Ask HN: Will you support XP when developing new software?
Now that XP has end of life, is it worth considering building software which will run on Windows XP if there's extra development overhead? Anybody care to share their web stats on XP visitors based on regions around the world?
8 comments
[ 1.9 ms ] story [ 30.3 ms ] threadI don't develop for Windows anymore, but if I had to wager a guess, I'd predict that it'll be like the older IE versions: officially dead, but persistently popular, and grudgingly supported.
BCC (very non-technical, international audience with strong Anglosphere focus): XP has 18% of Windows' 65% share. They index lower for actual purchases, but by quick eyeball it is 10% or so of revenue.
AR ("less technical" audience, overwhelmingly US/Canada): XP has 11% of Windows' 60% share. I am unable to quickly calculate their percentage of revenue, but know off the top of my head that my two largest accounts have hard requirements for XP.
This means that I'll probably continue supporting XP for those products, although given that 10% of BCC revenue isn't all that much I won't exactly tie myself in knots to do so.
By comparison, my blog (highly technical audience): XP has 6.5% of Windows' 40% share. That's below the noise floor to me. I don't track conversions on the blog (if I did, they'd be to email signups) but my sense of things is that if I 500ed everybody in that segment I'd be unable to detect any change in my business as a result.
By comparison, the web app has added 25k users over the same interval.
So, percentage-wise, more than 99% of new users use the web app exclusively.