[-1] Disapprove

The deal breakers:
1. The source tarball (tar.bz2 at least) contains main/MacOSXX64Env.Set and
main/MacOSXX64Env.Set.sh. I don't believe they belong in there (they're not
in trunk), and if they do, they're missing ASLv2 licenses and cause
unapproved license errors in the RAT report.
2. The source fails to build on 32 bit Xubuntu 14.04 both on VMs and
physical hardware (details later).

Otherwise:

Please can we see links to RAT reports, changelog, test results, code
quality metrics, and other useful info in the emails proposing a release.

Neither Google Test nor the unit test changes themselves were backported
from trunk to 4.1.2, and many old cppunit tests would fail and break the
build if --with-system-cppunit were to be used, so unit tests can't be
tested. I did run the bvt and fvt tests. Since it's badly documented, you
first "source LinuxX86Env.Set.sh" (or your equivalent) and then run the
tests in the test/ subdirectory with:
ant -Dtest.args="-tp bvt"
ant -Dtest.args="-tp fvt"
with -Dopenoffice.home=/path/to/AOO added if you do what I did, which is
ran the more up-to-date tests from trunk against 4.1.2. The bvt tests take
about 10 minutes, the fvt tests 1 hour.

================
Xubuntu 14.04, 32 bit
================

The binary package installs. Java 7 is automatically detected and used.

The source reproducibly fails to build (on both VMs and physical hardware)
in main/svl with this error I reported on the dev list months ago, which
can be worked around by copying and pasting the command that fails and
running it manually with "-lc" appended (as the missing __stack_chk_fail
symbol is present in libc, but the linker doesn't search there for some
reason):

--snip--
[ build LNK ] Library/passwordcontainer.uno.so
R=/home/user/aoo-4.1.2 && S=$R/main && O=$S/solver/412/unxlngi6 &&
W=$O/workdir &&  mkdir -p $W/LinkTarget/Library/ && g++ -shared
-Wl,-z,noexecstack '-Wl,-rpath,$ORIGIN:$ORIGIN/../ure-link/lib'
'-Wl,-rpath-link,$O/lib' -Wl,-rpath-link,/lib:/usr/lib -Wl,-z,combreloc
-Wl,-z,defs   -L$S/solenv/unxlngi6/lib -L$O/lib -L$S/solenv/unxlngi6/lib
-L/usr/lib/jvm/java-7-openjdk-i386/lib
-L/usr/lib/jvm/java-7-openjdk-i386/jre/lib/i386
-L/usr/lib/jvm/java-7-openjdk-i386/jre/lib/i386/client
-L/usr/lib/jvm/java-7-openjdk-i386/jre/lib/i386/native_threads
-L/usr/lib/i386-linux-gnu  -Wl,--hash-style=both
-Wl,--dynamic-list-cpp-new -Wl,--dynamic-list-cpp-typeinfo
-Wl,-Bsymbolic-functions
$W/CxxObject/svl/source/passwordcontainer/passwordcontainer.o
$W/CxxObject/svl/source/passwordcontainer/syscreds.o  -luno_cppu
-luno_cppuhelpergcc3 -luno_sal -lstdc++ -lucbhelper4gcc3 -lutl
-Wl,--start-group  -Wl,--end-group -o $W/LinkTarget/Library/
passwordcontainer.uno.so
/usr/lib/i386-linux-gnu/libc_nonshared.a(stack_chk_fail_local.oS): In
function `__stack_chk_fail_local':
(.text+0x10): undefined reference to `__stack_chk_fail'
collect2: error: ld returned 1 exit status
make: ***
[/home/user/aoo-4.1.2/main/solver/412/unxlngi6/workdir/LinkTarget/Library/
passwordcontainer.uno.so] Error 1
dmake:  Error code 2, while making 'all'

1 module(s):
    svl
need(s) to be rebuilt
--snip--

The test results:

5 bvt tests out of 51 total get errors in 4.1.2:
bvt.gui.ContinuePoint initializationError Test class should have exactly
one public constructor
bvt.gui.ContinuePoint initializationError No runnable methods
bvt.gui.FileExport testSaveAs String index out of range: -1
bvt.gui.TestType initializationError Test class should have exactly one
public zero-argument constructor
bvt.gui.TestType initializationError No runnable methods

All of these also fail on trunk. The bvt.gui.ContinuePoint and
bvt.gui.TestType classes are not tests, so they shouldn't be getting tested
in the first place. The bvt.gu.FileExport class contains mistakes like
using hardcoded slashes and backslashes instead of
java.io.File.separatorChar, fails because it constructs a path badly, and
is looking for a samples directory that doesn't exist. Trunk has 4
additional failures not present in 4.1.2.

