Dave,
You have lots of good and valid questions below. I'll respond below and
also cc RAI area mailing list. We have discussed this topic recently, in
particular with regards to concerns over scheduling for IETF-72. One
concern that the RAI area has, as well is some overlap and dependencies
on TSV area work that also makes our scheduling even more difficult and
contentious.
Thanks,
Mary.
-----Original Message-----
From: Dave Crocker [mailto:dcrocker(_at_)bbiw(_dot_)net]
Sent: Friday, July 18, 2008 11:58 AM
To: Barnes, Mary (RICH2:AR00)
Cc: ietf(_at_)ietf(_dot_)org
Subject: Re: Proposed Experiment: More Meeting Time on Friday for IETF
73
Mary Barnes wrote:
Dave,
There are a few topics for which mailing list discussion has failed to
reach consensus and would really benefit from f2f time. You can look
at SIP WG archives for example for a couple of the hot topics.
As chairs, we do try to push for completion of work on the mailing
lists, but this isn't always successful ...
...
The other issue is just the sheer volume of work incoming to RAI -
it's over 20% of overall IETF drafts per Jari's stats:
Mary,
Thanks for the quick response. From my perspective, your opening words,
above, do suggest exactly the right underlying process and the right
role for f2f meeting time.
From what you posted before, however, it sounds as if there is a
chronic lack of meeting time. This means either that "too many" items
can't be solved on the list -- that is, that there is something about
mailing list exchanges that isn't working well enough.
Another possibility is what's suggested by your later text: there are
too many things being attempted. In other words, perhaps RAI is trying
to do too much too quickly? (I mean that as an honest question; I don't
really have an opinion of my own. ) RAI is certainly a broad area, but
it's hard to miss that even RAI folk take note of the massive workload.
Hence sipping, as a filter on proposals.
[MB] I don't know if I would use the adverb "quickly" for RAI work. And,
it's not too easy to figure out the "Whys?". Here's my first pass at
what I see as some of the whys:
1) Lots of requirements that come in from other SDOs.
2) Lots of potential for building on what's already been done in RAI
3) A broad range of interested parties (some overlap with 1, but less
than one would think): researchers, vendors and service providers (some
of whom have conflicting objectives, architectural views and
philosophies)
[/MB]
An industry can only absorb just so much out of standards arena. Is it
really absorbing (implementing, deploying, using) RAI output fast enough
to justify the current workload?
[MB] If we look at SIPit reports, I think the answer here is no at this
point. However, my view is that a lot of folks have done the bread and
butter for the core protocol (which as for any protocol this complex can
take a while to work out bugs). And, there are a lot of folks working on
adding some of the additional functionality albeit different bits based
on the "whys" above.
The area (in particular SIP WG) has attempted to highlight and suggest
the basics that should be implemented:
http://tools.ietf.org/html/draft-ietf-sip-hitchhikers-guide-05
I do feel that in a couple of years, we may well be in a far different
situation in terms of the amount of industry uptake and the current
situation may just be a natural part of the process for a maturing
protocol/industry.
In one sense there have been some improvements. SIPPING has completed a
lot of work items (and may be the reason SIP is so overloaded right
now). In addition, RAI has created more focused/tightly chartered WGs to
work on some of the key items such as Emergency Services and Service
Interworking, both of which will be important for the industry uptake.
So, we have been trying to solve the problems in the area, but some
solutions do create other problems (i.e., more WGs is one of the key
contributors to the scheduling issue).
[/MB]
Also, note that 6 of the RAI WGs make up 50% of the top 12 WGs based
on volume of documents: http://www.arkko.com/tools/stats/wgdistr.html
It's pretty easy to have lots of documents get generated. What is
difficult is building real energy (support and effort) to pursue them,
adopt them, and use them. Hence my question about industry uptake.
[MB] IMHO, we have had some energy issues in various WGs. Part of the
issue is that it is often the same people involved across the board.
Again, Jari has good stats on that in terms of number of docs/author and
range of WGs - RAI folks on the top of the list again. So, I don't know
that IETF can solve this one given the voluntary nature of the job and
the current industry/day job pressures on many of the key contributors.
[/MB]
d/
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf