Without some additional information, these numbers are meaningless. For example, what is the margin of error for this poll? How were the developers selected? Was the sample sufficiently "random"?
One of the most difficult problems of basing predictions on polls is making sure you have an unbiased sample. Two American presidents were elected, despite polls that showed them well behind their opponents (Franklin Roosevelt and Truman).
And besides all that, I really don't care much about who may or may not be using Perl. I like it, and I'll be using it.
|