platform/upstream/glib.git
19 years agoInitialize variables to not make gcc nervous.
Matthias Clasen [Tue, 9 Aug 2005 02:45:20 +0000 (02:45 +0000)]
Initialize variables to not make gcc nervous.

2005-08-08  Matthias Clasen  <mclasen@redhat.com>

* glib/gutils.c (_g_compute_locale_variants): Initialize
variables to not make gcc nervous.

19 years agoremove stray -V option to grep so it actually works.
Manish Singh [Tue, 9 Aug 2005 00:25:18 +0000 (00:25 +0000)]
remove stray -V option to grep so it actually works.

2005-08-08  Manish Singh  <yosh@gimp.org>

        * glib/abicheck.sh: remove stray -V option to grep so it actually
        works.

19 years agoEnable the endianness test.
Matthias Clasen [Mon, 8 Aug 2005 19:50:38 +0000 (19:50 +0000)]
Enable the endianness test.

2005-08-08  Matthias Clasen  <mclasen@redhat.com>

* tests/convert-test.c: Enable the endianness test.

* glib/gconvert.c: Make the caching of iconv descriptors
optional.

* configure.in: Add an --enable-iconv-cache option, and
default to disabling iconv caching on new enough glibc.
Somebody with access to Solaris systems will need to test
if opening/closing of iconv descriptors is enough of
a performance problem to warrant the caching on that
platform. Note that the caching is causing correctness
problems in some corner cases, thus turning it off
is desirable unless it has severe performance implications.

19 years agoFix up the testcase
Matthias Clasen [Mon, 8 Aug 2005 17:02:33 +0000 (17:02 +0000)]
Fix up the testcase

19 years agoAdd a test for endianness handling.
Matthias Clasen [Mon, 8 Aug 2005 16:48:23 +0000 (16:48 +0000)]
Add a test for endianness handling.

2005-08-08  Matthias Clasen  <mclasen@redhat.com>

* tests/convert-test.c: Add a test for
endianness handling.

19 years agoAdded Telugu translation done by Vikram Phaneendra <vikramphaneendra@yahoo.co.in...
Sunil Mohan Adapa [Mon, 8 Aug 2005 09:54:32 +0000 (09:54 +0000)]
Added Telugu translation done by Vikram Phaneendra <vikramphaneendra@yahoo.co.in>, Dandu Prasad <dandu_prasad2000@yahoo.com> and Ramana Sai <rmn_sai@yahoo.co.in>

19 years agoAdd convert-test here.
Matthias Clasen [Mon, 8 Aug 2005 05:30:20 +0000 (05:30 +0000)]
Add convert-test here.

2005-08-08  Matthias Clasen  <mclasen@redhat.com>

* tests/Makefile.am: Add convert-test here.

* tests/convert-test.c: Add the beginning of a testsuite
for g_convert() and friends.

19 years agoImprove ABI checks
Matthias Clasen [Sat, 6 Aug 2005 15:34:38 +0000 (15:34 +0000)]
Improve ABI checks

19 years agoImprove ABI checking
Matthias Clasen [Sat, 6 Aug 2005 15:26:49 +0000 (15:26 +0000)]
Improve ABI checking

19 years agoSmall fixes.
Matthias Clasen [Sat, 6 Aug 2005 03:40:07 +0000 (03:40 +0000)]
Small fixes.

2005-08-05  Matthias Clasen  <mclasen@redhat.com>

* gobject/tmpl/gparamspec.sgml:
* glib/tmpl/types.sgml:
* glib/tmpl/option.sgml: Small fixes.

19 years agoremove unused n_threads variable.
Manish Singh [Sat, 6 Aug 2005 00:33:49 +0000 (00:33 +0000)]
remove unused n_threads variable.

2005-08-05  Manish Singh  <yosh@gimp.org>

        * tests/refcount/closures.c: remove unused n_threads variable.

19 years agoShut up CVS
Manish Singh [Sat, 6 Aug 2005 00:32:33 +0000 (00:32 +0000)]
Shut up CVS

19 years agoSmall fixes
Matthias Clasen [Fri, 5 Aug 2005 20:40:52 +0000 (20:40 +0000)]
Small fixes

2005-08-05  Matthias Clasen  <mclasen@redhat.com>

* gobject/tmpl/generic_values.sgml: Small fixes

19 years agoBump version
Matthias Clasen [Fri, 5 Aug 2005 16:10:41 +0000 (16:10 +0000)]
Bump version

19 years ago2.7.7 GLIB_2_7_7
Matthias Clasen [Fri, 5 Aug 2005 16:09:42 +0000 (16:09 +0000)]
2.7.7

19 years agoAlways export g_atomic_int_get and g_atomic_pointer_get as functions, even
Matthias Clasen [Thu, 4 Aug 2005 17:21:03 +0000 (17:21 +0000)]
Always export g_atomic_int_get and g_atomic_pointer_get as functions, even

2005-08-04  Matthias Clasen  <mclasen@redhat.com>

* glib/glib.symbols:
* glib/gatomic.[hc]: Always export g_atomic_int_get and
g_atomic_pointer_get as functions, even if we have macros,
to avoid changing the ABI depending on configuration.

