Contributed code, etc. (was [Maxima] Teaching differential equations with Maxima)

--- James Amundson <amundson@fnal.gov> wrote:
> Instead of responding to various messages in this thread
> individually,
> I've created one large response to various issues raised in it:
> 
> 1) I am very happy to have people seek out existing
> Macsyma/Maxima-related code. We have not yet fixed up the packages in
> the share directory, however, so it does not yet make sense to seek
> out new code for its own sake. Even so, seeking out code that someone

> has a particular interest in always makes sense.

Hmm.  I guess I sort of disagree, primarily because as people
retire/technology ages it will get harder and harder to locate such
code.  Plus, at least in my case, I don't know what is/was out there. 
But, I see your point.  OK.
 
> 2) If there is a Macsyma package available on the web under
> conditions like "free for scientific use", I have no problem with 
> someone contacting the author to see if his/her code may be 
> incorporated into Maxima. Doing so does not require my permission. 
> However, the author has to know that Maxima is distributed under the 
> GPL, which does not restrict distribution to scientific use, 
> non-commercial use, etc. 

Right.

> 2.1)If the author does not want to remove those restrictions, we
> should be respectful of his/her right to do so. We could definitely 
> set up a web page at the Maxima site that links to such code.

That sounds like a very good idea.
 
> 2.2) I *do not* think it behooves us to try to get contributions from
> people who have a financial interest in their code. This includes the
> current owner of the commercial Macsyma code, CRC Press, etc. I
> discussed this in an earlier message, but let me add this: a
> *baseless* legal challenge from an even modestly-funded commercial 
> entity could very well be the end of the Maxima project. We simply 
> have zero resources to fight such an action.

My sentiments exactly.

> 3) There was some discussion about what to do with contributed code
> that is not necessarily up to the standards of the maxima 
> distribution. It has been suggested that we put it somewhere other 
> than "share." This has come up before. I created the share/contrib 
> directory expressly for that purpose. I consider that problem solved.

OK, I'd forgotten about that.  So if we have a new unevaluated package
we'd like to upload for storage/future consideration, it's OK to upload
it there?  I've got a couple misc. packages I'd like to get into cvs so
I'm not dependant on my backups to keep copies of them.  Some are darn
hard to find, and if I lose my current information tracking them down
again is nontrivial.

> It should also be noted that share packages that work are currently
> the exception, not the rule. Fixing them is one of the milestones 
> between where we are now and 6.0.

Yep :-).  That'll be a biggie.

> 4) I am not clear on the legal status of the files that were
> originally distributed as "DOE Macsyma." We need to determine whether

> there really are functions that are missing (as opposed to simply 
> superseded), and, if so, exactly what they are. In the end, it may be

> simpler to re-implement them then to locate the old versions and 
> determine their legal status. I don't know.

That'd be my guess.  Wouldn't the original authors be able to OK the
release, as long as they weren't employees of Macsyma Inc. or some
other company?

> 5) Vadim suggested a file describing all the share packages. I think
> it would be very good to have someone work on such a file.
>
> 6) After much discussion of current and past ODE implementations in
> Maxima, please let me say what I would really like to see for ODE's
> in Maxima. There is an excellent ODE package at
> <http://lie.uwaterloo.ca/odetools.htm>;. The package is written for
> Maple. The code's license is such that it wouldn't be directly usable
> for us, anyway. However, the techniques are documented in a series of
> papers available from that site. A Maxima implementation of the
> techniques should be a straightforward exercise.

For arbitrarily large definitions of "straightforward" ;-).  I agree
though, that their ODE/PDE setup is very impressive and would be quite
an asset to Maxima.  Maybe I'm making it more complex than it is - how
would the implimentation work?  Work with our current ODE code or
replace it totally?

CY

__________________________________________________
Do you Yahoo!?
Yahoo! Tax Center - forms, calculators, tips, more
http://taxes.yahoo.com/