On 25.02.23 00:56, Peter Saint-Andre wrote:
On 2/24/23 8:47 AM, Tedd Sterr wrote:The original sender of a message stanza SHOULD give it id=UUID. Unfortunately, this wasn't a requirement in the RFCs, so now we have various hacks to try to deal with that because we can't just fix the problem while maintaining compatibility.At some point we will want to published updated RFCs - probably once the SASL2 spec (etc.) is stable and widely deployed. We could then change RECOMMENDED to REQUIRED for the 'id' attribute and specify that it MUST be a UUID. In the meantime we could strongly suggest this to all implementors...
We probably will do another RFC round with the nice new things you teased. :)
However, it is still unclear to me how changing the RFC 'id' attribute specification from "must be unique within the scope of the stream id" to "must be globally unique, for example by using UUID" solves much we discussed in this thread.
It would just solve the "how to reference a stanza in a 1:1 chat" problem. But with a questionable backwards interoperability story. Therefore -- but please consider that I am biased here -- origin-id feels like the better solution.
In general, it appears that the topics discussed in this thread are more related to design consideration for current and future groupchat protocols, but much less for future RFC iterations.
- Flow
OpenPGP_0x8CAC2A9678548E35.asc
Description: OpenPGP public key
OpenPGP_signature
Description: OpenPGP digital signature
_______________________________________________ Standards mailing list Info: https://mail.jabber.org/mailman/listinfo/standards Unsubscribe: standards-unsubscr...@xmpp.org _______________________________________________
