Home Original page

I (don' ) will kill my Python

Tim Peters tim.one at home.com
Sat Jan 20 17:35:15 EST 2001 
[rturpin at my-deja.com, recites various good reasons for why commercial
 users want to package Python with their apps]

> ...
> Once you start thinking about this, you quickly believe
> that the sensible thing is to create a new Python distribution
> on each platform, regardless of whether one already exists.

Yes, and that's what people *do*, too.  Nothing is stopping them.  To the
extent that the core Python distribution doesn't actively *help* them,
that's a reflection of how seldom anyone who goes through this pain
contributes their work back to the project.  BTW, there are many reasons for
people not giving the work back, such as:

    lack of time
    they didn't think it was hard enough to bother
    they thought it was sooo hard that they gain a competitive
        advantage by not sharing their work
    they had unique needs they don't think will make sense for
        others, and don't want to bother ripping that out or
        explaining it
    they had simple needs and don't want to bother trying to
        become all things to all people
    they're waiting for the Python Distutils SIG to solve it
    the corporate lawyers got in the way

Those are some of the reasons I've personally heard.

>> .. but they [businesses] can afford to hire Gordon.

> Is he job hunting?

Gordon's a contractor -- he's *always* job hunting <wink>.  To judge from
his relative silence lately, though, he's currently got more business than
he can handle.  In the meantime, there's also nothing stopping anyone who
cares from taking his installer work and updating it to 2.0 (or, come
Monday, 2.1a1).  Also nothing stopping people from looking at how, e.g.,
PythonWorks-- or even PySol --package Python, and copying their schemes.

Heck, I have a friend who built a highly specialized Python app for his
father, and packaged it all up in one afternoon, starting with *no*
knowledge of how Python is structured:  he just kept copying Python library
files into a work directory until the ImportErrors stopped.  Low tech but
effective, and it probably would have taken him just as long to figure out
how to *use* a all-things-to-all-people packaging tool.

it's-not-like-a-business-ships-a-new-release-60-times-each-day-ly y'rs
    - tim
More information about the Python-list mailing list