[Axiom-mail] address@hidden: Re: Axiom Journal]

[root]

------- Start of forwarded message -------
Date: Fri, 29 Nov 2002 09:54:51 -0500
From: root <address@hidden>
To: address@hidden
CC: address@hidden
Subject: Re: Axiom Journal
Reply-to: address@hidden
X-RAVMilter-Version: 8.3.1(snapshot 20020109) (mail.idsi.net)
X-UIDL: EgU!!;Qj!!"H>!!U"R"!

> While I'm strongly in support of your views of literate programming
> and your ideas for an Axiom Journal, I think that the issue is hard to
> solve: people mostly don't want to publish in niche journals,
> electronic journals have not existed long enough to know whether any
> particular one will survive, and there are already so many journals
> that no one really claim to can keep up.  So far electronic journals
> exist as a labor of love by their founders and supporters: what
> happens when those people lose interest, retire, expire, change jobs,
> ...?

Ah, there's the rub... In fact, it is the driving point behind the
literate programming effort. At the moment I'm one of probably three
people in the world who know how to build Axiom and I'm working hard
to permanently code myself out of that position. That way I can retire :-)

The Journal issue is somewhat different though. I can set up the
machinery to support the Journal but I can't force people to 
contribute. My current plan of attack is to work with people who
have already written Axiom code to convert existing code into 
Journal pamphlets and update them with new theory. I have talked
to a few people who seem willing to do this provided I'm willing
to handle most of the grunt work like final rewrites into the
proper format, etc.

Axiom has about 370 algebra files. Some percentage of these will
eventually be backed by existing published papers. I'm hoping that
these can be used as a demonstration (even to me) of the usefulness
of publishing pamphlets. I've already gotten permission to use some
of the original publish source materials.

I'm also in discussions about actually proving the algorithms
with some formal logic which would make the publications unique.
If I can build support in the program proof community for proving
submitted algorithms as part of the "review" process then the 
Journal would offer something unique and useful as an incentive.

Beyond that, it depends on how active the Axiom community becomes.
There is a fair-sized community of teachers in France who have
an interest in using Axiom for teaching. And another community
interested in using it for research. I'm hoping that since this
Journal has a non-exclusive policy that people will be willing to
consider this as a secondary path for publishing.

The only permanent "cure" for the "labor of love" problem is either
the Journal is picked up by an organization like the ACM or the
community of mathematicians using Axiom come to see it as a legitimate
"publish" path with real value. Beyond the startup issues and 
maintaining the background machinery I'm hoping to have an editorial
board and reviewers running that portion of the website.

The concept is still evolving. In any case, I figure it is worth
having a discussion about it and use the "group mind" rather than
muttering to myself.

Tim
------- End of forwarded message -------