@bstacey@brainblasted (mediawiki is hard to reason about, especially when you add a lot of traffic to the mix - or at least the parts i have to figure out who to beg for help with after they break in production are hard to reason about.
that said, it's also built by a lot of smart & experienced people, and it's not the kind of doomed, eldritch horror i used to encounter taking one-off craigslist webdev jobs or working on haphazard forks of early 2000s e-commerce platforms.)
@legoktm yeah. i'm not sure if anything is too important to care about code quality, exactly, but above some baseline threshold of possible-to-improve, maybe code quality usually isn't really the problem except in a narrow mechanical sense, much the same way that language choice is rarely the real problem.[0]
a heretical thesis i should probably explore more.
(of course, the "narrow mechanical sense" makes up most of my job, but still.)
@danielhglus@brennen@bstacey@brainblasted shrug, MediaWiki's impact on the world is too great to care about code quality tbh, it certainly has its warts, but at least we can slowly make it better.
@brennen jumping off the Rust bandwagon for a second, I totally agree with your take on language choice. Your post reminds me of what someone from Etsy said (paraphrased, possibly misremembered) at one of the Wikimedia dev summits: People only pick PHP when they're accidentally going to change the world c.f. Wikipedia, Facebook, WordPress.