12 fvt tests out of 858 total fail in 4.1.2:
fvt.gui.sc.validity.ValidityDialogSetting testAllowDateNotBetween
expected:<[12/31/07]> but was:<[]>
fvt.gui.sc.validity.ValidityDialogSetting testAllowTimeBetween
expected:<[09:00:00 PM]> but was:<[21:00:00]>
fvt.gui.sc.validity.ValidityDialogSetting testAllowTimeGreaterThan
expected:<07:30:00[ AM]> but was:<07:30:00[]>
fvt.gui.sw.table.TableGeneral testTableBorderLineStyle expected:<0.50 ["]>
but was:<0.50 [cm]>
fvt.gui.sw.table.TableGeneral testTableRowHeight expected:<0.50 ["]> but
was:<0.50 [cm]>
fvt.gui.sw.table.TableGeneral testTableColumnWidth expected:<2.00 ["]> but
was:<2.00 [cm]>
fvt.uno.sc.chart.ChartLegend testLegendPosition[0] Incorrect chart legend
position X got in .ods file. expected:<14009.0> but was:<14007.0>
fvt.uno.sc.chart.ChartLegend testLegendPosition[2] Incorrect chart legend
position X got in .ods file. expected:<9470.0> but was:<9468.0>
fvt.uno.sw.puretext.CharacterLocale testCharacterLocaleSetting assert
character language expected:<[US]> but was:<[ZA]>
fvt.uno.sd.file.CheckFileProperties testGeneralCreationDate CreationDate
should be the same as set
fvt.uno.sd.file.CheckFileProperties testGeneralEditingDuration Totally
editing time should be 60 expected:<60> but was:<61>
fvt.uno.sd.shape.ShapeProperties testShapeRotation RotateAngle is not 2500
expected: java.lang.Long<2500> but was: java.lang.Integer<2500>

All these fail on trunk as well, and trunk has 15 additional fvt test
failures and several errors not present in 4.1.2 :-(.

===================
Windows XP SP3, 32 bit
===================

The binary package installs. Java 7 is automatically detected and used.

With bvt tests, I get these 3 failures in addition to the 5 errors on Linux:
bvt.gui.FileTypeTest testSaveNewODP The typed text is saved!
expected:<[@AOO]> but was:<[]>
bvt.gui.FileTypeTest testSaveNewOTP The typed text is saved!
expected:<[@AOO]> but was:<[]>
bvt.gui.FileTypeTest testSaveNewPOT The typed text is saved!
expected:<[@AOO]> but was:<[]>

I think those tests are unreliable on Windows, as testSaveNewSXI and
testSaveNewSTI fail too when I run bvt.gui.FileTypeTest by itself.

With fvt tests, I get 3 errors and 3 failures, only 1 being common with
Linux:
fvt.gui.sc.sort.SortDialogSetting testSortOptionsIncludeFormats Timeout to
execute the dispatch!
fvt.gui.sd.headerandfooter.DocumentWithHeaderFooter
testOpenAOO34WithHeaderFooter Timeout to execute the dispatch!
fvt.gui.sd.headerandfooter.DocumentWithHeaderFooter
testOpenPPTWithHeaderFooter Timeout to execute the dispatch!
fvt.gui.sd.shape.ShapeTypes testCalloutShapes expected:<3> but was:<1>
fvt.gui.sd.shape.ShapeTypes testStarsShapes expected:<3> but was:<1>
fvt.uno.sd.shape.ShapeProperties testShapeRotation RotateAngle is not 2500
expected: java.lang.Long<2500> but was: java.lang.Integer<2500>

Wasn't able to compare 4.1.2's test results with trunk's: Windows is the
slowest and most problematic platform to build AOO on.

================
FreeBSD 10.2 64 bit
================

No binary package provided.

Test wise, I had hoped #126586 was the only issue with the tests on
FreeBSD, but it isn't, not by a long way :-(. The fvt.uno.db.DBAccess test
deadlocks AOO, freezing the UI so badly it needs to be killed, on both
trunk and 4.1.2. Every test that doesn't "discard()" or forcefully close
the last remaining document or dialog, will stay open, causing all
subsequent tests to fail with errors. I even began committing patches like
r1707970 to force-close the last dialog, before I realized this isn't
necessary on other platforms! FreeBSD thus couldn't be tested.

Otherwise thank you everyone for this RC, and hopefully the next one will
be released.
Damjan

On Fri, Oct 23, 2015 at 1:50 PM, Andrea Pescetti <pesce...@apache.org>
wrote:

> We have been in the Release Candidate stage for three weeks and it is now
> time to start a VOTE for releasing OpenOffice 4.1.2.
>
> Source and binary packages for RC3 are available at
> https://dist.apache.org/repos/dist/dev/openoffice/
> and the reference revision is r1709696. Many people helped in making this
> available, but let me think Juergen and Herbert again for their assistance
> and guidance.
>
> I'm initially keeping the VOTE just a bit longer than the minimum 72
> hours: please vote by Monday 26 October, 14:00 UTC. Extensions are surely
> possible if there is a need for them, but I hope that the very long testing
> phase we've gone through helps in shortening the final approval phase.
>
> So please VOTE on releasing RC3 (as defined above) as OpenOffice 4.1.2:
> [  ] +1 Approve
> [  ]  0 Abstain
> [  ] -1 Disapprove, with explanation
>
> Regards,
>   Andrea.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscr...@openoffice.apache.org
> For additional commands, e-mail: dev-h...@openoffice.apache.org
>
>

Reply via email to