[horde] Active sync: multiple calendars?

[Gunnar Wrobel]

Quoting Michael J Rubinsky <mrubinsk at horde.org>:

> Quoting Gunnar Wrobel <wrobel at horde.org>:
>
>> Quoting Michael J Rubinsky <mrubinsk at horde.org>:
>>
>>> Quoting Andreas Ntaflos <daff at pseudoterminal.org>:
>>>
>>>> Hi,
>>>>
>>>> is it possible to access and sync multiple Horde calendars (on the same
>>>> user account) with the new ActiveSync plugin? Or is this not supported
>>>> by design or specification?
>>>
>>> By specification. The AS protocol does not support multiple  
>>> calendars for a single user.
>>
>> But shouldn't there be ways to aggregate calendars? I'm pretty  
>> certain the ActiveSync support on the Kolab server allows selecting  
>> several.
>
> Sure, it would be possible to essentially combine all events from  
> multiple calendars into a single AS collection. However, there are a  
> number of issues that I ran into while trying to do this when I  
> originally wrote the library:
>
> There is no way to differentiate the calendars on the client (AS  
> only supports a single "calendar" folder) so no calendar names, no  
> horde-centric calendar colors (in fact some clients might color the  
> events based on the tags) - all events look like they are in the  
> user's "main" calendar.
>
> There is no way to mark events as read only, no way to map back an  
> event to the calendar it came from in the case that multiple  
> calendars may have the same event uid (though this is not specific  
> to AS), and no way on the client to select the calendar to add a new  
> event to (all events are assumed to be on the user's "main"  
> calendar). This last point _might_ also cause issues with both event  
> invitations and recurring event exceptions on calendars other than  
> the user's "main" calendar (don't have code or protocol docs in  
> front of me at the moment to verify this one).
>
> There would also be issues when a user changes which calendars they  
> want included after the initial setup (would basically require the  
> user or admin to explicitly reset the AS account to force a complete  
> replacement sync). For example - a user initially sets up two  
> calendars, but then decides to remove one. All the events from the  
> second calendar will still be present on the device. Similarly,  
> adding a second calendar would not sync the events that were  
> created/modified prior to the last successful sync time.
>
> Yes, some of these issues could be at least partially worked around  
> by e.g., limiting the calendars allowed to be synched to calendars  
> owned by the user, or adding user pref for the calendar to use when  
> creating new events on an AS device, but these rather feel like  
> hacks instead of implementations.
>
> If in fact kolab does allow this, I'd be interested in both how the  
> kolab/activesync connector implements this and how robust this  
> actually is.

This is definitely on my list once I can say that Horde works fully  
with Kolab again (hopefully soon). And your list of problems will be  
what I will work through with their solution. Thanks for posting it! I  
didn't test ActiveSync with the Kolab code either yet.

Cheers,

Gunnar

> If you have any links to code or similar, I'd love to take a look.  
> This is a sore point for me as well though even if we get minimal  
> support for this, would likely not help me since it's calendars I  
> have read only access to that I need to include (see note above  
> about read only issues).
>
> Also, it's worth noting that I'm aware that Google's "activesync"  
> support allows this - but this is not true activesync; they have  
> added features to the protocol and these only work if using a Google  
> client, not a generic activesync client.
>
> -- 
> mike
>
> The Horde Project (www.horde.org)
> mrubinsk at horde.org
>
> -- 
> Horde mailing list
> Frequently Asked Questions: http://horde.org/faq/
> To unsubscribe, mail: horde-unsubscribe at lists.horde.org

-- 
Core Developer
The Horde Project

e: wrobel at horde.org
t: +49 700 6245 0000
w: http://www.horde.org

pgp: 9703 43BE
tweets: http://twitter.com/pardus_de
blog: http://log.pardus.de