Bazaar Explorer crashes during Branch operation

Bug #493644 reported by David Roberts
6
This bug affects 1 person
Affects Status Importance Assigned to Milestone
Bazaar Explorer
Confirmed
High
Unassigned

Bug Description

To recreate:

Create an empty Win32 directory.

Open that directory with BE as a virtual repo.

Initialise a plain branch in the previously empty directory (= virtual repo)

create a single minimal test file in that branch and commit it.

From the virtual repo viewer pane, select the newly populated branch, select BRANCH (bottom right), and
branch to a second location at the same level within the virtual repo. Decline the option to create a
shared repo.

The branch operation appears to proceed normally, then throws a windows error dialog box which states
"Friendly distributed version control system has encountered a problem and needs to close. We are sorry for the inconvenience."
The 'more information' relates to an error in bzr.exe, ModName: qtgui4.dll

This is the logging for the commit and branch -

Mon 2009-12-07 16:32:23 +0000
0.047 bzr arguments: [u'qsubprocess', u'--bencode', u'l6:commit2:-m22:commit to plain branch6:READMEe']
0.047 looking for plugins in C:/Documents and Settings/djr/Application Data/bazaar/2.0/plugins
0.047 looking for plugins in C:/localapp/Program Files/Bazaar/plugins
0.125 encoding stdout as osutils.get_user_encoding() 'cp1252'
0.187 bzr arguments: [u'commit', u'-m', u'commit to plain branch', u'README']
0.187 encoding stdout as osutils.get_user_encoding() 'cp1252'
0.234 opening working tree 'F:/Branchbug/Plainbranch'
0.312 preparing to commit
[ 4028] 2009-12-07 16:32:23.749 INFO: Committing to: F:/Branchbug/Plainbranch/
0.359 Selecting files for commit with filter [u'README']
[ 4028] 2009-12-07 16:32:23.765 INFO: added README
[ 4028] 2009-12-07 16:32:24.265 INFO: Committed revision 1.
0.953 return code 0
19.876 return code 0
414.170 check paths: None
422.904 opening working tree 'F:/Branchbug/Plainbranch'
498.656 encoding stdout as sys.stdout encoding 'cp850'

Mon 2009-12-07 16:33:50 +0000
0.047 bzr arguments: [u'qsubprocess', u'--bencode', u'l6:branch24:F:/Branchbug/Plainbranch23:F:/Branchbug/copybranch18:--use-existing-dire']
0.047 looking for plugins in C:/Documents and Settings/djr/Application Data/bazaar/2.0/plugins
0.047 looking for plugins in C:/localapp/Program Files/Bazaar/plugins
0.141 encoding stdout as osutils.get_user_encoding() 'cp1252'
0.188 bzr arguments: [u'branch', u'F:/Branchbug/Plainbranch', u'F:/Branchbug/copybranch', u'--use-existing-dir']
0.188 encoding stdout as osutils.get_user_encoding() 'cp1252'
0.235 opening working tree 'F:/Branchbug/Plainbranch'
0.282 opening working tree 'F:/Branchbug/Plainbranch'
0.516 creating repository in file:///F:/Branchbug/copybranch/.bzr/.
0.860 Using fetch logic to copy between CHKInventoryRepository('file:///F:/Branchbug/Plainbranch/.bzr/repository/')(<RepositoryFormat2a>) and CHKInventoryRepository('file:///F:/Branchbug/copybranch/.bzr/repository/')(<RepositoryFormat2a>)
1.235 creating branch <bzrlib.branch.BzrBranchFormat7 object at 0x00ED4BD0> in file:///F:/Branchbug/copybranch/.bzr/
1.735 created new branch BzrBranch7('file:///F:/Branchbug/copybranch/')
1.953 opening working tree 'F:/Branchbug/copybranch'
[ 5516] 2009-12-07 16:33:53.204 INFO: Branched 1 revision(s).
2.485 return code 0
501.234 opening working tree 'F:/Branchbug/copybranch'

So, it seems to be the re-opening of the new branch that fails, not the branching as such.

Re-running BE, re-entering the virtual repo and re-trying the Branch operation to a new destination in the same location
SUCEEDS without error. (Also, the 'failed' branch appears to be intact and usable.)

This is with BE 0.10.0-dev [r326] and Bazaar 2.1.0b3 installed from the setup.exe Win32 installer. It also occured
with the BE 0.9.0 code bundled with that installer. QBzr is 0.16.0

Changed in bzr-explorer:
importance: Undecided → High
status: New → Confirmed
Revision history for this message
David Roberts (smartgpx) wrote :

This problem does not seem to be reproducable with bzr-2.1.0rc1 and explorer 1.0.0dev [r395]

Revision history for this message
Ian Clatworthy (ian-clatworthy) wrote :

My guess is that this may have been the segfault/freeze bug. I'll mark it as a duplicate. Please undo that if you think otherwise and can make it happen again.

To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.