bmzero3

that, or whatever you used to rotate image did so by editing exif orientation data, and that got stripped/lost somewhere along the way when posted, so the image unrotated itself (I guess?)

this is still somewhat speculative but if you're running this behind a proxy, you might want to configure either nginx realip module or equivalent, or this: https://symfony.com/doc/current/deployment/proxies.html so it could see and work with real user ip, if you haven't done so already (only do one of them, not both)

currently suspecting that this is the big part on why it hit 429 rate limit when it shouldn't be: the rate limiter probably only sees the proxy address, not the users, and so it ended up rate limiting the entire proxy instead of individual clients/users

Agreed, I also hope for this to be the case (and tried my besht to facilitate it), but as you can see you could say that mbin's community building is ...not so great right now.

there are far fewer but there are some (fedidb list), notable ones include karab.in (polish instance) and artemis.camp (special instance that's used for Artemis app development, also the one that @pewg7 uses)

as for why there're way more lemmy instances, my opinion is that it mostly comes down to:

• lemmy comes first -- first lemmy release dates to around 2019 while kbin's first commit is at 2021 and apparently it was not until sometimes in 2023 when the main instance came up
• had time to be tested and build its own reputation, and overall looks more matured and stable compared to kbin (I've seen quite a few opinions that although they find kbin nice, it's still needs more development to be more usable/comparable to lemmy)
• using more modern tech stack and organization -- rust backend, inferno (react-like) frontend, split frontend and backend design; vs using php which is perceived as dated and it seems almost a rite of passage to make a "haha php bad" joke at least once, and I don't think using Symfony is going to save that
• having a working API implementation -- when the reddit thing happened it didn't take long to see new lemmy app developments exploded like crazy, some were from those who made third party reddit apps and they had rework/adapt their apps into talking lemmy, and then there's Artemis where kbin didn't have an API implementation yet when development started, and the earliest version was resorting to scraping kbin's web for data (from Artemis FAQ, second point)
• (partial information, take this with a bottle of salt) supposedly more healthy/inviting for developing/contributing -- I cannot say for lemmy, but kbin's development seems to be slowing down as of late and stopped to a screeching halt several weeks ago, apparently the head dev got some big irl problems and so they basically went missing, and things starts to fall apart (one instance is that the main instance went down for like half a day).
  • in fact it went so bad that one of the developer/contributor decided to make a fork (mbin, fedidb list) and some of the contributors/instance operators has since moved their instances/efforts to mbin instead
  • one instance of note: fedia.io, it used to be running kbin for quite some time, but things starting to fall apart, severe performance degrade and errors to the point that the instance owner was considering to shut down instance but has since migrated to mbin where the situation seems to be going far better, enough for the owner to retract the shutdown.

(also first serious post hope it goes out well)