There are several changes merged to 1.9.x branch since 1.9.0-rc2:
* r1686554, r1686557, r1686239, r1686541, r1686543, r1686802
Fix 'svnadmin hotcopy' for read-only FSFS repositories
Justification:
Format 7 repositories could not be hotcopied without write access to
the source repo - which is a regression vs. older formats. Because
the new test case also uncovered another regression with hotcopying
the rep-cache.db from r/o repos.
* r1682714, r1682854, r1683126, r1683135, r1683290
Fix segfaults in FSX's directory processing code.
Justification:
Despite its experimental state, FSX shall not segfault the server.
* r1686478, r1686888, r1686984
Make 'blame -g' work with old clients against new servers.
Justification:
Without this patch, old clients will "lose track" of what changes
happened in -g mode and produce wrong / worse blames than against
old servers.
The output of 'blame -g' is only an approximation. However, the
new server would cause much worse results in old clients especially
in simple cases where lines of development are kept in close sync.
* r1685085
Install svnbench as part of 'make install'.
Justification:
svnbench moved from tools/ to subversion/ so it should be installed
by default.
* r1683378
Prevent a possible FSFS repository corruption with power or network disk
failures during 'svnadmin pack'.
Justification:
Repository corruption/data loss.
* r1683303
Resolve a race condition in some test suite cleanup code.
Justification:
Without this patch running the testsuite on bdb x svnserve consistently
fails on at least some Windows test systems. bdb x serf sometimes
triggers the same problem, but in far less cases.
.
This specific test -unlike other tests- cleans up its own environment a
few times to retry some scenarios on the same paths.
* r1684325, r1684344
Fix an unintended doubling of error messages in 'svnadmin verify'
without the --keep-going flag.
Justification:
Fixes a bug that was introduced in r1683311, which was
already merged to 1.9.x.
* r1684412
Make JavaHL native code compile with a C++11 compiler.
Justification:
The upcoming MSVC14, part of Visual Studio 2015,
supports user-defined literals, which means that
trying to build JavaHL with VS2015 will fail.
* r1684077
Detect invalid svndiff data earlier.
Justification:
Do not waste memory on invalid user or server input.
* r1684322
Fix a minor omission in JavaHL's VersionExtended class:
the dispose() and finalize() methods were missing.
Justification:
Without the dispose() method, the user of the VersionExtended
class cannot avoid a memory leak in native memory.
* r1684034
Fix prop_tests.py#42 test failures with non-US default locale on
Windows [1].
Justification:
Tests should not fail.
* r1683311
Make error reporting from svn_repos_verify_fs3 consistent
with behaviour of svn_repos_verify_fs2 in 1.8.x.
Justification:
Errors from the FS implementation are squashed even in normal
mode without --keep-going. See: http://s.apache.org/j9Z
* r1683387
Fix abort() in svn-status handler on platforms that doesn't support C99
format specifiers for strftime(): %F and %z are new in C99 and C89
compiler/runtime doesn't support them.
Justification:
Server-side crash in new small feature. Simple fix.
* r1683071
Document the meaning of "XFAIL" for users building from source.
Justification:
Answers a FAQ.
These commits contain public API changes and the community guide
states that such changes should be considered as potentially
destabilizing.
So maybe it's time to release RC3 and restart the soak period?
--
Ivan Zhakov
