From: a.h.s. b. <spu...@no...> - 2002-11-16 22:46:15
|
Hey folks -- I run a PHP-based web site called the Radicalendar (http://www.radicalendar.org/), and since the release of Apple's iCal, I've been developing a new version that will enhances its compatibility with iCalendar format. (New version isn't publicly-accessible yet). Specifically, I'm 1) making it capable of output in iCalendar format, so users can "subscribe" to the various calendars, and 2) I'm using iCalendar's well-defined recurrence rules to define an interface for creating recurring events (which I didn't accommodate before). So I'm approaching iCalendar from the opposite direction of your excellent PHP iCalendar work, as I've not been concerned with _reading_ iCalendar format files. I thought we could perhaps share information or problems with each other, since I know of no other discussion lists for iCalendar (are there any? I'd love to know about them). In particular, I've run into one ambiguous situation regarding recurring events, where RFC2445 isn't clear, and Apple has chosen an interpretation that I can't find support for in the spec. To wit: If I create an event in iCal on Nov. 15th, and set it to recur DAILY every 1 DAY for 3 TIMES, iCal creates events on Nov 16, Nov 17, and Nov 18. So far so good. If I then modify (for "Only This Event") the subject of each one of the events on Nov 16,17,18, I am left with the parent event (on the 15th) and 3 _modified_ recurrence instances (which are thereafter unaffected by new changes to the parent). If I then select the parent, and change the recurrence schedule to WEEKLY, every 1 WEEK, still for 3 TIMES, iCal will add events on Nov 22, Nov 29, and Dec 6. I now have _6_ children for the parent event (3 earlier modified, 3 brand new). According to my interpretation of RFC2445, the COUNT parameter should define the TOTAL number of recurrence instances, so now we have too many events. (What "should" happen according to my understanding is that NO new events are created, despite the change in recurrence rule, because we already have 3 recurrence instances). iCal appears to understand this when, if you change the parent event's recurrence rule back to DAILY every 1 DAY for 4 TIMES, it adds only 1 new event, on the 19th, because the earlier 3 children fall on the proper days in the recurrence set. This exact situation is ambiguous in RFC2445, but did Apple intentionally make the decision to interpret it in this manner, or was this an oversight? How would you interpret the "proper" action to take, and why? Cheers, spud. ------------------------------------------------------------------- a.h.s. boy spud(at)nothingness.org "as yes is to if,love is to yes" http://www.nothingness.org/ ------------------------------------------------------------------- |