On Wed, Apr 9, 2014 at 3:19 AM, alister <alister.nospam.w...@ntlworld.com> wrote: > On Wed, 09 Apr 2014 01:48:52 +1000, Chris Angelico wrote: > >> I managed to convince him to let me use Pike for a lot of the work, >> though I suspect that - now that we're no longer working together - >> he's ripping a lot of it out in favour of either PHP or JavaScript. >> And that's a job I wouldn't wish on my worst enemy. >> >> ChrisA > > Why was your code that bad? ;-)
It was idiomatic Pike code, which is in many ways similar to idiomatic Python (definitely not exactly, but I'm guessing you're not familiar with Pike, so just pretend I'd written a good few KLOC of idiomatic Python). The code is expressive; in just a handful of lines, I can do bulk operations on entire arrays of values, and I can comfortably run callbacks efficiently, and (this part wouldn't work so easily in Python) on a SIGHUP, the code would smoothly switch over to the latest version on the disk, without breaking any currently-processing work. Now try rewriting that in PHP (or serverside JavaScript - probably node.js or something - but he hadn't gone anywhere beyond the "vague theory" stage with server JS, so PHP is more likely). Bear in mind that you have to handle multiple concurrent network connections, so you need to do either piles of threads/processes or asynchronous I/O (the Pike code used the latter). So, yeah. My code would make for terrible PHP. :) One day, out of morbid curiosity, I might try SSHing into one of the boxes to see whether any of the accounts are still there for which I know the passwords. (I was the primary network admin, so I knew the passwords for a lot of obscure accounts on obscure servers.) Would be interesting to see what's going on there. But more than likely the project's dead in the water. It's highly unlikely he'll find a decent programmer willing to work for the low wages I was getting there, and even less likely that he'll have the time to do it all on his own. ChrisA -- https://mail.python.org/mailman/listinfo/python-list
