432417 - calRecurrenceInfo::onStartDateChange leaves wrong RECURRENCE-IDs

Status: VERIFIED FIXED

(Calendar :: Internal Components, defect)

Description

[Daniel Boelzle [:dbo]]

Attachment: test ics file

spin-off bug from bug 368976

When adopting RECURRENCE-IDs of overridden items, those have to be tracked in timezone of the master's DTSTART timezone, otherwise they potentially don't match anymore.

To reproduce:
1. Subscribe to the attached test ics file.
2. Modify the whole series' to start e.g. 5 hours later.
=> in effect, the RECURRENCE-ID of the overridden item doesn't match anymore, you see two occurrences for 30th of march, the overridden one starting on 29th and a proxy starting on 30th.

Comment 1

[Daniel Boelzle [:dbo]]

Attachment: fix

Comment 2

[Philipp Kewisch [:Fallen][📱📆]]

Didn't you mention something about this not working for RDATEs?

Comment 3

[Daniel Boelzle [:dbo]]

Comment on attachment 319548 [details] [diff] [review]
fix

Yes, reading <http://tools.ietf.org/rfcmarkup?doc=draft-ietf-calsify-rfc2445bis-08#section-3.8.4.4> again, I think it's valid that RDATE instances could be overridden (RECURRENCE-ID == RDATE).

Comment 4

[Daniel Boelzle [:dbo]]

Attachment: fix - v2

this should do for RDATEs, too

Comment 5

[Philipp Kewisch [:Fallen][📱📆]]

Comment on attachment 330541 [details] [diff] [review]
fix - v2


>+        for (var i in ritems) {
>+            var ritem = ritems[i];
Any reason not to use for each (var ritem in ritems) ?


r=philipp

Comment 6

[Daniel Boelzle [:dbo]]

No, there is not (I just took over most of the code); changed to use for-each loop.

Checked in on HEAD and MOZILLA_1_8_BRANCH => FIXED.

Comment 7

[Andreas Treumann]

Checked in mxr.mozilla.org -> VERIFIED