hey. wiggle your tail.

don't have one? we can help~
big or small, fluffy, scaly,
feathers are nice too,
whether coming in slowly
or bursting out, swishy, flicky,
helpful, snuggable,
mind of its own,

then we could look at
paws, wings, ears, a snoot,
but for now:

wiggle your tail 💚

#tf

@gabboman

yeah they don’t have an equivalent failure mode

but i think that being able to mitigate the issue proactively in advance is nice, but outside of specific groups (users of social media that isn’t intrinsically tied to atproto, atproto power users, folks on black/northsky, etc) it still is effectively a full outage

i use bsky mainly to look at art. i don’t think any of the artists i follow on there use blacksky or northsky or wafrn or some client that can work around bsky.social appview issues. they probably don’t even know what the hell a pds is. and some of them also (used to) use fedi. so whatever education there is on bluesky on how it works clearly is working less than the billions of confusing fedi explanations

@gabboman it’s not really comparable. even with the mastodon.social focus (which i heard is being addressed) of the mastodon app you can still log into a different instance who wouldn’t have to release their own app in order for it to work outside the browser

and even then if mastodon.social goes down most fedi users can go about their day just without mastodon.social on their feed

@gabboman i am not denying that you could have just posted through the outage. i don’t think that this is something the average user knows is even an option

plus from what i have been reading this was an appview issue, no?

where you would need a specific app that doesn’t use the bluesky appview, and know how to configure it to not use the bluesky appview if applicable. this goes too deep into internal patform details for someone who just wants to use it. the sheer magnitude of the bluesky appview makes failures on it more akin to centralized social media outages even if technically you could just not be affected by the outage

@gabboman aiui wafrn has a more fedi-style approach to ATproto federation?

my point is more that there is a single node¹ that almost entirely controls the entire network. while it is certainly possible under certain conditions to continue using ATproto with specific server or client software, it’s not a representative experience of users that download the bluesky app to access their bluesky account on bluesky (which is by far the vast majority of users, not even close)

¹: collection of nodes under the same management, really

also despite those alternative appviews existing, bluesky only uses the official bluesky appview, not any of the other ones. understandable because they can enforce different or no moderation and because those projects could likely not handle the sudden influx of millions of users

fwiw i host my own PDS (which is the only practical part for someone to self-host). it did absolutely nothing about the bluesky outage yesterday. that’s because you’d need a bluesky appview, which has an ever growing storage and bandwidth requirements. there’s good community projects that host them, but the network is not going to have a significant amount of these in practice due to the costs involved.

bluesky outages taking out (within rounding error) the entire network is a good argument for how it’s not actually decentralized but it’s also where we get all of the spicy takes about it that would probably be posted on bluesky instead if it wasn’t down

1: trash can
2: dumpster
3: attic
4: city park
5: forest