OK, so Ohloh has fairly junky stats.
BUT, for the subset of the overall numbers they do sample, their RELATIVE trends should at least be fairly useful as a source of information.
While doing a little comparative graphing, I uncovered this interesting gem (no Ruby pun intended).
(As you can see, I also uncovered tinyurl's awesome vanity tinyurl feature)
Now, lets be clear what this graph shows.
Ohloh defines a "contributor" in these graphs as the number of distinct accounts who did at least one change to a file written in that language in that month.
The Ruby is there mainly as a curiosity, either it's jumped the shark now, or (more likely I suspect) a great big chunk of users has moved on to the Ruby'tastic github (which Ohloh doesn't know about yet, right?) this year.
More interesting are the two quite different trends.
One pair would seem to link the shell community with the Perl community, and the other links the Python community (apparently) to the PHP community.
Now CLEARLY (er, right?) the PHP and Python communities aren't inherently related as languages but they do represent the "new web/apps" to some degree.
And Perl and Shell may be somewhat related (although not necessarily tightly) but they are more accurately linked as the "sysadmin/command line" crowd.
Perhaps there's something going on here beyond just the languages themselves...
One positive note is that compared to the other languages in Perl's class (shell, c, etc) we've had a pretty good year, maintaining strong contributor levels when there's been some off crashes in the numbers for C, shell, Ruby etc.
Of course, being a egotistical bastard, I would prefer to attribute the detaching from our related languages to someone releasing a CPAN-capable Windows Perl distribution and drawing more non-Unix people into the maintainer pool.