[07:14:51] --- oliver88 has joined
[07:15:03] --- oliver88 has left
[09:27:52] --- mikeh has joined
[10:12:14] --- mikeh has left
[10:12:58] --- mikeh has joined
[10:13:32] --- pigdog has joined
[10:21:48] --- pigdog has left
[10:23:57] --- Eliot Lear has joined
[10:30:27] --- Eliot Lear has left: Logged out
[10:32:15] --- pigdog has joined
[10:38:44] --- bernard.desruisseaux has joined
[10:39:08] --- pigdog has left
[10:39:43] --- Eliot Lear has joined
[10:41:45] --- Eliot Lear has left: Logged out
[10:42:20] --- pigdog has joined
[10:50:36] <pigdog> hello. our session will begin in 10 minutes
[10:55:29] --- alexeymelnikov has joined
[10:56:09] <pigdog> we'll start in about 5 minutes
[10:58:28] --- Aki has joined
[10:58:56] <Aki> Hello all.
[11:00:41] --- cyrus_daboo has joined
[11:01:34] <alexeymelnikov> Hi
[11:01:50] --- thnetila has joined
[11:02:17] <pigdog> hi everyone
[11:02:22] <thnetila> HI
[11:02:26] <pigdog> i've been typing into alexey's window for a good minute or three
[11:02:32] <pigdog> that was really useful for all of you
[11:02:42] <pigdog> so welcome to our second jabber chat
[11:02:53] <pigdog> last week we went through the first 11 issues
[11:03:07] <pigdog> we didn't close all of them, of course, but we made good progress
[11:03:15] <pigdog> we need to review some of them, and then go on to issue number 12
[11:03:22] <Aki> Link: http://www.ofcourseimright.com/cgi-bin/roundup/calsify
[11:03:23] <pigdog> for those who want to follow along within the issue tracker
[11:03:31] <pigdog> Aki's just conveniently read my mind
[11:03:47] <pigdog> ok, so, before i continue anyone want to bash the agenda?
[11:04:03] <pigdog> ok then.
[11:04:16] <pigdog> moving on back through the issue list, starting with issue number 5
[11:04:52] <pigdog> i currently have this as "Cyrus to raise discussion"
[11:05:01] <pigdog> Cyrus, do you wish to raise any discussion?
[11:05:12] <pigdog> title is "proposed changes to RRULE / EXDATE"
[11:05:37] <cyrus_daboo> hang on a sec..
[11:05:41] <pigdog> hanging
[11:06:07] <cyrus_daboo> I think the proposed wording is fine.
[11:06:20] <cyrus_daboo> So close this one out
[11:06:24] <pigdog> ok, would you kindly send that note to the list? then i think we can move on
[11:06:53] <pigdog> ok, any other comments on this issue?
[11:07:16] <pigdog> ok, going back (hopefully briefly) to issue 11
[11:07:51] <pigdog> i've just discovered that bernard is not receiving messages
[11:08:00] <pigdog> perhaps we should do a role call
[11:08:04] <bernard.desruisseaux> The first msg I got from you is "(11:06:27) pigdog: ok, going back (hopefully briefly) to issue 11"
[11:08:16] <pigdog> Aki, still there?
[11:08:21] <Aki> I'm here.
[11:08:24] <pigdog> alexy?
[11:08:28] <pigdog> sorry- alexey?
[11:08:34] <alexeymelnikov> Yes, I am here
[11:08:41] <cyrus_daboo> ping from Cyrus
[11:08:48] <pigdog> anyone else?
[11:08:58] <pigdog> ok, authors and chairs are on. we're moving along
[11:09:00] <mikeh> yep
[11:09:09] <cyrus_daboo> also someone from Kerio
[11:09:39] <pigdog> on issue 11, there was a lot of discussion relating to whether or not there was an actual bug, but we have no text. i don't think we think there's a real bug in the normative text
[11:10:21] <pigdog> i think bernard was going to consider adding a clarifying example. this is where we left things
[11:10:27] <pigdog> is everyone okay with that resolution?
[11:10:48] <cyrus_daboo> YEs
[11:10:53] <alexeymelnikov> I am
[11:10:57] --- oliver88 has joined
[11:11:10] <pigdog> no objections? we'll move along. hi oliver and welcome
[11:11:11] <bernard.desruisseaux> yep
[11:11:20] <cyrus_daboo> FYI - whilst on RRULEs, can I do a quick report from Calconnect?
