В письме от 8 марта 2013 22:31:52 пользователь Alexander Neundorf написал:
> On Friday 08 March 2013, Yury G. Kudryashov wrote:
> > There was no FindLibUSB1.cmake in kdelibs. Do we mention source
> > incompatibilities with old unreleased ECM versions?
>
> Doesn't it repl
omatically generated e-mail. To reply, visit:
> > > http://git.reviewboard.kde.org/r/108954/#review27490
> > > ---
> > >
> > >
> > > This review has been submitted with commit
> > > 095d
Rex Dieter wrote:
> Yury G. Kudryashov wrote:
>
>> I don't think that adding REQUIRED is the right thing to do. Probably,
>> if(NOT TARGET nepomukcore)
>>if(KdepimLibs_FIND_REQUIRED)
>> find_package(NepomukCore REQUIRED NO_MODULE)
>>
Andreas Pakulat wrote:
> Hi,
>
> On Thu, Dec 20, 2012 at 2:47 PM, Christophe Giboudeaux
> wrote:
>> On Thursday 20 December 2012 14:12:17 Yury G. Kudryashov wrote:
>>> I don't think that adding REQUIRED is the right thing to do. Probably,
>>
mukCore NO_MODULE)
+endif()
to KdepimLibsConfig.cmake.in. NepomukCore is Required by kdepimlibs (look at
CMakeLists.txt in master and KDE/4.10, not in rc1).
I don't think that adding REQUIRED is the right thing to do. Probably,
if(NOT TARGET nepomukcore)
if(KdepimLibs_F
Stephen Kelly wrote:
> Yury G. Kudryashov wrote:
>
>> Look at git://gitorious.org/~urkud1/cmake/urkud-cmake.git w/export-set
>>
>> I have holidays till 10 May, and this is in my TODO list for these
>> holidays.
>
> Cool. The commits mostly look good so
Yury G. Kudryashov wrote:
> I've rebased my changes on trunk. It compiles. `make test` is still
> running.
Oops... ExportImport test fails. Trying to fix the problem
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem maili
Stephen Kelly wrote:
> Stephen Kelly wrote:
>
>> Yury G. Kudryashov wrote:
>>
>>> Stephen Kelly wrote:
>>>
>>>> Alexander Neundorf wrote:
>>>>
>>>>> Hi,
>>>>>
>>>>> AFAIK there is still
some uncommitted work in progress. I'll merge it with master
tomorrow, and push to gitorious.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
t it's not pressing, by any means. We can revisit
> for Frameworks 5, when we have more code using it.
Then it could be added either to extra-cmake-modules or cmake.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Alexander Neundorf wrote:
> On Sunday 11 March 2012, Yury G. Kudryashov wrote:
>> Alexander Neundorf wrote:
>> > Hi,
>> >
>> > I think things are looking quite good, and we got most of the things we
>> > wanted to get into cmake in.
>>
&g
by one, or download all using "meta" repo, and build the whole package.
I think that each component (libstreamanalyzer etc.) should install its own
Export file. But if you build the whole project, you get a library in export
B that depends on a library in export A.
--
Yury G.
some feedback here before
talking to digikam developers.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Le 10 февраля 2012 20:12:16 Alexander Neundorf a écrit :
> On Tuesday 17 January 2012, Yury G. Kudryashov wrote:
> > Alexander Neundorf wrote:
> ...
>
> > > Before going too far in one direction, please post a proposal to
> > > cmake-
> > > developers.
&g
IN32 special cased ?
> >
> > `man cmake` says that cmake does not look into $prefix/lib/cmake/name on
> > Windows.
>
> Indeed.
> Our KDE-on-Windows never complained about that the Config.cmake files would
> not be found.
> Also, from a quick look at cmFindPackageCommand.c
ME@ @value@)
// Next line for each 'var' in EXTRA_VARS
set(@var@ @${var}@)
@EXTRA_CONFIG_FILE_CONFIGURED@
The PackageConfig.cmake will contain the same information adjusted for
"internal" usage.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
signature.asc
Description: This is
Alexander Neundorf wrote:
> On Wednesday 18 January 2012, Yury G. Kudryashov wrote:
>> Sune Vuorela wrote:
>> > Wouldn't it be a even better idea to get SDO to provide a
>> > SharedDesktopOntologiesConfig.cmake file ?
