[Council] Council Meeting Minutes for 2017-03-08

Dave Cridland dave at cridland.net
Thu Mar 16 12:36:34 UTC 2017


On 16 March 2017 at 12:31, Dave Cridland <dave at cridland.net> wrote:
> On 8 March 2017 at 17:50, Daniel Gultsch <daniel at gultsch.de> wrote:
>> Council agrees that this needs more list discussion on list on whether
>> this should be an inplace upgrade to XEP-0115. If it becomes a new XEP
>> depending XEPs should be updated to a more generalized wording like
>> "use XEP-0030 or any caching mechanism like ecaps2"
>>
>
> I don't think it does, actually.
>
> More specifically, I think that specifications ought to be talking in
> terms of XEP-0030 features, and this spec (and '115) are simply
> mechanisms for caching the features, and irrelevant to the
> specifications that use the underlying features.
>

Been rubbish and didn't answer the thought on whether this ought to be
an update to '115 or not.

It shouldn't:

1) There are known interop concerns with '115 clients which only use
one '115 caps in a given presence stanza.

2) I recall introducing the new, hash-based, caps, and finding it very
difficult to explain to people that there were two, radically
different, variants of '115 - and typical clients needed to support
both.

(1) means we cannot, unfortunately, use the same qualified element for
a new '115.

(2) means I'd like to keep the old specification around, even if we're
aiming to deprecate.

Dave.


More information about the Council mailing list