 Why would Meta implement ActivityPub? 1½ reasons are compelling, another frequently mentioned one is not. Blogged yesterday. So far, a bunch of nodding, no real disagreements. Still want your thoughts, particularly if you disagree or have something to add I missed. https://reb00ted.org/tech/20230625-meta-why-activitypub/ 14 comments
@J12t "Embrace and extend is an entirely ineffective strategy if what you want to embrace and extend has only a tenth of the users you want “in just a few months”. It is not worth to embark on such a complex strategy if it only works for a month or two at best." Yep! @J12t the only point I can find to disagree on is that I don’t know enough about the EU laws to be sure that explanation is actually plausible. Won’t their existing services still be subject to the same laws independent from any new service? What’s their plan to make money with a new fedi service? Etc. Not sure that really counts as a disagreement @ShadSterling The peculiar thing about all those new EU laws and regulations is that they are intentionally very vague, and what it means exactly is only being worked out over time between regulator and regulated. E.g. the EU sent a "tech ambassador" to reside in San Francisco for this purpose. But some intents are very clear: one of them is interoperability. @J12t in that case building things to influence regulation is very plausible, but expecting to evade the interoperability requirement for Facebook and Instagram by creating a separate interoperable service is not convincing @ShadSterling Not evading. Just delaying. On the grounds that 1) we are working on it, see, with our new app, from the ground app, yes we are changing our ways, and 2) this is all so new, nobody in the industry knows how to do this, we all need to learn before we can do anything for billions of users, and you dear regulator don't know either let's gradually learn together. @J12t with the goal of making as much money as possible until the regulators figure out they’re lying, that’s plausible. Not sure it’ll work as well as they expect, but I’d be surprised if the laws penalize that enough to make real compliance more profitable  @J12t My guess is that folk at Meta finally realized that they can maintain useful dominance even within an "open" SocialWeb. The reality is that there are still many network effects associated with large platforms. (Mastodon.social is a good example) While theory may argue that being open may result in fragmentation, in practice it doesn't. An "open" Meta encourages the troublesome folk to federate elsewhere while staying connected and thus somewhat co-opted.  @J12t Another possibility: Meta has good business reasons for doing a decentralized network (which I think is clearly the case), and ActivityPub is one of the best-known decentralized protocols , so it's a good PR/marketing approach to say they're starting there. Either they make it work with proprietary extensions, or they say "we tried it but it didn't work out, and our approach is better because ..." @jdp23 I'm not seeing the "good business reasons for doing a decentralized network" for Meta? @J12t see @darnell.moe's points here and the ensuing discussion for several good reasons @jdp23 Hmm. I don't think I understand what he means with "decentralized". Meta in the business of shipping installable server-side software, golden master, patch releases, that kind of thing? Obviously I don't know, nobody does, but I would consider this to be supremely unlikely. Offering an API for 3rd-party software that their biz customers and large-scale influencers can use? Sure. But that's hardly decentralized and hardly new. Using AP as part of that API? Yes for my reasons 2 and 3. @J12t "Decentralized" doesn't necessarily mean "everybody can install it and run it themselves." I can certainly see them offering something that orgs with IT staffs can host on their own for their communities. For individuals, I can see them partnering with approved hosting vendors -- in fact one way to think about their outreach to instance admins is as an experiment, a proof point if it succeeds or something that can be easily discarded and blamed on others if it doesn't. |
Gregory's Server
@J12t I agree with your analysis. Plausible makes sense.