[11:11:35] <pigdog> sure. we're between issues
[11:11:40] <pigdog> would you like to do so now?
[11:11:53] <pigdog> assuming nobody objects?
[11:12:00] <thnetila> I'm here, but I have a phone call...
[11:12:06] <cyrus_daboo> There was discussion at Calconnect in the Mobile technical committee about iCalendar simplicfication
[11:12:27] <cyrus_daboo> We had people from the OMA DS group presentg (symc ML + mobile devices)
[11:13:07] <cyrus_daboo> They made it clear that some form of simplification of RRULEs really was needed to help out mobile devices and they were very interested in Lisa's proposal.
[11:13:50] <cyrus_daboo> We were going to follow up with them in the next few weeks and try and get a better feel for what they think. It might be worth adding some time to the IETF meeting to discuss this again in more detail based on their feedback.
[11:13:50] --- pigdog has left
[11:14:18] <cyrus_daboo> At this point though, I think we were leaning more towards a 'profile' of rules as opposed to actually modifying the base spec.
[11:14:35] <cyrus_daboo> Anyway, just a heads-up on what was talked about.
[11:14:36] <Aki> Sure, I think we can afford agenda time for that.
[11:14:41] <cyrus_daboo> Side-show over!
[11:15:17] <cyrus_daboo> OK _ I will see if I can get a couple of slides together.
[11:15:32] <Aki> Seems we lost Eliot...
[11:16:15] --- pigdog has joined
[11:16:33] <Aki> Ok, he's back.
[11:16:52] <pigdog> wow. nothing like a chair getting blown away
[11:16:59] <pigdog> thanks cyrus for that.
[11:17:40] <pigdog> so, based on cyrus's discussion, we will open up a generic issue on simplification i think
[11:17:52] <pigdog> and we do need to more seriously discuss lisa's proposal
[11:17:59] <pigdog> (IMHO)
[11:18:14] <pigdog> does anyone have any additional questions or comments?
[11:18:17] <cyrus_daboo> Issue is simplification vs profiles
[11:18:55] <pigdog> cyrus: we can discuss that in the issue, but i see where you
[11:18:58] <pigdog> you're going
[11:19:01] <alexeymelnikov> Profiles are fine. If it gets better deployment, it might replace the original spec
[11:19:02] <cyrus_daboo> Profiles has implications for iTIP in terms of negotiating profiles between different end points. I will put all these issues together for the files.
[11:19:13] <cyrus_daboo> s/files/slides/
[11:19:21] <pigdog> classically the IETF likes to shy away from profiles
[11:19:33] <pigdog> this having been said, in this case it may well make sense
[11:19:43] <cyrus_daboo> True - but lemonade has gone down that route.
[11:19:52] <pigdog> but we could get ourselves into the following dilemna:
[11:20:27] <pigdog> if we do a profile and it turns out to be SORT OF popular then if someone else does ANOTHER profile we end up an interoperability mess
[11:20:46] <pigdog> because you get one group of people testing against one profile and another group of people testing against another
[11:20:49] <Aki> Agreed. Personally, I'm a bit on the fence about doing profiles, or "cherry-picking" in general. Had bad experiences in the past.
[11:20:52] <alexeymelnikov> IMHO, profiles should be nested
[11:21:11] <pigdog> so this means that we need to tread very carefully
[11:21:28] <pigdog> let's have this as a topic for the next jabber chat if people are up for another one
[11:21:30] <cyrus_daboo> Right - multiple profiles are bad. Just one profile for 'mobile' devices would be OK, and would give intermediaries (e.g. sync agents) a reasonable chance of ensuring interop.
[11:21:41] <cyrus_daboo> Ok - lets move on.
[11:21:43] <pigdog> ok
[11:21:47] <pigdog> this brings us to Issue 12
[11:21:56] <pigdog> VTIMEZONE error in RFC 2445?
[11:22:39] <bernard.desruisseaux> I don't believe the RRULE in the VTIMEZONE are broken.
[11:23:16] <bernard.desruisseaux> We just need to clarify how BYDAY works with FREQ=YEARLY;BYMONTH=xx
[11:23:32] <pigdog> and i think we sort of just discussed that in issue 11, right?
