1 Apr
2026
1 Apr
'26
11:50 a.m.
Hi Daniel,
Thanks for the report, and to you and Marvin for travelling all the way
to Shenzen (!) for IETF 125.
Your suggestion of a hackfest at IETF 126 in Vienna is a great idea,
since it will introduce more people to folks active at the IETF and how
the IETF works.
As to change control, if it's managed correctly the IETF standardization
process should not result in arbitrary changes. "We believe in rough
consensus and running code" implies that if the code is running and
someone asks for changes that are merely aesthetic or their personal
preference, there's no good reason to make those changes (after of
course reasonably considering the concerns that have been raised).
As to "bring in new people", I think that's a bit of a red herring.
There is no such thing as "XSF people" or "IETF people" - there are
just
people. If some people come to the IETF asking to do some work
(especially if they are are *coming back* in good faith to update a
technology of long standing at the IETF), there is no requirement to get
new people interested. In my various roles at the IETF I have seen this
repeatedly with work on technologies like FTP.
As to "align this work with the need for a common messaging standard",
that could be a rat hole of epic proportions, so we'd need to understand
what kind of scope people are thinking about. Although I have no doubt
that there are ecosystem gaps XMPP can fill (and arguably should have
filled long ago), quite possibly those gaps have come about because of
product and (roughly speaking) political decisions made by the major
players, which can't necessarily be solved by a standardization effort,
especially one to update XMPP Core rather than define new extensions or
applications of XMPP. If some of the requirements align that's great,
but I'd be careful about saddling the update effort with a whole bunch
of new requirements just to satisfy the longing for a common messaging
standard, which has existed for 30 years and might never be satisfied.
Just my two cents, of course.
Peter
On 3/31/26 2:31 AM, Daniel Gultsch wrote:
Hi,
I just wanted to give a short update on my position on this subject.
Marvin and I were at IETF125 in Shenzen last week and independently of
each other had very positive interactions with key people in the IETF.
Generally speaking the vibe was very positive and welcoming. The fact
that XMPP already is an RFC and we are 'just' updating it means that
the process of (re)establishing a working group should not be a big
deal.
Process wise our first step would be to produce a draft to demonstrate
that there is something substantial.
What I had pitched to people at the IETF - and which I believe sort of
reflects what I’m hearing from our community is to first update core.
I mean the way you connect to an XMPP server these days looks
fundamentally different to what we have in 6120. (I think if and how
we also want to update IM and/or what extensions should be combined in
that RFC is something we don’t need to worry about at this point in
time.)
What I've been told repeatably is that we would be giving up change
control - but we know that. By extensions this obviously means that we
also have to be willing to update our respective implementations to
what ever comes out the other end of the IETF process. I think this is
fine? But if major XMPP implementers stay on the XSF flavor that would
be bad. (I realize that nobody can make any promises but a general
vibe check in our community on whether or not people would be willing
to implement the RFC version - independently of whether they have
worked on the RFC - would probably be good to do.)
The other thing someone mentioned to me - not as discouragement or as
a show stopper - but as a friendly tip - was that if we don’t get
outside collaborators / new participants and if instead it's only XSF
people working on it anyway then why bother? Marvin and I have plans
to organize a side meeting at the next IETF to maybe gauge some
interest. I think generally speaking there must be some interest in
the IETF community for standardized, federated instant messaging.
Matrix isn’t it. MIMI isn’t it. So there is currently a gap that XMPP
can fil. But maybe a side meeting in Vienna can help to figure out who
else is out there.
Speaking of IETF126 in Vienna: The IETF has a Hackathon on the
Saturday and Sunday (July 18th and 19th) that is free to attend and
even has free catering. Given the proximity of a lot of XMPP developer
to Vienna I think it would be nice if a few of us could attend there
to work on XMPP. (Even those that are not usually attending IETF). On
one hand it would be a nice opportunity to have a regular XMPP sprint
with the location and the food sorted for us. On the other hand
suddenly having an entire table full of XMPP people at an IETF event
will certainly raise some attention. At the end of the Hackathon there
is a 90s-2 minute opportunity to present the Hackathon project and I
did a presentation at the Hackathon in Shenzhen and apparently that
turned some heads as well. If we want people to notice XMPP that's a
good starting point. IETF126 Hackathon would be a win-win. Free
location free food for a sprint we regularly do anyway. + Maybe some
free advertisements (and raise attention for the side meeting we are
also trying to do)
cheers
Daniel
_______________________________________________
Standards mailing list -- standards(a)xmpp.org
To unsubscribe send an email to standards-leave(a)xmpp.org