Update tqt-mt.pc to export the same parameters previously exported in tqt.pc
#200
Merged
MicheleC
merged 1 commits from drop/tqtinterface
into master
4 months ago
Reviewers
Request review
No reviewers
TDE/Core
Labels
Clear labels
General - need additional info from contributor
Pull request - do not delete branch after merging
Pull request - need fixing
Pull request - request for comments
Pull request - update to translation files needed
Pull request - work in progress
Related to R14.0.x series
Related to R14.1.x series
Severity level - critical
Severity level - major
Severity level - minor
Severity level - normal
Severity level - regression from previous version
Severity level - trivial
Severity level - wishlist request
Status - duplicate of another issue
Status - invalid report
Status - not our problem
Status - rejected
Status - won't fix
Status - works for me, unable to reproduce
Apply labels
GE/need-info
General - need additional info from contributor
PR/keep-branch
Pull request - do not delete branch after merging
PR/not-ok
Pull request - need fixing
PR/rfc
Pull request - request for comments
PR/update-trans
Pull request - update to translation files needed
PR/wip
Pull request - work in progress
RS/R14.0.x
Related to R14.0.x series
RS/R14.1.x
Related to R14.1.x series
SL/critical
Severity level - critical
SL/major
Severity level - major
SL/minor
Severity level - minor
SL/normal
Severity level - normal
SL/regression
Severity level - regression from previous version
SL/trivial
Severity level - trivial
SL/wishlist
Severity level - wishlist request
ST/duplicate
Status - duplicate of another issue
ST/invalid
Status - invalid report
ST/notourproblem
Status - not our problem
ST/rejected
Status - rejected
ST/wontfix
Status - won't fix
ST/worksforme
Status - works for me, unable to reproduce
No Label
GE/need-info
PR/keep-branch
PR/not-ok
PR/rfc
PR/update-trans
PR/wip
RS/R14.0.x
RS/R14.1.x
SL/critical
SL/major
SL/minor
SL/normal
SL/regression
SL/trivial
SL/wishlist
ST/duplicate
ST/invalid
ST/notourproblem
ST/rejected
ST/wontfix
ST/worksforme
Milestone
Set milestone
Clear milestone
No items
No Milestone
Assignees
Assign users
Clear assignees
No Assignees
3 Participants
Notifications
Due Date
The due date is invalid or out of range. Please use the format 'yyyy-mm-dd'.
No due date set.
Blocks
#383 DEB: drop tqtinterface module
TDE/tde-packaging
#37 Don't explicetly include tqt.h anymore
TDE/tde-common-admin
#120 Detects tqt-mt instead of tqt
TDE/tde-cmake
#74 Drop tqtinterface. Leave a short README for info only
TDE/tqtinterface
#7 Don't explicetly include tqt.h anymore
TDE/dbus-tqt
#28 Don't explicetly include tqt.h anymore
TDE/sip4-tqt
#315 Don't explicetly include tqt.h anymore
TDE/tdelibs
#545 Don't explicetly include tqt.h anymore
TDE/tdebase
#51 Don't look for libtqui anymore.
TDE/tdesdk
#56 Don't explicetly include tqt.h anymore
TDE/tdevelop
#70 Don't explicetly include tqt.h anymore
TDE/amarok
#81 Use tqt-mt instead of tqt library
TDE/koffice
#25 Don't explicetly include tqt.h anymore
TDE/rosegarden
#35 DEB compiz-trinity: adjust after dropping tqtinterface
TDE/extra-dependencies
Reference: TDE/tqt3#200
Reference in New Issue
There is no content yet.
Delete Branch 'drop/tqtinterface'
Deleting a branch is permanent. It CANNOT be undone. Continue?
No
Yes
Required to drop tqtinterface.
805696c12e
to48d6b0d862
4 months ago@ -1610,3 +1610,1 @@
t << "Libs: -L${libdir} -l" << lname.left(lname.length()-Option::libtool_ext.length()) << " ";
for(TQStringList::ConstIterator it = libs.begin(); it != libs.end(); ++it)
t << project->variables()[(*it)].join(" ") << " ";
t << "Libs: -L${libdir} -l" << lname.left(lname.length()-Option::libtool_ext.length()) << " -ltqui -L/usr/lib ";
Why hard-coded "-L/usr/lib" here?
