Python Popularity: Questions and Comments

[Paul Prescod]

Arthur Siegel wrote:
> 
> Paul Prescod writes -
> 
> >When one of Python's killer apps takes over the universe (Zope, Pygame,
> >Alice, Mojonation whatever), Python will take over the universe.
> 
> I am confused by the discussion only to the extent that it is my
> perception that Python slower but surely *is* taking over the universe.
> Steady as we go.

I'm concerned by the historical precedents. What language have you ever
heard of that took over the universe slowly and quietly? They tend to do
it in a big bang, following some trend like Unix or the Web or mobile
computing. If there is ever a "big bang" associated with a language in
Python's class, Python could be overshadowed. For instance if PHP had
been better designed as a general purpose programming language, and had
had support from a big corporation (e.g. Sun), it might be the "default
language" in the scripting world today.

Once a default language takes hold (e.g. C) it becomes quite difficult
to convince bosses to allow us to use other languages (e.g. Eiffel, Ada,
Sather).

> Precisely because of the reasons you point to. It is wonderfully general
> purpose, with an appeal to such a wide audience needing to solve
> widely divergent problems, and coming to *their* problem with widely
> divergent backgrounds and experience.

That isn't really the way to popularity. The way is to solve the problem
that many people need to solve right now. People need to be able to tell
their bosses that they are switching language because the new langauge
is demonstrably better at solving some problem than the one the company
has standardized upon. "I need to do CGI. You cannot do CGI in C.
(cough) You have to do it in Perl. I need to use Perl."

> The only misstep I can see at this point is to skew Python's development
> in a particular direction to solve a particular problem for a particular
> segment
> of its potential audience -  whether it be the novices, biochemists,
> mathematicians
> or  web application builders.

I would also advise against this, because it is essentially impossible
to know what horse to bet on. Most languages fall into their domains,
they aren't designed for it. Java was "supposed" to be for set-top
boxes, and then for applets before it became the default server-side app
language. Nevertheless, it is clear from history that being the default
language for an up-and-coming platform or application class is a much
more surefire path to success than being a really good programming
language.

 Paul Prescod