| Commit message (Collapse) | Author | Age | Files | Lines |
|\ |
|
| |
| |
| |
| |
| |
| |
| | |
https://bugzilla.gnome.org/show_bug.cgi?id=752812
* glib/glibmm/refptr.h: add warning in a comment, replace
warn_unused_result attribute with corresponding GLib macro.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The latest glibmm requires enough C++11 to be supported so that it can be
built, which is Visual Studio 2013 for the Visual Studio compilers. This
updates the 2010 projects (and their settings) to 2013. The folder in
which the projects reside in need to be updated, and we would preferably
need to make the projects compatible with Visual Studio 2015, from the 2013
projects, which is not too different from the 2010 ones.
|
| |
| |
| |
| |
| | |
This is not actually less efficient - the code is inline anyway.
Bug #752876
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* gio/src/inputstream.hg:
* gio/src/notification.hg:
* gio/src/outputstream.hg:
* gio/src/resolver.hg:
* gio/src/socketconnection.hg: Remove _IGNORE() directives that have become
unnecessary when the .defs files don't contain information from private.h
files. gmmproc warns about unnecessary _IGNORE().
|
| |
| |
| |
| | |
Regenerated without information from private.h files.
|
| |
| |
| |
| |
| |
| |
| |
| | |
* tools/gen_scripts/gio_generate_enums.sh:
* tools/gen_scripts/gio_generate_methods.sh:
* tools/gen_scripts/glib_generate_enums.sh:
* tools/gen_scripts/glib_generate_methods.sh: Don't collect information
from header files with names ending in private.h.
|
| |
| |
| |
| |
| |
| |
| | |
* gio/src/gio.defs:
* gio/src/filelist.am: Remove gio_unix_functions.defs.
* gio/src/gio_unix_functions.defs: Remove the whole file.
All data in this file is also included in gio_methods.defs.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
https://bugzilla.gnome.org/show_bug.cgi?id=753013
* tools/generate_wrap_init.pl.in: add support for nested namespaces in
--namespace parameter, i.e. --namespace=Gst::Bad. It requires to
store all classes from module at least in Gst::Bad namespace, but
allows to use this namespace in wrap_init.h class.
|
| |
| |
| |
| |
| |
| | |
The *with_parent* tests test much the same thing, but these ones
are simpler, based on Marcin Kolny's simpler tests for the
universal reference versions.
|
| |
| |
| |
| |
| | |
test_refptr_universal_reference_move_constructor(): Make even
more sure that its the copy constructor that the compiler uses.
|
| |
| |
| |
| |
| |
| |
| |
| | |
* glib/glibmm/refptr.h: add move constructor and move assignment
operator which allow to set underlying object to castable type.
* tests/glibmm_refptr/main.cc: add tests for move constructor and
move assignment operator containing universal reference as an
argument.
|
| |
| |
| |
| |
| |
| |
| |
| | |
https://bugzilla.gnome.org/show_bug.cgi?id=752812
* glib/glibmm/refptr.h: "release" method gives an access to managed
underlying object, releasing RefPtr's ownership. Method is useful
e.g. in wrappers of functions containing transfer-full arguments.
|
| |
| |
| |
| |
| | |
The && was const, which was a silly typo. Thanks to Marcin Kolny
for noticing.
|
| |
| |
| |
| |
| | |
To test refcounting, though I'm not sure that the test for the
move constructor is quite right.
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Because I recently find that the connection often, but not always,
times out. Maybe Google has changed something about its servers.
I was seeing this output when running the test:
$ ./giomm_tls_client/test
Successfully resolved address of test host 'www.google.com'.
First address of test host is 2a00:1450:4016:802::1011.
Could not connect socket to 2a00:1450:4016:802::1011:443. Exception: Connection timed out
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
* glib/src/glib_enums.defs.patch: G_IO_FLAG_IS_WRITEABLE is an enum constant
since 2013-02-04. No patch needed for that any more.
* gio/src/gio_enums.defs.patch: New file. Such a file has been useful for a
long time. It should have been stored in git long ago.
|
| |
| |
| |
| |
| |
| |
| |
| |
| | |
* tools/gen_scripts/gio_generate_enums.sh:
* tools/gen_scripts/gio_generate_extra_defs.sh:
* tools/gen_scripts/glib_generate_enums.sh:
* tools/gen_scripts/glib_generate_methods.sh: These scripts patch the
generated .defs files. Update them like gtkmm's gtk_generate_extra_defs.sh,
i.e. more comments and an option to regenerate the patch file.
|
| | |
|
| |
| |
| |
| |
| |
| |
| | |
This reverts commit 05610cec2ccbc54f20fcc3e995e41649f21c5714.
See the commit in the previous commit about gtkmm-2.24. This is a
temporary act of kindness to Inkscape.
|
| |
| |
| |
| |
| |
| |
| |
| | |
To avoid breaking the gtkmm-2.24 build with --enable-warnings=fatal,
and the build of apps that do this too.
However, those apps (Inkscape) need to stop using that ancient
deprecated version of gtkmm.
Bug #752797
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| | |
* tools/pm/Output.pm:
* tools/pm/DocsParser.pm: When a C++ parameter name differs from the
corresponding C parameter name, change the name in the documentation.
The translation of argument names works for _WRAP_METHOD() and _WRAP_SIGNAL().
Bug #752469
|
| |
| |
| |
| |
| |
| | |
I named the previous version 2.45.31 instead of 2.45.3.1, so we have to
use a number bigger than 31. Let's say that 40 is like 4 and we'll
use 50 for the next one instead of 5. Sorry.
|
| | |
|
| |
| |
| |
| |
| | |
For some reason, gmmproc didn't warn about this unwrapped signal.
Bug #752600 (Ben)
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| | |
This seems to be correct and allowed. It makes the code clearer.
|
| | |
|
| |
| |
| |
| |
| | |
I don't think these have been used since before gtkmm 3.0.
They were already hidden from the documentation.
|
| |
| |
| |
| | |
Though we cannot hide them via GIOMM_DISABLE_DEPRECATED.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Because this generally invites programming errors, though I am
less concerned about shadowing of method names by parameter or
variable names, which requires some tedious parameter renaming.
In MatchInfo::set_gobject() the confusion between take_ownership and
this.take_ownership does seem to have caused a programming error,
hopefully now corrected.
|
| |
| |
| |
| |
| | |
Although -Wsuggest-override isn't actually available yet in my version
of g++ (4.9.2).
|