Be sure to check out the poll, the poll’s HN comments, the raw results, the initial analysis, and the initial analysis’s HN Comments. Big thanks to Utkarsh for putting this together, making the results public, and expressing interest in the opinions of others regarding the data.
These were some of the metrics I wanted to see. You’ll notice that all of these are proportional comparisons; y is always out of 100%. This is because there was no consistency in the number of submissions per category, per user. It might be interesting to repeat something like this, but limit each user to one-one-one submissions, and then allow multiple submissions. Then we could really say things like, ‘Ruby devs like VIM’ and ‘objective-c users like xcode’. As it stands, I can say things like, ‘around 75% of activepython users did not indicate any alternative IDE usage’ and ‘fewer than 5% of devs using textpad indicate that they develop in ruby’. Not quite as meaningful, but the comparison tables are interesting (although predictable); and came with a couple of surprises.
I also think it would be good to devote some time to mining deeper to get a little bit more info about which results are coming out of multiple submissions and which stem from single data points (although there are very few submissions that have single values in multiple fields; most devs who submitted a single language, submitted multiple editors - or similar).
Some expected/unexpected (at least by me) results:
Make sure you check out the initial analysis for the actual values, they give the percentages a lot more context. That is, the VIM, Eclipse, JavaScript/Python numbers are likely to be more meaningful (and more accurate), since they have the higher absolute values. Let me know if you spot any errors or have questions/requests.