ntk/apt.git
9 years agoinitialize iPolicyBrokenCount in DepCache::Update
Warren He [Fri, 29 Aug 2014 09:15:30 +0000 (11:15 +0200)]
initialize iPolicyBrokenCount in DepCache::Update

All other counters are correctly initialized here, expect this one. The
practical effect is low as in apt we usually just do "!= 0" checks, but
only correct counters are good counters.

Closes: 758397

9 years agoPortuguese manpages translation update
Américo Monteiro [Fri, 29 Aug 2014 09:07:23 +0000 (11:07 +0200)]
Portuguese manpages translation update

Closes: 759608

9 years agoThe following command otherwise yields many blank lines:
Andreas Oberritter [Thu, 28 Aug 2014 22:04:20 +0000 (15:04 -0700)]
The following command otherwise yields many blank lines:
apt list -o APT::Cmd::use-format=true -o APT::Cmd::format=\${Package}

And even worse when adding "-o APT::Cmd::All-Versions=true".

Signed-off-by: Andreas Oberritter <obi@opendreambox.org>
9 years agoGerman program translation update
Holger Wansing [Thu, 28 Aug 2014 16:16:02 +0000 (18:16 +0200)]
German program translation update

Closes: 758837

9 years agoreleasing 1.0.7 1.0.7
Michael Vogt [Thu, 28 Aug 2014 00:24:03 +0000 (00:24 +0000)]
releasing 1.0.7

9 years agoprepare 1.0.7
Michael Vogt [Thu, 28 Aug 2014 00:14:51 +0000 (17:14 -0700)]
prepare 1.0.7

9 years agosupport versioned provides as implemented by dpkg
David Kalnischkies [Sat, 16 Aug 2014 14:33:46 +0000 (16:33 +0200)]
support versioned provides as implemented by dpkg

APT supported versioned provides for a long while in an attempt to get
it working with rpm. While this support is old, we can be relatively
sure that it works as versioned provides are used internally to make
Multi-Arch:foreign work.

Previous versions of apt will print a warning indicating that the
versioned provides is ignored, so that something which "Provides: foo (=
2)" doesn't provide anything.

Note that dpkg does allow only a equals-relation in the provides line
as anything else is deemed too complex. apt doesn't support anything
else either and such a support would require potentially big changes.

Closes: 758153

9 years agoadd dpkg::source-options for dpkg-source invocation
David Kalnischkies [Sat, 16 Aug 2014 11:29:10 +0000 (13:29 +0200)]
add dpkg::source-options for dpkg-source invocation

dpkg-source can be told to enforce signature checks with
--require-valid-signature, but while this isn't feasible as default for
Debian itself at the moment, a local admin should be able to use it.

This commit also fixes the size limit on the construction of the command
being called for dpkg-source and dpkg-buildpackage.

Closes: 757534

9 years agoCzech program translation update
Miroslav Kure [Fri, 15 Aug 2014 17:42:41 +0000 (19:42 +0200)]
Czech program translation update

Closes: 758208

9 years agoTurkish program translation update
Mert Dirik [Fri, 15 Aug 2014 17:37:39 +0000 (19:37 +0200)]
Turkish program translation update

Closes: 756710

9 years agorewrite and extend new README file
David Kalnischkies [Sat, 19 Jul 2014 11:30:48 +0000 (13:30 +0200)]
rewrite and extend new README file

Reuse description from the package descriptions and add various
subsection describing useful debug options and general information.

Git-Dch: Ignore

9 years agoremove obsolete documentation bits
David Kalnischkies [Sat, 19 Jul 2014 11:30:08 +0000 (13:30 +0200)]
remove obsolete documentation bits

Git-Dch: Ignore

9 years agoensure that all docs use all entities files
David Kalnischkies [Fri, 18 Jul 2014 14:42:40 +0000 (16:42 +0200)]
ensure that all docs use all entities files

Not all are needed for all files at the moment, but the new docbook
building hadn't available some of the entities it used as the files
weren't correctly copied around in all cases and having the same across
the bord makes working with all of them a little easier.

