guess they forgot the "don't make any mistakes" at the end of the prompt
niartenyaw
joined 2 years ago
5 9s means 99.999% uptime, which translates to a site only being down for ~0.86 seconds per day. This is the golden standard for online services
I can't seem to find any information on it, but how hard would if be to host your own Bridgy Fed? I feel a tool like this would ideally be decentralized.
and that's unfortunately only true because the greedy groups have destroyed all the non-greedy ones by slaughter or forced participation
we just need to make sure that we don't rely on their instance(s) too heavily so we only have minimal losses when they eventually do drop support.
it may be more comprehensive than trying to scrape other platforms, but the data that is available is absolutely nothing compared to what they collect on their own platforms. they're almost certainly not very concerned with collecting our relatively mild amount of data.
now that's an immaculate conception
hmm that sounds really interesting! if you end up remembering it and also remember to respond i'd love to know lol. yeah i hope if something good enough comes along two platforms will implement and then others will just follow suit.
that's all very fair, i guess i was just hoping federation was at least on a long term roadmap
yeah, i agree that i don't think federation of something like this would be through ActivityPub (matrix has its own). i guess it just feels unfortunate that if users want to access communities across multiple instances, users will have to have separate logins and identities for each one.
i looked into revolt and pretty concerned with these two stances of the project
We don't think federation is beneficial to Revolt
We have a variety of monetisation ideas lined up internally, with these, it is not my intention for us to paywall features and I find it unlikely we would ever do that considering it would contradict what we're trying to achieve.
like, to me it seems they want to get communities invested and then later monetize in ways those communities don't yet know about?? idk that sounds extremely sus. especially when competing instances will fight against network effects with no federation.
view more: next ›
and not only that. in your example of choosing to address the hull first over the engine, the engine problem is actually prescient. when taking time to debate about AGI, it is to debate a hypothetical future problem over real current problems that actually exist and aren't getting enough attention to be resolved. and if we can't address those, why do we think we'll be able to figure out the problems of AGI?