Bron: Reminder that this is a combined CalConnect/IETF session and is operating under the IETF NoteWell, not the standard CalConnect policies. No agenda bashing. Eventpub --- Mike lost focus over the past few months. Barry: issue with discusses that delay on. Mike - next few days, review where we got and re-post. Think we're pretty close. Barry: also Ben Kaduk's discuss. Daniel: believe we were addressing comments from Ben. ACTION: Mike to post a summary of where we are at for eventpub in the next few days. Server-side subscriptions --- Bron: I missed the call for adoption finishing on this sorry, it should be re-uploaded to calext. Mike: will upload as a calext draft. Ken: think Apple have done something similar, should check that. ACTION: Mike will post calext version of serverside subscription draft in the next day or two. ACTION: Ken will check with Apple about their serverside subscriptions. Icalendar Series --- Not much done. Bron: issue that almost everything has Mike as primary or only author. Mike: really want to move this along - it has lots of potential. Last thing - Neil came up with a good suggestion to reduce work needed. Have done some implementation, don't appear to be problems. Bron: be good to at least post again with an updated version number to refresh it. ACTION: Mike will refresh draft for icalendar series. jscalendar --- Robert: thanks Mike for all the feedback and implementation experience. Running gag - we're STILL in last call! Robert away in June, so early July target date for end of last call. EXRULE: feedback that it's complex, ambiguous, and nobody using it. Mike: most people won't create them, but for event publishing it gets quite complex and with the right tools, people could encode the rules as they understand them rather than having to manually fix things. Mike: ical4j does multiple RRULEs and multiple EXRULEs and handles them perfectly fine. Bron: libical is the other one. Does it support it? Ken: not sure Mike: probably does because 2245 handled it. Ken: know there's blocks of code that address them, but not sure what happens. Neil: could add extensions of multiple RRULEs plus EXRULEs as a separate spec. Mike: when doing mapping from ICALENDAR to JSCALENDAR, need to make it as easy as possible, which means not having things that are in the spec not have a representation. Mike: maybe a section of "currently deprecated properties". Can say "deprecated in icalendar, but here's how you do it in jscalendar. Neil: you have to define the semantics as well, which we can do - but it's weird to add it in a new spec... and what does "it's deprecated" mean? Will we remove it later? Mike: put it in the mapping draft - here's how you would store it for round trip purposes. Mike: most clients would represent it correctly in icalendar - we could check again. Robert: you can round trip unknown items into a jcal "blob" that's JSON, but if you want semantics in jscalendar then you need it to have a real mapping. Anyone else? Daniel: agree that it's weird to put a deprecated thing in. Need to understand why it was deprecated. Mike: appears to work fine in implementations which exist. Issues with RRULE. Neil: think we've fixed that now at least! We would need to define the semantics. Mike: don't want to bog JSCalendar down with the argument. Bron: Agree with the argument of "have the tool generate the EXDATEs for a published event, rather than having the EXRULE stored in the event itself". Neil: downside is that it's more complex and people won't implement. Neil: kind of think the same with multiple RRULEs. It's all implementable, but trying not to make it feel scary to people. Ken: libical supports multiple RRULEs and multiple EXRULEs. Ken: multiple RRULEs would make the size of VTIMEZONEs a lot smaller! Mike: argument against multiple RRULEs was based on what UIs do, which is bogus. It's easier to express complex rules as multiple RRULEs. Neil: if most of the client UIs support it, because the libraries do, then maybe it's OK. Ken: college class schedules for example could benefit from this. Neil: leaning towards include EXRULEs. Ken: may be worth looking at the holiday rule. Mike: maybe a good use for "related to" to point to an event that is used to provide exclusions. ACTION: Ken to test libical to see what it does with EXRULEs. ACTION: Neil/Robert to add EXRULEs to jscalendar format if needed. ACTION: more discussion on list about jscalendar and EXRULE Subscription Upgrade --- Mike: need to refresh it (both in mind and as a draft!) Next meeting is only a month from now! ACTION: Mike to refresh draft for subscription upgrade Valarm extensions --- Ken: Daniel did a a review Daniel: now that JSCalendar is close, do we add a note to all future drafts to say how to make it compatible with JSCalendar? Ken: depends where mapping draft is. Mike: probably need to check if there's anything. Ken: think most facilities in this are already in JSCalendar. Mike: yes. ACTION: Ken to read, review and submit changes to valarm extensions if needed. ACTION: Ken to double check if there's anything in valarm extensions that's not already covered by JSCalendar. vpoll: --- It's expired again. There was a big change for PARTICIPANTs instead of VVOTERs - Mike will double check. Still needs a lot of work in the iTIP area. Mike: will try to get a version out before the next meeting. There's also been some other interest in vpoll recently. ACTION: Mike to check if we've switched to PARTICIPANT rather than VOTER in vpoll, then refresh draft. Other business: --- Paul: Thunderbird is planning to move from libical to icaljs. Will ask to see what icaljs supports. Philipp is the author of icaljs. Milestones --- Scheduling Controls: ACTION: Bron to refresh before the next meeting. Jul 2020 - VALARM EXT Aug 2020 - JSCalendar Aug 2020 - Scheduling Controls Dec 2020 - Series Dec 2020 - Subscription Upgrade Dec 2020 - Serverside Subscriptions Mar 2021 - VPoll