Jabber log from calsch IETF meeting

[17:16:49] *** chatroom for calsch working group
[17:16:49] *** This room supports the MUC protocol.
[14:24:29] <Bruce_Kahn> Pardon my dust.  IBMs firewalls are wreaking 
havoc on my Jabber connection
[17:16:49] *** The topic has been set to: calsch working group
[17:16:49] *** nsb has joined
[17:18:29] <nsb> Hi, I'll be your friendly jabber scribe today
[17:19:00] <nsb> Proposed Agenda:
[17:19:03] <nsb> 1. Preliminaries
[17:19:05] *** andreas-b has joined
[17:19:11] <nsb> 2. WG status, proposed charter revision
[17:19:15] <nsb> 3. CAP
[17:19:19] <nsb> 4. Calconnect interop report
[17:19:23] <nsb> 5. Caldav draft
[17:19:26] <nsb> 6. Other issues
[17:19:47] <nsb> Preliminaries done.
[17:19:52] <nsb> (blue sheets, etc.)
[17:20:11] <nsb> Any comments on the revised proposed charter that Bob 
posted recently?
[17:21:04] <nsb> Note that the new charter doesn't include ical 
revisions; that is envisioned for another WG.
[17:21:05] <DougRoyer> I agree - keep CALSCH until CAP is out, then new 
WG for new issues.
[17:21:33] *** lisa has joined
[17:21:50] <nsb> NSB agrees, notes that they don't have to be sequential
[17:22:05] <nsb> Ted:  Should recharter be in this WG, or new WG?
[17:22:11] *** waynet has joined
[17:22:19] <nsb> Anyone want to advocate re-using this WG?
[17:22:20] <DougRoyer> CAP in this WG
[17:22:39] *** leg has joined
[17:23:07] <nsb> NSB:  Goal of new WG would be revision of existing 
(1998) ical++ specs
[17:23:36] <nsb> Martin from Oracle agrees w/new group
[17:23:47] <nsb> Bob says it seems like consensus
[17:24:11] <DougRoyer> Does there have to be  a BOF or can it just be 
started?
[17:24:14] <nsb> and obviously individuals can start working on docs 
any time; new WG is not a prerequisite.
[17:24:51] <nsb> Ted:  Clarify -- there will be a target date for CAP, 
right?  And it will be clear to the community that this WG is not 
undertaking additional tasks.
[17:25:18] <nsb> Bob:  Ted, is it true that RFC revisions may be done 
by individuals?
[17:25:35] <nsb> Ted:  Yes, but my sense of the scope of the task is 
that you'll probably need a big enough team to make a WG useful.
[17:26:18] <DougRoyer> (yes to BOF or yes to ted?)
[17:26:28] <nsb> Ted:  (answering Doug):  A BOF isn't really necessary. 
  He recommends setting up a new list and announcing it on the existing 
list
[17:26:44] <nsb> and working on a new charter for the new group there.
[17:26:52] <DougRoyer> Great!
[17:27:32] <nsb> (Ted's Yes was to Bob's question; the later answer was 
to Doug's)
[17:28:04] <nsb> Are there any CAP-related discussions we should be 
having today???

[17:28:49] <nsb> Bob recently posted a summary of remaining issues from 
bugzilla, mostly minor.
[17:29:01] <DougRoyer> There is going to be cleanup work that should be 
wording and typos, and such. No technical changes (other than those 
posted on WG today by Bob)
[17:29:09] <nsb> Bob is bringing the bugzilla stuff up on the screen
[17:29:50] *** pbh has joined
[17:30:16] <nsb> They include:  BEEP interaction issues -- incomplete 
sentences, etc. in the BEEP profile?  Bob has asked for a BEEP-er to 
give it a look.
[17:30:35] <DougRoyer> YES - BEEP experts needed
[17:30:44] <nsb> Any comments on that one?
