You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by David Anderson <da...@calixo.net> on 2006/03/14 09:28:07 UTC
1.3.1 tarballs up for testing/signing
Subversion 1.3.1 is up for testing and signing. The magic revision is r18866.
http://lolut.utbm.info/pub/subversion/1.3.1/
As usual, this code is unreleased until the links are posted on
announce@ and on the users list. Until then, the code is unsupported,
and if it breaks or is incorrect, then it is totally your fault.
As usual also, signatures from commiters back to me, feedback from
enthusiast non-commiters also welcome.
Again, distro package maintainers and network admins are not
encouraged to make this RC available to end users, because it may
cause many headaches if problems are found before the official 1.3.1
release. Until it is officially announced, please don't package this
code in any distribution whatsoever.
- Dave
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by David Anderson <da...@calixo.net>.
* Peter N. Lundblad <pe...@famlundblad.se> [2006-03-14 13:17:51]:
> It seems like the md5 and sha1 checksums are missing from this directory.
Woops, forgot to upload them to the correct dir. Done now.
- Dave.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Helio Chissini de Castro <he...@mandriva.com>.
Em Terça 14 Março 2006 09:17, David Anderson writes:
> Subversion 1.3.1 is up for testing and signing. The magic revision is
> r18866.
>
> http://lolut.utbm.info/pub/subversion/1.3.1/
make check is not LANG safe.
For package test, i did at my local environment, pt_BR.UTF8 ( Brazilian
Portuguese ), and suddenly lot of tests start to fail.
For correction, i was need to add LANG=C LC_ALL=C and everuthing is back.
Maybe some warning or force the language on test will avoid some annoyance
[]'s
--
Helio Chissini de Castro
Research and development
Mandriva Brasil Labs
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Helio Chissini de Castro <he...@mandriva.com>.
Em Terça 14 Março 2006 15:16, Garrett Rooney escreveu:
> > For package test, i did at my local environment, pt_BR.UTF8 ( Brazilian
> > Portuguese ), and suddenly lot of tests start to fail.
> > For correction, i was need to add LANG=C LC_ALL=C and everuthing is back.
> > Maybe some warning or force the language on test will avoid some
> > annoyance
>
> What kind of failures are you seeing? Just stuff resulting from the
> tests looking for particular error codes, or more substantial
> problems?
>
> -garrett
[helio@underdog subversion-1.3.1]$ LANG=pt_BR.UTF8 LC_ALL=pt_BR.UTF8 make
check
Running all tests in compat-test...success
.
.
.
Running all tests in translate-test...success
Running all tests in utf-test...success
Running all tests in target-test.py...FAILURE <---------------------
Running all tests in getopt_tests.py...FAILURE <----------------------
Failure on all remaining tests
[helio@underdog subversion-1.3.1]$ LANG=C LC_ALL=C make check
Running all tests in compat-test...success
.
.
.
Running all tests in translate-test...success
Running all tests in utf-test...success
Running all tests in target-test.py...success
Running all tests in getopt_tests.py...success <--------------------
Sucess in every test
--
Helio Chissini de Castro
Research and development
Mandriva Brasil Labs
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Garrett Rooney <ro...@electricjellyfish.net>.
On 3/14/06, Helio Chissini de Castro <he...@mandriva.com> wrote:
> Em Terça 14 Março 2006 09:17, David Anderson writes:
> > Subversion 1.3.1 is up for testing and signing. The magic revision is
> > r18866.
> >
> > http://lolut.utbm.info/pub/subversion/1.3.1/
>
> make check is not LANG safe.
>
> For package test, i did at my local environment, pt_BR.UTF8 ( Brazilian
> Portuguese ), and suddenly lot of tests start to fail.
> For correction, i was need to add LANG=C LC_ALL=C and everuthing is back.
> Maybe some warning or force the language on test will avoid some annoyance
What kind of failures are you seeing? Just stuff resulting from the
tests looking for particular error codes, or more substantial
problems?
-garrett
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Helio Chissini de Castro <he...@mandriva.com>.
Em Terça 14 Março 2006 09:17, David Anderson writes:
> Subversion 1.3.1 is up for testing and signing. The magic revision is
> r18866.
>
> http://lolut.utbm.info/pub/subversion/1.3.1/
make check is not LANG safe.
For package test, i did at my local environment, pt_BR.UTF8 ( Brazilian
Portuguese ), and suddenly lot of tests start to fail.
For correction, i was need to add LANG=C LC_ALL=C and everuthing is back.
