[Axiom-mail] Axiom Journal

[root]

>  Carlo Traverso wrote:
>
> I would like to use an existing journal, if it were not a journal with
> a restrictive policy on electronic access.
> 
> Journals (including SIGSAM, JSC and JCM) usually allow access only to
> subscribers, and often prohibit reposting and free access to the
> articles.
> 
> Creating a new journal has a sense if it is a GNU-licence journal. But
> of course it has to be journal with a good program committee, and a
> backing organization - that means some kind of funding. But the
> program committee is the difficult part.

re: Existing vs Free Journal

Since virtually every Journal I've ever received was paper and 
restricted I can't say what the actual dynamics of Journals are.

In some sense there is a direct analogy between commercial software
and current Journals. Both are centralized and both are expensive.

The free software model can be applied to Journals. 

re: Reviewers

It is clear that the quality of a Journal is related to the quality
of the reviewers as well as of the received papers. Quality review is
time-consuming work. Unpaid work, at least in my experience. While
the "work for free and give it away to everyone" model has clearly
infected the programming community I don't see it having spread
widely in the Journal community (except for the reviewers and
contributors, of course). One would have to build up a supporting
group of open-source reviewers just like building up a supporting
group of open-source developers. If the work is interesting and
worthwhile an open-source project can attract developers. So is
is unreasonable to expect that it could attract reviewers?

A key advantage for reviewing Pamphlet files would be that the
theory has been "reduced to practice" in the form of runnable code.
That would make checking the results somewhat easier and make the
review more concrete. Of course, the author has to work harder to
write quality code as well as the new theory but that's the glory
of it all, no?

re: Copyright

The copyright issue might be a problem for some University based
work. I'm not sure if there are Universities that insist on
holding some part of a copyright on published work. Clearly you
would not want publishing in an online Journal to preclude also
publishing elsewhere. So the Journal would have to have a license
policy more like a BSD-style license.

I know that it is bad form to publish the same paper in two Journals
(and probably violates current copyright agreements).  This might be a
problem since who, after all, would want to publish a paper online in
some iffy-startup-journal when there are more widely recognized
Journals available. Perhaps the early submissions will consist of
"republished" material where the author owns or has re-acquired
publication rights. Or they might consist of rewritten papers with
the embedded code which would hardly be a republish.

re: Distribution

An Axiom Journal, because it is composed of pamphlet files, would have
to have electronic distribution. And since virtually all parts of the
pamphlet would be absorbed into various parts of a running system
you'd have to have a "free-license-only" (e.g. BSD) policy. Otherwise
you couldn't compose the various pamphlets into booklets for teaching
or use the "proven code" as a theorem (the very notion of a
"copyrighted" theorem gives me pause...).

Given that the distribution is electronic there would be essentially
zero distribution cost. Since submission and review are already done
on a free basis I don't see much of a change, except that perhaps
reviewers would be identified by name and reviews could happen
continuously even after "publication". I suppose you'd have to keep
the Journal articles (pamphlets) under some sort of CVS control as
misprints and mistakes were corrected.

Reviewers, by the way, could have recognized areas of expertise.
So a submission could be "reviewed for theory", "reviewed for form",
"reviewed for algorithm proof", etc.

One could even have "threads" of topics in the Journal that go into
the details of discussion among reviewers and the author.  So an
online Journal would have a newsgroup quality to it also.  Later users
of the article (pamphlet) could contribute to the threads.

re: Funding

As to the funding part... well, funding is always welcome, of course.
However, money is only a form of "stored work" and the community has
so far been willing to store their work in running code instead. I
think we need to build it, see who comes, try it out, modify it and
extend it. I've searched around and haven't found a source of funding
for free computer algebra in general nor Axiom in the specific so
I don't expect to see financial support for a Journal. 

Thirty years from now nobody will care who funded it but the work will
still be in the CVS. I'm trying to "take the long view".

Tim 
address@hidden