>>
>> Look at the attached FindSh
Sune Vuorela wrote:
> On 2012-01-17, Yury G. Kudryashov wrote:
>>
>> --nextPart3022437.JoNrGlkKet
>> Content-Type: text/plain; charset="ISO-8859-1"
>> Content-Transfer-Encoding: 7Bit
>>
>> Hi!
>>
>> I propose to remove all the comp
Hi!
I propose to remove all the compatibility stuff from
FindSharedDesktopOntologies.cmake for KF5 (see attachment).
BTW, Where should one place mark_as_advanced(), in *Config.cmake or in
Find*.cmake?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru# - Try to find shared-desktop-ontologies
# The
Alexander Neundorf wrote:
> On Tuesday 17 January 2012, Yury G. Kudryashov wrote:
>> Alexander Neundorf wrote:
>> > On Tuesday 17 January 2012, Yury G. Kudryashov wrote:
>> >> so we need to ensure (or document?) that the
>> >> same directory used while w
ctory.
>
> Providing something which makes testing the installed files easier would
> be also good. I'm not sure yet how to do it.
What do you mean by "testing the installed files"?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
_
Alexander Neundorf wrote:
> On Tuesday 17 January 2012, Yury G. Kudryashov wrote:
>> so we need to ensure (or document?) that the
>> same directory used while writing MyLibConfig.cmake and installing it.
>
> Hmm, I don't understand exactly what you mean, can you p
Alexander Neundorf wrote:
> On Tuesday 17 January 2012, Yury G. Kudryashov wrote:
>> I have to know the DESTINATION at generation time to hardcode correct
>> relative paths.
>
> They can be calculated too, see the determine_installed_location() macro
> in the branch Impro
Alexander Neundorf wrote:
> On Monday 16 January 2012, Yury G. Kudryashov wrote:
>> Alexander Neundorf wrote:
>> > On Sunday 15 January 2012, Yury G. Kudryashov wrote:
>> >> 15 January 2012 20:06:51 Alexander Neundorf written:
>> >> The only com
Alexander Neundorf wrote:
> On Sunday 15 January 2012, Yury G. Kudryashov wrote:
>> 15 January 2012 20:06:51 Alexander Neundorf written:
>> The only comment: there are 4 possible combinations of
>> cmake -DLIB_INSTALL_DIR=relative_or_absolute -
>> DINCLUDE_INSTALL_DIR
Yury G. Kudryashov wrote:
> 15 January 2012 20:06:51 Alexander Neundorf written:
>> Hi,
> Hi,
>
> Try #1 awaits moderator approval and can be safely deleted.
>>
>> I create a new branch ImprovedConfigDotCMakeFile in the kdeexamples
>> repository, used by
BRARY_TARGETS@)
// Next line for each NAME=value in EXTRA_VARS
set(@NAME@ @value@)
// Next line for each 'var' in EXTRA_VARS
set(@var@ @${var}@)
@EXTRA_CONFIG_FILE_CONFIGURED@
The PackageConfig.cmake will contain the same information adjusted for
"internal" usage.
--
Yury G. Kud
discuss this on
> kde-buildsystem or k-c-d.
>
> - Alexander Neundorf
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Alexander Neundorf wrote:
> On Saturday 14 January 2012, Yury G. Kudryashov wrote:
>> Hi!
>>
>> Should I commit
>> http://commits.kde.org/ksecrets/ac27ebfe7a2cbb2ed3b0561b42003ab507a28d1c
>
> Where does PkgConfigGetVar come from actually ?
kde-workspace
-
Hi!