Git-Dch: Ignore

9 years agoparse debian/tests/control for test dependencies
David Kalnischkies [Fri, 18 Jul 2014 10:28:21 +0000 (12:28 +0200)]
parse debian/tests/control for test dependencies

Our integration tests need some additional dependencies to run and
function correctly, but while multiple places run them, there is no need
to also specify the these dependencies in multiple places.

Git-Dch: Ignore

9 years agoFix debListParser to accept "no" as a value for the Multi-Arch field
Julian Andres Klode [Sun, 24 Aug 2014 16:43:07 +0000 (18:43 +0200)]
Fix debListParser to accept "no" as a value for the Multi-Arch field

Seems this was missed somehow.

Closes: #759099

9 years agodoc/po/pt.po: updated, thanks to Américo Monteir
Michael Vogt [Tue, 29 Jul 2014 16:07:44 +0000 (18:07 +0200)]
doc/po/pt.po: updated, thanks to Américo Monteir

Closes: #756200

9 years agodoc/apt.8.xml: fix typo, thanks to Jakub Wilk
Michael Vogt [Tue, 29 Jul 2014 15:54:03 +0000 (17:54 +0200)]
doc/apt.8.xml: fix typo, thanks to Jakub Wilk

Closes: #756056

9 years agoFix SmartConfigure to ignore ordering of packages that are already valid
Michael Vogt [Tue, 29 Jul 2014 13:01:13 +0000 (15:01 +0200)]
Fix SmartConfigure to ignore ordering of packages that are already valid

With the change of SmartConfigure() in git commit 42d51f the ordering
code was trying to re-order dependencies, even when at this point in
time this was not needed. Now it will first check all targets of the
given dependency and only if there is not a good one try to reorder
and unpack/configure as needed.

Closes: LP: #1347721

9 years agol10n: vi.po (636t): Update one new string
Trần Ngọc Quân [Thu, 24 Jul 2014 08:03:45 +0000 (15:03 +0700)]
l10n: vi.po (636t): Update one new string

Signed-off-by: Trần Ngọc Quân <vnwildman@gmail.com>
9 years agoapt-pkg/acquire-item.cc: make pkgAcqDiffIndex more uniform
Michael Vogt [Thu, 17 Jul 2014 07:59:55 +0000 (09:59 +0200)]
apt-pkg/acquire-item.cc: make pkgAcqDiffIndex more uniform

9 years agoUse @builddeps@ in the debian/tests/control file
Michael Vogt [Thu, 17 Jul 2014 07:59:01 +0000 (09:59 +0200)]
Use @builddeps@ in the debian/tests/control file

9 years agoUpdate Japanese documentation translation
victory [Wed, 16 Jul 2014 12:17:09 +0000 (14:17 +0200)]
Update Japanese documentation translation

Closes: #754817

9 years agoStringToBool: only act if the entire string is consumed by strtol()
Michael Vogt [Wed, 16 Jul 2014 11:57:50 +0000 (13:57 +0200)]
StringToBool: only act if the entire string is consumed by strtol()

StringToBool uses strtol() internally to check if the argument is
a number. This function stops when it does not find any more numbers.
So a string like "0ad" (which is a valid packagename) is interpreted
as a "0". The code now checks that the entire string is consumed
not just a part of it. Thanks to Johannes Schauer for raising this
issue.

9 years agoMerge remote-tracking branch 'mvo/feature/README' into debian/sid
Michael Vogt [Thu, 10 Jul 2014 18:53:20 +0000 (20:53 +0200)]
Merge remote-tracking branch 'mvo/feature/README' into debian/sid

9 years agoadd REAMDE.md
Michael Vogt [Thu, 10 Jul 2014 15:46:26 +0000 (17:46 +0200)]
add REAMDE.md

9 years agoreleasing package apt version 1.0.6 1.0.6
Michael Vogt [Thu, 10 Jul 2014 09:46:56 +0000 (11:46 +0200)]
releasing package apt version 1.0.6

9 years agoprepare 1.0.6
Michael Vogt [Thu, 10 Jul 2014 08:52:34 +0000 (10:52 +0200)]
prepare 1.0.6

