Alright it's late and i need to go to bed, but here's a draft FEP to do full account migration with posts and whatever other kinda objects you want to bring with you.
Technical Discussion
54
Posts
11
Posters
3
Views
-
i think i'm supposed to direct discussion here, but i also linked back here from there, so idk in the spirit of everything being everywhere as far as i'm concerned go hogwild and yell at me online wherever you want
https://socialhub.activitypub.rocks/t/fep-1580-move-actor-objects-with-a-migration-collection/8111
jonny@neuromatch.social FWIW there's no need to use SocialHub. You could post a discussion thread anywhere (like ActivityPub.Space, GitHub, a WordPress blog, etc... and yes, even a Mastodon status can be a starting point for discussion.)
-
@jonny @mariusor from what i can tell the only reason the origin based security model and it's problems are a thing at all is because there is no mechanism to see who "owns" an object from a reference alone, which namespacing would solve
otherwise you have to fetch every single object ever referenced just to check for a backlink. not only is fetching over AP is already inefficient as it is with no batching and surprisingly slow crypto, nobody serves backlinks of any kind> see who "owns" an object from a reference alone, which namespacing would solve
i don't think that's desirable on its own (since id should be opaque) but you can get it anyway if you require actors to have their own dns name (and then tls does the rest). that's how the same-origin model is *supposed* to work in theory; the bit that breaks the assumption is when one origin has multiple tenants
re: actor namespaces, i am writing/exploring a fep for that actually...
-
> see who "owns" an object from a reference alone, which namespacing would solve
i don't think that's desirable on its own (since id should be opaque) but you can get it anyway if you require actors to have their own dns name (and then tls does the rest). that's how the same-origin model is *supposed* to work in theory; the bit that breaks the assumption is when one origin has multiple tenants
re: actor namespaces, i am writing/exploring a fep for that actually...
-
@jonny @kopper @mariusor i've had the same wish, except instead of "actorPublicKey" it's any uri
basically take a base uri for your contact in your address book, and append a relative reference. the did specs can do this with service + relativeRef, and i think https://w3id.org/fep/e3e9 uses that? i'd like to be able to do that with cids too, or probably even http redirects
-
> see who "owns" an object from a reference alone, which namespacing would solve
i don't think that's desirable on its own (since id should be opaque) but you can get it anyway if you require actors to have their own dns name (and then tls does the rest). that's how the same-origin model is *supposed* to work in theory; the bit that breaks the assumption is when one origin has multiple tenants
re: actor namespaces, i am writing/exploring a fep for that actually...
@trwnh the way I meant namespacing object IDs was simply related to being able to decompose an IRI into a clean and intuitive structure: example.com/~jdoe/outbox/123, means that the activity with that ID can be said to be contained in jdoe's outbox on example.com.
-
@mariusor @kopper @jonny the uri owner in this case is example.com and you cannot assume that "~jdoe" or "/outbox/" mean anything special
...however, if you knew that /~jdoe was an actor, then /~jdoe can tell you that you can implicitly trust any attribution claims to /~jdoe if the resource exists within some prefix like /~jdoe/ or /objects/
-
@mariusor @kopper @jonny the uri owner in this case is example.com and you cannot assume that "~jdoe" or "/outbox/" mean anything special
...however, if you knew that /~jdoe was an actor, then /~jdoe can tell you that you can implicitly trust any attribution claims to /~jdoe if the resource exists within some prefix like /~jdoe/ or /objects/
> you cannot assume
@trwnh but I can. :D At least in GoActivityPub I built some logic around composing IRIs using this kind of mechanism.
It's not guaranteed for handling external objects, but it's useful for creating internal ones.
And in the context of moving objects (which is what I was replying to initially), I feel like having a example.com/~jdoe namespace ensures that you can do a batch move (and not have collisions) for all objects that have them as author just by changing the relevant bits of the object IDs, instead of having to generate a new ID individually for each moved object.
-
> you cannot assume
@trwnh but I can. :D At least in GoActivityPub I built some logic around composing IRIs using this kind of mechanism.
It's not guaranteed for handling external objects, but it's useful for creating internal ones.
And in the context of moving objects (which is what I was replying to initially), I feel like having a example.com/~jdoe namespace ensures that you can do a batch move (and not have collisions) for all objects that have them as author just by changing the relevant bits of the object IDs, instead of having to generate a new ID individually for each moved object.
@mariusor well, internally you control the identifiers, so you can do whatever you want. the danger is in assuming meaning externally.