[17:31:37] <nsb> Next issue:  Longstanding issue of semicolon added to 
request-status
[17:31:42] <nsb> "World's most argued-over semicolon"
[17:32:00] <nsb> Chair position:  retain consistency with 2445 syntax, 
pending any further comment.  Any???
[17:33:12] <nsb> EXPAND issue -- closed pending objection
[17:33:13] *** hildjj has joined
[17:33:45] <Bruce_Kahn> Which was the EXPAND issue?  Didnt see any 
email on that.
[17:33:50] <nsb> Bruce's comments about revising status code -- 
substantial changes, anyone want to argue?
[17:34:09] <nsb> The EXPAND was your issue, Bruce, from a while back
[17:34:21] <Bruce_Kahn> Clarification: The original Feb 03 proposal was 
Dougs, not mine.  I simply remembered we forgot to actually do it
[17:34:37] <nsb> "The Use of EXPAND is still not sufficiently clear for 
all query cases"  -- Bruce Kahn, 5/13
[17:34:54] <nsb> Want to keep arguing, Bruce?
[17:35:02] *** ohm has joined
[17:35:10] <Bruce_Kahn> I think the 1 new line in CAP-13 fixed it.  
Will recheck now.
[17:36:08] <nsb> Defining the use of "*" to mean everything in the CAP 
ABNF -- 6/24 from Bruce.  Bob thinks this is a good idea.  Anyone 
opposed?
[17:36:32] <nsb> Nobody seems to be opposed.  That's the whole bugzilla 
list, folks!

[17:36:32] <DougRoyer> Reading again...
[17:36:44] <Bruce_Kahn> EXPAND fixed by "The results will be bounded by 
any date range or other limits in the query."
[17:37:29] <nsb> Bob:  Doug, any guesses how soon CAP 14 will be ready?
[17:37:34] <DougRoyer> Use of '*', it looks to me to be missing ABNF 
only - correct?
[17:37:43] <Bruce_Kahn> Clarification: The original REQUEST-STATUS 
rework propsal (Feb 03) was Dougs, not mine.  We had agreement on it I 
thought but never got done.
[17:38:07] <DougRoyer> Depending on the changes decided today in a week 
or so.
[17:38:24] <nsb> Yes, the only issue with * is missing ABNF
[17:38:49] <nsb> Bruce -- clarifying your clarification -- there is 
still an issue here or not?
[17:39:05] <Bruce_Kahn> For "*" we should add at least 1 line of text 
that says something like "All components are specified by the special 
value "*""

[17:39:12] <nsb> Bob:  Hope that CAP 14 will be ready for last call 
soon.
[17:39:21] <nsb> Doug, I assume Bruce's * comment is OK?
[17:39:28] <Bruce_Kahn> Only still an issue if the issues the propsal 
was to fix are not resolved.
[17:39:41] <nsb> Huh?  I don't understand your comment Bruce
[17:39:44] <Bruce_Kahn> CAP does not define what each "root level" 
REQUEST-STATUS class are.  it should.
[17:39:53] <Bruce_Kahn> What is a 7.x value?
[17:39:56] <DougRoyer> I think so. I'll comment on the list of that 
conflicts with anything ("*" issue)
[17:40:08] <Bruce_Kahn> etc.
[17:40:48] <Bruce_Kahn> Also: There are still 'gaps' in the numbering.  
For example, there is no 6.0 defined but we have a 6.1.  There is no 
10.0 thru 10.3 yet there is a 10.4.
[17:41:07] <nsb> Is that a problem?
[17:41:11] <Bruce_Kahn> These should be pretty straight forward to fix 
though.  A few lines in the right place should do it.
[17:41:15] <nsb> Joe H:  Space for future enhancements
[17:41:15] <DougRoyer> What do you mean by "root level" ?
[17:41:33] <Bruce_Kahn> Well, gaps are not good if you are trying to 
make a general hierarchy of code like we did in iTIP.
