First all the bs with Twitter and Elon, then Reddit having an exodus to Lemmy (not complaining lol), then Twitch. Are we like, in an alternate self healing dimension or something?
First all the bs with Twitter and Elon, then Reddit having an exodus to Lemmy (not complaining lol), then Twitch. Are we like, in an alternate self healing dimension or something?
Every request is tied to some functionality. Databases and storage is laughably cheap these days.
The complex queries and all the overhead features is where the real expense is. Crafting a personal, ad-optimized timelines is what’s costing Twitter the most money. The public/subscribed feeds of mastodon are incredibly efficient even on something super slow like ruby on rails.
Lemmy is having the same problem and doesn’t have ads.
Does it though? This instance has thousands of users and interactions already and is running on just few dollars a month.
It’s running on a few hundred dollars a month, if I recall correctly, and it has only about 450 users per day. (The sidebar statistics don’t include a figure for peak concurrent users, unfortunately, and that’s what we really need to know.)
Ah didnt see that increase though decentralized systems are inheritly very inefficient unfortunately
The issue I saw was with answering user requests for content—which post do you want to see, which community was it posted to, which comments are on that post, who wrote them, what are their vote scores, and so on. Nothing to do with decentralization. Reddit would have had the same problem, and judging from early Reddit’s performance and reliability woes, it probably did have the same problem.
That’s just bad engineering tbh.
Nobody’s born knowing how to optimize database queries. Optimization is hard and arcane, and SQL is no exception.
I didn’t look into this but this looks way more complicated than it should be. The query in the issue looks absolutely ridiculous tbh.