather than pouring money out of state. There is a
wellspring of net accessible volunteer talent waiting to be tapped.
If I fail to bring about this transition (from the non-free to the free)
from the bottom of government upward, then I will ask the Legislature to
declare a moratorium on all software and software training expenditures
until the matter can be investigated. I believe that tens of millions are
being needlessly spent each biennium and I am pretty sure I can prove it,
if given the chance. I prefer bottom up solutions, rather than top-down
solutions imposed from orders on high, and I hope to persuade one
department to, at least, do a fair cost-benefit analysis and develop a
prototype. If that gentle approach fails, I will go for the cold turkey,
Legislature mandated withdrawal.
Let me say very clearly that I am in no way intending to disparage the fine
work being done by any government employee. Any informed person you talk
to in government is gravely concerned about the escalating costs of
technology and wants viable solutions. They didn't sit down one day, a
decade ago, and decide that they were going to become addicted to the
latest whiz-bang technology. It just happened, fix by fix, and now they
are addicted. I admit that there are people in the game to build their
resumes. For the most part, I have observed that government managers have
long been gravely concerned but have believed they had no alternative. It
used to be that nobody in government could ever be reprimanded for
recommending IBM. Today the same holds true for Microsoft. Tomorrow, I
want the same to hold true for GNU software. This is the quintessential
issue.
Nobody wants to stand out there and face the risks of recommending an
unproven technology solutions. Any recommended solution must consider many
things like: the stability and usability of the software, the cost of the
software, the availability of adequate support and the cost of adequate
support. I contend that GNU/Linux is proven technology and that the
GNU/Linux development and support infrastructure is the best. It has the
advantage that much support is donated by the GNU volunteers. It has the
advantage that commercial support is available for the same price as the
alternatives. If I can prove these assertions, then I may be able to
persuade one manager to stand out there a little bit. I am recommending
two steps to precede the final choice: 1) perform a cost-benefit study of
the Common Desktop alternatives and 2) create a prototype of the primary
alternative Common Desktop solutions (Windows NT, GNU/Linux, platform
independent Java clients, and other reasonable alternatives).
Just so you know, I am also recommending consideration of a thin client,
platform independent Java enabled browser solution. The department has
lots of old hardware around that would get a lot of extra life with this
solution. Also, I am trying to persuade them, in terms of their
information and processes, to open themselves to the taxpaying public and
the public they serve and the advocates. Many members of the public have
older technology that could utilize a Java enabled browser to conduct a
dialogue with the department's mainframe computer system. To my knowledge,
only Java has this particular ability. This is a side issue to this
posting but I thought I should disclose it.
That's kind of a thumbnail of this initiative. Those who remember my other
writings <http://www.freedomain.org/u/lls/writings/lls-portal.html> will
see this as a small piece of a bigger work that intends: 1) to bring
copyleft <http://www.gnu.org/copyleft/copyleft.html> to government
information and software development, 2) to connect up the free software
community with government, 3) to open photonic government to the people and
4) to help build photonic democracy. (Note: photonics is merely my
preferred transport mechanism; you can substitute "electronic" if you
prefer.)
Thank you for your time spent considering this matter and for any help you
can provide.
August 27, 1998, posted to < comp.os.linux.misc> regarding <The GNU/Linux
Desktop Alternative> by < "Lyno Sullivan" <lynosull@maroon.tc.umn.edu>>
Posted to <debian-user@lists.debian.org>
--
Copyright (c) 1998 Lyno Sullivan; this digital object is free and
may be copied, modified and distributed under the GNU General
Public License (GPL) at http://www.gnu.org/copyleft/gpl.html and
it comes with absolutely NO WARRANTY; mailto:lynosull@maroon.tc.umn.edu
Reply to: