Re: [Sheflug] cospa project

[Alex Hudson <home [at] xxxxxxxxxxxxxx>]

On Tue, 2004-03-09 at 17:02, Andres Baravalle wrote:
> I'm working at the department of computer science of the University of 
> Sheffield for a project related to the use of open source software in 
> the public administration (COSPA, http://cospa.case.unibz.it/).

Ah, an FP6 project. 

> Our aim is to verify if using open source software and open data 
> standards could be an effective option. Open source has already proved 
> to be effective on the server side, and the core of our investigation 
> would be on the desktop use.

In this country, it is already an option - Government procurement
guidelines as currently set by OGC say it should be considered alongside
proprietary solutions on a value-for-money basis. There are similar
requirements for open standards in publicly funded projects; the eGif
requirements are an example.

> Have you got any comments/suggestions?

The Association For Free Software are already interested in this area:

	http://www.affs.org.uk/government/index.html

AFFS would definitely be interested in both helping with your research,
and the results of it. We recently hosted a conference (FLOSSIE), which
was aimed (partially) at increasing the interest in free software on the
desktop within education, which is another public sector.

One problem that currently hinders the take-up of free software in terms
of Governmental procurement is the natural tendency towards collective
negotiation on a national basis. Government is attempting to realise
discounts via bulk purchase of software, which often means that even
though free software is perfectly adequate in many areas, it could not
be purchased because there is already a local or national licence in
place for a proprietary equivalent. AFFS is currently looking at ways of
exposing that cost - you'll see on the Government pages above, for
example, we link to the record of the public accounts committee when
they looked into the cost of proprietary licences. There are also
interesting statements in Hansard, where Paul Boateng MP said something
to the effect of, "it's impossible to say how much proprietary software
is costing us; we don't keep the records and it would be too expensive
to find out". 

The question of whether or not free software is price or feature
competitive with proprietary software is not really an interesting
question any more; it is demonstrably feature competitive, and the price
of the software really depends on how you want to calculate it. For
example, you can make proprietary upgrades look quite cheap when you
factor in the cost of retraining when you move away from the package.
But, you can make free software look quite cheap when you take into
account that the current marketplace has no choice, there is no version
upgrade treadmill, and similar issues.

"Open standards" is quite a weak topic. It's very difficult to define
what constitutes an open standard, for a start. Is published
documentation good enough? No, not really - for any real-world
interchange format, the documentation is not likely to be accurate
enough to implement it. What about patents? Many formats are covered by
patents - MPEG is a good example - so even if you have 100% accurate
documentation, you still cannot implement the standard. Is encryption
used, and who has the keys? If you cannot decrypt a document, you cannot
read it, even if you know the structure inside. Etc., etc., etc. For me,
the only real practical definition of an open standard is "there is some
free software which talks this standard". But, most people will not
recognise the huge numbers of problems that my definition attempts to
address, and assume weaker definitions, such as "you can download a
specification document", or (even worse) "the proprietor will allow you
to licence the format under Reasonable And Non-Discriminatory (RAND)
terms" which is usually complete b******t.

Cheers,

Alex.

___________________________________________________________________

Sheffield Linux User's Group -
http://www.sheflug.co.uk/mailfaq.html

  GNU the choice of a complete generation.