When accepting an ics invitation send by mail in KMail and putting it in a CalDAV resource, the HTTP request is either dropped (e.g. with owncloud) or causes a complete de-sync (e.g. with tine20). Example recurrence id causing described behaviour: RECURRENCE-ID;TZID=W. Europe Standard Time:20141218T160000 Resulting HTTP request: 192.168.1.11 - user1 [05/Dec/2014:10:23:41 +0000] "PUT /calendars/da39a3ee5e6b4b0d3255bfef95601890afd80709/40/e3b0c442-98fc-1c14-9afb-f4c8996fb9242014-12-04T16:00:00+01:00.ics HTTP/1.1" 500 2219 "-" "Mozilla/5.0 (X11; Linux x86_64) KHTML/4.13.3 (like Gecko) Konqueror/4.13" Example of recurrence id not causing malformed HTTP request: RECURRENCE-ID; Resulting HTTP request: 192.168.1.11 - user1 [05/Dec/2014:10:23:41 +0000] "PUT /calendars/da39a3ee5e6b4b0d3255bfef95601890afd80709/40/e3b0c442-98fc-1c14-9afb-f4c8996fb924.ics HTTP/1.1" 204 1976 "-" "Mozilla/5.0 (X11; Linux x86_64) KHTML/4.13.3 (like Gecko) Konqueror/4.13" Reproducible: Always Steps to Reproduce: 1. Send yourself email with ics attachment containing "RECURRENCE-ID;TZID=W. Europe Standard Time:20141218T160000" 2. Accept invitation in KMail and choose a CalDAV calendar to record the event. 3. Instant de-sync (open "Akonadi Configuration" KDE Configuration Module to see for yourself) Actual Results: CalDAV calendar resource de-syncs. Expected Results: A correctly formed HTTP request being sent, resulting in an event recorded in the CalDAV calendar. I don't know how to honestly assess the severity. It kind of makes me lose data, so could be considered "critical", but is in my opinion at least more sever than "normal".
Related: https://forge.tine20.org/view.php?id=10602
Hey, Could you attach an ICS with which to reproduce this? I can't craft one with owncloud and vim it seems :/ Cheers, Grégory
Before I do that an update: 4.14.6 does not show this behaviour anymore. There is still a problem, though. As I understand it it is the following: You can have recurring events which would be ics files with an RRULE. You might create an exception to the rule, which would be an ics file with a RECURRENCE-ID. Following the RFCs an ics with a RECURRENCE-ID is nothing if it does not have a corresponding ics with an RRULE. Some Microsoft services send ics files which have an RECURRENE-ID without providing the main recurrent event. All RFC obeying software has problems dealing with that. A possible solution would be if Akonadi would automatically create a base event whenever it encounters an ics file with RECURRENCE-ID that has none, yet. This would also ensure consistency when those two incidences are pushed to other services via CalDAV.
Created attachment 93351 [details] recurrence-id ics This is an ics file which led to the described behaviour. As stated it's not as bad anymore as of 4.14.6, but an event as such still does not show up in the calendar view.
Hey, I'm not really a big fan of having the resource creating a new event that will never be sent back to the server. It's not how Akonadi is supposed to work and the design doesn't support this easily. Re-assigning to KOrg to see if something can be done at this level to have the event displayed. Cheers, Grégory
And also updating the status to Confirmed, sorry for the noise :)
As this bug changed quite considerably, I opened a new one at https://bugs.kde.org/show_bug.cgi?id=349993 There is also a python script attached which filters out RECURRENCE-IDs for quick mitigation.
This bug has never been confirmed for a KDE PIM version that is based on KDE Frameworks (5.x). Those versions differ significantly from the old 4.x series. Therefore, I plan to close it in around two or three months. In the meantime, it is set to WAITINGFORINFO to give reporters the oportunity to check if it is still valid. As soon as someone confirms it for a recent version (at least 5.1, ideally even more recent), I'll gladly reopen it. Please understand that we lack the manpower to triage bugs reported for versions almost two years beyond their end of life.
Just as announced in my last comment, I close this bug. If you encounter it again in a recent version (at least 5.1 aka 15.12; preferably much more recent), please open a new one unless it already exists. Thank you for all your input.