[Standards] Proposed XMPP Extension: Multi-User Chat Light

Piotr Nosek piotr.nosek at erlang-solutions.com
Mon Dec 14 16:04:22 UTC 2015


Hi Dave,

a) It retains some level of compatibility, please see Implementation Notes.
It is possible to use 0045 protocol for most of the functionality in
transition period. "Substantial chunk of work" is not very precise. In our
case the initial implementation that did not support 0045 compatibility
took about 2-3 weeks. Please note it included refining the protocol in the
meantime, so it was not pure implementation of existing, ready standard.

b) It is a compilation of requirements of mobile chat providers. I can't
see why being useful only for mobile clients is a reason to treat is as
useless. It is a common belief amongst many developers that XMPP is not
very attractive for mobile environments, why can't we make several
extensions that are specifically mobile-friendly?
Yes, there is no possibility of sending IQs but the thing is - what
IQ-based functionality we would need in groupchats? File transfer? It's a
common practice nowadays to upload files to external storage like Amazon S3
and then just send a message with a link. (extra benefit: it can get
archived by MAM).

Regards,
Piotr

On Wed, Dec 9, 2015 at 10:48 AM, Dave Cridland <dave at cridland.net> wrote:

> This is quite a substantial protocol, but has, I think, two issues which
> mean it is problematic to accept in my opinion:
>
> a) It is not just presence-less MUC. It's an entirely new protocol which
> is incompatible with existing XEP-0045. Even the room affiliation model is
> different, allowing for example only one owner. This is problematic because
> it's not reusing much (if anything) of the existing infrastructure. As such
> it's going to be a substantial chunk of work for server developers to
> implement, and difficult to offer a transitional approach into XEP-0045
> based services.
>
> b) It is only presence-less MUC. It's not offering anything beyond simple
> fan-out of chat messages, and as a result there is no incentive for
> non-mobile, or non-chat, clients to adopt it. As an example, there's no way
> to send any IQ traffic through the system, due to a combination of no
> visibility of either presence or full jids, meaning there's no possibility
> of, for example, file transfer.
>
> I appreciate there is a degree of not wanting to accept it because we're
> expecting a MUC2 protoXEP to arrive, however I'm trying not to let that
> influence my thinking here, since there's currently no XEP.
>
> On 8 December 2015 at 17:39, XMPP Extensions Editor <editor at xmpp.org>
> wrote:
>
>> The XMPP Extensions Editor has received a proposal for a new XEP.
>>
>> Title: Multi-User Chat Light
>>
>> Abstract: This specification provides a presence-less standard for
>> Multi-User Chats. Its feature set is a response to mobile XMPP applications
>> needs and specific environment.
>>
>> URL: http://xmpp.org/extensions/inbox/muc-light.html
>>
>> The XMPP Council will decide in the next two weeks whether to accept this
>> proposal as an official XEP.
>>
>>
>
> _______________________________________________
> Standards mailing list
> Info: http://mail.jabber.org/mailman/listinfo/standards
> Unsubscribe: Standards-unsubscribe at xmpp.org
> _______________________________________________
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.jabber.org/pipermail/standards/attachments/20151214/2c1edb7e/attachment.html>


More information about the Standards mailing list