Recurring events with timezone

Hendrik Helwich h.helwich at tarent.de
Mon Oct 11 09:53:13 CEST 2010



Am Sonntag, 10. Oktober 2010 21:59:37 schrieb Gunnar Wrobel:
>[..]
> I don't see a direct limitation there. The conversion to UTC is
> certainly possible and valid while you are doing it to and from one
> timezone. I see a problem once you are looking at the same event with
> a client in a different timezone.
>
> And to be honest I believe there are still some significant problems
> when it comes to timezones on the Kolab server.
>
> But maybe you can detail what the exact problem that you are having
> with the conversion is.

Hi Gunnar,

ok i should have made an example:
* person A has a client which stores events in iCalendar data (like evolution) 
and creates a recurring event which occurs every Monday at 8:00 in the german 
time zone
* person B has a kolab client like kontact and needs to attend this event. The 
time of this event need to be converted to UTC format due to the kolab 
format. So in the kolab server the start time is stored with 6:00 UTC due to 
the german time zone difference of two hours from UTC time.
* person B is also in the german time zone. The kolab client of person B will 
add 2 hours to the stored UTC time and the shown start time will be 8:00. So 
everything is fine till now
* In germany at 31.10.2010 the difference of the local time to UTC changes 
form 2 to 1 hour due to the daylight saving time.
* the shown start time of the event will be still 8:00 for person A but 7:00 
for person B because the kolab client will only add one hour to the stored 
UTC time 6:00.

> I could imagine that you might also be able to solve your specific
> problem by adding a custom attribute to the event XML.

It would be possible to preserve the time zone information of calendar data by 
adding custom tags.
Unfortunately this would not solve the problem because the time zone 
information would not be considered by a kolab client if time zone data is 
nor part of the kolab format.

Greets,
Hendrik


> Cheers,
>
> Gunnar




More information about the format mailing list