~bzr-pqm/bzr/bzr.dev : changes

~bzr-pqm/bzr/bzr.dev » Changes from revision 4543.2.15

Rev	Summary	Authors	Date
4543.2.15	NEWS about fixing bug #393349 NEWS about fixing bug #393349	John Arbash Meinel	15 years ago
4543.2.14	Clarify some comments, fix up a debugging change. Clarify some comments, fix up a debugging change.	John Arbash Meinel	15 years ago
4543.2.13	Remove an pdb statement. Remove an pdb statement.	John Arbash Meinel	15 years ago
4543.2.12	Always sort the inventories in pure topological order Always sort the inventories in pure topological order The revisions may be in a slightly different sort order because the target wants to come last, but the inventories can be different (hopefully).	John Arbash Meinel	15 years ago
4543.2.11	Use a fixed file-id for the root to make the tests more stab... Use a fixed file-id for the root to make the tests more stable.	John Arbash Meinel	15 years ago
4543.2.10	Clarify a comment. Clarify a comment.	John Arbash Meinel	15 years ago
4543.2.9	Down to 2 failing tests. Down to 2 failing tests. Rather than using get_inventory_xml, switch to get_record_stream, but use a helper because otherwise it just looks crummy. The get_invalid_bundle code seems to be a bit fragile, we might consider doing it differently.	John Arbash Meinel	15 years ago
4543.2.8	Add a custom get_invalid_bundle and allow BadBundle to be Add a custom get_invalid_bundle and allow BadBundle to be a valid way to handle an invalid bundle. The code was already checking that applying the mpdiff was giving the correct sha1sum, and computing a custom value is very difficult.	John Arbash Meinel	15 years ago
4543.2.7	It turns out CHKSerializer was inheriting from xml5 It turns out CHKSerializer was inheriting from xml5 However, xml5 does not support rich roots. So it was inherently a bad xml format for trasmitting rich-root information. I don't believe the XML format was used elsewhere (there was some code in StreamSource that used it for a while, but that is currently replaced by the InterDifferingSerializer code path, and in the future we will be switching to inventory deltas anyway.)	John Arbash Meinel	15 years ago
4543.2.6	redefinning self.bzrdir_format() automatically sets the defa... redefinning self.bzrdir_format() automatically sets the default format.	John Arbash Meinel	15 years ago
4543.2.5	Fix issues with keys/ids and ghost handling. Fix issues with keys/ids and ghost handling.	John Arbash Meinel	15 years ago
4543.2.4	Start working on code that will use Repository._serializer.w... Start working on code that will use Repository._serializer.write_inventory_to_strig. The idea is that we upcast the CHK inventory back into an XML representation, and then we shove that into an mpdiff based bundle (same as current) The main reason is so that we can maintain cross-format compatibility for bundles.	John Arbash Meinel	15 years ago
4543.2.3	Change the name to test_merge_directive Change the name to test_merge_directive Start using the 'support_altered_by_hack' as it applies to the same formats as a hypothetical 'inventory fulltexts are complete' flag. (alternatively, we could use repository._format.supports_chks, etc.) Almost there.	John Arbash Meinel	15 years ago
4543.2.2	work out some tests that expose that bundles don't work w/ 2... work out some tests that expose that bundles don't work w/ 2a formats. Turns out you have to disable/clear the chk_map cache, because otherwise the target repository pulls the root chk page out of the cache, rather than seeing it doesn't have the page in the repository.	John Arbash Meinel	15 years ago
4543.2.1	Fix some comments in the send code. Fix some comments in the send code.	John Arbash Meinel	15 years ago
4543	(robertc) Fix bzr mv to not take out branch locks, (robertc) Fix bzr mv to not take out branch locks, fixing it when the branch is readonly. (Robert Collins, bug 216541)	Canonical.com Patch ...	15 years ago
4542	(andrew) Improve error message in osutils.file_relpath. (andrew) Improve error message in osutils.file_relpath.	Canonical.com Patch ...	15 years ago
4541	(andrew) Fix spurious 'does not support stacking' warning wh... (andrew) Fix spurious 'does not support stacking' warning when pushing.	Canonical.com Patch ...	15 years ago
4540	(andrew) Make test_remote a little clearer by adding an (andrew) Make test_remote a little clearer by adding an assertFinished method.	Canonical.com Patch ...	15 years ago
4539	(robertc) Fix bugs 187207, (robertc) Fix bugs 187207, 282402 - small but annoying bugs in commit. (Robert Collins	Canonical.com Patch ...	15 years ago

Older »