4251
Comment:
|
3951
|
Deletions are marked like this. | Additions are marked like this. |
Line 17: | Line 17: |
=== assert svn_revprops.has_key(properties.PROP_REVISION_DATE) === | === Empty parent '...' added, but child '...' wasn't added !? === * https://code.launchpad.net/~vcs-imports/fun3compiler/trunk * https://code.launchpad.net/~vcs-imports/kdesupport/trunk === CHKInventoryRepository(...) has no revision ... === * https://code.launchpad.net/~vcs-imports/subvertpy/trunk-svn Fixed in bzr-svn trunk. === bzrlib.errors.InconsistentDeltaDelta === * https://code.launchpad.net/~gary/zc.buildout/trunk == broken svn server == === [codeplex.com] assert svn_revprops.has_key(properties.PROP_REVISION_DATE) === |
Line 40: | Line 54: |
== Launchpad code bugs == === Launchpad uses "guessed" layout rather than a layout that matches the requested branch === * https://code.launchpad.net/~vcs-imports/soprano/trunk * https://code.edge.launchpad.net/~vcs-imports/project-neon/trunk |
|
Line 52: | Line 72: |
=== revprop_list PROPFIND failed === Succeeds for me locally, so perhaps the issue is fixed in a newer subversion/subvertpy/bzr-svn than the importds are running. * https://code.launchpad.net/~vcs-imports/almidon/trunk |
== Needs investigation == |
Line 56: | Line 74: |
== Needs investigation == === Empty parent '...' added, but child '...' wasn't added !? === * https://code.launchpad.net/~vcs-imports/fun3compiler/trunk |
== timing out == |
Line 62: | Line 78: |
* https://code.launchpad.net/~vcs-imports/kdesupport/trunk At least some of these are fixed in bzr-svn 1.0.4. === CHKInventoryRepository(...) has no revision ... === * https://code.launchpad.net/~vcs-imports/subvertpy/trunk-svn Fixed in bzr-svn trunk. |
|
Line 77: | Line 85: |
=== bzrlib.errors.InconsistentDeltaDelta === * https://code.launchpad.net/~gary/zc.buildout/trunk === bzr-svn's hardcoded layout handler for the KDE repository is making unwarrantedly strict decisions on what constitutes a branch? === * https://code.launchpad.net/~vcs-imports/soprano/trunk * https://code.edge.launchpad.net/~vcs-imports/project-neon/trunk This is a bug in the way Launchpad opens bzr-svn branches. -- jelmer |
Overview / analysis of failing bzr-svn imports
An attempt to classify persistently failing bzr-svn code imports by cause.
Contents
Bazaar does not support something present in Subversion
Unable to convert Subversion path ... because it contains characters invalid in Bazaar.
https://code.launchpad.net/~vcs-imports/choco/trunk -- suspended... should we suspend a failure???
https://code.launchpad.net/~vcs-imports/enlightenment/trunk -- suspended... should we suspend a failure???
bzr doesn't support newlines in symlink targets yet
https://code.launchpad.net/~vcs-imports/freeswitch/trunk -- suspended... should we suspend a failure???
bzr-svn bugs
Empty parent '...' added, but child '...' wasn't added !?
CHKInventoryRepository(...) has no revision ...
Fixed in bzr-svn trunk.
bzrlib.errors.InconsistentDeltaDelta
broken svn server
[codeplex.com] assert svn_revprops.has_key(properties.PROP_REVISION_DATE)
This is caused by a server side bug present in the Microsoft Team Foundation Server to Subversion bridge running on codeplex.com.
I'm inclined to close these as wontfix since they are a server side issue. Perhaps we can look into importing natively from the team foundation server using bzr-tfs ? -- jelmer
Subversion bugs
get_log REPORT failed
https://code.launchpad.net/~vcs-imports/greenstone/greenstone3
https://code.launchpad.net/~vcs-imports/chromium-browser/trunk
get_dir PROPFIND failed
Launchpad code bugs
Launchpad uses "guessed" layout rather than a layout that matches the requested branch
Some analysis done, culprit uncertain
get_uuid PROPFIND failed
This failure mode seems to correspond with being unable to make a successful PROPFIND against the repository root URL. This can be because:
- There is some sort of erroneous redirect occurring:
- Only a subtree of the repository is visible, and accessing the root causes a 401:
Needs investigation
timing out
SSL Verification error: signed using insecure algorithm
Checksum mismatch... possible corruption in source repository?