Sorry for the catchy phrase/title. I find TIOBE a lot of fun, largely because we usually can say that it's a pretty bad way to measure anything. With per-monthly huge changes in the "interest" - I mean, these are not really realistic now are they? But ... all the criticism aside, I think that TIOBE, plotted over the longer range, narrowed a bit, +/- 2 percent or so, actually should be somewhat close to the "real" data, whatever that is. In this context, probably people being interested in the given language, and possibly also using that language (primarily, though of course people can use lots of different languages at the same time, like combine C and python or C++ and ruby). Here is the link to TIOBE: http://www.tiobe.com/tiobe-index/ Snapshot February 2017, just the relevant ones, that is the "scripting" family of programming languages: 5 Python 4.043% 6 PHP 3.072% 7 JavaScript 2.872% 10 Perl 2.171% 11 Ruby 2.153% There are also some other sites; google rank chart, well, the trends. Also not ideal for my search terms, due to some terms being unrelated to programming, but I think the overall trend may be somewhat similar: https://trends.google.com/trends/explore#q=Ruby%20... The latter chart also has a somewhat similar trend, with python gaining the most, PHP losing a little bit, javascript growing slowly but steadily. Ruby being about constant, and perl losing a tiny bit - in about the last 6 months or so. So I think that, from within the "scripting" families, perhaps we can learn a bit what python does better. I do not know python reallyso I have no real idea. One difference may be the docs, python probably has better ones, but I don't think that this can be the only explanation. Anyway, there is another reason why I mention the above. I think that the "scripting" programming languages fill just about the same niche. Someone else wrote this on his blog some years ago and I agree. While there are many individual differences, I think that in the long run, the family of scripting languages up there, is very, very similar. We can also include lua and R. And actually, I kind of did that - just summing up the above. That way we reach about .. 16% or so. Hmmmm. Can this be a fair and accurate distribution or estimate? Java alone at TIOBE sits at 16.676% so I think we can assume that there must be a whole army of Java hackers right? I mean, 16% is not bad at all. Perhaps we could also count swift and nim to it, partially, even though they are not quite scripting languages; but at the least nim has a "scripty" touch and swift also has a semi-scripty touch. Also crystal - it has a ruby-like syntax though it is compiled. Elixir too. But even then... we can reach here... 20% or so? Are the "scripting" languages so bad or are the compiled languages just so much more prevalent and dominating? It would be kind of nice to compare the trend, say, in 3 years and see whether anything has changed. Or at the least I find it interesting, no idea if anyone else does find this interesting. :)
on 2017-03-03 23:08