You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by "Hyrum K. Wright" <hy...@mail.utexas.edu> on 2010/12/01 19:14:30 UTC
1.5.8 up for signing/testing
1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
http://people.apache.org/~hwright/svn/1.5.8/
To sign the release, please input your signatures using the script here:
http://work.hyrumwright.org/pub/svn/collect_sigs.py
(The script worked pretty well for 1.6.15, but if you discover any
bugs please let me know.)
Testing by enthusiastic users is also welcomed (but remember that this
is not yet a blessed release, with all that that implies). If you are
a package maintainer, please do not included this release in your
distribution until after it has been formally released.
I'd like to collect all the signatures in time to do a release by December 3.
-Hyrum
Re: 1.5.8 up for signing/testing
Posted by Stefan Sperling <st...@elego.de>.
On Thu, Dec 02, 2010 at 08:54:35AM -0500, C. Michael Pilato wrote:
> On 12/02/2010 07:38 AM, Stefan Sperling wrote:
> > You need to use serf 0.3.x with Subversion 1.5.
> > Subversion 1.6.x includes changes to make it compatible with newer serf
> > versions, but those haven't been merged to 1.5.x.
> > 1.5.x works fine with serf 0.3.x.
>
> Great, thanks. Now, if we wind up rolling a new release anyway, then can we
> enforce this serf version cap at build time for 1.5.x?
Makes sense to me.
Re: 1.5.8 up for signing/testing
Posted by "C. Michael Pilato" <cm...@collab.net>.
On 12/02/2010 07:38 AM, Stefan Sperling wrote:
> You need to use serf 0.3.x with Subversion 1.5.
> Subversion 1.6.x includes changes to make it compatible with newer serf
> versions, but those haven't been merged to 1.5.x.
> 1.5.x works fine with serf 0.3.x.
Great, thanks. Now, if we wind up rolling a new release anyway, then can we
enforce this serf version cap at build time for 1.5.x?
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: 1.5.8 up for signing/testing
Posted by Stefan Sperling <st...@elego.de>.
On Thu, Dec 02, 2010 at 01:11:44AM -0500, C. Michael Pilato wrote:
> On 12/01/2010 02:14 PM, Hyrum K. Wright wrote:
> > 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
> > http://people.apache.org/~hwright/svn/1.5.8/
>
> Summary:
>
> +0 to release, pending review of the ra_serf situation (see below).
>
> Platform:
>
> Ubuntu 9.04 (jaunty) Linux.
>
> Tested:
>
> 1.5.8 tarball with local dependencies (not from deps tarball)
> ((neon,serf,local,svn) x (fsfs,bdb)) + py + pl + rb + javahl
>
> Verified that 1.5.8 deps were logically equivalent to 1.5.7 deps.
>
> Results:
>
> ra_serf tests (fsfs and bdb) fail miserably due to a SEGFAULT.
> svn_ra_serf__conn_setup() is dereferencing a NULL conn->session
> when trying to pass conn->session->context to
> serf_context_bucket_socket_create(). I'm using serf
> ^/branches/0.6.x@1421 (which works fine for Subversion's trunk).
> ra_serf didn't change at all in 1.5.8, so I suspect that this
> problem exists in other Subversion 1.5.x releases, too. As ra_serf in
> 1.5.x is still highly experimental, I'm disinclined to allow this to be
> a release blocker, and would prefer to release-note that there's some
> maximum serf version that's known to work okay with this release line.
You need to use serf 0.3.x with Subversion 1.5.
Subversion 1.6.x includes changes to make it compatible with newer serf
versions, but those haven't been merged to 1.5.x.
1.5.x works fine with serf 0.3.x.
Stefan
Re: 1.5.8 up for signing/testing
Posted by "Hyrum K. Wright" <hy...@mail.utexas.edu>.
