3 ms·

>I think these possibilities are some of the strongest reasons to be intrigued by AT Protocol Genuine question, considering you can already do what you're talk

by S0y 2y ago

>I think these possibilities are some of the strongest reasons to be intrigued by AT Protocol Genuine question, considering you can already do what you're talking about with ActivityPub what makes you stick around on ATproto/Bsky?

steveklabnik 2y ago
Important context: long ago, my friends and I built a federated Twitter alternative based on ostatus, the predecessor to ActivityPub. I believe the architectural model of ActivityPub is bad. I think the architectural model of atproto is good. Everything is downstream of that. I don’t have time to elaborate at this exact moment, but figured I’d at least give you a preliminary answer.
- Diti 2y ago
  > I believe the architectural model of ActivityPub is bad. I think the architectural model of atproto is good. Everything is downstream of that. ActivityPub is a W3C standard which uses standard, “boring”, sometimes “complex” (JSON-LD) technologies. The fact the AT protocol does things smarter doesn’t mean it should, or will, be used on the WWW. There’s a reason why we still use e-mail and JSON for interoperability, instead of smarter technology. We should favor standards IMHO, even if they’re imperfect. Anyway, I’m really curious about your analysis in your future comments!
  - steveklabnik 2y ago
    Bluesky also uses W3C and other standard technologies, for example, DID, the most centralized part, is a W3C standard https://www.w3.org/TR/did-core/ https://www.w3.org/TR/did-core/. Lexicons use JSON-schema. dag-cbor is another example, though not W3C. CAR is another example. Being fit for purpose is more important than being "standard", if a standard doesn't do what it needs to do, that it's standard is irrelevant. Bluesky/atproto uses a ton of standard approaches where appropriate, and innovates where appropriate. Standards are borne out of usage, and a need to interoperate. Being blessed by a standards body doesn't inherently mean something is good. I can represent some data in JSON, or in HTML, but because both are standards doesn't mean that they're equally good. Standardization is a means to an end, not an inherent good in and of itself. People need to try new things in order to eventually produce new standards.
  - diggan 2y ago
    > There’s a reason why we still use e-mail and JSON for interoperability, instead of smarter technology. Ironically, JSON came out of one language, and didn't come from any standards body, like ActivityPub. It wasn't until ECMA-404 (2013) that JSON got a standard from one of the bodies. It's a good example of technology that came from outside the typical standards, but was eventually adopted by a large portion of the ecosystem, and eventually became a standard. Just like HTML :)

by S0y 2y ago

steveklabnik 2y ago

Important context: long ago, my friends and I built a federated Twitter alternative based on ostatus, the predecessor to ActivityPub. I believe the architectural model of ActivityPub is bad. I think the architectural model of atproto is good. Everything is downstream of that. I don’t have time to elaborate at this exact moment, but figured I’d at least give you a preliminary answer.

Diti 2y ago
> I believe the architectural model of ActivityPub is bad. I think the architectural model of atproto is good. Everything is downstream of that. ActivityPub is a W3C standard which uses standard, “boring”, sometimes “complex” (JSON-LD) technologies. The fact the AT protocol does things smarter doesn’t mean it should, or will, be used on the WWW. There’s a reason why we still use e-mail and JSON for interoperability, instead of smarter technology. We should favor standards IMHO, even if they’re imperfect. Anyway, I’m really curious about your analysis in your future comments!
- steveklabnik 2y ago
  Bluesky also uses W3C and other standard technologies, for example, DID, the most centralized part, is a W3C standard https://www.w3.org/TR/did-core/ https://www.w3.org/TR/did-core/. Lexicons use JSON-schema. dag-cbor is another example, though not W3C. CAR is another example. Being fit for purpose is more important than being "standard", if a standard doesn't do what it needs to do, that it's standard is irrelevant. Bluesky/atproto uses a ton of standard approaches where appropriate, and innovates where appropriate. Standards are borne out of usage, and a need to interoperate. Being blessed by a standards body doesn't inherently mean something is good. I can represent some data in JSON, or in HTML, but because both are standards doesn't mean that they're equally good. Standardization is a means to an end, not an inherent good in and of itself. People need to try new things in order to eventually produce new standards.
- diggan 2y ago
  > There’s a reason why we still use e-mail and JSON for interoperability, instead of smarter technology. Ironically, JSON came out of one language, and didn't come from any standards body, like ActivityPub. It wasn't until ECMA-404 (2013) that JSON got a standard from one of the bodies. It's a good example of technology that came from outside the typical standards, but was eventually adopted by a large portion of the ecosystem, and eventually became a standard. Just like HTML :)