[Kde-pim] KDEPIM 4.0

[Reinhold Kainhofer]

Am Freitag, 23. März 2007 schrieb mark at easymailings.com:
> Isn't this a problem of implementation rather than interface?

No, it's a more fundamental problem of the way the whole resource system is 
designed.

> I remember adding code to kTimeTracker that was reloaded the resource when
> korganizer modified the file--there were methods provided for this in the
> resource framework.

Yes, that's implemented in the resource framework, and that is causes the 
problems: The resource will be automatically reloaded when it detects a 
changed file. Now, KPilot needs to loop through all events to sync them. 
While it is in that loop, it
(1) needs to ensure that original data stays unchanged, otherwise we'll have 
data loss on the next sync (as the changes in the original data will not be 
detected, and
(2) The resource must not reload the file in the background, as this deletes 
all events (and makes all existing pointers invalid). Now, since KPilot is 
already looping through the existing events, after a reload all pointers are 
invalid, but KPilot has no way to detect this. So it will happily access the 
invalidated pointers and crash...

The other problem is with loading the whole calendar. There is currently 
nothing -- no signal, no flag, or anything else -- that indicates when the 
resource calendar has finished loading all resources. In particular, some 
resources use sync loading and simply return true for the load method to 
indicate they were successfull, while others use async loading, return true 
for the load method to indicate that the async loading was triggered. When it 
has finished, a signal is sent out.
Now the problem is that sync resources don't send out that signal, and there 
is no way to detect whether a resource is sync or async....


So, in short, there are fundamental problems in the framework design which 
should really be tackled for 4.0. Unfortunately, we all lack the time (and 
maybe also the courage?) to do that. But if we don't fix these problems for 
4.0, we'll probably have to stick with the same discouraging problems/flaws 
for a very long time...

Cheers,
Reinhold



-- 
------------------------------------------------------------------
Reinhold Kainhofer, Vienna University of Technology, Austria
email: reinhold at kainhofer.com, http://reinhold.kainhofer.com/
 * Financial and Actuarial Mathematics, TU Wien, http://www.fam.tuwien.ac.at/
 * K Desktop Environment, http://www.kde.org, KOrganizer maintainer
 * Chorvereinigung "Jung-Wien", http://www.jung-wien.at/
_______________________________________________
kde-pim mailing list
kde-pim at kde.org
https://mail.kde.org/mailman/listinfo/kde-pim
kde-pim home page at http://pim.kde.org/