I’m not sure I completely understand what Twitter Cards are — “frozen caveman developer who has drifted away from publishing software, etc.”. I tend to think of them as the stuff Twitter automatically adds to a tweet when you include a link.
One thing I do know about Twitter Cards is if you accidentally publish something that’s a mistake, (like a bunch of PHP errors added to a page and then statically published to your site, for example), the content of these Twitter cards are cached, with unclear-to-me rules on how often that cache is refreshed.
Some googling about led me to a professional-looking-but-maybe-content-farm-but-maybe-not-2019-amirite article which, in turn, led me to Twitter’s own validator tool for cards, which looks pretty useful AND (per the author’s claims) should clear the cache of those cards.
So I tried that, reposted my link and — the bad summary was still showing in the card. So now what? Was this a problem on Twitter’s end? Does the validator not actually clear the cache? Was the cache purge request making its way through what I presume is a cavalcade of microservices at Twitter? Or was some frontend caching holding on to the bad description?
I started thinking about the meetings but decided I didn’t have the mental bandwidth for that today. So then I thought about changing the URL slug of my post to trick Twitter’s cache, but 20 year old cache busting strats made me sad. So I poked at the internet some more, tried one more time, and the cache was cleared. The system works?
I’m not sure there’s a lesson here other than old man yells at cloud, but I do miss the days where we all had to do a little more work but had more rational behavior out of our publishing systems.