pleroma.site

for what it’s worth, I mostly agree with @denschub about his post. especially acknowledging that litepub has had a lot of difficulty gaining traction, but some good has come out of it.

a large part of why litepub failed to gain traction is because it didn’t go far enough, but it is still an activitypub dialect, and is probably the most widely deployed one verses the mastodon dialect. and, for the most part, these are cross-compatible.

i am probably going to write some response to his blog at some point, but i need to think about what i want to say about it.

what i will say is this: of all the AP implementations, I think Pleroma and some of the smaller implementations like Kroeg are the only ones that are seriously trying to deliver a universal experience. in my opinion, Mastodon and many other implementations either drop or degrade non-native content to the point that they have less functionality than RSS readers.

in particular, i don’t understand how the Mastodon developers can say:

The social network that is Mastodon isn’t really Mastodon. It’s bigger. It’s any piece of software that implements ActivityPub. That software can be wildly different in how it looks and what it does! But the social graph–what we call the people and their connections–is the same.

… when rich media objects (or hell, even Article objects) are degraded into Note objects. after all, what good is a social graph if you can’t fully interact with all of the nodes in it?

i will also say that Pleroma hasn’t been out there promising the entire world the moon. we have not been saying “yeah, implement ActivityPub and then you will automatically interoperate with us in a nice way,” because it simply isn’t true.

the main problem with ActivityPub and ActivityStreams is that the designers failed to observe or at least mention that the interpretation of objects induces side effects: each implementation has to know what a Video or an Article or a Page object are in order to be able to present them in a useful way to the user. there’s a couple of approaches that can be used to do this, you can degrade the unknown objects to an object type you understand and hope for the best, or you can define a way to present the object that makes sense. Mastodon does the former, and Pleroma does the latter.

ask yourself: who should be the ones actually pushing projects to link into the fediverse? the ones implementing a project that degrades everything to a Note, or the ones implementing a project that tries to be a universal client of everything the fediverse has to offer?

the implementations which are working hard to provide a rich experience are staying quiet about ActivityPub, and the implementations which degrade the experience are talking up a serious game about ActivityPub. i think that speaks for itself.

anyway, you should definitely read dennis’s blog. i’ll expand on this some point later.

https://schub.io/blog/2019/01/13/activitypub-final-thoughts-one-year-later.html

replies
0
announces
14
likes
21

one last thought: in fact, generally when presented with "I'm going to implement ActivityPub in my new project," if it's something completely different than what is going on right now, I have been suggesting they look at IndieWeb technologies instead where such an architecture makes more sense. for example with the alternative to forgefed I proposed earlier.