Merge lp:~vila/bzr/1020007-branchformat4-lock into lp:bzr
| Status: | Merged |
|---|---|
| Approved by: | Jelmer Vernooij on 2012-07-02 |
| Approved revision: | 6535 |
| Merged at revision: | 6534 |
| Proposed branch: | lp:~vila/bzr/1020007-branchformat4-lock |
| Merge into: | lp:bzr |
| Diff against target: |
149 lines (+69/-11) 4 files modified
bzrlib/branch.py (+1/-11) bzrlib/plugins/weave_fmt/branch.py (+56/-0) bzrlib/tests/per_branch/test_config.py (+8/-0) doc/en/release-notes/bzr-2.6.txt (+4/-0) |
| To merge this branch: | bzr merge lp:~vila/bzr/1020007-branchformat4-lock |
| Related bugs: |
| Reviewer | Review Type | Date Requested | Status |
|---|---|---|---|
| Jelmer Vernooij (community) | 2012-07-02 | Approve on 2012-07-02 | |
|
Review via email:
|
|||
Commit Message
Properly save ``branch.conf`` changes when unlocking a BzrBranch4.
Description of the Change
I ran into bug #1020007 while working on a plugin: the all-in-one format has
an issue with the way the last unlock is recognized for ``branch.conf``
saving.
This went unnoticed as no test was explicitly testing BranchFormat4.
I've added such a test and split the lock_{read_write} and unlock methods
betwenn Branch and BzrBranch4.
This makes the weave_format plugin sole responsible for the specific lock
strategy and leave the Branch class cleaner.
Since the bug was introduced with the bug #948339 fix, only bzr.dev needs to
be fixed.
Since BzrBranch4 is not widely (to say the least) used anymore, I doubt
anyone encounter the bug. Yet, it has the potential to break backwards
compatibility and looks like a good cleanup to have (and completes the work
to isolate weaves into their decicated plugin).
| Vincent Ladeuil (vila) wrote : | # |
| John A Meinel (jameinel) wrote : | # |
sent to pqm by email
| Vincent Ladeuil (vila) wrote : | # |
sent to pqm by email
| John A Meinel (jameinel) wrote : | # |
sent to pqm by email

sent to pqm by email