Re: [Visualpython-users] new files

[Arthur <ajs...@op...>]

> Would the Python people be uncomfortable/unhappy if we made such a bundled
> thing, containing a stripped-down Python, or even a full Python? Right now
> there isn't even a copy of Python at vpython.org, just a link to a file
> housed at python.org. That link gets you an official Python, complete with
> libraries, etc. (and documentation in the Windows case).

I honestly don't think so.

This is not at all unusual.  When the clear intention is to focus on a
specific application - like VPython - I see nothing controversal about
creating an environment appropriate to that focus.  I see it as within the
intentions of the Python creators and maintainers. The Python license is,
IMO, specifically designed to accommodate this kind of distribution.

The hotter political issue is, I suspect, not including IDLE in a standalone
VPython distribution - as VPython is sort of part of the history of IDLE, as
the move toward the merge of IDLE and IDLEfork proceeds.  Perhaps the
"official" VPython standalone should include IDLE.  SciTE for VPython could
be available as a separate download for those who might want to give it a
whirl.  On the contributions page - from me.

As to docs, my idea is to keep the VPython index.html as is, except to point
the Python help to the online url at python.org, rather than to a local copy
of the docs.  It assumes online access, of course.

As a general statement -

I aan't help  wondering why you feel this concern about Python in the
proposed distribution, and not about Numeric, in the current distribution.
The situations are fully analogous, in my mind.

As I've said before, its the "neither fish nor fowl" that might raise
issues. It seems to me that a stand-alone distribution, that is clearly
understood to be a stand-alone environment - handles, rather than escalates
these issues.

All this is admittedly one guy's take on it.  Others might have reason to
disagree.

>
> A practical issue is that the download traffic from vpython.org would
> increase significantly due to the much larger installer containing Python.
> This isn't necessarily terrible but I should opt for a higher download
limit
> from that site. (Recently traffic has grown to where some increase is
going
> to be needed in any case.)

The VPython standalone setup.exe (which includes ScITE) is about 3.5 megs -
the current VPython about 1.6 megs.  Your call about how significant the
traffic issue is. Personally I think VPython should be more ubiqitous than
it is.  Find it hard to see this as a big issue at what I expect are the
current levels of trafiic.

>
> It would be important for vpython.org to be very clear in its
instructions,
> something along the lines of "If you don't have Python, install this
bundle.
> If you already have Python, here's an add-on package."
>

Plan2 -

I am - no surprise - working toward a standalone PyGeo implementation. It,
of course, will include the full VPython.  I guess you could just point
folks who might like a stand-alone environment towards it, instead of having
anything at vpython.org.  But, of course, I will be trying to highlight
PyGeo as well as VPython, in such a distribution.

Art