9 years agoOnly show packages as upgradable if the have a CandidateVer != 0
Michael Vogt [Tue, 8 Jul 2014 14:06:56 +0000 (16:06 +0200)]
Only show packages as upgradable if the have a CandidateVer != 0

Closes: #753297

9 years agodoc: Unfuzzy DocBook translations
Guillem Jover [Wed, 2 Jul 2014 00:38:27 +0000 (02:38 +0200)]
doc: Unfuzzy DocBook translations

9 years agodoc: Convert from DebianDoc SGML to DocBook XML
Guillem Jover [Wed, 2 Jul 2014 00:22:32 +0000 (02:22 +0200)]
doc: Convert from DebianDoc SGML to DocBook XML

9 years agobuild: Convert from DebianDoc SGML to DocBook XML
Guillem Jover [Wed, 2 Jul 2014 02:10:37 +0000 (04:10 +0200)]
build: Convert from DebianDoc SGML to DocBook XML

9 years agobuild: Set the XSL parameter through the command line instead of sed
Guillem Jover [Wed, 2 Jul 2014 23:56:52 +0000 (01:56 +0200)]
build: Set the XSL parameter through the command line instead of sed

Use the xsltproc --stringparam option instead of replacing it inside the
file.

9 years agopo: Fix format specifier order in translation
Guillem Jover [Thu, 3 Jul 2014 22:54:15 +0000 (00:54 +0200)]
po: Fix format specifier order in translation

9 years agopo: Fix encoding issues
Guillem Jover [Fri, 4 Jul 2014 06:14:07 +0000 (08:14 +0200)]
po: Fix encoding issues

Remove trailing CR characters, and fix Content-Transfer-Encoding field
value.

9 years agopo: Fix or add missing email addresses
Guillem Jover [Fri, 4 Jul 2014 06:11:06 +0000 (08:11 +0200)]
po: Fix or add missing email addresses

Fill Report-Msgid-Bugs-To field, remove spurious email line junk and
make Last-Translator and Language-Team fields contain valid emails.

9 years agopo: Fix Plural-Forms fields
Guillem Jover [Fri, 4 Jul 2014 06:09:16 +0000 (08:09 +0200)]
po: Fix Plural-Forms fields

9 years agopo: Fill Project-Id-Version with correct project id and version
Guillem Jover [Fri, 4 Jul 2014 06:03:31 +0000 (08:03 +0200)]
po: Fill Project-Id-Version with correct project id and version

9 years agopo: Remove fuzzy from file msgid header
Guillem Jover [Wed, 2 Jul 2014 15:52:34 +0000 (17:52 +0200)]
po: Remove fuzzy from file msgid header

9 years agopo: Fill or add missing Language field
Guillem Jover [Wed, 2 Jul 2014 15:52:34 +0000 (17:52 +0200)]
po: Fill or add missing Language field

9 years agoDo not clean "/" in pkgAcquire::Clean/pkgArchiveCleaner
Michael Vogt [Tue, 8 Jul 2014 09:27:01 +0000 (11:27 +0200)]
Do not clean "/" in pkgAcquire::Clean/pkgArchiveCleaner

Having "/" here is most likely a user configuration error and
may cause removal of import symlinks like /vmlinuz

Closes: #753531

9 years agoImprove description how to turn off the caches
Cédric Barboiron [Tue, 8 Jul 2014 09:24:46 +0000 (11:24 +0200)]
Improve description how to turn off the caches

Closes: #753531

9 years agouse printf instead of echo in testing framework
Michele Orrù [Mon, 7 Jul 2014 18:48:16 +0000 (20:48 +0200)]
use printf instead of echo in testing framework

The behaviour of echo "\tA\t" differs between dash/zsh which interprets
the \t as tab and bash which prints it literally. Similar things happen
for other escape sequences – without the -e flag.
Switching to printf makes this more painless^Wportable, so that the
tests are also working correctly with bash as sh.
(commit message by committer, patch otherwise unmodified)

9 years agoproperly handle (currently unused) dpkg pass-through
David Kalnischkies [Mon, 7 Jul 2014 18:45:28 +0000 (20:45 +0200)]
properly handle (currently unused) dpkg pass-through

Git-Dch: ignore

9 years agouse exit instead of incorrect return in test wrapper
Michele Orrù [Mon, 7 Jul 2014 18:43:45 +0000 (20:43 +0200)]
use exit instead of incorrect return in test wrapper

bash as sh doesn't like it, too.