Should I commit
http://commits.kde.org/ksecrets/ac27ebfe7a2cbb2ed3b0561b42003ab507a28d1c
and
https://projects.kde.org/projects/kde/kdeedu/kalzium/repository/revisions/6a53035397998db107a9ff4fcd47e82c5586fa49
to KDE/4.8?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
Andreas Pakulat wrote:
> On 12.01.12 22:48:10, Alexander Neundorf wrote:
>> On Thursday 12 January 2012, Yury G. Kudryashov wrote:
>> > Alexander Neundorf wrote:
>> > > I think FindLibUSB1.cmake looks good.
>> >
>> > From libusb-1.0.pc:
&
Alexander Neundorf wrote:
> On Thursday 12 January 2012, Yury G. Kudryashov wrote:
>> Alexander Neundorf wrote:
>> > I think FindLibUSB1.cmake looks good.
>>
>> From libusb-1.0.pc:
>> includedir=${prefix}/include
>> Cflags: -I${includedir}/libusb-1.0
&g
en somebody will need it, he will just copy the old FindUSB.cmake. What do
you think about "review/" subdir in e-c-m?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
nside with the second approach is that this FindTaglib will be
unable to find "old" taglib libraries. Is it OK for ECM?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/m
08130.html
http://mail.kde.org/pipermail/kde-buildsystem/2011-September/008131.html
Should I commit these files to ECM/find-modules/ or should I fix some other
issues?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru# - Try to find msgfmt
# Once done this will define
#
# MSGFMT_FOUND - system has m
n at package-generation time.
Is it possible to let installer rewrite prefixes in files like d-bus
services, FooConfig.cmake etc.?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
k for
> Windows...
You can answer yourself by cloning cmake.git... CMake now has FindFLEX.cmake
that superseeds FindFlex.cmake from kdelibs.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://
Yury G. Kudryashov wrote:
> Alexander Neundorf wrote:
>
>> On Sunday, September 18, 2011 10:59:25 AM Yury G. Kudryashov wrote:
>>> Hi!
>>>
>>> I think that presense of FindLibUSB1.cmake will slightly simplify
>>> transition from libusb-0.1 to
Alexander Neundorf wrote:
> On Sunday, September 18, 2011 10:59:25 AM Yury G. Kudryashov wrote:
>> Hi!
>>
>> I think that presense of FindLibUSB1.cmake will slightly simplify
>> transition from libusb-0.1 to libusb-1.0. Review the attached file,
>> please.
>
Alexander Neundorf wrote:
> On Sunday, September 18, 2011 10:40:59 AM Yury G. Kudryashov wrote:
>> Yury G. Kudryashov wrote:
>> > Hi!
>> >
>> > Reviewboard failed to understand `git diff -M` patch, so I'm sending
>> > both the patch and the
Hi!
Review the attached version of FindMsgfmt, please.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru# - Try to find msgfmt
# Once done this will define
#
# MSGFMT_FOUND - system has msgfmt
# MSGFMT_EXECUTABLE - full path of msgfmt
# MSGFMT_VERSION - the version as reported by msgfmt
Hi!
I think that presense of FindLibUSB1.cmake will slightly simplify transition
from libusb-0.1 to libusb-1.0. Review the attached file, please.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru# - Try to find the libusb-1.0 library
# Once done this defines
#
# LIBUSB1_FOUND - system has libusb
Yury G. Kudryashov wrote:
> Hi!
>
> Reviewboard failed to understand `git diff -M` patch, so I'm sending both
> the patch and the new FindUSB.cmake here.
>
> Should I add OPTIONAL to include(FeatureSummary) until cmake-2.8.6 is not
> out?
BTW, should we either rena
Hi!
Reviewboard failed to understand `git diff -M` patch, so I'm sending both
the patch and the new FindUSB.cmake here.
Should I add OPTIONAL to include(FeatureSummary) until cmake-2.8.6 is not
out?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ruFrom: Yury G. Kudryashov
Subject: [PATCH] R
found;
* GSTREAMER_LIBRARIES contains ONLY libgstreamer.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Alexander Neundorf wrote:
> On Saturday, September 17, 2011 04:39:59 AM Alexander Neundorf wrote:
>> On Friday, September 16, 2011 08:40:14 PM Yury G. Kudryashov wrote:
>> F**k, there is a FindFLEX.cmake in cmake, and we have a FindFlex.cmake.
