[Bug-gnupedia] My two "centimes"...

[Mehdi TIBOUCHI]

Hello,

I've been reading this mailing list (and its archives) for several
days now and I've been working on a French translation of RMS's
Project Announcement meanwhile (that I hope I can finish some time
next week), which forced me to read it more carefully.

This makes me think that what Stallman has in mind is *actually*
much broader than an encyclopedia. Even the title of the announcement
suggests that: GNUPedia is supposed to be a free "learning resource",
as well as a universal encyclopedia. The idea of a learning resource
covering all areas of knowledge at each and every level, from the most
elementary to the most involved, is what excited me so greatly about
GNUPedia (or whatever it turns out to be called). No real-world
encyclopedia has ever had an ambition even close to that.
On the other hand, the learning resource is quite different from a
library: RMS explicitly excludes such things as fiction and art from
its scope, and I indeed feel it would be inappropriate. IMHO, GNUPedia
should be something everyone would *learn from* about any subject they
like. Something between the 1 and 2 of Seth Nickell's summary. The
description that fits my conceptions best so far is that of Philippe
Aigrin (la proximité culturelle, peut-être? :-). [Clearly, you can
learn something from a novel, but not in the sense I'm using "learn"
here].

I'm very concerned about censorship issues as well. As most seem to
agree by now, I am of the opinion that there should be as little
as possible. Ideally, there should be none, i.e. any submitted
material should be made available to the public at once, and
classified and/or rated afterwards, but never rejected as long as
it meets the required freeness criteria. Now even this is a problem.
Who should be entrusted with checking whether some submitted article
isn't plain rubbish and complies with our freeness standards? As
Bob Dodd noticed, this requires at the very least that the person
in question reads the language. But she must also be liberal enough
as to what "plain rubbish" is, for example, and this is a weekness
in the system from a paranoid freedom fighter's point of view
(suppose a lobby of some sort infiltrates the pool of editors).
OTOH, if all submitted articles are all transferred to a publicly
accessible archive of unreviewed and unclassified material until
a random user does the job, several problems will arise:
* the archive might be spammed.
* someone might send non-free material by lack of knowledge (or even
  mischieviously). This shouldn't be a problem if we reject any
  article not licensed under GFDL or not containing appropriate
  disclaimers, but I'm not a lawyer, and even less an international
  one (and laws seem to benefit to people with money rather than
  people with ideals).
* you don't want an author to classify his article himself or have
  it reviewed by some friend. That's always an issue though.
* probably more.
I don't claim to have any reasonable solution to all this. I'm
rather optimistic usually ("GNU idealistically assumes people to be
trustworthy. Suprisingly enough they turn out to be trustworthy."),
but I'm not sure I would be a truly liberal editor myself, so let
alone others ;-).

Finally, I deem multilinguism important, but I don't share the
opinion of Philippe Aigrin about discouraging people to provide
an English translation when they submit articles in their native
language. Much of the non-English material in the encyclopedia will
probably come from translations, and it is unlikely that there will
be many direct translations from say Swahili to Mandarin--an English
intermediate is to be expected in-between. A "lingua franca" is not
necessarily a bad thing. Hopefully people will care to translate as
much as they can in their mother tongue, though. As for mailing lists,
there will probably be enough people speaking local languages to
setup lists in other languages, as there are for GNU/Linux.

Now, it's getting late on this side of the Ocean.

Mehdi.