On Thu, 8 Nov 2018 at 07:49, Patrik Fältström <paf=40frobbit.se@xxxxxxxxxxxxxx> wrote:
On 8 Nov 2018, at 11:41, Randy Bush wrote:
>> Perhaps the IETF needs to admit defeat here...
>
> my memory is that xmpp was originaly brought into the ietf to be rubber stamped.
Being Area Director of the time, my memory is the following:
Step 1:
- Work on coming up with a chat protocol
- Outcome: four proposals, most interesting based on BEEP
- People start using Jabber
Step 2:
- No agreement, allow all four to become RFCs (I think all four ended up being RFCs, I might be wrong)
- Discussion with Jabber, that did not want change control in IETF
Step 3:
- Jabber people came to IETF and said we can have a good constructive discussion
- Jabber discussed
Step 4:
- XMPP be finished
So, "rubber stamped" was a bit to brutal _expression_ maybe ;-)
Jabber/XMPP was treated and was managed the same / similar way as all protocols that at first are dealt with outside of IETF and then brought to the IETF community. A quite successful example I must say. Although the understanding of how IETF works to start with was, also as always, a bit confusing.
Patrik
I would also add:
Step 5:
- The XMPP community formed a standards group of its own, essentially modelled on the IETF's processes.
Step 6:
- The XMPP community revised the specifications within the IETF.
I think the relationship has been generally successful.
Dave.