>> Probably due to the differe
find_package_handle_standard_args(MyPackage DEFAULT_MSG
MyPackage_INCLUDE_DIRS MyPackage_LIBRARIES)
?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Alexander Neundorf wrote:
> On Friday, September 16, 2011 08:47:52 PM Yury G. Kudryashov wrote:
>> Alexander Neundorf wrote:
>> > Hi,
>> >
>> > same as with many others:
>> >
>> > * unnecessary shortcut at the top
>> > * unnecessary
Hi!
macro_log_feature can display only a basic information (name+description)
for found features, and display the detailed information for not found
packages. Are there any plans to add such feature to feature_summary?
P.S.: Where should I submit typo fixes for FeatureSummary.cmake?
--
Yury G
ugh.
* Use lowercase.
Should we add GSTREAMER_INCLUDE_DIRS to follow cmake/Modules/readme.txt
recommendations?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru# - Try to find GStreamer
# Once done this will define
#
# GSTREAMER_FOUND - system has GStreamer
# GSTREAMER_VERSION - version as repor
l
FAM package or by gamin.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Hi!
Should I just commit to extra-cmake-modules/attic/modules, or submit patches
to reviewboard?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde
DIRS, ATTR_INCLUDE_DIRS to $acl_prefix/include;
$attr_prefix/include.
* No documentation for ATTR_LIBS.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
atibility layer.
> Are FindDNSSD.cmake and FindAvahi.cmake (which looks weird/bad) in some
> way related ?
>
> Beside that, it should use FPHSA().
And it should call
find_path(DNSSD_INCLUDE_DIR dns_sd.h PATH_SUFFIXES avahi-compat-libdns_sd)
instead of the current hardcoded /usr/include/ava
ant_prefix/include/enchant to the user?
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
Alexander Neundorf wrote:
> Hi,
>
> very basic, no problems.
> Could be changed to all-lowercase, if somebody feels like doing that.
>
It seems that usually we pass the library path as the first argument to
REQUIRED_VARS in fphsa. Here I see FAM_INCLUDE_DIR first.
--
Yur
o probably one never need only
one of these components. In this case, just ignore this point.
--
Yury G. Kudryashov,
mailto: ur...@mccme.ru
___
Kde-buildsystem mailing list
Kde-buildsystem@kde.org
https://mail.kde.org/mailman/listinfo/kde-buildsystem
eloper, I just submit patches
sometimes...
As far as I understand, kauth can (does?) use different backends on
different platforms, and polkit-qt-1 is "the" backend on Linux.
I think that it should be clear that kauth does not depend on other tier1
libraries except for polkit-qt-1.
; which we do not have much contorl (Raptor, Virtuoso etc)?
As far as I understand, the proposed scenario for projects with cmake-based
buildsystem is:
* the project itself installs ProjectNameConfig.cmake;
* ecm (or cmake) distribute a short FindProject.cmake that calls
fin
Git commit 0b24b899fe92008ebdeeecd4871b63233d7f3839 by Yury G. Kudryashov.
Committed on 14/08/2011 at 22:53.
Pushed by kudryashov into branch 'master'.
Display feature log
CC:kde-buildsystem@kde.org
M +1-0CMakeLists.txt
http://commits.kde.o
Friedrich W. H. Kossebau wrote:
> Hi,
>
> I have blindly copied this for cagibi but since then wonder if this is
> correct in FindKDE4Internal.cmake:
>
>_set_fancy(SYSCONF_INSTALL_DIR "${CMAKE_INSTALL_PREFIX}/etc"
> "The sysconfig install dir (default ${CMAKE_INSTALL_PREFIX}/etc)")
>
>
Harald Sitter wrote:
> On Sat, Mar 12, 2011 at 9:11 PM, Alexander Neundorf
> wrote:
>> The pages here: http://www.vtk.org/Wiki/CMake/Tutorials explain how to do
>> this.
>
> There is a branch for that (SaroEngels was working on it), never got
> merged thought *shrug*.