* glib/gatomic.c: Fix the s390 implementations of
g_atomic_pointer_compare_and_exchange.

19 years agoMove the G_BEGIN_DECLS/G_END_DECLS pair outside the #if/#else/#endif
Tor Lillqvist [Thu, 4 Aug 2005 17:07:21 +0000 (17:07 +0000)]
Move the G_BEGIN_DECLS/G_END_DECLS pair outside the #if/#else/#endif

2005-08-04  Tor Lillqvist  <tml@novell.com>

* glib/gstdio.h: Move the G_BEGIN_DECLS/G_END_DECLS pair outside
the #if/#else/#endif block. Otherwise we had G_BEGIN_DECLS without
matching G_END_DECLS on Unix, and G_END_DECLS without matching
G_BEGIN_DECLS on Win32.

19 years agoBump version
Matthias Clasen [Wed, 3 Aug 2005 21:03:17 +0000 (21:03 +0000)]
Bump version

19 years ago2.7.6 GLIB_2_7_6
Matthias Clasen [Wed, 3 Aug 2005 20:58:37 +0000 (20:58 +0000)]
2.7.6

19 years agoSmall doc updates
Ross Burton [Wed, 3 Aug 2005 20:24:18 +0000 (20:24 +0000)]
Small doc updates

19 years agoMark the vint member as volatile, which seems to be necessary to make
Matthias Clasen [Wed, 3 Aug 2005 20:20:47 +0000 (20:20 +0000)]
Mark the vint member as volatile, which seems to be necessary to make

2005-08-02  Matthias Clasen  <mclasen@redhat.com>

        * gclosure.c (union ClosureInt): Mark the vint member as
        volatile, which seems to be necessary to make atomic operations
        work on s390.

19 years agoAdd native atomic operations for s390.
Matthias Clasen [Wed, 3 Aug 2005 20:18:15 +0000 (20:18 +0000)]
Add native atomic operations for s390.

2005-08-03  Matthias Clasen  <mclasen@redhat.com>

        * glib/gatomic.c: Add native atomic operations
        for s390.

        * configure.in: ... and use them on s390.

19 years agoFix Docbook tag typo
Ross Burton [Wed, 3 Aug 2005 19:44:52 +0000 (19:44 +0000)]
Fix Docbook tag typo

19 years agoAdd G_BEGIN_DECLS/G_END_DECLS. (#312424)
Matthias Clasen [Wed, 3 Aug 2005 13:59:43 +0000 (13:59 +0000)]
Add G_BEGIN_DECLS/G_END_DECLS. (#312424)

2005-08-03  Matthias Clasen  <mclasen@redhat.com>

* glib/gstdio.h: Add G_BEGIN_DECLS/G_END_DECLS.  (#312424)

19 years agoName fixes
Matthias Clasen [Tue, 2 Aug 2005 21:30:49 +0000 (21:30 +0000)]
Name fixes

19 years agoDon't go in an infinite loop if the input text ends in an incomplete
Matthias Clasen [Tue, 2 Aug 2005 21:10:16 +0000 (21:10 +0000)]
Don't go in an infinite loop if the input text ends in an incomplete

2005-08-02  Matthias Clasen  <mclasen@redhat.com>

* glib/gconvert.c (g_convert_with_iconv): Don't go in an
infinite loop if the input text ends in an incomplete multibyte
character.  (#312402, Sebastian Bacher)

19 years agoBump version
Matthias Clasen [Tue, 2 Aug 2005 16:22:17 +0000 (16:22 +0000)]
Bump version

19 years ago2.7.5 GLIB_2_7_5
Matthias Clasen [Tue, 2 Aug 2005 16:17:48 +0000 (16:17 +0000)]
2.7.5

19 years agoRemove accidental documentation redundancies
Matthias Clasen [Tue, 2 Aug 2005 15:45:08 +0000 (15:45 +0000)]
Remove accidental documentation redundancies

19 years agoCleanup converter state after the conversion. Document streaming
Matthias Clasen [Tue, 2 Aug 2005 14:58:20 +0000 (14:58 +0000)]
Cleanup converter state after the conversion. Document streaming

