Re: Thoughts on Project Structure



--- Richard Fateman <fateman@cs.berkeley.edu> wrote:
> I may be more qualified by previous experience to help
> with lisp, but not with the various other aspects.
>   I am not familiar with the Sourceforge setup and
> I am frankly scared of the prospect of integrating
> code written by persons of various skills into something
> that works.

Can't blame you there :-)  For the first, I know a little about
Sourceforge and would be willing to learn whatever else about that I
need to to help with that aspect of things. 

>    In my view one of the major problems
> of the 1970-1980 coding (the last done at MIT) was
> that a self-appointed energetic undergraduate who thought
> he was really smart and knew how to code "better"...
> substantially raised the barrier to understanding the
> system. He also inserted bugs because he didn't know
> what some of the algorithms were supposed to do.

We obviously want to avoid that, but we also don't want to discourage
such energy.  That's why the idea of having an experienced leader is so
good - that person can direct the energies of interested people,
propose needed tasks, etc.  I know better than to try and code for
Maxima, so I'll be restricted to documentation and testing (hopefully
useful tasks) but there may be some people who can start out by
handling small jobs, which can be checked over by the experienced
coders, and then critiqued until everyone feels that the code is worth
including.  We could stress that while we appreciate all efforts to
improve maxima, certain standards must be met before it will be
included, and work with people as long as they are willing to work in
order to make their code usable in Maxima.
 
> I am quite familiar with undergraduate projects
> here at Berkeley, and getting such a project to
> really work typically requires a substantial additional
> effort by staff or graduate students or faculty.

True.  However, if well defined minor tasks can be assigned, we still
might get some useful work done.  People must learn to walk before thay
can run.

> If someone really wants to make something better, I
> think he/she should make a proposal to this list and
> let us all look at it.  The guiding principle should
> probably be, "first, do no harm".

Absolutely. But by the same token, we should be very willing to listen
to such proposals.  We need interest.  If we get a lot of proposals
that aren't so good we might start a document detailing them and why
they aren't good ideas.  That in itself might be very educational for a
lot of people.

>   If all or some of us think it
> is a bad idea, but someone still wants to do it, all
> that is needed is for the author to keep the code in
> a separate file which can be loaded on request on top of the
> system.  Note that almost any piece of code can be
> replaced in a lisp system merely by loading a file
> that redefines functions and data.  So the main
> point would be to keep two people from independently
> implementing the same new feature.  Is there a "to-do"
> list concept in sourceforge?

Yes there is, although I don't have too much experience with using it. 
(Anyone?)  It is called the Task list.  I've set up a basic example
here: http://sourceforge.net/pm/?group_id=22406



__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com