Could you please provide a
Andreas Pakulat wrote:
> On 19.02.11 00:29:57, Yury G. Kudryashov wrote:
>> set(POLKITQT-1_POLICY_FILES_INSTALL_DIR "share/polkit-1/actions")
>>
>> in polkit-qt-1/PolkitQt-1Config.cmake.in. And, of course, adjust the
>> warning.
>
> And how do I do
John Layt wrote:
> On Friday 18 February 2011 06:48:54 Yury G. Kudryashov wrote:
>> >> 2) If path different from polkit-qt AUTH_POLICY_FILES_INSTALL_DIR then
>> >> display an appropriate warning
>> >> 3) Install to the requested location, it's then up
Alexander Neundorf wrote:
> Hi,
>> The way I think this should work is:
>> 1) Use KDE4_AUTH_POLICY_FILES_INSTALL_DIR if specified, else use KDEDIR
Why KDEDIR, not CMAKE_INSTALL_PREFIX?
>> 2) If path different from polkit-qt AUTH_POLICY_FILES_INSTALL_DIR then
>> display an appropriate warning
>> 3)
Alexander Neundorf wrote:
> On Wednesday 09 February 2011, Yury G. Kudryashov wrote:
>> Offtopic: I'm playing with translating qmake to cmake (mostly to learn
>> cmake). What is the cmake way to do what
>> qt/tools/linguist/shared/formats.pri does?
>>
>> IN
Alexander Neundorf wrote:
> Is that what QMAKEPATH is for, i.e. a list of directories where qmake will
> search ?
Qt up to 4.7.1 doesn't load QMAKEPATH/mkspecs/features/qt_*.pri
automatically. If you want qt-4.7.2 to load these files, ask qt guys to
backport 210726453ff3ca0f16f036683e17097e1b62d
Michael Jansen wrote:
> Hi
>
> We currently have some problem for people that compile phonon with a
> distro qt. Phonon tries to install a file into the qt prefix. The file is
> named qt_phonon.pri and is according to thiago a sign of things to come
> namely some modularization in qt. But it is n
Akarsh Simha wrote:
> Thanks for your replies Yuri and Andreas. Sorry for re-initiating this
> discussion.
>
> Yes, I notice that I don't have CMAKE_PREFIX_PATH set, although I have
> "$KDEDIR" set.
KDEDIRS, not KDEDIR
___
Kde-buildsystem mailing list
Alexander Neundorf wrote:
> Uh, that doesn't look too good, it hardcodes the "lib" prefix and the
> ".so" suffix. Does that work on Windows ? OSX ?
>
> How can it be done better:
> 1.) use exported targets, then this is done automatically. For this the
> targets have to be put into an "export set
Akarsh Simha wrote:
> I have a question regarding CMAKE_MODULES_PATH.
>
> So I tried exporting CMAKE_MODULE_PATH, appropriately, but it didn't
> work, and I attribute that to a line in kdeedu/CMakeLists.txt that
> says:
>
> set(CMAKE_MODULE_PATH ${CMAKE_SOURCE_DIR}/cmake/modules )
>
> Shouldn't
Andreas Hartmetz wrote:
> On Sunday 26 December 2010 10:05:40 Yury G. Kudryashov wrote:
>> 2. Create a kdesupport project for cmake macroses.
>>
> Then we still have to find a good install path. May or may not be a
> problem.
This package can install all macroses to $prefix/
Andreas Hartmetz wrote:
> On Saturday 25 December 2010 21:06:03 Yury G. Kudryashov wrote:
>> Andreas Hartmetz wrote:
>> > I think FindPhonon.cmake being installed by kdelibs, into a
>> > KDE-specific path, is the underlying problem that needs to
>> >
Andreas Hartmetz wrote:
> I think FindPhonon.cmake being installed by kdelibs, into a
> KDE-specific path, is the underlying problem that needs to
> be fixed. Looks like a circular dependency of sorts to me.
