5 ms·
I have to question how valuable it is to include stack overflow in equation. It implies that all languages suffer the same rate of problems for which people nee
by defertoreptar 8y ago
I have to question how valuable it is to include stack overflow in equation. It implies that all languages suffer the same rate of problems for which people need to seek out help.
- ne9xt 8y agoagreed. interesting data point, but maybe not worthy of an entire axis.
- axaxs 8y agoIMO, it's extremely reckless for this type of ranking. Most people who only seek 'coding' as a means to an end end up in JS, which would lead to an explosion in that section from folks still learning. Most ASM and to a lesser extent C programmers see it as a last resort...at least I do.
- hnra 8y agoThe graph shows SO vs GH rank, and if I'm reading it correctly it seems like C is slightly more popular on SO relative to GH while JS falls pretty much on the line. Wouldn't that imply that C programmers resort to SO more often than they publish code on GH compared to JS?
- axaxs 8y agoSure, I was a bit surprised by that. But, there are things to consider. JS is by nature(well, until Node) open source and has a generally younger base, both which would pad its GitHub numbers. C has two big areas I think - Linux/open source software, and embedded work. A lot of open source projects in Linux land are for whatever reason not in Github. Heck I still see sourceforge a lot. Embedded work is likely done for a company and of course wouldn't be in Github.
- raz32dust 8y agoWhile not 100% reliable, more questions in stackoverflow does indicate higher probability of the language being more popular, if you think in Bayesian network terms. And they call out all such caveats very clearly. Any "ranking" of programming languages is going to be flawed in some way or the other. Using only the github data might skew it towards more recreational languages and hide the enterprise languages, for example.
- Insanity 8y agoNot to forget that many students will use stackoverflow to ask questions. Which means that languages being taught at uni have a higher ranking (Java, Python and JS come to mind). both GH and SO are probably not representative of the industry.
- ksec 8y agoBut for things like Rails, most of the questioned were answered already and likely wont grow in any place. I wonder if PageView for questions ( How many people were checking in for each PL ) and a mix of number of questions would be a better indicator for popularity of languages.
- Gatsky 8y agoI also think there is some correlation between how easy a language is to install and the number of questions on stack overflow.
- flukus 8y agoJust because something is tagged with $lang doesn't mean it's an issue with $lang, it can be a library specific library in $lang or "how to do X in $lang" for someone new to it.
- H1Supreme 8y agoI'm building internal apps with Javascript (Vue + vanilla) and Go. Both Vue and Go have good documentation. As a result, I'm not using Stack Overflow nearly as much as I once did. With Go especially, my default move when I have a question, is to go to golang.org/pkg and find find the package I'm having issues with. Assuming it's a standard package problem, of course. The built in editor for examples is really great. I can run a little test quickly, and move on.
- robdachshund 8y agoThey also included CSS as a language... which is a "stylesheet language." HTML isn't included, so why is CSS? Neither are programming languages. You could make the case for SCSS and things like that, but it just says CSS. Meh.