Re: [Visualpython-users] History and Status?

[Bruce S. <bas...@un...>]

Excellent, Arthur! Many thanks for volunteering to work on these issues. The
excessive use of "I" and "me" in some of my postings was due to the fact
that no one had previously offered to contribute to the task of building
installers, whether to their liking and needs or to my liking and needs. And
I felt defensive of the needs of my students because I had seen little
understanding in the Python community that not every potential user of
Python, even Linux users if they are new to Linux, is comfortable with the
following incomprehensible statement from the download page at the Python
site: "All others should download Python-2.3a1.tgz, the source tarball, and
do the usual "gunzip; tar; ./configure; make" dance."

It is already the case that http://vpython.org has a growing set of links to
the contributions of many people (including your own PyGeo), and one
feasible scheme is that as you create new and better installers they can
live at your site with links to them (with those links increasingly
replacing the current installers). Alternatively they can be moved to the
VPython site and stored there -- whichever would work best for you. FYI:
VPython is housed at pair.com, a very good and surprisingly inexpensive
host.

I'd also love to see a consensus on improved placement of the files. I'm
quite certain that I made wrong choices, especially in the Linux case,
partly out of ignorance and partly because there has been something of a
moving target in the last few years about community expectations of where
various pieces should go.

As you work with installer schemes, perhaps you could treat the needs of my
students as a kind of superset of your needs. That is, there could be two
distutil schemes that differ only in that one of them has some extra stuff
added to make a single bundle.

Note that while the Visual source code and the demos are in CVS at
sourceforge.net, the html reference manual is not but could be if and when
others would like to contribute to its care and feeding. Currently it's just
a zip file at http://vpython.org.

Bruce Sherwood

----- Original Message -----
From: "Arthur" <ajs...@op...>
To: "Bruce Sherwood" <bas...@un...>; "John Keck"
<joh...@ho...>; <vis...@li...>
Sent: Monday, February 17, 2003 9:11 AM
Subject: Re: [Visualpython-users] History and Status?


> [Bruce]
> > Since so far Arthur is the
> > only person who has complained about this, I haven't been able to make
> this
> > a priority, as it does involve extra work. Until very recently there has
> not
> > been sufficient people power to address issues such as this, but I
> certainly
> > don't rule out providing more installation options in the future.
>
> If there is to be an open source community approach, then there has to be
> some sense of openness to community based decision making.  I am willing
to
> undertake responsility for creating and maintaining a distutils
> installation - source and binary, Windows and Linux - for VPython as a
stand
> alone distribution.  The fact is that I am so unhappy about the current
> distribution, that I was pursuing this in any case, and was going to make
it
> available for PyGeo users. Though I would much prefer to have it on the
> VPython site.  And I have no objection to offering PyGeo users the
> "all-in-one" distribution that you prefer for the strudents you focus on
as
> your prime concern.  Though there would have to be some coordination to
> assure that the distributions are consistent -  in terms of where files
are
> place,d for example.  I  have a sense of what I would call general
> expectations within the community of Python users of how a third party
> module should install (and uninstall).  It is important to me that those
> expectations are met. Because I think that the general community of active
> Python users is an important audience for VPython.
>
> So I can help you solve the manpower issue as to the distribution, but
only
> assuming "we" - and I think the we here goes beyond Bruce and myuself -
can
> agree as to what is and is not appropriate in terms of how the
distribution
> behaves.
>
> But let me say directly, I have problems with your wording above. On one
> hand we are talking about a more community based approach to VPython and
its
> future, on the other hand you are talking about what *you* are willing to
> rule in or out. Without a clear idea of what your criteria might be.
>
> If your own focus is narrowly on the constituency of physics students, and
> you understand their needs quite well, is it possible that someone like
> myself who has an idea that VPython could and should have a broader focus,
> go his own way with it.  In other words, offer my own distribution - and
> essentially fork VPython.
>
> It seems to me that should be totally unnecessary and unfortunate.  But I
> also shouldn't need to send you reports of issues as to the current
> distribution by certified mail.  I submit you don't see the issues because
> you have a narrow focus on a particular constituency.  Looking at it from
> the prespective of a broader constituency, ths issues I rasie are quite
> obvious.  There is no good reason why all constinuencies needs cannot be
> addressed at www.vpython.org.  But it is difficult to work cooperatively
if
> you seem to think other people  judgment on these matters cannot be relied
> upon.
>
>
> Art