>If you have a complex, infinite RRULE I think it would be very difficult to convert
>from one VTIMEZONE to another.
>
>Even if it is possible, do we really want iCalendar to be this complex?Hence my suggestion to restrict the range of the VTIMEZONE to only that which bounds the affected entries. For the more common "Lets meet for the next 25 Mondays" case this works fine. I think the "Lets meet every Monday and Thursday from now until Armagedon" is less likely. As such, matching is going to be something that should be tried and will probably have a good chance of working; esp. if we tightly restrict VTIMEZONEs from being infiinte (or even 500 years longer than the entries repeat set).
Anyone got a prelim copy of Dougs tool that did Olsen->iCalendar? I looked and couldn't find it. If so, I suggest they post it to the list so we can all start w/a common set of values. Perhaps you can post any tools you have already made to do this...Something which the olsen tzdata mailing list is all too familiar with is that timezone definitions continually change. My particular burden is Brazil which changes the dates on which they switch to DST on a yearly basis. So, if there was a conference call in Sao Paulo on Mondays at 2pm that I want to dial into regardless of what time it is where I am, I am going to have to realize that the VTIMEZONE rules will change during the life of the event. I sure hope you didn't convert the event into UTC when storing it...
dmadeo
Attachment:
smime.p7s
Description: S/MIME Cryptographic Signature