Closed
Bug 641665
Opened 13 years ago
Closed 10 years ago
Caldav synchronization error in cached calendars
Categories
(Calendar :: Provider: CalDAV, defect)
Calendar
Provider: CalDAV
Tracking
(Not tracked)
RESOLVED
FIXED
3.3
People
(Reporter: fjmulero, Assigned: Fallen)
Details
(Whiteboard: [calconnect25])
Attachments
(1 file, 2 obsolete files)
2.85 KB,
patch
|
mmecca
:
review+
|
Details | Diff | Splinter Review |
User-Agent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_6_6; es-es) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27 Build Identifier: Mozilla/5.0 (X11; U; Linux i686; es-ES; rv:1.9.2.11) Gecko/20101013 Ubuntu/9.04 (jaunty) Firefox/3.6.11 Build Identifier: Mozilla/5.0 (X11; U; Linux i686; es-ES; rv:1.9.2.15) Gecko/20110303 Lightning/1.0b2 Thunderbird/3.1.9 If thunderbird is closed before a cached calendar is fully sincronized (not all events are retrieved from server) if nothing is changed in the server, the client don't request the rest of events. Reproducible: Always Steps to Reproduce: 1. Create a new cached caldav calendar over a high collection (1000 events for example) 2. Close Thunderbird before the load finished. 3. Open Thunderbird and reload remote calendars. Actual Results: The caldav scheduling don't follow, don't try to recover the rest of events. Expected Results: Calendar continues loading events from server. Its posible that the "ctag" of the caldav collection is saved before all the calendar reports are finished.
Reporter | ||
Updated•13 years ago
|
OS: Linux → All
Reporter | ||
Comment 1•13 years ago
|
||
This patch includes a posible solution to the bug. This solution consists in use a tentative ctag until the updateItems function finalize. When this function finalizes correctly calls to finalizeUpdatedItems, where the ctag is confirmed and saved.
Reporter | ||
Updated•13 years ago
|
Attachment #519871 -
Flags: review?(browning)
Assignee | ||
Updated•13 years ago
|
Attachment #519871 -
Flags: review?(browning) → review?(nomisvai)
Reporter | ||
Updated•13 years ago
|
Summary: Caldav scheduling error in cached calendars → Caldav synchronization error in cached calendars
Updated•13 years ago
|
Assignee: nobody → fjmulero
Status: UNCONFIRMED → ASSIGNED
Ever confirmed: true
Hardware: x86 → All
Comment 2•13 years ago
|
||
Comment on attachment 519871 [details] [diff] [review] Patch over comm-central Very sorry for the long delay in reviewing this. I think you have the right idea in how to fix this issue, however, I think the usage of mOldCtag should be completely removed in favor of the new mProposalCtag, please submit another patch and I will review it faster this time :) Thanks.
Attachment #519871 -
Flags: review?(nomisvai) → review-
Comment 3•13 years ago
|
||
Does this look more like what you're after? I still get a case where this doesn't do what I want, if the calendar retrieval fails due to an internal bug (ie, not all items actually load, but did get an appropriate 207 response) then things still go haywire in that unless there is actually a change on the server the client won't try to retrieve the calendar again.
Comment 4•13 years ago
|
||
Ok, I just noted that my whitespace is stuffed in the attached patch, not going to roll a new one without also fixing the issue. As it stands it seems calDav_finalizeUpdatedItems gets called irrespective of whether there was actually updated items or not, at least in my case (the initial sync) none of the items from the server gets loaded into the local calendar (Warning: CalDAV: Fatal Error doing multiget for JKroon), but still when next I try and force a resync I get "CalDAV: ctag matches, no need to fetch data for calendar JKroon". The output from the finalizeUpdatedItems entry code: aChangeLogListener=undefined calendarURI=http://atlantis.local.uls.co.za/caldav/jkroon/ iscached=false this.mQueuedQueries.length=13 How can I detect this case and deal with it accordingly (by either doing a fail-all and not proceeding with the sync, failing it entirely), or possibly by stopping at the first failure and "calculating" a ctag at that point?
Assignee | ||
Updated•13 years ago
|
Attachment #575775 -
Flags: review?(nomisvai)
Comment 5•13 years ago
|
||
Hi, just FYI, the "success" came from a bug in a different section as per bug 704012#c6 - so I think this patch (bar the whitespace) should be OK.
Assignee | ||
Updated•12 years ago
|
Whiteboard: [calconnect25]
Updated•10 years ago
|
Attachment #575775 -
Flags: review?(nomisvai) → review?(philipp)
Assignee | ||
Comment 6•10 years ago
|
||
I've updated the patch to the latest comm-central. It can be tested with a server that doesn't use webdav-sync, but uses ctags. For example, Zimbra servers fulfill this criterion, so if you have a yahoo account you might be able to test it.
Attachment #519871 -
Attachment is obsolete: true
Attachment #575775 -
Attachment is obsolete: true
Attachment #575775 -
Flags: review?(philipp)
Attachment #8394775 -
Flags: review?(matthew.mecca)
Comment 7•10 years ago
|
||
Comment on attachment 8394775 [details] [diff] [review] Fix - v3 Untested, but looks good codewise. r=mmecca
Attachment #8394775 -
Flags: review?(matthew.mecca) → review+
Assignee | ||
Updated•10 years ago
|
Keywords: checkin-needed
Comment 8•10 years ago
|
||
https://hg.mozilla.org/comm-central/rev/043437216602
Updated•10 years ago
|
Assignee: fjmulero → philipp
Target Milestone: --- → 3.3
You need to log in
before you can comment on or make changes to this bug.
Description
•