On Thu, Dec 2, 2010 at 8:06 AM, C. Michael Pilato <cm...@collab.net> wrote:
> On 12/02/2010 01:11 AM, C. Michael Pilato wrote:
>> On 12/01/2010 02:14 PM, Hyrum K. Wright wrote:
>>> 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
>>> http://people.apache.org/~hwright/svn/1.5.8/
>>
>> Summary:
>>
>> +0 to release, pending review of the ra_serf situation (see below).
>
> In light of discussion elsethread, I'm upgrading this to +1, but with the
> *recommendation* that we scuttle this release and retry with a version that
> includes the 'blame -g' fixes (and perhaps some 'configure'-time warning
> about building against a Serf newer than 0.3.1.
Let's make it so.
-Hyrum
Re: 1.5.8 up for signing/testing
Posted by "C. Michael Pilato" <cm...@collab.net>.
On 12/02/2010 01:11 AM, C. Michael Pilato wrote:
> On 12/01/2010 02:14 PM, Hyrum K. Wright wrote:
>> 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
>> http://people.apache.org/~hwright/svn/1.5.8/
>
> Summary:
>
> +0 to release, pending review of the ra_serf situation (see below).
In light of discussion elsethread, I'm upgrading this to +1, but with the
*recommendation* that we scuttle this release and retry with a version that
includes the 'blame -g' fixes (and perhaps some 'configure'-time warning
about building against a Serf newer than 0.3.1.
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: 1.5.8 up for signing/testing
Posted by "C. Michael Pilato" <cm...@collab.net>.
On 12/01/2010 02:14 PM, Hyrum K. Wright wrote:
> 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
> http://people.apache.org/~hwright/svn/1.5.8/
Summary:
+0 to release, pending review of the ra_serf situation (see below).
Platform:
Ubuntu 9.04 (jaunty) Linux.
Tested:
1.5.8 tarball with local dependencies (not from deps tarball)
((neon,serf,local,svn) x (fsfs,bdb)) + py + pl + rb + javahl
Verified that 1.5.8 deps were logically equivalent to 1.5.7 deps.
Results:
ra_serf tests (fsfs and bdb) fail miserably due to a SEGFAULT.
svn_ra_serf__conn_setup() is dereferencing a NULL conn->session
when trying to pass conn->session->context to
serf_context_bucket_socket_create(). I'm using serf
^/branches/0.6.x@1421 (which works fine for Subversion's trunk).
ra_serf didn't change at all in 1.5.8, so I suspect that this
problem exists in other Subversion 1.5.x releases, too. As ra_serf in
1.5.x is still highly experimental, I'm disinclined to allow this to be
a release blocker, and would prefer to release-note that there's some
maximum serf version that's known to work okay with this release line.
All other tests pass.
MD5 Checksums:
afac4ff062a75bc851b0d6cf8bc65e93 subversion-1.5.8.tar.gz
cf7a4b95b0a4f8b55070cf91a73975dd subversion-deps-1.5.8.tar.gz
c67de7f0787fde742a473b73225fa8f9 subversion-1.5.8.tar.bz2
ce02461eb31c358e5e2a1b5346af778c subversion-deps-1.5.8.tar.bz2
GPG Signatures (left-aligned because Hyrum likes it that way):
::: subversion-1.5.8.tar.gz :::
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
iEYEABECAAYFAkz2pU8ACgkQokEGqRcG/W4aFgCfdH+4X7TGomqRj51N3zP1zheM
2gUAn3SkboTA9oql9p+40mOUkYQYe4CH
=XGhj
-----END PGP SIGNATURE-----
::: subversion-1.5.8.tar.bz2 :::
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
iEYEABECAAYFAkz2pVQACgkQokEGqRcG/W4VzwCeKHiFjVcNiTrRkLR6Mvtguzf2
FKQAoJ3S8xiJ37Lx3tjt9AwXwB50tQeZ
=bZ9S
-----END PGP SIGNATURE-----
::: subversion-deps-1.5.8.tar.gz :::
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
iEYEABECAAYFAkz2pVcACgkQokEGqRcG/W66gQCdEW8cBfzL3tm3OGqusMap1WrR
yIMAn3pK2lEStFDz/POTHcJOXeDf40u3
=6WC2
-----END PGP SIGNATURE-----
::: subversion-deps-1.5.8.tar.bz2 :::
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
iEYEABECAAYFAkz2pVgACgkQokEGqRcG/W4bXgCgi18X0esPWowFkxtNXbVTBi9q
kKkAnjQ23LV4Ip/9SI1WafLvv9kY1+u0
=0+IO
-----END PGP SIGNATURE-----
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: 1.5.8 up for signing/testing
Posted by Paul Burba <pt...@gmail.com>.