Maybe some warning or force the language on test will avoid some annoyance
[]'s
--
Helio Chissini de Castro
Research and development
Mandriva Brasil Labs
Re: 1.3.1 tarballs up for testing/signing
Posted by "Peter N. Lundblad" <pe...@famlundblad.se>.
David Anderson writes:
> Subversion 1.3.1 is up for testing and signing. The magic revision is r18866.
>
> http://lolut.utbm.info/pub/subversion/1.3.1/
>
It seems like the md5 and sha1 checksums are missing from this directory.
Thanks,
//Peter
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Mark Phippard <ma...@softlanding.com>.
Malcolm Rowe <ma...@farside.org.uk> wrote on 03/14/2006 11:32:15
AM:
> On Tue, Mar 14, 2006 at 11:25:00AM -0500, Mark Phippard wrote:
> > > > 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that
0.26.0
> > was
> > > > just released as well.
> > > >
> > >
> > > We should probably ship with 0.25.4, which is what 1.3.0 shipped
with.
> > > This is supposed to be a stable release.
> >
> > 0.25.4 has some very bad behavior for Windows users if they are using
SSPI
> > authentication. I believe it made it impossible to use Subversion in
some
> > cases without disabling SSPI. So I would recommend going with 0.25.5
> > which is the current stable version for 0.25.x. I believe that is
what
> > Brane used when he built the Windows binaries, although it might have
been
> > 0.25.4 + patch for SSPI.
> >
>
> Yes, I'd forgotten about those. 0.25.5 makes sense in that case (I just
> thought you'd mistyped/read the version number).
>
> I don't think going to 0.26.0 makes sense, though: I don't even think
> trunk builds with it at the moment :-)
That is why I pointed out 0.26.0. I had not been aware it was even
coming, so I figured the build system in trunk would not support it and
someone would see this and work on it.
On the APR issue, this made me realize that when I do my builds, I am
using APR from the HTTP Server so that I can build mod_dav_svn. I never
even use the APR in the zip file. Should I change that in terms of
providing a signature?
I currently pass --with-httpd to gen-make.py. Is there a way to use the
APR from the zip file, but still build mod_dav_svn?
Mark
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Malcolm Rowe <ma...@farside.org.uk>.
On Tue, Mar 14, 2006 at 11:25:00AM -0500, Mark Phippard wrote:
> > > 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0
> was
> > > just released as well.
> > >
> >
> > We should probably ship with 0.25.4, which is what 1.3.0 shipped with.
> > This is supposed to be a stable release.
>
> 0.25.4 has some very bad behavior for Windows users if they are using SSPI
> authentication. I believe it made it impossible to use Subversion in some
> cases without disabling SSPI. So I would recommend going with 0.25.5
> which is the current stable version for 0.25.x. I believe that is what
> Brane used when he built the Windows binaries, although it might have been
> 0.25.4 + patch for SSPI.
>
Yes, I'd forgotten about those. 0.25.5 makes sense in that case (I just
thought you'd mistyped/read the version number).
I don't think going to 0.26.0 makes sense, though: I don't even think
trunk builds with it at the moment :-)
Regards,
Malcolm
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Mark Phippard <ma...@softlanding.com>.
Malcolm Rowe <ma...@farside.org.uk> wrote on 03/14/2006 11:21:41
AM:
> > 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0
was
> > just released as well.
> >
>
> We should probably ship with 0.25.4, which is what 1.3.0 shipped with.
> This is supposed to be a stable release.
0.25.4 has some very bad behavior for Windows users if they are using SSPI
authentication. I believe it made it impossible to use Subversion in some
cases without disabling SSPI. So I would recommend going with 0.25.5
which is the current stable version for 0.25.x. I believe that is what
Brane used when he built the Windows binaries, although it might have been
0.25.4 + patch for SSPI.
Mark
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Malcolm Rowe <ma...@farside.org.uk>.
On Tue, Mar 14, 2006 at 10:41:31AM -0500, Mark Phippard wrote:
> I am just starting the testing process, but a couple of quick comments in
> regards to the Windows zip file.
>
> 1) It is 700 K smaller than the 1.3.0 zip which seems odd.