-
@julian @jonny SocialHub is not a bad place though - FEP category is federated
https://socialhub.activitypub.rocks/ap/object/bee813d28b0947ffce97f8c2bbebb955
-
@julian @jonny SocialHub is not a bad place though - FEP category is federated
https://socialhub.activitypub.rocks/ap/object/bee813d28b0947ffce97f8c2bbebb955
Of course, not trying to say otherwise. Just that there is a preconception that jonny@neuromatch.social had to post there as part of the FEP process, which even I was under the assumption of.
We chatted about that and you disabused me of that notion :smirk:
-
@kopper i was looking for an instance metadata item that was just some list of tokens that was "the list of things that the instance supports," and i could have sworn it existed, but i couldn't find it when i looked. most fedi apps (and even most LD apps) don't actually dereference the URIs in a context and treat them as tokens anyway, but yes failure to resolve terms is a big problem and am not a fan of DNS-based linked data.
jonny@neuromatch.social kopper@not-brain.d.on-t.work Perhaps Capability Discovery is what you might be looking for? There is some implementor support for this one.
-
Alright it's late and i need to go to bed, but here's a draft FEP to do full account migration with posts and whatever other kinda objects you want to bring with you. It's a trivial expansion of existing ActivityPub/streams systems and supports gradual migration as it's implemented and after an account migration. It should be possible to migrate pretty much everything this way, both private and public objects.
criticism, feedback, revisions, etc. welcome - i don't think this is a "final version" and there are certainly things i overlooked.
https://codeberg.org/fediverse/fep/pulls/692
#MoveAllPosts #FediDev #FEP #FEP_1580 #FullMigration #AccountMigration
@jonny you're amazing 😍
-
Alright it's late and i need to go to bed, but here's a draft FEP to do full account migration with posts and whatever other kinda objects you want to bring with you. It's a trivial expansion of existing ActivityPub/streams systems and supports gradual migration as it's implemented and after an account migration. It should be possible to migrate pretty much everything this way, both private and public objects.
criticism, feedback, revisions, etc. welcome - i don't think this is a "final version" and there are certainly things i overlooked.
https://codeberg.org/fediverse/fep/pulls/692
#MoveAllPosts #FediDev #FEP #FEP_1580 #FullMigration #AccountMigration
This looks fine to me, great to see many edge cases considered.
-
i'm almost disappointed bc it's so simple and obvious. there are a lot of caveats and requirements there because i was trying to make it clear enough to be implementable, but really it's as simple as doing this: https://neuromatch.social/@jonny/115338330944599542
@jonny I'm seeing notes here for the technical challenge of performing a migration of objects between two willing hosts given a user in good standing with sufficient credentials. I think it's great that you have written that down.
But I think the reason we don't have this yet is because that's only 20% of the problem. The other 80% of the challenge is contained in that first sentence of mine.
- How do we signal this intention to migrate to both hosts?
- How do we validate that both hosts are willing?
- How do the hosts validate credentials and good standing of both the migrating user and each other?
- What obligations does each party have to propagate and maintain this effective URL redirect?
- What controls does each party have in order to manage this transition process?
- How do we manage moderation of the incoming content?
- How will this impact current user agreements on most instances?
Any specification here needs to address admins and moderators as key players.
-
Alright it's late and i need to go to bed, but here's a draft FEP to do full account migration with posts and whatever other kinda objects you want to bring with you. It's a trivial expansion of existing ActivityPub/streams systems and supports gradual migration as it's implemented and after an account migration. It should be possible to migrate pretty much everything this way, both private and public objects.
criticism, feedback, revisions, etc. welcome - i don't think this is a "final version" and there are certainly things i overlooked.
https://codeberg.org/fediverse/fep/pulls/692
#MoveAllPosts #FediDev #FEP #FEP_1580 #FullMigration #AccountMigration
@jonny Could I possibly encourage you to add a brief 1-2 sentences-or-so "Social Considerations" section -- or, well, the way you organized your headings a "Social" heading in your "Discussion" section -- next to "Privacy" and "Security".
I think in this case the basic Social motivation and implication for this specifically is reasonably obvious to everyone in the discussion right now, but because that obviousness won't be as available to people looking back at this during a more future implementation and won't be as available to people reading this looking for interactions with their own drafting FEP at a later point, it's worth including in the text itself.
If you don't want to or if this isn't the sort of feedback you're looking for no worries and sorry for imposing.
(Alternatively if you'd be up for it but have many other things atm, and if you trust me enough, I could try to grab time from dodging between the holidays at the moment and do a brief one for you, if you want.) -
@jonny I'm seeing notes here for the technical challenge of performing a migration of objects between two willing hosts given a user in good standing with sufficient credentials. I think it's great that you have written that down.
