21 comments

[ 3.7 ms ] story [ 68.9 ms ] thread
Substitute jQuery for Javascript and the graph's trend becomes very different.
Yes, but I tried to focus on Languages, not libraries / frameworks.

The reason I posted it was

1) most people think (at least I did) that Scala has no jobs (well, absolute job count is still low, but it's encouraging to know it's growing fast)

2) my surprise, as it seems that Indeed changed the algorithm of the graph, few months back it was a different picture (Scala was behind Clojure for example)

3) seems the trend is faster than Ruby, which is interesting

Try adding Objective-C to to the chart and then see which looks the highest.

http://www.indeed.com/jobtrends?q=scala%2C+ruby%2C+objective...

Edit: I'm sure there are tons of other examples of this, too, but Objective-C seemed like the grossest omission.

Oops you are right, good point, I am not a mobile developer so only web and desktop crossed my mind (also Indeed already has iOS on top 3 top trends on the front page of /trends) How should I change the title to reflect it? (though I think for non mobile developers it's still an important insight)

Edit: changed the title to be a little more correct, feel free to suggest how to make it better

Android totally breaks the relative graph for all others.
The absolute version of the graph, as opposed to the growth version, shows a different story:

http://www.indeed.com/jobtrends?q=scala%2C+ruby%2C+python%2C...

The fastest growing languages also happen to be the languages with the least jobs. I imagine Scala's growth will tail off a long time before it can compete in sheer number of jobs with javascript or java.

Yep, known and painful, but the point of the post is to show it's relatively growing, and doing so very fast. Many people (such as myself) look at the absolute number and get discouraged, so this is a little good news to Scala fans
Most of the search results for "scala" in my area (seattle) are not actually Scala jobs, they just mention it, as in "experience with scala and 3 other languages we don't actually use is a plus", so not sure how accurate the graph is.
Yeah this is pretty important to note, it seems like most Clojure this job posting are like this too unfortunately.
That chart is statistically meaningless.

It just shows % change since an arbitrary date in the past. Obviously, since different languages are on different parts of the growth curve, it results in a pretty unfair comparison. In particular, it will give very high values for any language that happened to be starting from a very small base at the beginning of the time period under consideration.

log of absolute values would be a much more interesting chart.

ColdFusion, a language derided by many, has more absolute job postings than Scala, Clojure, and Groovy:

http://www.indeed.com/jobtrends?q=scala%2C+coldfusion%2C+clo...

(Of course, the obvious growth trend shows that won't always be the case)

This means .. absolutely nothing. The original linked chart showed little more. I doubt Scala developers are excited by the day they finally catch up with ColdFusion. The relative trend chart might lead one to believe that they'd be better focusing on Scala than Java, even though there's about 100X more jobs postings for Java. (per Indeed's data)

(comment deleted)
Good point, those percentages are people with those skills, so it means Clojure has higher supply growth than Scala, and (according to Indeed) Scala has more demand than Clojure (both in absolute numbers and in relative growth) As much as I love Clojure, this moves the needle a bit toward Scala regarding where I should put my focus on.
No, this numbers are year by year change, so putting your efforts on Clojure seems the best option.
Still if Clojure has higher growth than Scala in people knowing it, but lower growth than Scala in companies wanting to hire people knowing it, then sadly I still have to disagree. Is my logic completely flawed?
Sorry, I don't think this is a 100% correct logic, but may be the grow in LinkedIn shows some information that is not gathered at Indeed and the reverse is also true.

LinkedIn shows that more people were adding the Clojure skill, it can be a new skill within people who was working on a company for long and it doesn't appear on Indeed. At the same time Indeed captures recruiting alternatives but when the job was published (not on alternative and important channels like presenting an acquaintance)

Since the length of the sample is small the variance is big, so the numbers obtained for the conclusion are noisy, (random effect has a big impact).

Sea reference hear for some comments are the law of small numbers, the most dangerous equations and the like http://news.ycombinator.com/item?id=4893258