Python's popularity statistics

[Aaron K. Johnson]

In message <MRLN246641E232 at merlin.envox.local.hr>, "Hrvoje Nezic" wrote:
> > I said the same thing elsewhere in this thread. I cited Perl for its,
> well,
> > plain terrible design (I wasn't specific, though---mainly, it's a good old
> > syntax problem). I wouldn't imply popularity=quality, ever. Look at pop
> music
> > compared to Chopin,Bach, Bartok and friends, for example. (Yes-I'm a
> classical
> > music snob). And I would say more often than not it indicates lack of
> quality
> > when something's popular. But we have to be careful.
> >
> > Python is that rare instance of a language which deserves popularity (it's
> the
> > Beatles of programming languages).
> >
> > For me though, programming language popularity does say something of its
> > usefulness and power, if not of its difficulty.
> 
> I don't think so. I think in many cases "big players" (M$, Sun, etc.) are
> pushing
> some languages, and this is why they are so popular. If C# was invented not
> by Microsoft, but by someone who is not a big player, I don't think it would
> achieve popularity it has.

Yes, that's true. The almighty $ helps to give a language leverage. But I would
say its not entirely true. Sometimes a language being free is why its popular
(perl, python)--in fact, its remarkable that perl and python are as popular as
they are without a major corporation pushing them. I think that what you say
apllys mostly to Java and Sun--- C# is a corporate extension of a pre-existing
ubiquitous paradigm, and from what I can tell, is only being used by dedicated
microsoft developers. And let's not forget that Apple (a big player) developed
Dylan, which went nowhere fast.

> Availability of libraries and tools is just a consequence of popularity, but
> of
> course, it has an effect of positive feedback to further rise of popularity.

That's a good clarification of my point.
 
> I don't think that C or C++ are powerful languages, especially when compared
> with others. Of course, the question is what we mean by this. In my view,
> real object-oriented languages are powerful, but C++ is not one of them.

If power is flexibility, generality, and performance, I disagree; If powerful
means ease of use, quick development, and beautiful syntax, I agree.
 
> When I mentioned languages that are not popular, while they deserve it,
> I meant languages like Eiffel. Eiffel beats its competitors, like C++, Java,
> C#, Smalltalk and others in many respects. It is a state-of-the-art
> object-oriented programming language. It has genericity, design-by-contract
> (assertions: preconditions, postconditions, invariants), it has genericity
> (constrained and unconstrained), it has multiple inheritance, carefully
> designed and implemented, it has elegant and clear syntax, it is *pure*
> object-oriented language, which means it has consistent and uniform
> treatment of all types, basic and others, and it is very efficient.
> All these characteristics means that it is very powerful.
> Yet it is not popular.

I don't know Eiffel (I've heard of it). Besides having the technical merits you
mention, does it come with batteries included (meaning a great set of
libraries?) Most languages that are touted as being clean and beautiful in an
academic sense never catch on because their libraries suck, or they have no
libraries, or too many implementations to have 1 standard (like scheme)

 
> I think that languages with static typing and dynamic typing belong to
> different categories. I wouldn't use Python or any language
> with dynamic typing for really serious and critical tasks,
> like writing software for the rocket Ariane, or for banks,
> where on launching the rocket you could get run-time errors
> which would never pass compile phase of languages with static typing.

Interesting, though, is that NASA uses python all the time. I don't know
whether it's for critical stuff though.

Best,
Aaron.