[17:41:42] *** hardie has joined
[17:41:54] <Bruce_Kahn> I think we called them "classes" in iTIP or 
iCalendar.
[17:42:00] <nsb> Bob nods sagely & inscrutably
[17:42:18] <Bruce_Kahn> We called 'em "classes" in iCalendar.
[17:42:36] <Bruce_Kahn> Section 4.8.8.2 Request Status, pp 134-135
[17:43:05] <nsb> There is concern that there are only 2 people who have 
an opinion on this issue.
[17:43:10] <Bruce_Kahn> Any "1.xx " has a particular meaning  as does 
any "2.xx", etc
[17:43:53] <Bruce_Kahn> Would at least like a definition for the 6.xx, 
7.xx, 8.xx .. 10.xx classes CAP is adding so its clear where future 
changes shoudl go and what they mean as a class.
[17:43:56] <nsb> Roy Fielding, JSoftware:  No way this will be a valid 
ABNF -- missing brackets, missing "or" characters.  Asks authors to do 
formal BNF checking, follow appropriate references
[17:44:16] <nsb> There are formal tools for this -- Harald Alvestrand 
has an ABNF checker.
[17:44:32] <nsb> There's a web page where you paste the grammar, get 
back results.
[17:44:42] <nsb> www.apps.ietf.org has a validator, but that's a 
different one
[17:45:04] <nsb> alvestrand.no has a tool
[17:45:18] <nsb> www.apps.ietf.org/abnf.html
[17:45:42] *** paf has joined
[17:46:11] <nsb> How hard will it be to clarify this?  (nsb)
[17:46:15] <paf> On the apps.ietf.org/abnf.html there are links to the 
abnf parser harald has.
[17:47:06] <nsb> Is there a well-formed proposal for the request-status 
changes?  Doug, are you opposed to this or merely awaiting a more 
detailed proposal?
[17:47:33] <DougRoyer> I am not opposed.

[17:47:56] <nsb> Bruce, are you willing to produce a more detailed 
proposal?
[17:48:02] <DougRoyer> I'll make a X.y section like itip.
I do not care if there a holes or gaps in the sequences.

[17:48:12] <nsb> That is, how it should be as opposed to what's wrong 
with the current v ersion.
[17:48:14] <Bruce_Kahn> Isnt the one from Dougs posting "Subject: CAP 
REQUEST-STATUS - update and proposals" sufficient?
[17:48:23] <Bruce_Kahn> Or do you mean the Class descriptions?
[17:48:42] <nsb> I'm not sure -- Doug, do you know?
[17:49:18] <DougRoyer> I think that I went through all of the codes and 
looked for duplicated and added needed ones.

[17:49:39] <DougRoyer> In the process I removed some duplicates may be 
the reason for the 'holes'.
[17:49:56] <nsb> So is this just a need for a simple "rationalization" 
of the codes, Bruce?
[17:49:56] <Bruce_Kahn> Part of that proposal was:
[17:49:57] <Bruce_Kahn> I also propose that we call all 6.x codes CMD 
or CS codes
and renumber all 6.x, 7.x, 8.x, 9.x, and 10,x codes to
be 6.x codes.
[17:51:06] <DougRoyer> I can do that, does anyone care what the final 
numbers are (6.3 vs 6.12)?
If so, submit a proposal.
[17:51:35] <nsb> NSB asks AD for clarification on how document status 
will be decided
[17:51:45] <Bruce_Kahn> No preference as to any ordering/values.
[17:53:13] <nsb> Ted explains the meaning of Proposed, Experimental, 
Informational
[17:54:53] <nsb> Ted:  The WG really ought to decide what status it 
wants.  Is there a community that finds this document & the ones it 
depends on ready to be implemented by multiple interested parties?
[17:55:34] <nsb> Barry Leiba:  Another way of putting it:  Is the WG 
proposing this as standard?
