Google Translate:
First of all, the ActivityPub system is not suited to something like Misskey, which flows at lightning speed.
It was originally designed to connect blogs, small-scale SNSs, and wikis.
How can it handle the TL hell where tens of thousands of requests fly in per second?
It’s based on the idea that it would be nice if various small services could send each other updates, so it’s quite costly.
I sorry, but I find this hard to believe. Either they’re counting all requests their service gets (not just the APub ones), or Misskey is doing something really weird. Lemmy is notoriously busy compared to most Activity Pub stuff (PixelFed had to rewrite their activity handling as Lemmy overwhelmed the queue) and feddit.uk gets about 10-20K APub requests an hour.
yea, the Threadiverse probably gets a relatively high number of requests since the upvotes/downvotes are also requests
but everything seems to be working fine over here
I took it as a translation issue and misskey was the one that was designed for less traffic. The next sentence kind of clarifies. “it was designed…” for slower things like wiki updates means misskey.
I was confused at first too but that’s the only way it makes sense. I don’t know the history of misskey though so I could be wrong.
Thanks for posting the translation though! Super helpful
I don’t know if I would say it’s working fine. If you don’t clear your cache regularly the main page takes like a full minute to load.
Hopefully it’s fixable but this is a level of performance most internet users would not tolerate.
if a client side action (clearing cache) affects the issue, then it’s not related to the ActivityPub protocol
Maybe not. I just don’t think saying threadiverse works fine is completely true. We have lots of issues here and there.