On Thu, Dec 2, 2010 at 8:57 AM, Philip Martin
<ph...@wandisco.com> wrote:
> Stefan Sperling <st...@elego.de> writes:
>
>> On Wed, Dec 01, 2010 at 01:14:30PM -0600, Hyrum K. Wright wrote:
>>> 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
>>> http://people.apache.org/~hwright/svn/1.5.8/
>>
>> 1.5.8 is missing the critial blame -g server-side memory leak crash fix.
>> The trunk revisions merge cleanly into 1.5.x (apart from the property
>> conflicts mentioned in the STATUS entry), so I've nominated those
>> revisions for 1.5.x.
>>
>> I am running tests on 1.5.8 anyway, but would prefer cutting 1.5.9
>> with the blame -g fix instead of releasing 1.5.8.
>
> The blame fix does look like something we should release.
Ugh, agreed, I thought somebody already had...I obviously ran svn
mergeinfo --show-revs merged on the *wrong* branch. Cue much cursing
on my part. I just provided the approving votes if we do...
...and if we don't, the 1.5.8 tests do pass for me [fsfs | bdb] x
[ra_local | ra_svn | ra_neon | ra_serf] using serf 0.3.0.
Paul
> I've just finished running the tests on 1.5.8, so for the record:
>
> Platform:
>
> Linux (Debian/stable).
>
> Verified:
>
> signatures, MD5 checksums, SHA1 checksums
> only expected difference compared to tags/1.5.8
>
> Tested:
>
> 1.5.8 tarball with local dependencies (not deps tarball)
> (local, svn, serf, neon) x (fsfs, bdb) + swig-pl/py/rb + javahl + svn+sasl
>
> Dependencies:
>
> subversion : 1.5.8
> apache2-threaded-dev : 2.2.9-10+lenny8
> libapr1-dev : 1.4.2-3~bpo50+2
> libaprutil1-dev : 1.2.12+dfsg-8+lenny5
> libdb4.6-dev : 4.6.21-11
> libneon27-dev : 0.28.2-6.1
> libsasl2-dev : 2.1.22.dfsg1-23+lenny1
> libsqlite3-dev : 3.6.21-2~bpo50+1
> swig : 1.3.36-1
> perl : 5.10.0-19lenny2
> python2.5-dev : 2.5.2-15+lenny1
> ruby1.8-dev : 1.8.7.72-3lenny1
> openjdk-6-jdk : 6b11-9.1+lenny2
> serf : tags/0.3.1
>
> Results:
>
> All tests pass.