Git-Dch: Ignore

9 years agohandle moved mmap after UniqFindTagWrite call
David Kalnischkies [Mon, 7 Jul 2014 15:19:29 +0000 (17:19 +0200)]
handle moved mmap after UniqFindTagWrite call

A call to UniqFindTagWrite can trigger the need for a bigger mmap, which
is usually done by moving it, but with this move all pointers into it
become invalid (and have to be remapped). The compiler calculates the
pointer before the execution of the call though, so it tries to store
the returned value at the old location, resulting in a segfault.
We solve this by use of a temprorary variable as we did in the other
instances of this problem before.

Closes: #753941

9 years agoDanish program translation update
Joe Hansen [Mon, 7 Jul 2014 10:18:58 +0000 (12:18 +0200)]
Danish program translation update

Closes: 753979

9 years agoTry not to parse invalid translation files (LP: #756317)
Michael Vogt [Thu, 3 Jul 2014 16:20:44 +0000 (18:20 +0200)]
Try not to parse invalid translation files (LP: #756317)

9 years agoredo two typo fixes lost in German update collision
David Kalnischkies [Wed, 25 Jun 2014 11:37:58 +0000 (13:37 +0200)]
redo two typo fixes lost in German update collision

In 037fada40db175e95f44c0cb039474c6dc518963 two typos were fixed which
collided with an update ongoing for German po which was committed in
48388911a734ed9ce10ab05043fadba3eed09a1f, so this commit just merges the
first commit again.

Git-Dch: Ignore

9 years agol10n: vi.po: Update 3 new messages
Trần Ngọc Quân [Sun, 22 Jun 2014 02:20:48 +0000 (09:20 +0700)]
l10n: vi.po: Update 3 new messages

Signed-off-by: Trần Ngọc Quân <vnwildman@gmail.com>
9 years agomethods/http.cc: use Req.str() in debug output
Michael Vogt [Tue, 24 Jun 2014 13:45:09 +0000 (15:45 +0200)]
methods/http.cc: use Req.str() in debug output

9 years agoGerman translation reviewed by Erik Pfannenstein
Chris Leick [Thu, 19 Jun 2014 20:01:48 +0000 (22:01 +0200)]
German translation reviewed by Erik Pfannenstein

9 years agoreleasing package apt version 1.0.5 1.0.5
Michael Vogt [Wed, 18 Jun 2014 11:36:36 +0000 (13:36 +0200)]
releasing package apt version 1.0.5

9 years agoEDSP doc: some typo and wording fixes
David Kalnischkies [Tue, 17 Jun 2014 17:25:03 +0000 (19:25 +0200)]
EDSP doc: some typo and wording fixes

Git-Dch: Ignore

9 years agoEDSP doc: (minor) consistently use 2 blank lines before headings
Stefano Zacchiroli [Tue, 17 Jun 2014 15:04:09 +0000 (17:04 +0200)]
EDSP doc: (minor) consistently use 2 blank lines before headings

Git-Dch: Ignore

9 years agoEDSP doc: clarify that Install/Remove packages are arch-qualified
Stefano Zacchiroli [Tue, 17 Jun 2014 15:03:27 +0000 (17:03 +0200)]
EDSP doc: clarify that Install/Remove packages are arch-qualified

9 years agodo not call resolver twice on (dist-)upgrade
David Kalnischkies [Tue, 17 Jun 2014 17:05:53 +0000 (19:05 +0200)]
do not call resolver twice on (dist-)upgrade

9 years agoshow our broken packages message in 'apt' solver
David Kalnischkies [Tue, 17 Jun 2014 15:45:33 +0000 (17:45 +0200)]
show our broken packages message in 'apt' solver

9 years agofix SubstVar to be usable as a replace_all method
David Kalnischkies [Tue, 17 Jun 2014 14:55:19 +0000 (16:55 +0200)]
fix SubstVar to be usable as a replace_all method

The name suggests that it is supposed to substitute a variable with a
value, but we tend to use it in a more liberal replace_all() fashion,
but this breaks if either of the parameters is empty or more importantly
if two "variable" occurrences follow each other directly.