> We're going to see similar problems in the future with kdesupport
>
luxInteg wrote:
> Greetings,
>
> I keep stalling on line 100 of FindOOoSDK.cmake shipped with the current
> version of koffice (of kde.org) (A copy is attached) I have all the
> prerequisites installed in /opt/openoffice.org. I would be grateful if
> someone more aufait with cmake modules co
Michael Jansen wrote:
> On Friday 29 October 2010 15:03:44 Flavio Castelli wrote:
>> Hello I'm qjson maintainer and I have a problem with cmake.
>> Some times ago Michael Jansen reported qjson had some issues with cmake:
>> http://goo.gl/gKw6
>
> We had that discussion with a result i personally
Allen Winter wrote:
> Adding the KDE Buildsystem to this discussion..
>
> My only concern is that pkg-config files are typically installed in
> locations (like /usr or /usr/local) where the regular user doesn't have
> write-access. Thereby causing problems when installing KDE as a regular
> user.
David Faure wrote:
> Recently I made all application .desktop files executable in trunk.
> But at install time, the permissions are not copied over, so the files
> don't end up being executable.
Why system-wide .desktop files should have executable bit? I thought that
this bit is not required for
Alexander Neundorf wrote:
> On Friday 08 October 2010, Yury G. Kudryashov wrote:
>> You can ask cmake
>> developers to look in CMAKE_LIBRARY_PATH as well.
>
> No, I wouldn't do that. CMAKE_PREFIX_PATH should be used.
>
> You can add this as a feature request
Ingomar Wesp wrote:
> Alexander Neundorf wrote:
>
> If FindAkonadi.cmake is the only culprit, it would certainly be nice if
> someone could fix it as well - that is, if we still aim to support more
> exotic install hierarchies. Unfortunately, I currently don't have the
> necessary knowledge to do
Ingomar Wesp wrote:
> You might have guessed that I'm not too familiar with cmake, so let me
> outline the problem I ran into:
>
> I tried to build and install the following from trunk in the following
> sequence: phonon, kdesupport/soprano, kdesupport/akonadi,
> kdesupport/attica, kdelibs, kdepi
Kåre Särs wrote:
> On Sunday 03 October 2010 22:14:22 Alexander Neundorf wrote:
>> On Thursday 30 September 2010, Kåre Särs wrote:
>> > Hi,
>> >
>> > Alexander Neundorf noted that it is not such a good idea to have
>> > FindKSane.cmake in the module that provides the library. So now I would
>> >
Alexander Neundorf wrote:
> To make sure that the FindPackageHandleStandardArgs.cmake from kdelibs is
> found, CMAKE_PREFIX_PATH should be set so that the kdelibs
> share/apps/cmake/modules/ directory comes before the cmake standard
> directory (which should not be listed in this variable at all).
Alexander Neundorf wrote:
> On Monday 16 August 2010, Yury G. Kudryashov wrote:
> They are outside trunk/KDE/, I can't care about them:
>> libktorrent-1.0.2
>> qjson-0.7.1
> It would be nice if you could send the authors an email explaining that
> that's wro
Alexander Neundorf wrote:
> On Thursday 02 September 2010, Yury G. Kudryashov wrote:
>> Hi!
>>
>> Check for acl.h fails on my laptop because acl.h contains C++-style
>> comment, and cmake uses too strict compiler flags.
>>
>> What do you think
Hi!
Check for acl.h fails on my laptop because acl.h contains C++-style comment,
and cmake uses too strict compiler flags.
What do you think about replacing check_include_files by
check_include_file_cxx?
___
Kde-buildsystem mailing list
Kde-buildsyst
Paulo Rômulo wrote:
> Hello.
>
> I'm intending to add the attached file to kdelibs/cmake/modules in order
> to conditionally build the Solid UPnP backend. This file tries to find the
> HUPnP library on the system. Is everything ok with that? Can I commit?
I don't think that it is a good idea to
Maciej Mrozowski wrote:
> With that mind, runtime-only dependency handling could be simplified to:
> (see attachment):
>
> macro_log_runtime_dependency("FOO" "File Ownership Obstuctor"
> "http://foo.org"; TRUE "" "Sophisticated file ownership obstructor.