2005-08-02  Matthias Clasen  <mclasen@redhat.com>

* glib/gconvert.c (g_convert_with_iconv, g_convert_with_fallback):
Cleanup converter state after the conversion. Document streaming
conversion pitfalls.  (#311337)

19 years agoForgotten log entry
Matthias Clasen [Tue, 2 Aug 2005 13:42:41 +0000 (13:42 +0000)]
Forgotten log entry

19 years agotests/refcount/objects.c tests/refcount/properties.c Use g_usleep()
Tor Lillqvist [Tue, 2 Aug 2005 06:55:38 +0000 (06:55 +0000)]
tests/refcount/objects.c tests/refcount/properties.c Use g_usleep()

2005-08-02  Tor Lillqvist  <tml@novell.com>

* tests/refcount/objects.c
* tests/refcount/properties.c
* tests/refcount/signals.c: Use g_usleep() instead of sleep() for
portability.

19 years agotest high contention on closure reference counts to trigger and catch
Tim Janik [Mon, 1 Aug 2005 21:47:15 +0000 (21:47 +0000)]
test high contention on closure reference counts to trigger and catch

Mon Aug  1 23:33:47 2005  Tim Janik  <timj@imendio.com>

        * tests/refcount/closures.c: test high contention on closure
        reference counts to trigger and catch non-atomic updates.

        * tests/refcount/objects.c:
        * tests/refcount/objects2.c:
        * tests/refcount/properties.c:
        * tests/refcount/properties2.c:
        * tests/refcount/signals.c:
        fixed up test and threading fundamentals. variables accessed from all
        threads need to be volatile. context switches are enforced by using
        g_thread_yield(), not g_usleep(1) which may result in busy waits on
        some platforms. for testcode, always consider all warnings and
        critical messages fatal. issue the currently running program on
        stdout. improved progress indicators.

        * tests/refcount/properties.c:
        * tests/refcount/objects.c:
        don't overdo the number of testing threads to keep the testing machine
        usable, 2 threads can produce as much contention as 20 if executing the
        same code.

        * tests/refcount/signals.c: only start 1 thread per object. GObject
        doesn't provide mutually exclusive object access, but only mutually
        exclusive reference count modification.

        * tests/Makefile.am: added closures test.

19 years agoturned all modifications to the first 32 integer bits in a closure into
Tim Janik [Mon, 1 Aug 2005 21:17:50 +0000 (21:17 +0000)]
turned all modifications to the first 32 integer bits in a closure into

Mon Aug  1 23:00:42 2005  Tim Janik  <timj@imendio.com>

        * gclosure.c: turned all modifications to the first 32 integer bits in a
        closure into atomic accesses. wrapped write accesses into special macros
        to keep the atomic modification logic in a single place. comment cleanups.

        * gclosure.h: made all atomicly accessed closure fields volatile.

        * gobject.h: made ref_count field volatile.

19 years agoMinor simplification in safe_strcmp_filename().
Tor Lillqvist [Mon, 1 Aug 2005 19:25:37 +0000 (19:25 +0000)]
Minor simplification in safe_strcmp_filename().

19 years agoUgh, fix braino.
Tor Lillqvist [Mon, 1 Aug 2005 19:23:10 +0000 (19:23 +0000)]
Ugh, fix braino.

19 years agoMake it pass on Win32. (from_uri_tests[]): Take into consideration that on
Tor Lillqvist [Mon, 1 Aug 2005 19:17:35 +0000 (19:17 +0000)]
Make it pass on Win32. (from_uri_tests[]): Take into consideration that on

2005-08-01  Tor Lillqvist  <tml@novell.com>

* tests/uri-test.c: Make it pass on Win32.
(from_uri_tests[]): Take into consideration that on Win32 we don't
return "localhost" hostnames.
(safe_strcmp_filename): New function that considers slash and
backslash equal on Win32.
(run_roundtrip_tests): Use safe_strcmp_filename().

19 years agomoved G_DATALIST_*_POINTER() macros here, because proper use requires the
Tim Janik [Sun, 31 Jul 2005 00:05:39 +0000 (00:05 +0000)]
moved G_DATALIST_*_POINTER() macros here, because proper use requires the

Sun Jul 31 01:50:20 2005  Tim Janik  <timj@gtk.org>

        * glib/gdataset.c:
        moved G_DATALIST_*_POINTER() macros here, because proper use requires
        the global g_dataset_global mutex to be acquired.
        g_datalist_id_get_data():
        g_datalist_unset_flags():
        g_datalist_set_flags(): properly acquire and release dataset mutex.