>
> Signatures:
>
> subversion-1.5.8.tar.bz2
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.9 (GNU/Linux)
>
> iQEcBAABCAAGBQJM959UAAoJEHbXiOHtGlmcRi0H/3ulIEf/O0Gdo+IAJ71f9WbG
> FnNUzOt1AWRi9GXo5EqQRciuSy7atKF+N3gQOLcXLVEiox6B/xnv2kErKmSpFBOn
> O9RGbd1QUxW4IKo6XtvR7L5taVQqWLeXdbhaMF9xmoVkOeMKDy22/hUsFpoizmIN
> 3GYt/2Clu0PJRy4JOdirM/VtJyeEDJXlpgbANvXipFXvzdwFWLR8fJzuyNOSrxTT
> Pclj7czadJDOklxqmDO26DIbCp6PPqVPVpXZAaDKV+XSiqcG0uq5hXxYfxY0yY++
> gGouXDEBG3JXkxPjXyQkmQnwtTfDoYqMty7hh4YIN5jVHqgdUv0mzb7fCUbiqjY=
> =YvJv
> -----END PGP SIGNATURE-----
>
> subversion-1.5.8.tar.gz
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.9 (GNU/Linux)
>
> iQEcBAABCAAGBQJM959wAAoJEHbXiOHtGlmcotoH/j3nZ8+PzJcBGwrCG+6qhl3y
> PxM54qe8hscV6Um7UUND0xL0SdrcKqbW10LMcBbquAXrpIPY51nvT3DumzZGf1wO
> iTYNj0FiyOUM0OSMEqcG8xD3qf8B2WRivEqC+fhbHP4byLIYK6fZf7wGS7I/v2+B
> 86Ro/3TnQGJhvasKlyMU+HX850rvPt3mlsGzKmVBR0mBM+RCdslHYmtA5fPQoKxQ
> YyFCKrbRR6r69dEqHmv4Xa/zRO+IgvoF4FVCqObtK1EXg+ZHYbD+h2xLuwx8yIfz
> 9yGzeH/d4OImT1t6iHvSgD+CEfnF1/dLXhTI53BZmFF0rvvSXT32fxqtmUL9VgQ=
> =5mKA
> -----END PGP SIGNATURE-----
>
>
> --
> Philip
>
Re: 1.5.8 up for signing/testing
Posted by Philip Martin <ph...@wandisco.com>.
Stefan Sperling <st...@elego.de> writes:
> On Wed, Dec 01, 2010 at 01:14:30PM -0600, Hyrum K. Wright wrote:
>> 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
>> http://people.apache.org/~hwright/svn/1.5.8/
>
> 1.5.8 is missing the critial blame -g server-side memory leak crash fix.
> The trunk revisions merge cleanly into 1.5.x (apart from the property
> conflicts mentioned in the STATUS entry), so I've nominated those
> revisions for 1.5.x.
>
> I am running tests on 1.5.8 anyway, but would prefer cutting 1.5.9
> with the blame -g fix instead of releasing 1.5.8.
The blame fix does look like something we should release.
I've just finished running the tests on 1.5.8, so for the record:
Platform:
Linux (Debian/stable).
Verified:
signatures, MD5 checksums, SHA1 checksums
only expected difference compared to tags/1.5.8
Tested:
1.5.8 tarball with local dependencies (not deps tarball)
(local, svn, serf, neon) x (fsfs, bdb) + swig-pl/py/rb + javahl + svn+sasl
Dependencies:
subversion : 1.5.8
apache2-threaded-dev : 2.2.9-10+lenny8
libapr1-dev : 1.4.2-3~bpo50+2
libaprutil1-dev : 1.2.12+dfsg-8+lenny5
libdb4.6-dev : 4.6.21-11
libneon27-dev : 0.28.2-6.1
libsasl2-dev : 2.1.22.dfsg1-23+lenny1
libsqlite3-dev : 3.6.21-2~bpo50+1
swig : 1.3.36-1
perl : 5.10.0-19lenny2
python2.5-dev : 2.5.2-15+lenny1
ruby1.8-dev : 1.8.7.72-3lenny1
openjdk-6-jdk : 6b11-9.1+lenny2
serf : tags/0.3.1
Results:
All tests pass.
