Re: WG Agendas to submit only 2 days after -00 submission cut-off date

[Mary Barnes <[e-mail]>]

On Sun, Jul 21, 2013 at 12:55 PM, DRAGE, Keith (Keith) <
[e-mail]> wrote:

> ********
>
> What I am not buying is your contention that confusion does not exist
> because dispatch has different deadlines.****
>
> ** **
>
> I’d note the words ****
>
> ** **
>
> “in cases where documents are clearly under the purview for a specific
> WG, then it's not an issue to go directly to the WG” ****
>
> ** **
>
> Do not actually appear in the DISPATCH charter – which refers generally to
> “new work” without any further clarification. ****
>
> ** **
>
> So we are left having to interpret the overlap between charters of
> existing WG, which presumably have some ownership and DISPATCH.****
>
> ** **
>
> So for discussions that fall in this overlap, the fact that DISPATCH has
> different deadlines adds to the confusion.****
>
> ** **
>
> And I can list a number of documents in the past where the author has
> wavered between DISPATCH and other working groups, and apparently ended up
> going for the wrong working group.
>
[MB] And, that's why if you are uncertain, then bring it to DISPATCH.  It's
also *extremely* useful to follow the naming convention and include
-dispatch - it makes it much easier for the chairs to notice and keep an
eye on new work.

DISPATCH has had different deadlines for 4+ years.  The reason we
implemented DISPATCH is because the SIPPING WG clearly showed us that it is
extremely ineffective to have a kitchen sink WG.  And, WG shopping just
wastes everyone's time, which is what we had ALOT of before we chartered
DISPATCH  We've done quite well with the WGs that we've chartered for
discrete work items.

You are correct that there can be lack of clarity when things *might* look
like SIPCORE or MMUSIC or AVT/EXT, etc.  In those cases, unless you've
talked to the ADs and they've said to take the item to a specific WG, you
should bring it to DISPATCH.   The motivation behind the early deadline is
there is often little value in discussing a draft that is submitted at the
-00 deadline as there is rarely enough time to determine if there is any
interest in the topic.  The whole point was to get away from allocating
agenda time just because someone submitted a draft.  Perhaps you've
forgotten the insanity we had in SIPPING prior to DISPATCH - it was insane.
 We had WG items with 5 minutes of agenda time, we had work items that
totally languished, etc.  If you have better suggestions as to how to
manage the volume of new work that comes into the RAI area, including work
items that overlap WGs, then please post those on the RAI area mailing list
where that discussion has already been started.
[/MB]