> Required by KIO.")
+1
_
Friedrich W. H. Kossebau wrote:
> Hi,
>
> following up to my and your emails on "How to test for a D-Bus run-time
> dependency?" last tuesday here please find and then comment on my proposal
> to test for the existance of cagibi during the build, so there is some
> hint at every build of kdebase
Alexander Neundorf wrote:
> On Wednesday 18 August 2010, Yury G. Kudryashov wrote:
>> Hi!
>>
>> Sometimes I see
>>
>> m_o_f_p(Foo) # or find_package(Foo)
>> if (NOT Foo_FOUND)
>> display_feature_log(...)
>> endif (NOT Foo_FOUND)
>>
>>
Hi!
Sometimes I see
m_o_f_p(Foo) # or find_package(Foo)
if (NOT Foo_FOUND)
display_feature_log(...)
endif (NOT Foo_FOUND)
In this case nothing is displayed when dependency is found. It looks strange
to me, and I'd like to remove these if..endif.
I'm new to KDE buildsystem, so I prefer to ask b
Alexander Neundorf wrote:
> On Monday 16 August 2010, Yury G. Kudryashov wrote:
>> Alexander Neundorf wrote:
>> > Hi,
>> >
>> > Once I figured that out I tried to make sure that kdelibs and I think
>> > kdepimlibs stay the only modules which do that.
Sune Vuorela wrote:
> On 2010-08-17, Friedrich W. H. Kossebau wrote:
>> So what could I do to test at build-time the correct version of cagibi,
>> so cagibi can be listed as usual as (missing) dependency?
>
> In general, you should not fail nor 'not install' things based on which
> runtime compo
Alexander Neundorf wrote:
> On Tuesday 17 August 2010, Andreas Pakulat wrote:
>> On 16.08.10 23:54:51, Michael Jansen wrote:
>> > And absolutely and totally makes no sense, breaks stuff and is a
>> > surefire symbol for the upcoming apocalypse judging from most comments
>> > i read today.
>>
>> Th
Alexander Neundorf wrote:
> On Monday 16 August 2010, Yury G. Kudryashov wrote:
>> >
>> > Looks good from a quick view.
>> > Maybe the messages from the macro_log...() should contain "4" (so it's
>> > very clear that this is about KDE4).
&g
Alexander Neundorf wrote:
> On Monday 16 August 2010, Yury G. Kudryashov wrote:
>
> Do you have one at hand ?
In qjson sources.
> Do you actually have commit access ?
No.
>
> Alex
___
Kde-buildsystem mailing list
Kde-buildsy
Alexander Neundorf wrote:
> Hi,
>
> Once I figured that out I tried to make sure that kdelibs and I think
> kdepimlibs stay the only modules which do that.
Some apps that install files to $prefix/share/apps/cmake/modules:
kdebase-workspace-4.5.0
kdegraphics-4.5.0
libktorrent-1.0.2
qjson-0.7.1
kd
Alexander Neundorf wrote:
> On Monday 16 August 2010, Yury G. Kudryashov wrote:
>> Hi!
>>
>> kdenetwork/kget/CMakeLists.txt tries to find libkonq and libkworkspace in
>> the kdelibs install prefix (with NO_DEFAULT_PATH) and doesn't
>> macro_log_feature on
Alexander Neundorf wrote:
> Hi all, Lukas,
>
>> kdenetwork calls macro_optionally_find_package(KTorrent), and it doesn't
>> stop processing if it fails to find FindKTorrent.cmake.
>
> This is clearly wrong.
> http://websvn.kde.org/trunk/KDE/kdenetwork/CMakeLists.txt?view=log
> Uh, so we have tha
nd it haven't told me about this
failure in the feature log.
I propose the following changes (see attached patch):
1. Add macro_log_feature.
2. Use standard FindLibKonq and FindKDE4Workspace instead of find_library.
3. Use macro_optional_find_package.
--
Yury G. Kudryashov--- a/kget/CMakeList
1 - 100 of 128 matches
Mail list logo
