[Jingle] <iq/> versus <message/> (Was: Tomorrow's meeting)

Emil Ivov emcho at jitsi.org
Thu Jul 25 11:21:22 UTC 2013



On 24.07.13, 14:42, Kevin Smith wrote:
> On Wed, Jul 24, 2013 at 1:37 PM, Emil Ivov <emcho at jitsi.org> wrote:
>>
>>
>> On 24.07.13, 13:23, Kevin Smith wrote:
>>>
>>> Let's try that again, with less fatfingering.
>>>
>>>    On Wed, Jul 24, 2013 at 12:17 PM, Emil Ivov <emcho at jitsi.org> wrote:
>>>>
>>>> On 24.07.13, 12:22, Dave Cridland wrote:
>>>>>
>>>>> Right, and adoption is something we've got to work on.
>>>>
>>>>
>>>>
>>>> Agreed and just deprecating all existing implementations with a wave of
>>>> our
>>>> collective hand does not sound very much as nurturing adoption.
>>>
>>>
>>>    I don't think that adding the option of using messages instead of iqs
>>>    does this. I assume that we would add another disco for
>>> jingle-message, and that if both end points have this they'd use
>>> messages, otherwise iqs.
>>
>>
>> Which means that we are keeping IQs as MTI. Is this what you are saying?
>
> I would imagine so. Completely changing the transport mechanism with
> no backwards compatibility would take some persuading me that it's a
> good idea.

Well ... I could live with it that way. Once we roll the first 
implementations of the experimental XEP we could decide whether and how 
to pursue. Maybe one day all implementations will have message support 
and we could even decide to drop the IQs.

Emil

P.S. Has anyone tried sending arbitrary (i.e. non-body) message content 
through Facebook? Does it have the same fate as IQs or do they let it 
through?


-- 
https://jitsi.org


More information about the Jingle mailing list