Python's popularity statistics

[Aaron K. Johnson]

Oops, I should retract my statement that there is a 'strong' correlation
between the two data sets. Change strong to 'rough', in that the exact order is
not observed, but the grouping of 'more used' and 'more talked about' seem the
same in chunks of about 5 languages or so.....

-A.

In message <3DF803BE.5175A7EC at alcyone.com>, Erik Max Francis wrote:
> "Aaron K. Johnson" wrote:
> 
> > Yes...let me be more precise. I'm not saying that there is a
> > coefficient x by
> > which you can multiply the number of posts to comp.lang.python and get
> > the
> > magic number of python programmers. What I am saying is that in some
> > real way,
> > the greater the number of comp.lang.python posts, the greater its
> > overall
> > presence and use. Dig?
> 
> Probably.  Greater activity (non-spam activity, anyway; the raw numbers
> don't give that) probably indicates greater popularity.  But what people
> are rightly telling you is that the relation between newsgroup traffic
> and the popularity of its language may not be the same for each
> newsgroup.
> 
> In other words, even if there's a strong positive correlation between a
> language and its newsgroup traffic, that doesn't mean that the raw
> traffic figures for different newsgroups are comparable in a way that
> correlates to their relative popularity.
> 
> So that, for an example, comp.lang.c++ (I presume that it's
> comp.lang.c++ itself and not its moderated cousin) gets 2.8 times as
> much (raw) traffic as comp.lang.python doesn't really tell you anything
> quantitatively about the relative popularity of C++ to Python, because
> the factors involved with correlating a language's popularity to its
> newsgroup probably aren't (and I'd say almost certainly aren't) the same
> for different languages.
>