Signatures:
subversion-1.5.8.tar.bz2
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
iQEcBAABCAAGBQJM959UAAoJEHbXiOHtGlmcRi0H/3ulIEf/O0Gdo+IAJ71f9WbG
FnNUzOt1AWRi9GXo5EqQRciuSy7atKF+N3gQOLcXLVEiox6B/xnv2kErKmSpFBOn
O9RGbd1QUxW4IKo6XtvR7L5taVQqWLeXdbhaMF9xmoVkOeMKDy22/hUsFpoizmIN
3GYt/2Clu0PJRy4JOdirM/VtJyeEDJXlpgbANvXipFXvzdwFWLR8fJzuyNOSrxTT
Pclj7czadJDOklxqmDO26DIbCp6PPqVPVpXZAaDKV+XSiqcG0uq5hXxYfxY0yY++
gGouXDEBG3JXkxPjXyQkmQnwtTfDoYqMty7hh4YIN5jVHqgdUv0mzb7fCUbiqjY=
=YvJv
-----END PGP SIGNATURE-----
subversion-1.5.8.tar.gz
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
iQEcBAABCAAGBQJM959wAAoJEHbXiOHtGlmcotoH/j3nZ8+PzJcBGwrCG+6qhl3y
PxM54qe8hscV6Um7UUND0xL0SdrcKqbW10LMcBbquAXrpIPY51nvT3DumzZGf1wO
iTYNj0FiyOUM0OSMEqcG8xD3qf8B2WRivEqC+fhbHP4byLIYK6fZf7wGS7I/v2+B
86Ro/3TnQGJhvasKlyMU+HX850rvPt3mlsGzKmVBR0mBM+RCdslHYmtA5fPQoKxQ
YyFCKrbRR6r69dEqHmv4Xa/zRO+IgvoF4FVCqObtK1EXg+ZHYbD+h2xLuwx8yIfz
9yGzeH/d4OImT1t6iHvSgD+CEfnF1/dLXhTI53BZmFF0rvvSXT32fxqtmUL9VgQ=
=5mKA
-----END PGP SIGNATURE-----
--
Philip
Re: 1.5.8 up for signing/testing
Posted by Stefan Sperling <st...@elego.de>.
On Wed, Dec 01, 2010 at 01:14:30PM -0600, Hyrum K. Wright wrote:
> 1.5.8 tarballs are up for testing and signing. The magic revision is r1041089:
> http://people.apache.org/~hwright/svn/1.5.8/
1.5.8 is missing the critial blame -g server-side memory leak crash fix.
The trunk revisions merge cleanly into 1.5.x (apart from the property
conflicts mentioned in the STATUS entry), so I've nominated those
revisions for 1.5.x.
I am running tests on 1.5.8 anyway, but would prefer cutting 1.5.9
with the blame -g fix instead of releasing 1.5.8.
Stefan
Re: 1.5.8 up for signing/testing
Posted by Mark Phippard <ma...@gmail.com>.
SUMMARY:
+1 to release
PLATFORM:
Windows 7
VS 2008 SP1
Java 1.6
COMPONENTS:
Apache 2.2.15
APR 1.4.2
APR-UTIL 1.3.9
OpenSSL 1.0.0a
Neon 0.29.5
ZLib (from deps)
VERIFIED:
md5 and sha1 sums
gpg signature verified
TESTED:
JavaHL
ra_local | ra_svn | ra_neon X fsfs
RESULTS:
All passed
SIGNATURES:
subversion-1.5.8.zip
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (Darwin)
iEYEABECAAYFAkz2oiIACgkQJl34oANalqn7OACgkXfcbCr17XRqvYSFIkwGrqst
3V0AniZnJGzvU3o5ePj4k1u/OO8dMFhI
=B+uM
-----END PGP SIGNATURE-----
subversion-deps-1.5.8.zip
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (Darwin)
iEYEABECAAYFAkz2oi0ACgkQJl34oANalqmpgACgmek/6rW9+JJpAfKJ7wx0H4BV
i2oAn0oTiXZea3TsnlcYpaE/TSg5AOeG
=d74O
-----END PGP SIGNATURE-----
--
Thanks
Mark Phippard
http://markphip.blogspot.com/