> ****
>
> ** **
>
> Keith****
>
> ** **
>   ------------------------------
>
> *From:* Mary Barnes [mailto:[e-mail]]
> *Sent:* 18 July 2013 20:28
> *To:* DRAGE, Keith (Keith)
> *Cc:* Flemming Andreasen; [e-mail]
>
> *Subject:* Re: WG Agendas to submit only 2 days after -00 submission
> cut-off date
> ****
>
>  ** **
>
> On Thu, **Jul 18, 2013** at **2:10 PM**, DRAGE, Keith (Keith) <
> [e-mail]> wrote:****
>
>  Sorry – not buying this.****
>
>  [MB] Then you should raise the issue to the ADs.  I'm the implementer of
> a process that is in place for managing work in their area. [/MB] ****
>
>   ****
>
> Your comment is only applicable if you buy the argument that all new
> proposals, whether in scope or out of scope of another RAI working group,
> go to DISPATCH, and that currently does not appear to be the case. So there
> is the case of a document being submitted, and then some evaluation occurs
> that maybe it ought to have been in DISPATCH, after the DISPATCH deadlines
> have passed.****
>
>  [MB] As I said previously, in cases where documents are clearly under
> the purview for a specific WG, then it's not an issue to go directly to the
> WG.  BTW, I have no idea what document you're referring to. And, keep in
> mind that the process requires notifying the chairs, so it may not be
> visible to you what discussions might occur offline. [/MB] ****
>
>  Even I am not sure if a document should be in DISPATCH or not. If I read
> follow the DISPATCH charter alone, then there are a number of other items
> on the MMUSIC agenda that have never been dispatched to MMUSIC (or anywhere
> else), and therefore ought to be in DISPATCH. They do fall in scope of the
> MMUSIC agenda. I could point this out in a number of other working groups.
> ECRIT even has stuff on the agenda that is not in the scope of ECRIT, has
> never been through DISPATCH and so on.****
>
>  [MB]  I only see one document on the MMUSIC WG agenda that probably
> should have first gone to DISPATCH and actually, the MMUSIC chairs could
> have raised that issue and suggested such to the document authors rather
> than give agenda time when they certainly do have really high priority,
> critical items to discuss.   The document I am referring to relates to at
> least two different WGs (including one that is now closed) and wasn't at
> all an obvious fit for MMUSIC.    In addition, as I said, it should have
> been obvious that there was a fairly clear expectation that  the websocket
> and alternate transport docs are first discussed in  DISPATCH - e.g.,
> BFCPbis, mrsp websocket document, etc.     ****
>
> [/MB]****
>
>   ****
>
> If the deadlines for DISPATCH had not been different, then maybe there
> would have been different issues, but at least it is the different DISPATCH
> deadlines that have currently ruled out the discussion of this document in
> ****Berlin****.****
>
>  [MB] That's absolutely correct - that's the whole intent of the DISPATCH
> deadlines - these authors simply had to send an email to the DISPATCH WG
> chairs by June 24th indicating they were putting together a draft that was
> an alternative to a previously discussed draft.  They didn't do so.  If
> they had submitted the draft by July 1st, there would have been some
> flexibility, but that wasn't the case.  There were other people that did
> so, that also requested agenda time that were not given agenda time because
> it wasn't deemed necessary. AGain, you can progress work without discussing
> at a f2f meeting. ****
>
> ** **
>
>  If you have issues with the DISPATCH process, then please bring those to
> the DISPATCH WG or better yet RAI area mailing list.  We've had this
> process in place for 4 years so the concept of notifying the chairs several
> weeks before draft deadlines has been there the whole time. Indeed, we had
> the dates for IETF-87 much later than usual.   Keep in mind as well that
> there is NOTHING in the DISPATCH/RAI process that requires a document to
> have f2f discussion.  Additionally, keep in mind that the document in
> question has normative dependencies such that this document cannot be
> published until those docs are completed. [/MB] ****
>
>   ****
>
> Regards****
>
>  ****
>
> Keith****
>
>  ****
>   ------------------------------
>
> *From:* [e-mail] [mailto:[e-mail]] *On
> Behalf Of *Mary Barnes
> *Sent:* 18 July 2013 19:28
> *To:* Flemming Andreasen****
>
>
> *Cc:* [e-mail]
> *Subject:* Re: WG Agendas to submit only 2 days after -00 submission
> cut-off date****
>
>  ****
>
> I have on comment below on an example cited as to the problems with the
> single deadline. ****
>
>  ****
>
> Mary.****
>
>  ****
>
> On Wed, **Jul 17, 2013** at **6:27 PM**, Flemming Andreasen <
> [e-mail]> wrote:****
>
>
> On **7/17/13** **12:14 AM**, Ted Lemon wrote:****
>
> On **Jul 16, 2013**, at **9:03 PM**, Flemming Andreasen <
> [e-mail]> wrote:****
>
> You risk a lot of other pushback when you start making up your own
> WG-specific rules (and rightfully so).****
>
> It's hard to understand what you might mean by this.****
>
> When you have a single set of consistent rules that are written down, it's
> easy for everybody to understand and accept. When the rules are WG specific
> and are created at the discretion of the WG chairs it becomes more of a
> challange:
> 1) Chairs now have the ability to define rules that influence desired
> outcomes (which is not to say they will)
> 2) People may believe that chairs in fact exercise 1), especially when
> things don't go as they like.
>
> Neither is desirable.****
>
>  ****
>
> All true in theory, but this is not how it happens and works in practice.
> I prefer being pragmatic. It worked reasonably well the way it was before;
> it's not working well the way it's being done now.****
>
> Can you cite an example of an actual problem that has occurred because of
> this change?****
>
>  Sure. I'm travelling this week and had to scramble last night to look at
> several -00 drafts that were requested to be on the agenda. I had to get on
> a ** 5:30 am** call this morning to coordinate the draft agenda with my
> co-chair due to the impending draft agenda deadline. The result feels a bit
> rushed with little time to dig into exactly what the issues are in the
> drafts and hence what to prioritize and with how much time allocated.
>
> Another example involves a -00 draft that was submitted a little more than
> a week ago to DISPATCH. The authors weren't quite sure which group it
> belonged to, asked for an MMUSIC slot, and it then bounced around between
> several WG chairs. By the time we all agreed it indeed did belong to
> DISPATCH, it was all too late because DISPATCH hadn't received an agenda
> request in time.****
>
> [MB] This one is not a particularly good example in terms of problems
> caused by the deadlines.  This is an example of someone not following the
> RAI/DISPATCH WG process.  It should have been no surprise to the authors
> that the document should be discussed first in the DISPATCH WG, as that is
> our basic process unless it's beyond obvious that it is related
> specifically to an existing chartered WG.  In addition, these authors (one
> of whom is not an inexperienced IETFer) should have been aware of how we
> were handling documents of a similar nature as we dispatched several at the
> last IETF meeting.  I think this is an example of WG shopping and not of
> too late deadlines.  In particular given that the DISPATCH deadline
> required that notification of intent to submit a document and desire to
> discuss a topic in dispatch was June 24th, with charter proposals/problem
> statements due July 1st. This document was submitted well over a week after
> the last deadline.  There's nothing wrong with a WG chair flat out saying
> that a document wasn't submitted in time to garner enough consideration to
> warrant agenda time at the f2f meeting.  Again, the RAI/DISPATCH process
> allows progression of documents without ever having been discussed at a f2f
> meeting.  We should continue discussion of situations like this on the RAI
> area list where there is discussion of process/structural issues that are
> currently being encountered:****
>
> http://www.ietf.org/mail-archive/web/rai/current/msg01384.html****
>
> It is unfortunate for you (and not at all your fault) that your WG is in
> the middle of the concerns raised, but I think the draft deadlines are the
> least of your problems. ****
>
> [/MB]****
>
>  ****
>
>
> We can all agree that none of this would happen if people submitted their
> drafts earlier. Human nature and work commitments being what they are,
> earlier draft deadlines seem to work a lot better at facilitating that.
>
> Thanks
>
> -- Flemming
>
>
>
> ****
>
>   ****
>
>  ** **
>