[17:55:51] <DougRoyer> Its ready to try. Like 2445-7, I suspect there 
will be changes and issues. But we have to start
[17:56:47] <nsb> Poll:  How many people have scrutinized the spec 
enough to have a strong understanding / assessment of the CAP protocol?
[17:57:03] <nsb> Two hands went up in the room, both from Oracle.
[17:57:09] <nsb> And half a hand from Cyrus Daboo.
[17:57:18] <nsb> Question:  Which finger was that?  [17:57:35] <nsb> 
Obviously Doug, you are on that list.  Anyone else on Jabber?
[17:57:48] *** hardie has left: Disconnected
[17:57:49] <DougRoyer> Me, Novel and INET in Italy in implementing.
[17:58:27] <Bruce_Kahn> Waffling...
[17:58:43] <DougRoyer> Is Lotus/IBM doing cap?
[17:59:25] <nsb> Ted:  It seems that despite the interest showed by the 
interop, there doesn't seem to be enough attention to the spec.
[17:59:50] <DougRoyer> The interop never included CAP
[18:00:24] <nsb> Oracle:  Can imagine CAP-26 before we finally get to 
RFC
[18:00:45] <nsb> Ted:  Reiterating, there are only todays' 4 issues 
plus cleanup, right?
[18:00:57] <lisa> Roy's from Day Software, not Oracle [18:01:04] <nsb> 
Oops, sorry Roy
[18:01:27] <nsb> ted makes all stand up
[18:01:58] <nsb> stay standing if you are willing to spend 5 hours on 
spec before Sept or currently have a CAP implementation or plan one.
[18:02:04] <nsb> Only people standing are two folks from Oracle.
[18:02:23] <nsb> Ted:  This seems a strong statement
[18:02:51] <nsb> Ted:  Q:  Is ten hours from Oracle folks, plus Doug's 
efforts & Bruce's kibbitzing, enough to finish CAP with a high 
confidence level?
[18:02:52] <DougRoyer> CAP is heavy. Wait until last call, that's when 
people will read it.
[18:03:08] *** hildjj has left: Disconnected
[18:03:11] <nsb> Barry:  If we think that what will come from next WG 
is better, we shouldn't make this a proposed standard.
[18:04:03] <nsb> NSB:  Is "Informational" status more appropriate?
[18:05:05] <nsb> ted:  you could do that, as a "this is how far we got" 
document.  But he advises that we first go to the mailing list with the 
same question about willingness/interest to work on CAP.  If we can't 
get enough people to seriously review the doc, we can/should go for 
informational or experimental.
[18:05:14] <nsb> Ted:  This makes me think we *should* start with a BOF 
for the new effort.
[18:05:42] <nsb> Ted:  I don't want to discourage anyone, but if we 
can't get the needed amount of effort...
[18:07:25] <nsb> Oracle:  Everyone but MS is suffering from the 
poorstandardization
[18:07:38] <DougRoyer> yes.
[18:07:47] <nsb> MS:  We're suffering too.  At least 5 different 
implementations of ical, all struggling to interoperate
[18:09:37] <DougRoyer> Still in session?
[18:09:45] <nsb> Martin/Oracle:  We're working hard to try to get a 
sense of what it will take to advance the state of the standards.  Can 
we at least agree that an access protocol is important?  Discussion 
focuses on the protocol rather than the problems the protocol is trying 
to solve
[18:10:16] <nsb> MS:  Before we can get a protocol we all agree on, we 
may need a schema we all agree on, which is why it is hard to do CAP 
before fixing ical
[18:11:11] <DougRoyer> I disagree, CAP transports iCal objects. No 
matter what their  for.
And CAP also transport scheduling (iTIP) iCal objects, not matter what 
their sequencing or content.
[18:11:20] <nsb> Cyrus:  We're talking about ical problems, but we 
haven't yet heard the interop results.  Can we hear about them?
[18:11:49] <nsb> Ted:  Before that, please agree to repeat the "who 
will commit energy" question on the mailing list.
