Python's popularity statistics

[Aaron K. Johnson]

In message <N5yK9.3539$4W1.2012 at nwrddc02.gnilink.net>, "Raymond Hettinger"
wrote:
> Theory:
> > For all who are interested, I wrote a couple of python scripts that helped
> > extract newsgroup activity from the comp.lang.* hierarchy, and this can be
> a
> > rough guide to the popularity and usage of the big programming languages.
> 
> New Fact:
> This thread has generated a lot of activity on c.lang.py.
> 
> Conclusion:
> This thread has made Python more popular than ever.

All joking aside, it would be interesting to collect data on the number of
unique posters to a number of different comp.lang.* groups over say, a one or
two-week period. We could safely assume from the law of large numbers that a
fairly accurate estimate of the relative (no absolute number is ever possible)
user base of each language could be established from that, no? Unless you
somehow think that different languages for some reason have different likely
percentage of users to posters....my hunch is that the normal distribution
curve applies, i.e. its not likely, although not impossible, that different
langs will have radically varied posting ratios.

Someone with more knowledge of statistical interpretation of data than I might
suggest a revised technique. For instance, can we do a 'student's t' test on
the data, so that the data itself point to their own distribution?

Anyway, increase of thread activity is probably random, and balanced by periods
of quiet activity. Which makes me think you missed my point. Anyway, i wouldn't
ever say that increasing an effect increases a cause (i.e. Python is popular
increases newsgroup activity; the reverse doesn't apply.....)

-Aaron.