>
We're completely missing the following files (which were present in the
1.3.0 zipfile):
subversion-1.3.1/apr/apr.dep
subversion-1.3.1/apr/apr.mak
subversion-1.3.1/apr/build/libtool.m4
subversion-1.3.1/apr/build/ltmain.sh
subversion-1.3.1/apr/configure
subversion-1.3.1/apr/include/arch/unix/apr_private.h.in
subversion-1.3.1/apr/libapr.dep
subversion-1.3.1/apr/libapr.mak
subversion-1.3.1/apr-iconv/apriconv.dep
subversion-1.3.1/apr-iconv/apriconv.mak
subversion-1.3.1/apr-iconv/libapriconv.dep
subversion-1.3.1/apr-iconv/libapriconv.mak
subversion-1.3.1/apr-util/aprutil.dep
subversion-1.3.1/apr-util/aprutil.mak
subversion-1.3.1/apr-util/build/apr_common.m4
subversion-1.3.1/apr-util/build/config.guess
subversion-1.3.1/apr-util/build/config.sub
subversion-1.3.1/apr-util/build/find_apr.m4
subversion-1.3.1/apr-util/build/install.sh
subversion-1.3.1/apr-util/configure
subversion-1.3.1/apr-util/include/private/apu_config.h.in
subversion-1.3.1/apr-util/libaprutil.dep
subversion-1.3.1/apr-util/libaprutil.mak
subversion-1.3.1/apr-util/uri/gen_uri_delims.dep
subversion-1.3.1/apr-util/uri/gen_uri_delims.mak
subversion-1.3.1/apr-util/xml/expat/aclocal.m4
subversion-1.3.1/apr-util/xml/expat/config.h.in
subversion-1.3.1/apr-util/xml/expat/configure
subversion-1.3.1/apr-util/xml/expat/conftools/ltmain.sh
subversion-1.3.1/apr-util/xml/expat/lib/xml.dep
subversion-1.3.1/apr-util/xml/expat/lib/xml.mak
We're also missing quite a lot of Neon documentation (which seems quite
large).:
> 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0 was
> just released as well.
>
We should probably ship with 0.25.4, which is what 1.3.0 shipped with.
This is supposed to be a stable release.
Regards,
Malcolm
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by "C. Michael Pilato" <cm...@collab.net>.
David Anderson wrote:
> Damn. I'd looked up the version of Neon shipped with 1.3.0, not
> remembering. And I picked up 0.24.7 in one of the files. But it's a
> 0.25.4.
>
> Gaghh. Talk about a stupid, stupid way to burn a version number
> :-(. If I remember the discussions about it all, next tarballs I post
> are going to be 1.3.2 ?
Don't sweat it, Dave. We still appreciate that you're wearing the RM hat. :-)
(IIRC, 1.3.2 is correct. But do grab Neon 0.25.5 for it, as seen elsewhere
in this thread.)
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: 1.3.1 tarballs up for testing/signing
Posted by Molle Bestefich <mo...@gmail.com>.
David Anderson wrote:
> Gaghh. Talk about a stupid, stupid way to burn a version number
> :-(. If I remember the discussions about it all, next tarballs I post
> are going to be 1.3.2 ?
This must be an infinitely stupid question since noone has asked it.
Apologies in advance! :)
Here goes anyway: Instead of burning version numbers like this, why
don't you post a couple of release candidate zips first?
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
On 3/14/06, Malcolm Rowe <ma...@farside.org.uk> wrote:
> If a release or candidate release needs to be quickly re-issued due to
> some non-code problem (say, a packaging glitch), it's okay to reuse
> the same name, as long as the tarball hasn't been blessed by signing
> yet. But if it has been uploaded to the standard distribution area with
> signatures, or if the re-issue was due to a change in code a user might
> run, then the old name must be tossed and the next name used.
Yes, since our code is not changing at all, it's fine to reuse the
version number. All we're changing is the neon (and probably APR)
bundled with the tarball. If anything in our tree gets changed, poof
- it's a new version number. -- justin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Malcolm Rowe <ma...@farside.org.uk>.
On Tue, Mar 14, 2006 at 05:18:35PM +0100, David Anderson wrote:
> Gaghh. Talk about a stupid, stupid way to burn a version number
> :-(. If I remember the discussions about it all, next tarballs I post
> are going to be 1.3.2 ?
>
Hacking say:
If a release or candidate release needs to be quickly re-issued due to
some non-code problem (say, a packaging glitch), it's okay to reuse
the same name, as long as the tarball hasn't been blessed by signing
yet. But if it has been uploaded to the standard distribution area with
signatures, or if the re-issue was due to a change in code a user might
run, then the old name must be tossed and the next name used.
