I know that “rate limiting” (in the sense of the x-rate-limit-* headers) doesn’t apply to write requests such as /statuses/destroy. Instead the overall write limits documented here apply across all clients.
However that page doesn’t mention tweet deletion. Is deleting a tweet subject to the 2400/day tweet creation limit? Or does it belong to some other bucket?
In general, if I exceed the write limit do I get a 429, as I would for rate limited reads? If not, how do I detect this situation on the client?
Basically I’m trying to figure out how to be a good API citizen (and not get blacklisted). Is it considered acceptable to just keep sending writes until you hit a 4XX? That would leave the user unable to use some functionality even on twitter.com, for up to 24 hours. I’d rather back off before that happens, but that requires knowing the daily limits for /statuses/destroy.