9 years agodon't send pkg from an unknown architecture via EDSP
David Kalnischkies [Tue, 17 Jun 2014 11:47:01 +0000 (13:47 +0200)]
don't send pkg from an unknown architecture via EDSP

APT's cache can include packages from architectures dpkg has no
knowledge about and can therefore not be installed for e.g. to allow
easy lookups. There is no point in telling external solvers about them
though and some of them might even be really talkative about ignoring
them if we do.

9 years agouse P_ instead of ngettext to compiling with --disable-nls
Fredrik Fornwall [Tue, 17 Jun 2014 09:33:00 +0000 (11:33 +0200)]
use P_ instead of ngettext to compiling with --disable-nls

Closes: 751857

9 years agoadd missing comma in SEE ALSO of apt-secure manpage
Konstantin Manna [Sun, 15 Jun 2014 16:35:15 +0000 (18:35 +0200)]
add missing comma in SEE ALSO of apt-secure manpage

Closes: 748506

9 years agofix two german manpage spelling mistakes
Konstantin Manna [Sat, 14 Jun 2014 22:53:04 +0000 (00:53 +0200)]
fix two german manpage spelling mistakes

Closes: 751635

9 years agoMerge remote-tracking branch 'mvo/debian/sid' into debian/sid
Michael Vogt [Wed, 18 Jun 2014 06:32:59 +0000 (08:32 +0200)]
Merge remote-tracking branch 'mvo/debian/sid' into debian/sid

9 years agoTell the user if no updates are available after apt update
Michael Vogt [Wed, 18 Jun 2014 06:25:16 +0000 (08:25 +0200)]
Tell the user if no updates are available after apt update

Thanks to Jakub Wilk for the suggestion.

Closes: #751388

9 years agoMerge remote-tracking branch 'mvo/debian/sid' into debian/sid
Michael Vogt [Wed, 18 Jun 2014 06:14:08 +0000 (08:14 +0200)]
Merge remote-tracking branch 'mvo/debian/sid' into debian/sid

9 years agoUpdated translation of german documentation
Chris Leick [Tue, 17 Jun 2014 19:05:41 +0000 (21:05 +0200)]
Updated translation of german documentation

9 years agoCheck for gtest's header before building.
Michele Orrù [Tue, 17 Jun 2014 09:42:31 +0000 (11:42 +0200)]
Check for gtest's header before building.

Add AC_CHECK_HEADER directive for gtest, and report a fatal error message if not
found.

9 years agotest/integration/test-essential-force-loopbreak: fix on non-amd64 systems
Michael Vogt [Thu, 12 Jun 2014 08:09:24 +0000 (10:09 +0200)]
test/integration/test-essential-force-loopbreak: fix on non-amd64 systems

9 years agofix test-apt-ftparchive-cachedb-lp1274466 and apt-internal-solver tests
Michael Vogt [Wed, 11 Jun 2014 16:17:45 +0000 (18:17 +0200)]
fix test-apt-ftparchive-cachedb-lp1274466 and apt-internal-solver tests

9 years agofix autopkgtest tests
Michael Vogt [Wed, 11 Jun 2014 12:50:48 +0000 (14:50 +0200)]
fix autopkgtest tests

9 years agoreally release 1.0.4 1.0.4
Michael Vogt [Tue, 10 Jun 2014 16:27:34 +0000 (18:27 +0200)]
really release 1.0.4

9 years agorelease 1.0.4
Michael Vogt [Tue, 10 Jun 2014 13:59:15 +0000 (15:59 +0200)]
release 1.0.4