[11:24:22] <bernard.desruisseaux> yes
[11:24:24] <pigdog> and i believe the first sentence in the 2nd to the last paragraph is true-
[11:24:44] <cyrus_daboo> The table I did of when a BYxxx part expands or contracts the rule would help here. Werer you planning on having soemthing like that in 2445bis, Bernard?
[11:25:06] <pigdog> "If this is correct and can be generalized, then it implies that BYDAYmodifiers apply to the current set of evaluated occurrences, not to theFREQ"
[11:25:13] <bernard.desruisseaux> Yes. We should add the table in the spec
[11:25:39] <pigdog> ok, that brings us back to issue 11. shall i annotate it accordingly and are there any objections?
[11:25:49] <cyrus_daboo> Perhaps with this RRULE as an example of showing how BYMONTH contracts in one case, but expands in the other.
[11:25:57] <bernard.desruisseaux> I think you can close issue 12 as a duplicate of issue 11.
[11:26:07] <pigdog> i think i agree.
[11:26:31] <cyrus_daboo> Agreed.
[11:27:11] <pigdog> ok, issue closed.
[11:27:22] <pigdog> Issue 13
[11:27:35] <pigdog> confusion in DTSTART text about "if specified" and BY rules
[11:27:41] <pigdog> this smells like a duplicate as well
[11:27:58] <bernard.desruisseaux> Not sure.
[11:28:38] <bernard.desruisseaux> In my opinion "if specified" doesn't make sense given that DTSTART should be REQUIRED whenever RRULE is used.
[11:28:51] <pigdog> ok, but let's also look at issue 25.
[11:29:21] <bernard.desruisseaux> Well... issue 25 was not entered properly in the tracker ;-)
[11:29:34] <bernard.desruisseaux> Issue 25 is about EXRULE and not RRULE.
[11:29:43] <pigdog> ah. operator error by your's truly? never happens ;-)
[11:29:54] <bernard.desruisseaux> Can you please fix the title of issue 25?
[11:29:57] <pigdog> ok, i'll fix that later. back to issue 13
[11:30:25] <cyrus_daboo> IMHO - just remove "if specified"
[11:30:28] <bernard.desruisseaux> I agree that issue 13 and 25 are related. You should add a note to that effect
[11:30:42] <pigdog> i agree with cyrus AND bernard ;-)
[11:30:46] <pigdog> objections?
[11:31:28] <pigdog> ok, moving on
[11:31:37] <pigdog> Issue 14
[11:31:38] <bernard.desruisseaux> hold on.... how do we proceed for issue 13?
[11:31:47] <pigdog> Remove "if specified"
[11:31:52] <bernard.desruisseaux> :-)
[11:31:57] <bernard.desruisseaux> I mean about the mailing list...
[11:32:12] <bernard.desruisseaux> The URL in the tracker points to the ietf-calendar mailing list... and is broken.
[11:32:41] <pigdog> i'll fix the url but i believe the text indicates the problem. am i wrong?
[11:32:41] <bernard.desruisseaux> Listen... I'll send an email to calisify with this proposal...
[11:32:47] <pigdog> ok
[11:32:47] <bernard.desruisseaux> and Cyrus will reply with +1
[11:32:58] <pigdog> i will be sending minutes of this and the last chat tomorrow
[11:33:23] <pigdog> if there is a complaint about any particular action we can revisit
[11:33:37] <pigdog> in ietf mailing list rules :-)
[11:33:42] <pigdog> issue 13, please
[11:33:45] <pigdog> 14 that is
[11:34:03] <pigdog> meaning of DTSTART/DTEND with recurrence rules
[11:35:11] <pigdog> he's right but i don't know what text he wants changed
[11:35:42] <pigdog> suggestions as to how to proceed?
[11:36:27] <pigdog> am i still live?
[11:36:34] <alexeymelnikov> Yes you are
[11:36:54] <pigdog> thanks alexey - now i'm paranoid about it... "Is this thing on? can you hear me now??"
[11:37:23] <pigdog> ok, i'm going to annotate this that we need a more clearer textual change before proceeding
[11:37:24] <cyrus_daboo> We need to precisely specify the algorithm for generating the set of recurrences given DTSTART, RRULE, RDATE, EXRULE, EXDATE.
[11:37:32] <bernard.desruisseaux> Issue 14 is a duplicate of issue 25.
[11:37:37] <cyrus_daboo> Right now 2445 uses text description for that.
