[Axiom-developer] Re: [SIGSAM_OFFICERS] ACM/ISSAC Active Journal

[Emil Volcheck]

Tim,

I like both ideas: a potential journal as well as for the
ISSAC CD.  ACM has the wherewithal to launch a journal like
this, but we'd have to do prototypes like you suggest.
Possible starting points are the ISSAC Proceedings, or
a special issue of either Communications in Computer
Algebra or Transactions on Mathematical Software.

How would you propose to handle submissions that include
a Mathematica or Maple literate program?  Should we ask
commercial vendors to include trial versions?  Or perhaps
we could link this to Manuel Bronstein's idea of a SIGSAM
server running commercial and free computer algebra software
that members could connect to.

How do you think we should start this?  What's the first step?

The idea of a journal accepting software as well as articles
has been around for a while.  I remember when I was at
RISC-Linz over ten years ago, people were talking about getting
the JSC to accept and test programs to accompany articles.

--Emil

On Mon, Sep 05, 2005 at 10:28:03PM -0400, root wrote:
> Emil, et. al.
> 
> For the past two years I've been building the proceedings CD that gets
> distributed at ISSAC. This has been interesting but not really breaking
> any new ground. And the process starts at the last minute so there is
> very little time to do anything but get the CD assembled.
> 
> This year I'd like to consider something more interesting and to do that
> I need to start much earlier (now).
> 
> Consider the following idea:
> 
>   You are giving a talk about a new algorithm. During the talk you give
>   out the URL which contains a copy of your paper. The paper is a literate
>   program which contains latex plus the full source code.
> 
>   A person in the audience can put a special Live CD in their laptop (which
>   boots linux but does not change their hard drive). They open a browser.
>   They surf to the given URL. They "drag and drop" your paper onto their
>   machine, it compiles, installs, and is set up to run.
> 
>   Now the person in the audience can actually execute your algorithm while
>   you are giving your talk.
> 
>   This is the "Doyen" idea. It can be applied in all of the science
>   conferences but ISSAC is a perfect test case.
> 
> For background some of you need to know that Carlo Traverso has been trying
> to develop an "Active Journal" which would accept "Literate Programs" for
> publication.
> 
> Literate programs (my definition) are papers that contain the complete,
> runnable source code for the published algorithm as well as things like
> proof of termination, order of the algorithm, etc. If these are published
> in an Active Journal then people will be able to use the algorithms directly.
> This has been a problem in the past (e.g. a new, faster Groebner basis
> algorithm that does not have associated code). Given that we are doing
> computational mathematics I claim that running code is as important as
> a proof. Axiom has been rewritten to be all literate programs (that is,
> all of the files are now documents and there is no lisp, boot, spac, C, or
> Makefile files anymore).
> 
> Another effort is the Doyen Computational Science Platform CD project 
> (http://sourceforge.net/projects/doyencd)
> which has created a Live CD (that is, one that will boot linux on a
> computer from the CD without changing your hard drive). The goal of
> the Doyen project is to set up a Live CD that contains several computer
> algebra systems, a web server, a browser, and a special Wiki, a website
> user's can change. Thanks to Bill Page Axiom already has such a wiki
> (http://page.axiom-developer.org) 
> 
> You should be able to "drag and drop" a literate program onto a Doyen
> CD and have it automatically compile, install, latex, and set up to run
> the program in the paper. Doyen is not quite there yet but that is the
> goal.
> 
> If we can create a working demonstration of a Doyen CD for the next
> ISSAC then we can encourage people to write algorithms that will work
> on the CD software and publish the resulting literate programs in 
> Carlo's Active Journal.
> 
> All of this technology exists in prototype form. ISSAC would be the 
> perfect forum to demonstrate a working prototype and it would certainly
> make the ISSAC CDs much more than just an electronic copy of a paper
> journal. If this works then the ACM can spread the technology to the
> other computational science areas.
> 
> Comments?
> 
> Tim Daly

-- 
Emil Volcheck
address@hidden
http://acm.org/~volcheck