Closed Bug 326332 Opened 19 years ago Closed 18 years ago

Event creation in ics calendar fails after backup problems.

Categories

(Calendar :: Internal Components, defect)

defect
Not set
normal

Tracking

(Not tracked)

VERIFIED FIXED

People

(Reporter: ssitter, Unassigned)

Details

(Whiteboard: [swag:.5d][maybe have patch])

Attachments

(1 file)

Event creation in ics calendar fails after backup problems.

1. Start Sunbird with a new profile.
2. Delete the default storage calendar.
3. Create an ics calendar (I tested with a remote calendar at file:///c:/test.ics)
4. Create an event via drag'n'drop in week view. The following error appeared in console:

Backup failed in asyncOpen:[Exception... "Component returned failure code: 0x80520012 (NS_ERROR_FILE_NOT_FOUND) [nsIChannel.asyncOpen]"  nsresult: "0x80520012 (NS_ERROR_FILE_NOT_FOUND)"  location: "JS frame :: file:///D:/Programme/sunbird/components/calICSCalendar.js :: anonymous :: line 683"  data: no]

5. Create more events. 
In %SunbirdProfile%\backupData backup files are created named calBackupData_xxxx_edit-1.ics to calBackupData_xxxx_edit-3.ics. After creation of the next event the following error is shown four times in JavaScript console:

Error: [Exception... "Component returned failure code: 0x80520015 (NS_ERROR_FILE_ACCESS_DENIED) [nsIFile.remove]"  nsresult: "0x80520015 (NS_ERROR_FILE_ACCESS_DENIED)"  location: "JS frame :: file:///D:/Programme/sunbird/components/calICSCalendar.js :: purgeBackupsByType :: line 548"  data: no]

Afterwards it is not possible to create new events. Restart of Sunbird required.

Tested on windows with 2006-02-07 sunbird trunk build.
I've seen at least the access denied thing not too long ago; confirming.
Status: UNCONFIRMED → NEW
Ever confirmed: true
Is this still a bug?  I can't reproduce on linux using comment #0.
(In reply to comment #2)
I can still reproduce using Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9a1) Gecko/20060713 Sunbird/0.3a2+.

Error: [Exception... "Component returned failure code: 0x80520015 (NS_ERROR_FILE_ACCESS_DENIED) [nsIFile.remove]"  nsresult: "0x80520015 (NS_ERROR_FILE_ACCESS_DENIED)"  location: "JS frame :: file:///D:/sunbird/components/calICSCalendar.js :: purgeBackupsByType :: line 620"  data: no]
The bugspam monkeys have struck again. They are currently chewing on default assignees for Calendar. Be afraid for your sanity!
Assignee: base → nobody
I typically have to restart Sunbird on Windows and Mac on the order of 10 times a day.  This generally happens I edit multiple events in quick successionon a WebDAV  calendar.  There are really two bugs here: one is probably in nsLocalFile* and the other is that we should catch any errors thrown by nsLocalFile* and do something reasonable. I believe the bug in nsLocalFile is already filed.
Flags: blocking0.3?
OS: Windows 2000 → All
Hardware: PC → All
Flags: blocking0.3? → blocking0.3+
Whiteboard: [swag:.5d]
Attached patch maybe patch — — Splinter Review
I still can't reproduce this bug, so this is just a best-guess at a patch.  The idea here is that failing to remove a file is the least severe kind of error we can have.  (When were more backups not a good thing?)  Also, we'll eventually get to clear it up, on the next round of backups.
Attachment #233028 - Flags: second-review?(dmose)
Attachment #233028 - Flags: first-review?(mattwillis)
Whiteboard: [swag:.5d] → [swag:.5d][maybe have patch]
Comment on attachment 233028 [details] [diff] [review]
maybe patch

This at least catches the failure. That's good.

Should we at least send this to the error console rather than swallowing it entirely?
(In reply to comment #7)
> (From update of attachment 233028 [details] [diff] [review] [edit])
> This at least catches the failure. That's good.
> 
> Should we at least send this to the error console rather than swallowing it
> entirely?
> 
As I said in my comment, this strikes me as one of the absolutely least critical failure modes imaginable.  At best, I think we should dump it to the regular console.  I'm open to arguments why that's wrong.
Comment on attachment 233028 [details] [diff] [review]
maybe patch

Okay. I don't think this will hurt anything, and if it keeps people from having to restart, then it's at least forward progress.

r=lilmatt
Attachment #233028 - Flags: first-review?(mattwillis) → first-review+
Comment on attachment 233028 [details] [diff] [review]
maybe patch

r=dmose
Attachment #233028 - Flags: second-review?(dmose) → second-review+
Patch checked in.  I'm marking this fixed since it's on the blocking radar.  If there are other problems, please file followups.  Let's reserve this bug for backup-errors of the severity listed here.
Status: NEW → RESOLVED
Closed: 18 years ago
Resolution: --- → FIXED
VERIFIED with Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.9a1) Gecko/20060901 Calendar/0.3a2+
Status: RESOLVED → VERIFIED
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Creator:
Created:
Updated:
Size: