A bit OT, but: checking popularity with indeed.com is a fishy thing. First there is a trend even for simple PHP-only shops to post ads with the phrase "Haskell, Lisp, Scala, Clojure knowledge is a pre" (Yes, the "beating the averages" effect reached recruting years ago)

Haskell as an example: Haskell County healthcare and the company Haskell & White and more like these are occupying the first three pages of haskell results before there is a Java dev ad using the "at least one of { Scala, Erlang, SML, Haskell, Prolog, Lisp, Clojure }" phrase.

Clojure: on the first page I count three real clojure jobs and seven "at least one of ..." mentionings.

For Scala the indeed search turns out to be more relevant.

Conclusion: Get out of my face with indeed.com, their data is next to unusable for PL connected job trends.

The most useful thing we can learn from indeed.com is that they are all baseline noise compared to Java. So you're wiser to let your own priorities dictate your choice and forget popularity at this early stage.

While there are many corner cases which you have outlined, in the grand scheme of things job advert data is a reasonable indicator (and one of the few we have). In the case of languages with limited commercial adoption (Clojure, Haskell) there will be more noise due to the small sample size and low adoption. The indeed.com data indicates that Java and C# are popular languages. Do you believe that to not be an accurate reflection of reality?

What is interesting about the blog is that it also correlates the results with results from polls among developers.

The big three are Java, C# and C(++). But I do not need to go to indeed to know that.

Indeed is used often to show that language a has more adoption than language b. a and b being languages in the non- or pre-commercial domain. I'm only arguing that statements like "haskell overtakes prolog, statical typing > declarative programming" cannot be inferred from indeed data.

As for the "reasonable indicator": if the average java/php job points out that "haskell/erlang/prolog is a pre" , job ads might only be a reasonable indicator for the languages we already know to be at the top.

I assume nothing. I'm only wondering about the tone sneaking in. That aside:

I just notice: <programming language> is on the rise according to indeed.com trend stats.

I follow the link "watch <programming language> jobs"

I see that virtually none of the ads offer jobs where this pl is used.

What do you make of that?

According to my reading of the blog commercial adoption of Clojure is low despite its popularity amongst programmers. Your comments regarding the lack of valid jobs for Clojure supports this argument rather than refutes it. Also, you noted that Scala ads on indeed.com were in fact valid (which the blog seems to argue is being adopted commercially). In both cases the data you observed was in agreement with the conclusions drawn.

I'm just noticing, reading this, that I find "get out of my face with indeed.com" to be more offensive than if you said, "screw indeed.com", even though the former seems less dirty. Language is a funny thing!

Add to that my non-native-speaker-ness ... I will watch my language better in the future. But actually "screw indeed" would have been wrong: there is nothing wrong with indeed: they just aggregate job openings.

If Java itself were included in the poll, then a logarithmic scale would be required :-)

Personally, I really enjoy Clojure development, but Java is still the elephant in the room. Also, even though JRuby doesn't rate as high as the "big three," JRuby is incredible useful for merging work done in Ruby and Java, and for taking Ruby web apps and deploying them on the JVM.

Have they considered what the languages are being used for? Scala with its intensive static typing is more likely to be used for serious systems, while Groovy for quick scripts to test stuff. And then there's JRuby not mentioned, because its lumped under Ruby: who knows what proportion of that is JRuby and what is Matz's Ruby?

I find it unfortunate that we now equate "intensive static typing" with "serious systems." I know of almost as many serious systems written in Clojure as I do in Scala, and I think it's fair to say that they're pretty much polar opposites as far as typing goes.

> while Groovy for quick scripts to test stuff.

I doubt groovy would be listed in job ads just for scripting.

Grails has pretty significant adoption, both among startups, and to a larger extend among the enterprises, e.g. Java Shops.

This is a great summary of the sate of the Groovy / Grails ecosystem: http://burtbeckwith.com/blog/?p=1126

I am curious with Grails 2.0 out for a few months now, and Groovy itself nearing 2.0, whether the adoption will reach an inflection point (and I do anecdotally see more activity of late).

Knowing both (though much heavier experience with Grails), I don't see significant differences in productivity between the two.

For the Java shops, Grails vs Rails, is not really a contest - Grails should win on pure technical merits, although of course it not always does for other reasons.

The poll results indicate that on the JVM JRuby trails Scala, Groovy and Clojure significantly. The indeed.com discussion probably excludes JRuby for the reasons you mention.

Interesting to see how three JVM languages have emerged from the pack. Scala and Groovy proving popular amongst Java developers. Clojure also with a substantial following, but less commercial adoption.