[11:37:56] <cyrus_daboo> How about a list with a set of steps?
[11:38:29] <pigdog> Okay, but I think the confusion in this issue is what does DTSTART in the text refer to?
[11:38:34] <cyrus_daboo> 1. Generate the set of dates from each RRULE using DTSTART as the initial instance for each.
[11:38:42] <cyrus_daboo> 2. Add each RDATE.
[11:39:12] <cyrus_daboo> 3. Generate the set of dates for each EXRULE using DTSTART as the starting instance, but excluding DTSTART from that set.
[11:39:25] <cyrus_daboo> 4. Subtract the set in (3) from the total set from (1) & (2).
[11:39:42] <cyrus_daboo> 5. Remove all EXDATEs from the total set.
[11:39:53] <cyrus_daboo> Does that cover it?
[11:40:17] <pigdog> ok, now that *is* issue 25, and perhaps if we add what you've just outlined, cyrus we close both and then some
[11:40:20] <cyrus_daboo> I think that makes sure that EXRULE does not always exclude the DTSTART, but does allow EXDATE to exclude the DTSTART instance.
[11:40:34] <bernard.desruisseaux> I think so, but it should be noted by all that in #3 the part (, but excluding DTSTART from that set.) is NEW.
[11:40:56] <cyrus_daboo> Right - that portion of #3 addresses your original issue.
[11:41:15] <bernard.desruisseaux> But I'm still concerned that it might open some other issues! :-)
[11:41:37] <pigdog> does the text cause a change in behavior in today's implementations?
[11:41:39] <cyrus_daboo> Such as?
[11:41:52] <bernard.desruisseaux> pigdog: Yes, this is a change.
[11:42:03] <pigdog> what is the change?
[11:42:14] <cyrus_daboo> FYI we should request that Calconnect does an interop test for this case of EXRULE vs DTSTART and see what breaks or does not conform to how we think it should work.
[11:42:37] <bernard.desruisseaux> Cyrus, I'm not concerned about breaking existing implementations...
[11:42:44] <pigdog> (i am)
[11:42:58] <pigdog> are there any that this would break?
[11:43:06] <pigdog> i think cyrus your suggestion is a good one
[11:43:11] <cyrus_daboo> Who actually generates EXRULEs today?
[11:43:19] <cyrus_daboo> I suspect very few if any.
[11:43:35] <bernard.desruisseaux> I don't have a use case right now...but when I talked about this change to people here somebody came up with a weird use case (which I forgot) which was ambiguous...
[11:43:52] <bernard.desruisseaux> Cyrus: I think you are correct.
[11:44:03] <pigdog> what i suggest is that you cyrus propose the text on the mailing list, and let's see if anyone objects
[11:44:11] <bernard.desruisseaux> Oh yes!
[11:44:17] <bernard.desruisseaux> I remember!!!
[11:44:25] <bernard.desruisseaux> It had to do with INTERVAL
[11:44:28] <cyrus_daboo> Alternative: just do away with EXRULE as a simplification.
[11:44:40] <bernard.desruisseaux> If EXRULE doesn't take DTSTART as the first instance...
[11:44:49] <bernard.desruisseaux> we need to clarify how to handle INTERVAL with EXRULE
[11:45:36] <pigdog> if few implementations use EXRULE, perhaps that is worth deprecating
[11:45:48] <pigdog> also closing 25
[11:45:53] <bernard.desruisseaux> I've been told that there are VTIMEZONE that are using EXRULE...
[11:46:09] <pigdog> right
[11:46:11] <pigdog> i could see that
[11:46:18] <pigdog> well
[11:46:21] <cyrus_daboo> Ugh - this is going to get nasty - dumping EXRULE would really be better at this point. I don't think anyone will be sorry to see it go...
[11:46:21] <pigdog> i take it back
[11:47:00] <pigdog> it sounds like we have a candidate for dumping, but this is also a question for calconnect that they have probably already answered in fact
[11:47:21] <pigdog> how about this?
[11:47:23] <cyrus_daboo> The VTIMEZONEs could always be re-written using some other rules.
[11:48:35] <cyrus_daboo> What we would do is deprecate use of EXRULE in 2445bis with a strong statement to no longer generate it because of ambiguities as to how DTSTART etc should be treated. Implementations can still accept it and do what they currently do with the understanding that they may not interoperate because of the problems.
