On Freitag, 23. Februar 2018 12:36:01 CET Georg Lukas wrote:
> * Jonas Wielicki <[email protected]> [2018-02-23 11:39]:
> > - This is a larger one: If we’re changing XEP-0077 flow in a backward-
> > incompatible manner, wouldn’t it make more sense to do this in a wholly
> > new
> > namespace, with a new stream feature to advertise support? That would save
> > a round-trip (to detect that it *won't* work) and allows to passively
> > detect support.
> 
> The client advertises suport by adding a <preauth> element into the
> initial IQ. AFAIU IBR, the initial IQ needs to be sent anyway, so we are
> piggy-backing here without additional round-trips.

Right. I forgot for a moment that it doesn’t make sense to attempt <preauth/> 
with any domain except the domain where the token comes from, and that domain 
should better be supporting <preauth/> ;-)

> The server either ignores the element or returns a modified data form /
> IBR pseudo-form if supported.

Having this as a stream feature and in a separate, new namespace still feels 
cleaner to me.

kind regards,
Jonas

Attachment: signature.asc
Description: This is a digitally signed message part.

_______________________________________________
Standards mailing list
Info: https://mail.jabber.org/mailman/listinfo/standards
Unsubscribe: [email protected]
_______________________________________________

Reply via email to