[18:12:08] <nsb> Ted:  I think there has been a model shift, and that 
expectations have changed out from under ical.
[18:12:12] <lisa> Doug, from having looked at CAP's model earlier, 
which appears unchanged at core, I can't concur with your statement 
that CAP simply transports iCal objects.
[18:13:30] <nsb> nsb:  Lotus definitelky wants a CAPpish protocol, but 
has much more u rgent need for ical cleanup
[18:14:10] <nsb> Martin/Oracle:  It's a shame that we haven't been able 
to leverage the client efforts such as Mozilla, and that we get so many 
crappy PIM implementations
[18:14:35] <nsb> Martin:  But don't forget how much the community has 
already benefitted from ical.
[18:16:02] <nsb> Ted exits.  We're nearing the end.  But Bob asks, and 
most of us are willing to stay a bit longer.
[18:16:16] <nsb> Dave Thewlis, head of C&S Consortium, will now report 
on the recent interop event.
[18:16:36] <nsb> Held interop last Thusday & Friday.  Participants:  
IBM & Oracle.
[18:16:53] <nsb> Unlike previous interops, we tested EVERY MUST, MUST 
NOT, SHOULD, SHOULD NOT in the RFCs
[18:17:10] <nsb> Eye towards comprehensive review & their distance from 
Draft Status
[18:17:30] <nsb> Results:  Seems pretty clear that new drafts of the 
three RFC's are needed.
[18:17:48] <nsb> Pat Egen has volunteered to be editor for new drafts, 
has several volunteers to help.
[18:18:31] <nsb> Detailed results/report will be posted on the calsch 
mailing list
[18:18:45] <DougRoyer> With only two participating, the results are not 
that conclusive as to implemented features.
[18:19:22] <nsb> Results: we believe that new drafts can be written 
relatively easy that solve most of the open problems.
[18:20:54] <nsb> nsb:  I think we mostly need to chop things out & 
remove ambiguities, correct a few errors
[18:21:19] <nsb> Dave:  Pat has hopes of posting new drafts in Sept/Oct 
time frame.
[18:21:30] <nsb> ...and then another interop early next year.
[18:22:17] <nsb> nsb:  Will we really need a working group?
[18:22:33] <nsb> Bob:  Maybe not, if there's very little controersy.
[18:22:46] <nsb> Martin:  Should really be a separate list, to avoid 
complicating the CAP discussion.
[18:22:53] <nsb> nsb, Dave, Bob all agree
[18:23:02] <nsb> Bob asks Lisa about status of Caldav
[18:23:14] <nsb> Lisa;  Now that I work for OSAF, I have official 
sanction for working on caldav.
[18:23:31] <nsb> Lisa:  I've started to get feedback on the proposal & 
design choices
[18:23:46] <nsb> Lisa;  will not be just webdav, will have 
calendar-specific protocol logic
[18:24:12] <DougRoyer> Lisa: Is there a new draft coming out soon?
[18:24:41] <nsb> Lisa did one just before the deadline, got a few 
comments.  Available now.
[18:25:05] *** waynet has left
[18:25:07] <nsb> Session ends.  Thanks for playing!
[18:25:07] <DougRoyer> I did not see it on the CALSCH WG list, did it 
make the WG list?
[18:25:53] <lisa> It didn't go to the WG list
[18:25:54] *** ohm has left
[18:25:54] <nsb> Doug -- Better send emamil to Lisa and she will give 
you a copy
[18:26:05] <lisa> It's not a WG product and I didn't want to step on 
toes there
[18:26:08] <rlbob> it wasn't posted to the list, but it is in the I-D 
repository
[18:26:16] <lisa> It's an individual submission and the I-D people 
wouldn't know who to notify
[18:26:17] <paf> It's in the repository as draft-dusseault-caldav-01.txt
[18:26:22] *** lisa has left: Disconnected
[18:26:41] *** rlbob has left
[18:28:04] *** pbh has left
[18:31:05] *** andreas-b has left