This is shamelessly linkjacked from here:

http://www.tiobe.com/index.php/content/paperinfo/tpci/index....

But in any case, this one's worth a look too: http://langpop.com :-)

Thank you for pointing this out. It is almost exact copy. I am not associated with the blog, but I just found the link from my Friendfeed feed.

This reminds me a of a joke: A boss walks into the accountant's office and asks, "How is the data looking?". To which the accountant replied, "How would you like it to look?"

And reminds me of the Hillary Clinton primary campaign on how it kept redefining "winning".

Sadly, the methods are very stupid (see http://www.tiobe.com/index.php/content/paperinfo/tpci/tpci_d... ). They use YouTube as a search engine (eh?) when something more specialized like DZone or SourceForge or even delicious would be better. They use only one query and counting mentions of it, hardly a measure of real-world popularity.

Some better metrics: number of new projects written in a language. Number of servers hosting sites written in a language.

Anything really. Google hit counts doesn't cut it.

Though I find the TIOBE index interesting, the error margins are huge.

The top 5 seem fairly accurate. Considering the error margins, you'll notice there's a massive drop-off after position 5. (#1 - 21%, #5 - 9% and #6 - 5%). You can safely forget about accuracy after that.

I am a little surprised at C#'s low rating.

JavaScript's position seem reasonable - not everyone is building web apps. I do a ton of JS development myself and we're using it to squeeze every last drop out of the browser. Recruitment efforts on our end for this tend to focus on keywords like C, Haskell, LISP, assembly. Most JS positions in the market are for DOM manipulation and AJAX responses. I keep my CV up-to-date, but don't emphasize the JS aspect.

The 10:1 ratio of JS to ActionScript seems a little off, but I guess that's because most job ads look for 'Flash Developer' as opposed to 'ActionScript Developer'.

Delphi is up? Uh, what? And PowerShell is in the top 20 - higher than bash/csh/ksh and even higher than Lua? (I know a lot of game developers - Lua is fairly big in that sector).

Don't take lists like this as a guideline for your personal development. They're an indication of the current state and an error-prone one at that.

That said...

I think you should all be reasonably comfortable with C, Java, C# and two of Ruby, Python or Perl. And then throw in one or more of OCaml, Scheme, LISP, Haskell, what have you.

Oh, and assembly of course. Pick an architecture but get down and dirty with the a malloc implementation or two.

In closing, language popularity doesn't mean much.

Notice how highly "Visual Basic" places in the index. This is silly because if you've looked for a technical job using the conventional avenues (you aren't a demi-god fending off job offers left and right) recently you'll know that no one uses pre-.NET Visual Basic anymore. Most .NET shops use C#. VB.NET is not a popular programming language in professional software development. The reason it's there is because there's tons of legacy content from the late 90s and early 2000s before everyone switched en masse to C# or Java.

Really the TIOBE Index is pointless and just serves as a way for people to come up with phony "numbers" that lend credence to what they want to believe.

More important than total popularity is the popularity in a particular category. What's the most popular language for a certain type of task? C is certainly not exactly ideal for quickly setting up a dynamic website, and no sane person is going to use something like Java to write a video encoder. A better question would be, when there is a reasonable choice between many languages for a given task, what is the relative popularity of those languages?

If PHP remained at the same place, and Ruby went up, JavaScript, which in most cases comes along with those two languages should have remain at the same position and not going down.

Let alone, Python and C# which in many cases used for web development.

It took me a couple of times through to grok your intention...but I agree, JavaScript is extremely underrepresented in almost every language popularity study I've seen. It's so pervasive that everybody needs to know it. We build systems management software, and I've had to learn JavaScript. It's definitely not becoming less important over time...the "View" in MVC is eventually going to be nothing but JavaScript (by some definition of "view" and some definition of "MVC"--probably technically more accurate to say that there will be an MVC, with a lightweight MC, in JavaScript and an MVC, with a lightweight data structure V, on the server and they'll communicate via JSON or something). Anyway, Yegge is sorta fruity on some things, but I think he was spot on when he called JavaScript the "Next Big Language". We're only a few years into development of a "library culture" in JavaScript, but it's already obvious, to me anyway, that the future of the UI for most applications is JavaScript.

There is a nice article on MVC<->MVC where you have your MVC server setup like rails or merb and you have an MVC in javascript as well. Good read: http://welcome.totheinter.net/tutorials/model-view-controlle...

Speaking of the "library culture" of Javascript, JSAN has been receiving more and more attention these days.

Someone should calculate the stats for mentions of different programming languages in HN submissions (titles) and comments. (But without getting noticed the wrong way by YC.)

LangPop looks at reddit (and some other sites too), which has a bit broader user base than HN.

Go Perl, go!