Good point, I have updated the PR
Here is the second thing to solve: Why is there forced linking of
tqui
library? The forced linking of this library was not part oftqt.pc
, because it is a separate library that is definitely not needed for everyone. After all, that is why it is a separate library. This does not seem to be right to be forced to linked now.There are a few reason behing this choice.
tqtinterface
was providing two different.pc
files (tqt.pc
andtqtqui.pc
) but only one shared objectlibtqt.so
. This was already a design I didn't like.libtqt-mt.so
andlibtqui.so
are both provided bytqt3
in the same package, so they will always be present together in the systemwhile it is true that
tqui
is not always needed, providing it to the linker won't do much harm: if it is not needed, the linker will simply do nothing with it and the resulting executables/so files won't be any biggerGiven the above points, I saw an opportunity to simplify the detection code in cmake (no more need for detecting and linking
tqui
on its own) and that is why I implemented this way.If we prefer to keep detection and linkage of
tqui
separate from that oftqt-mt
, I will rework the code, but IMO it adds extra code for not real benefit.As per discussion on chat, we will keep single detection of tqt while linking
tqui
manually where needed. PR has been updated accordingly.48d6b0d862
to85a5377226
4 months ago85a5377226
to5b5cf7a81f
4 months ago@ -1609,4 +1609,1 @@
libs << "QMAKE_LFLAGS_THREAD"; //not sure about this one, but what about things like -pthread?
t << "Libs: -L${libdir} -l" << lname.left(lname.length()-Option::libtool_ext.length()) << " ";
for(TQStringList::ConstIterator it = libs.begin(); it != libs.end(); ++it)
t << project->variables()[(*it)].join(" ") << " ";
Why other
libs
libraries are ignored? just redundant?They would cause FTBFS, for example
-lgthread-2.0
.libtqt.pc
provided bytqtinterface
was exporting the same libraries that this PR is now exporting, so in terms of linkage for TDE applications, nothing has changed. Once you build a program using TQt, you need to link againsttqt-mt
. The extra libraries should probably used when buildingtqt
(as needed) but there should be no need to expose them to all the programs using TQt. Perhaps a left over code from past.There are some comments.
@ -1623,0 +1619,4 @@
// << varGlue("PRL_EXPORT_DEFINES","-D"," -D"," ")
// << project->variables()["PRL_EXPORT_CXXFLAGS"].join(" ")
// << varGlue("DEFINES","-D"," -D"," ")
<< " -DTQT_NO_ASCII_CAST -DTQT_NO_STL -DTQT_NO_COMPAT -DTQT_NO_TRANSLATION -DTQT_THREAD_SUPPORT -D_REENTRANT -I${includedir}";
I have tested that it does not have an impact on ABI, but it is strange that there are excluded definitions determined by configuration – like
QT_SHARED
,TQT_NO_DEBUG
. And I hesitate whether it is right or not.BTW, one space is at the end of
"Cflags: "
, second at the beginning" -DTQT_NO_ASCII_CAST...
, so there is a doubled space.For
QT_SHARED
,TQT_NO_DEBUG
, I looked into build logs and tried to mirror the exact variables as before. I had initially included these variables, then troubleshooting a FTBFS on a package I realized they were not used in the old build, so I removed them.As for the double space, I will fix that.
@ -9,0 +61,4 @@
#include <tqdom.h>
#include "tqobject.h"
#include <tqdrawutil.h>
#include "tqbrush.h"
Is there any sense that the use of includes with quotes and sharp parentheses is mixed? I suppose it makes sense to use only sharp parentheses.
@ -14,0 +274,4 @@
#if defined( QT_MOC_CPP ) || defined( QT_H_CPP ) || defined( Q_OS_MACX )
#include <private/tqcom_p.h>
#include <private/tqucom_p.h>
#include "private/tqcom_p.h"
Two lines higher is the same include in a variant with sharp parentheses.
@ -14,0 +293,4 @@
#include "private/tqtextcodecinterface_p.h"
#include "private/tqpsprinter_p.h"
#include "private/tqtitlebar_p.h"
#include "private/tqucom_p.h"
A few dozen lines higher is the same include in a variant with sharp parentheses.
5b5cf7a81f
tob6f43f9a98
4 months agoOther that the point about defined variables, the rest has been updated.
All seems good.
b6f43f9a98
toc489c62c17
4 months agoc489c62c17
into master 4 months agoReviewers
c489c62c17
.