When discussing programming language popularity, most people point to the tiobe index. But personally I think there are some serious flaws to their method, the main thing being that they arbitrarily decide which languages to track and which to ignore. I’ve been playing around on stackoverflow lately and noticed today that if you hover on a tag, the site tells you how many followers that tag has. You can also see how many questions are asked under a partucilar language tag. So I decided to create a list of popular languages (or dev environments) on their site and how many followers and questions each languauge has. I included ColdFusion at the bottom for reference.
Name
|
Questions
|
Followers
|
C#
|
149,600
|
23,015
|
Java
|
100,276
|
19,136
|
PHP
|
89,720
|
17,279
|
Javascript
|
77,954
|
13,432
|
jQuery
|
66,446
|
13,618
|
iPhone
|
64,943
|
11,086
|
ASP.Net
|
64,426
|
12,812
|
C++
|
59,445
|
13,983
|
Python
|
49,209
|
16.142
|
Android
|
43,304
|
9,983
|
…
|
…
|
…
|
ColdFusion
|
1,906
|
1,202
|
Is this scientific? No, but it’s interesting. That said, I do think it’s more scientific that tiobe’s, which is based on a secret search results based algorithm. This at least is based on data from a real programmer Q&A community, and a very popular one at that.