But I think the reason we don't have this yet is because that's only 20% of the problem. The other 80% of the challenge is contained in that first sentence of mine.
- How do we signal this intention to migrate to both hosts?
- How do we validate that both hosts are willing?
- How do the hosts validate credentials and good standing of both the migrating user and each other?
- What obligations does each party have to propagate and maintain this effective URL redirect?
- What controls does each party have in order to manage this transition process?
- How do we manage moderation of the incoming content?
- How will this impact current user agreements on most instances?
Any specification here needs to address admins and moderators as key players.
@gatesvp
All these questions are addressed in the FEP except moderation, and I'm talking with masto devs about what would be good there -
@jonny Could I possibly encourage you to add a brief 1-2 sentences-or-so "Social Considerations" section -- or, well, the way you organized your headings a "Social" heading in your "Discussion" section -- next to "Privacy" and "Security".
I think in this case the basic Social motivation and implication for this specifically is reasonably obvious to everyone in the discussion right now, but because that obviousness won't be as available to people looking back at this during a more future implementation and won't be as available to people reading this looking for interactions with their own drafting FEP at a later point, it's worth including in the text itself.
If you don't want to or if this isn't the sort of feedback you're looking for no worries and sorry for imposing.
(Alternatively if you'd be up for it but have many other things atm, and if you trust me enough, I could try to grab time from dodging between the holidays at the moment and do a brief one for you, if you want.)@gaditb
I briefly addressed this in the intro but I agree its worth expanding on -
@gaditb
I briefly addressed this in the intro but I agree its worth expanding on@jonny (My low-key agenda is, I kinda think every FEP should have a section about it. It's like, part of the discussion and the years of prior context people are discussing from,,)
-
@jonny (My low-key agenda is, I kinda think every FEP should have a section about it. It's like, part of the discussion and the years of prior context people are discussing from,,)
@jonny Okay actually more directly-specufic-to-content thought. Although one that I don't know the right answer to:
> Constitutively private objects like bookmarks, blocks, and mutes SHOULD be ingested during the ingestion routine for the purposes of creating a seamless migration from the perspective of the migrating Actor, but MUST NOT be included in the migration collection.
For Blocks in particular -- In cases where, as is allowed by SHOULD, a migration (either source or destination) implements migrating posts but /not/ Blocks,
that silently opens up old posts to now no longer be blocked from interaction, which feels particular significant currently (where we don't have shared blocklists) since it means primarily accounts that have been specifically blocked on an individual basis -- likely (? this is an assumption) for recurringly-applicable reasons. -
@jonny Okay actually more directly-specufic-to-content thought. Although one that I don't know the right answer to:
> Constitutively private objects like bookmarks, blocks, and mutes SHOULD be ingested during the ingestion routine for the purposes of creating a seamless migration from the perspective of the migrating Actor, but MUST NOT be included in the migration collection.
For Blocks in particular -- In cases where, as is allowed by SHOULD, a migration (either source or destination) implements migrating posts but /not/ Blocks,
that silently opens up old posts to now no longer be blocked from interaction, which feels particular significant currently (where we don't have shared blocklists) since it means primarily accounts that have been specifically blocked on an individual basis -- likely (? this is an assumption) for recurringly-applicable reasons.@gaditb
Really good point, I have some ideas for language here, and there does need to be a bit more specificity about how to handle visibility for e.g. instance software that may not have blocks importing from software that does - basically that visibility must be at least as constrained as the source object
Gli ultimi otto messaggi ricevuti dalla Federazione
-
Currently whenever a POST is sent to an inbox the response body is ignored. The http status code is checked for errors but that's it. Perhaps some more complex information about the state of the receiving ingress queue could be conveyed in the body.
-
@gaditb can i add you to acknowledgements in the FEP?
-
@evan does AP have a way to send control messages, just meant for the receiving server?
-
rimu@piefed.social I think it might have been ignored because I don't serve a moderator collection as per 1b12.
I'll try to get that set up this week and test again :see_no_evil:
-
fentiger@mastodon.social yes I was also thinking of the venerable 429.
evan the thing is, both approaches rely on the receiving end to support it. I would make the argument that a 429 would have nominal support from some implementations, but then again I wouldn't actually bet money on it.
Either way you're looking at an FEP to establish flow control signalling or 429 handling. I'd then make the argument that putting together guidance for 429 (+ exponential back off) is easier than specifying flow control logic.
-
@evan No, you haven't. You just need to back off and retry, just as you do for an ACK timeout in TCP.
-
fentiger@mastodon.social yes, but by that point you've failed. What I want is to have more adaptive sending, so you don't get to the 429 code.
