[Standards] Council Minutes 2020-02-26

Kevin Smith kevin.smith at isode.com
Wed Mar 4 12:27:44 UTC 2020


On 4 Mar 2020, at 10:22, Daniel Gultsch <daniel at gultsch.de> wrote:
> 
> Am So., 1. März 2020 um 01:14 Uhr schrieb Tedd Sterr <teddsterr at outlook.com>:
>> 4b) Advance XEP-0198 (Stream Management) - https://xmpp.org/extensions/xep-0198.html
>> Georg is unsure, but it's doing its job, expect for the unclear resume host connection mechanism.
>> Dave noted a comment on s2s, possibly from MattJ, which he has yet to consider, but s2s is under-specified at best.
>> Jonas doesn't think it's possible to move forward if there are zero s2s implementations; Dave doesn't think any were explicitly mentioned, which would itself be a procedural reason for not advancing.
>> Zash mentions that mod_smacks for Prosody does support XEP-0198 in s2s, though not resumption, and it's disabled by default - Dave thinks it's unclear what resumption would do for s2s.
>> 
>> Jonas: [on-list] (yet to catch up on the thread)
>> Daniel: -1 (people have brought up valid, but fixable, concerns)
>> Zash: -1 (haven't read that thread yet)
>> Dave: -1 (lack of clarity on s2s implementations)
>> Georg: -1
> 
> 
> Do we have a way forward with this that isn’t just ignoring it?
> Are any of the authors who are still active in the community (looking
> at you Matt and Dave) willing to incorporate whatever the consensus is
> on how to proceed?
> 
> Do we want to eliminate s2s from the XEP?
> 
> Are server developers interested in implementing 198 s2s?
> 
> Personally I would hope for the latter. But I’m not a server
> developer; Maybe it's more complicated or unnecessary as it seems.

FWIW, M-Link implements 198 acks (but not resumption) for X2X sessions (I don’t remember if we do S2S or not - I *think* we do, and if we don’t we’ll end up doing so).

/K


More information about the Standards mailing list