9 years agoapt-private/private-output.cc: fix cppcheck output
Michael Vogt [Tue, 10 Jun 2014 13:22:53 +0000 (15:22 +0200)]
apt-private/private-output.cc: fix cppcheck output

9 years agofix test/integration/test-apt-ftparchive-cachedb-lp1274466 in travis
Michael Vogt [Tue, 10 Jun 2014 13:21:30 +0000 (15:21 +0200)]
fix test/integration/test-apt-ftparchive-cachedb-lp1274466 in travis

9 years agouse pkgSrcRecords::Step() instead of Next()
Michael Vogt [Tue, 10 Jun 2014 13:07:15 +0000 (15:07 +0200)]
use pkgSrcRecords::Step() instead of Next()

9 years agoprepare 1.0.4 upload
Michael Vogt [Tue, 10 Jun 2014 12:59:58 +0000 (14:59 +0200)]
prepare 1.0.4 upload

9 years agoMerge remote-tracking branch 'mvo/feature/apt-update-info' into debian/sid
Michael Vogt [Tue, 10 Jun 2014 12:49:58 +0000 (14:49 +0200)]
Merge remote-tracking branch 'mvo/feature/apt-update-info' into debian/sid

9 years agoMerge remote-tracking branch 'mvo/bugfix/apt-get-source-unauthenticated-warning'...
Michael Vogt [Tue, 10 Jun 2014 12:47:03 +0000 (14:47 +0200)]
Merge remote-tracking branch 'mvo/bugfix/apt-get-source-unauthenticated-warning' into debian/sid

9 years agoMerge remote-tracking branch 'mvo/bugfix/lp1274466-cache' into debian/sid
Michael Vogt [Tue, 10 Jun 2014 12:46:02 +0000 (14:46 +0200)]
Merge remote-tracking branch 'mvo/bugfix/lp1274466-cache' into debian/sid

9 years agoMerge remote-tracking branch 'mvo/feature/srcrec-enum2' into debian/sid
Michael Vogt [Tue, 10 Jun 2014 12:43:56 +0000 (14:43 +0200)]
Merge remote-tracking branch 'mvo/feature/srcrec-enum2' into debian/sid

9 years agosupport Acquire::GzipIndexes in dumpavail
David Kalnischkies [Tue, 10 Jun 2014 12:12:12 +0000 (14:12 +0200)]
support Acquire::GzipIndexes in dumpavail

Closes: 742835

9 years agoseparate implementation from declaration of TryTo{Install,Remove}
David Kalnischkies [Mon, 9 Jun 2014 13:50:20 +0000 (15:50 +0200)]
separate implementation from declaration of TryTo{Install,Remove}

Git-Dch: Ignore

9 years agoadd pkgSrcRecords::Next() to step through all the pkgSrcRecords
Michael Vogt [Mon, 9 Jun 2014 07:06:07 +0000 (09:06 +0200)]
add pkgSrcRecords::Next() to step through all the pkgSrcRecords

9 years agodo not revert candidate for protected packages
David Kalnischkies [Sat, 7 Jun 2014 20:46:37 +0000 (22:46 +0200)]
do not revert candidate for protected packages

In commit 21b3eac8 I promoted the check for installable dependencies to
a pre-install check, which also reverts to a known good candidate (the
installed version) if it fails. This revert was done even for user
requested candidate switches which disabled our Broken detection so that
install requests which are impossible to satisfy do not fail anymore,
but print an (incomplete) solution proposal and then exit successfully.

Closes: 745046

9 years agoItalian program translation update
Milo Casagrande [Sat, 7 Jun 2014 15:17:20 +0000 (17:17 +0200)]
Italian program translation update

Closes: 750009

9 years agocmdline/apt-helper.cc: use less generic description/short-description
Michael Vogt [Fri, 6 Jun 2014 09:24:44 +0000 (11:24 +0200)]
cmdline/apt-helper.cc: use less generic description/short-description

9 years agoAdd compat mode for old (32bit FileSize) CacheDB (LP: #1274466)
Michael Vogt [Wed, 4 Jun 2014 10:39:36 +0000 (12:39 +0200)]
Add compat mode for old (32bit FileSize) CacheDB (LP: #1274466)

