For example: C++ programmers apparently don't mention SQL at all, while it's very popular with PHP (which doesn't have a built-in ORM). There is also no overlap between C++ and JavaScript programmers.

Rust is obviously very influenced by C++ and Haskell, but the C++ community doesn't even know about its existence. Somewhat naturally, the Matlab and PHP communities really don't have much in common.

Not that surprising I think since an ORM is probably better in a library since that allows more flexibility across systems.

http://en.wikipedia.org/wiki/Microsoft_Dynamics_AX#MorphX_an...

edit I guess not, since it's a nuget package now. Maybe LINQ to SQL could have been considered a built-in ORM?

It might be more accurate to say LINQ is built into the languages, but LINQ to SQL is a .NET platform feature, not a language feature. (Of course, the languages are so tied to the Microsoft distribution and the .NET platform that distinctions between language features, language standard library features, and platform features are, for .NET languages, somewhat academic in practice [inasmuch as there is a practical difference that is likely to matter, "does it work on Mono" is probably more important than strict language/library/platform distinctions].)

[Edit: Sorry, I was dragged away to a meeting before I could finish.]

Specifically, what is the blue language between C# and Haskell? It has a strangely strong link to Rust and Python.

This is answered by the mentions relative to TIOBE graph. They use VB, but they are careful not to talk about it.

Which is just another word for "conspiracy."

That said, I don't think the findings for C was shocking either.

You're not going to be writing kernel-modules in Clojure. It's going to be some sort of C. And the chances of it involving hardware (performant code or not) is significantly higher than other non-systems languages.

In addition to the reasons you mentioned it's also possible their are other demographic differences among languages in terms of age, experience etc. (I suspect though can't prove that the average C programmer is older than say the average PHP programmer).

"better" is subjective. While there are good things about the Java/JSP stack, its templating language is IMO cumbersome and unsafe, and its test-edit cycle can be quite long.

There are also MVC frameworks for PHP. The most popular, though, are those for scripting languages like Ruby and Python, namely Rails and Django.

What you shouldn't be doing is interleaving data access, business logic, and presentation all in one file, which is what straight PHP and JSP usually consist of. This is called "spaghetti code" and it's been widely regarded as a really terrible practice for around 10 years.

You can do client-side apps in Javascript that fetch JSON or XML from the server, but you want to be using a framework with some structure to serve the data, still not just straight PHP or JSP.

"spaghetti code" actually has a much broader definition than that particular kind of lack of structure (though I agree that that example might be viewed as a fairly modern example of "spaghetti code") and has been considered a terrible practice for much longer than PHP, JSP, or even the web itself has existed. [1]

[1] http://en.wikipedia.org/wiki/Spaghetti_code

I still agree that PHP is probably the fastest way to get a couple dynamic HTML pages that include server data. The limitations of PHP as a template language quickly reveal themselves as projects grow.

Considering that would be considered a growth path of a developing programmer, I find that fascinating.

PS: Go will always have problems with stuff like this. They should have just named the language string or var or something equal toxic to machine collection.

Then again marketing people would have made choices like RadPython, RubyGrunge, and CFlowerPower...

1. Naming things 2. Cache coherence 3. Off-by-one errors.

Nobody's ever going to find them through Google, but oh well. Nobody was going to use them anyway. May as well name them the way I like, right? ;)

This is very exciting to me, because I believe Rust is going to be a great replacement for C/C++, and if anyone understands how to create quality programming languages, Haskell programmers do.

TL;DR Rust is going to be the new C++, and the new Lisp. These are exciting times.

e.g.

"I think function application is super easy, you just do f(x,y,z)!"

vs.

"I think Haskell style function application is super easy, you just do f x y z!"

But I also suspect Haskell is just getting more attention these days -- it's quite trendy for sure.

    Forum X spends a great percentage of their time
    talking about Y than visa versa so it's colored X

Kind of a spinoff of Stroustrup's, "there are languages everybody complains about, and languages nobody uses".

clojure users seem very happy with what they use!

On the mutual mentions java users have near zero mention of haskell?

Java mentions Haskell quite a lot, Haskell mentions Java almost not at all.

PHP offers a wealth of creative possibilities in that regard which other languages seem to lack ;)