[11:48:58] <pigdog> does anyone object to cyrus' proposal?
[11:49:22] <pigdog> cyrus can you take it to the list?
[11:49:38] <cyrus_daboo> OK - I will start a thread on removing EXRULE.
[11:49:46] <Aki> Cyrus: SOunds good to me
[11:49:49] <pigdog> thank you. this covers issue 25
[11:50:22] <pigdog> as for issue 14. no text no movement
[11:50:57] <pigdog> ok
[11:51:00] <pigdog> moving on
[11:51:16] <pigdog> Issue 16
[11:52:23] <pigdog> do people agree with his interpretation?
[11:53:58] <pigdog> anyone have a comment on this issue?
[11:54:04] <bernard.desruisseaux> Yes!
[11:54:10] <cyrus_daboo> There are several ways to interpret this, but I think the current 2445 way is good enough if its clear. An exampe such as the one he presented is fine.
[11:54:11] <bernard.desruisseaux> THe example he gives is foo bar!
[11:54:21] <cyrus_daboo> Why is it foo bar?
[11:54:23] <bernard.desruisseaux> > DTSTART;TZID=Europe/Vienna: 20050201T120000 > RRULE:FREQ=MONTHLY;BYMONTHDAY=30;FREQ=3 > results in occurrences on 1 Feb 2006, 30 May 2006, 30 Aug 2006, 30 Nov 2006, > 30 May 2006.
[11:54:30] <bernard.desruisseaux> FREQ=3 ?
[11:54:40] <cyrus_daboo> He meant INTERVAL=3
[11:54:52] <pigdog> let's assume that he did mewan INTERVAL
[11:55:10] <bernard.desruisseaux> and you are missing a COUNT or UNTIL
[11:55:33] <pigdog> sorry - COUNT, not INTERVAL
[11:55:39] <cyrus_daboo> Sure, the example was not complete - it can be fixed.
[11:55:55] <pigdog> that's an interesting example.
[11:56:33] <pigdog> do we want to include a fixed version?
[11:57:01] <cyrus_daboo> I think it helps.
[11:57:16] <pigdog> aha! a volunteer to clean it up!
[11:57:24] <pigdog> fair enough, cyrus?
[11:57:59] <pigdog> "is this thing on?"
[11:58:15] <Aki> pong
[11:58:23] <alexeymelnikov> Yes
[11:58:56] <pigdog> ok, we'll need to stop here. i'm going to mark cyrus as the owner of this one, but if he wants to pawn it off on someone else that's fine too.
[11:59:01] <cyrus_daboo> Cleaned up:
[11:59:04] <cyrus_daboo> Example: DTSTART;TZID=Europe/Vienna: 20060201T120000 RRULE:FREQ=MONTHLY;INTERVAL=3;COUNT=5;BYMONTHDAY=30 results in occurrences on 1 Feb 2006, 30 May 2006, 30 Aug 2006, 30 Nov 2006, 30 May 2007.
[11:59:23] <pigdog> ok
[11:59:26] <pigdog> thanks all for coming
[11:59:34] <bernard.desruisseaux> I think we need more than an example. We need to clarify that what it means to ignore rule part with respect to COUNT *and* INTERVAL.
[11:59:35] <pigdog> try again this time next week?
[11:59:53] <cyrus_daboo> OK TO NEXT WEEK.
[11:59:56] <bernard.desruisseaux> Fine with me.
[12:00:14] <Aki> Thanks all!
[12:00:16] <pigdog> thanks!
[12:00:18] <alexeymelnikov> Ok to next week as well
[12:00:21] <oliver88> bye everybody
[12:00:23] <pigdog> bye for now!
[12:00:25] --- pigdog has left
[12:00:43] --- Aki has left
[12:00:43] --- oliver88 has left
[12:00:50] <cyrus_daboo> See ya!
[12:01:15] <cyrus_daboo> PS Since I just got the notice, anyone planning on going to the IETF social?
[12:02:21] <alexeymelnikov> I haven't decided yet.
[12:03:20] --- mikeh has left
[12:08:59] --- alexeymelnikov has left: Replaced by new connection
[12:18:47] --- cyrus_daboo has left
[12:30:35] --- thnetila has left
[13:27:43] --- oliver88 has joined
[13:36:29] --- oliver88 has left
[19:49:57] --- oliver88 has joined
[19:50:09] --- oliver88 has left