9 years agoShow unauthenticated warning for source packages as well
Michael Vogt [Fri, 30 May 2014 12:47:56 +0000 (14:47 +0200)]
Show unauthenticated warning for source packages as well

This will show the same unauthenticated warning for source packages
as for binary packages and will not download a source package if
it is unauthenticated. This can be overridden with
--allow-unauthenticated

Closes: #749795

9 years agouse 'native' instead of 'amd64' as pkg arch
David Kalnischkies [Fri, 30 May 2014 21:15:28 +0000 (23:15 +0200)]
use 'native' instead of 'amd64' as pkg arch

Git-Dch: Ignore

9 years agoupdate symbols files
David Kalnischkies [Fri, 30 May 2014 21:10:30 +0000 (23:10 +0200)]
update symbols files

Git-Dch: ignore

9 years agoshow current/total testcase statistic also in concise mode
David Kalnischkies [Fri, 30 May 2014 17:18:25 +0000 (19:18 +0200)]
show current/total testcase statistic also in concise mode

Git-Dch: Ignore

9 years agosupport parsing EDSP requests Architecture{,s} stanza
David Kalnischkies [Fri, 30 May 2014 16:01:47 +0000 (18:01 +0200)]
support parsing EDSP requests Architecture{,s} stanza

Adds also a small testcase for EDSP

Git-Dch: Ignore

9 years agowithout a filename we can't stat pipes
David Kalnischkies [Fri, 30 May 2014 15:14:13 +0000 (17:14 +0200)]
without a filename we can't stat pipes

EDSP code uses pipes opened via an FD as sources and later for those
files modification times and filesize are read - but never really used
again. The result we get from FileFd is probably wrong, but as we don't
use it anyway, we just don't fallback if we have nothing to fallback to

Git-Dch: Ignore

9 years agoprint error stack on failure of 'apt' solver
David Kalnischkies [Fri, 30 May 2014 14:52:19 +0000 (16:52 +0200)]
print error stack on failure of 'apt' solver

Git-Dch: ignore

9 years agocheck exit status of external solvers
David Kalnischkies [Fri, 30 May 2014 13:58:30 +0000 (15:58 +0200)]
check exit status of external solvers

Solvers are supposed to exit successfully even if they haven't found a
solution, but a solver which fails drastically (like e.g. segfaults)
should be detected and dealt with accordingly instead of ignored.

9 years agoif Resolver fails, do not continue even if not broken
David Kalnischkies [Fri, 30 May 2014 11:42:36 +0000 (13:42 +0200)]
if Resolver fails, do not continue even if not broken

This can happen if the request is already a well-formed request all by
itself (e.g. the package has no dependencies), but the resolver found
a reason to not accept it as solution. Our edsp 'dump' solver e.g.
shouldn't be able to trigger install, which it does otherwise.

9 years agoMerge EDSP 0.5 w/ multi-arch support for external solvers
David Kalnischkies [Fri, 30 May 2014 11:13:03 +0000 (13:13 +0200)]
Merge EDSP 0.5 w/ multi-arch support for external solvers

"I am going to merge it tomorrow…"

9 years agoMerge remote-tracking branch 'mvo/debian/sid' into debian/sid
Michael Vogt [Thu, 29 May 2014 10:23:51 +0000 (12:23 +0200)]
Merge remote-tracking branch 'mvo/debian/sid' into debian/sid

9 years ago(try to) fix travis-ci build failures
David Kalnischkies [Mon, 12 May 2014 19:25:43 +0000 (21:25 +0200)]
(try to) fix travis-ci build failures

dpkg on Ubuntu 12.04 does not seem to support parsing arch-specific
dependencies, so we try to detect if we face such a dpkg in the test.

In the other test the order depends on libdb, which changes per arch, so
we just run it through our sorting binary and be happy (hopefully).

Git-Dch: Ignore

9 years agoFix warning about uninitialized variable
Michael Vogt [Wed, 28 May 2014 08:00:52 +0000 (10:00 +0200)]
Fix warning about uninitialized variable

Reported-By: clang++ -Werror