I'd say that including the wrong version of Neon is a packaging glitch,
not a change to the Subversion code itself. I'm not sure what the
missing APR files problem is due to, but it smells like packaging as well.
So I guess you should stick with 1.3.1. It's not like it was even around
long enough to attract a single signature.
Regards,
Malcolm
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by Mark Phippard <ma...@softlanding.com>.
David Anderson <da...@calixo.net> wrote on 03/14/2006 11:18:35
AM:
> * C. Michael Pilato <cm...@collab.net> [2006-03-14 11:04:19]:
>
> > Mark Phippard wrote:
> > > I am just starting the testing process, but a couple of quick
comments in
> > > regards to the Windows zip file.
> > >
> > > 1) It is 700 K smaller than the 1.3.0 zip which seems odd.
> > >
> > > 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0
was
> > > just released as well.
> >
> > Hrm. The gz and bz files have Neon 0.24.7, too. That seems ungood.
But
> > they aren't interestingly different size-wise from the 1.3.0 stuffs.
>
> Damn. I'd looked up the version of Neon shipped with 1.3.0, not
> remembering. And I picked up 0.24.7 in one of the files. But it's a
> 0.25.4.
Make sure you grab 0.25.5 when you redo it. It had a very relevant fix
that was needed for Windows.
Thanks
Mark
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by David Anderson <da...@calixo.net>.
* C. Michael Pilato <cm...@collab.net> [2006-03-14 11:04:19]:
> Mark Phippard wrote:
> > I am just starting the testing process, but a couple of quick comments in
> > regards to the Windows zip file.
> >
> > 1) It is 700 K smaller than the 1.3.0 zip which seems odd.
> >
> > 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0 was
> > just released as well.
>
> Hrm. The gz and bz files have Neon 0.24.7, too. That seems ungood. But
> they aren't interestingly different size-wise from the 1.3.0 stuffs.
Damn. I'd looked up the version of Neon shipped with 1.3.0, not
remembering. And I picked up 0.24.7 in one of the files. But it's a
0.25.4.
Gaghh. Talk about a stupid, stupid way to burn a version number
:-(. If I remember the discussions about it all, next tarballs I post
are going to be 1.3.2 ?
So, report any other packaging errors you find. I'm investigating the
size discrepancy, but anything else that should go into the next
rolling, please notify.
the size difference for the zip seems to be coming from the apr*
bundles. In the 1.3.0 tarball, there are some directories which look
autogenerated, that aren't in 1.3.1: LibR, Release. There are also
more visual studio looking files in the root of all three libs.
- Dave, feeling mighty stupid.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: 1.3.1 tarballs up for testing/signing
Posted by "C. Michael Pilato" <cm...@collab.net>.
Mark Phippard wrote:
> I am just starting the testing process, but a couple of quick comments in
> regards to the Windows zip file.
>
> 1) It is 700 K smaller than the 1.3.0 zip which seems odd.
>
> 2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0 was
> just released as well.
Hrm. The gz and bz files have Neon 0.24.7, too. That seems ungood. But
they aren't interestingly different size-wise from the 1.3.0 stuffs.
--
C. Michael Pilato <cm...@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Re: 1.3.1 tarballs up for testing/signing
Posted by Mark Phippard <ma...@softlanding.com>.
David Anderson <da...@calixo.net> wrote on 03/14/2006 04:28:07
AM:
> Subversion 1.3.1 is up for testing and signing. The magic revision is
r18866.
>
> http://lolut.utbm.info/pub/subversion/1.3.1/
>
> As usual, this code is unreleased until the links are posted on
> announce@ and on the users list. Until then, the code is unsupported,
> and if it breaks or is incorrect, then it is totally your fault.
>
> As usual also, signatures from commiters back to me, feedback from
> enthusiast non-commiters also welcome.
>
> Again, distro package maintainers and network admins are not
> encouraged to make this RC available to end users, because it may
> cause many headaches if problems are found before the official 1.3.1
> release. Until it is officially announced, please don't package this
> code in any distribution whatsoever.
I am just starting the testing process, but a couple of quick comments in
regards to the Windows zip file.
1) It is 700 K smaller than the 1.3.0 zip which seems odd.
2) It includes Neon 0.24.7 as opposed to 0.25.5. I see that 0.26.0 was
just released as well.
Mark
_____________________________________________________________________________
Scanned for SoftLanding Systems, Inc. and SoftLanding Europe Plc by IBM Email Security Management Services powered by MessageLabs.
_____________________________________________________________________________
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org