        * glib/gdatasetprivate.h: implement G_DATALIST_GET_FLAGS() via atomic
        pointer access to account for memory barriers. moved all other macros.

19 years agouse g_datalist_set_flags() and g_datalist_unset_flags() to allow proper
Tim Janik [Sun, 31 Jul 2005 00:05:22 +0000 (00:05 +0000)]
use g_datalist_set_flags() and g_datalist_unset_flags() to allow proper

Sun Jul 31 02:04:23 2005  Tim Janik  <timj@gtk.org>

        * gobject.c: use g_datalist_set_flags() and g_datalist_unset_flags() to
        allow proper locking around &qdata modifications.

19 years agodocument non-thread-safety for _foreach() functions.
Tim Janik [Sat, 30 Jul 2005 21:41:18 +0000 (21:41 +0000)]
document non-thread-safety for _foreach() functions.

Sat Jul 30 23:40:35 2005  Tim Janik  <timj@gtk.org>

        * glib/tmpl/datalist.sgml:
        * glib/tmpl/datalset.sgml: document non-thread-safety for
        _foreach() functions.

19 years agoreverted notify_mutex introduction, since this prevents parallelized
Tim Janik [Sat, 30 Jul 2005 19:46:11 +0000 (19:46 +0000)]
reverted notify_mutex introduction, since this prevents parallelized

Sat Jul 30 21:10:26 2005  Tim Janik  <timj@gtk.org>

        * gobject.c: reverted notify_mutex introduction, since this prevents
        parallelized setting of object properties on different objects, and
        serves no apparent purpose (to me at least).
        g_object_real_dispose(): removed non-atomic reference count
        modifications.
        g_object_unref(): make sure the closures array is destroyed when
        destroying signal handlers.

        * gparam.c: cosmetic changes.

        * gsignal.c: comment fixup. allow 16bit blocking count.

        * gsignal.c: reverted GClosure related changes.

        * gclosure.c: reverted premature commit of atomic reference
        counting attempt.

19 years agoUpdated Finnish translation
Ilkka Tuohela [Thu, 28 Jul 2005 17:56:34 +0000 (17:56 +0000)]
Updated Finnish translation

19 years agoFixed format specifiers. Thanks to Murray Cumming for pointing this out.
Christian Neumair [Wed, 27 Jul 2005 12:16:37 +0000 (12:16 +0000)]
Fixed format specifiers. Thanks to Murray Cumming for pointing this out.

2005-07-27  Christian Neumair  <chris@gnome-de.org>

* de.po: Fixed format specifiers. Thanks to Murray Cumming for
pointing this out.

19 years agoInclude glibintl.h, not gi18n.h, noticed by Dan Winship.
Matthias Clasen [Tue, 26 Jul 2005 19:41:05 +0000 (19:41 +0000)]
Include glibintl.h, not gi18n.h, noticed by Dan Winship.

2005-07-26  Matthias Clasen  <mclasen@redhat.com>

* glib/goption.c: Include glibintl.h, not gi18n.h, noticed
by Dan Winship.

19 years agoTry -mt as compiler flag for threads, needed for the HP C compiler on
Matthias Clasen [Tue, 26 Jul 2005 19:37:34 +0000 (19:37 +0000)]
Try -mt as compiler flag for threads, needed for the HP C compiler on

2005-07-26  Matthias Clasen  <mclasen@redhat.com>

* configure.in: Try -mt as compiler flag for threads, needed
for the HP C compiler on HP-UX.  (#163051, Paul Cornett)

19 years agoFix a typo in the docs. (#311569, Ross Burton)
Matthias Clasen [Tue, 26 Jul 2005 15:49:15 +0000 (15:49 +0000)]
Fix a typo in the docs. (#311569, Ross Burton)

2005-07-26  Matthias Clasen  <mclasen@redhat.com>

* glib/ghash.c (g_hash_table_foreach): Fix a typo in the
docs.  (#311569, Ross Burton)

19 years agoUpdated German translation.
Hendrik Brandt [Tue, 26 Jul 2005 15:40:53 +0000 (15:40 +0000)]
Updated German translation.

2005-07-26  Hendrik Brandt  <heb@gnome-de.org>

        * de.po: Updated German translation.

19 years agoOne more improvement.
Matthias Clasen [Tue, 26 Jul 2005 15:37:22 +0000 (15:37 +0000)]
One more improvement.

2005-07-26  Matthias Clasen  <mclasen@redhat.com>

* glib/tmpl/arrays.sgml: One more improvement.

19 years agoUpdated Polish translation by GNOME PL Team.
Artur Flinta [Tue, 26 Jul 2005 08:52:31 +0000 (08:52 +0000)]
Updated Polish translation by GNOME PL Team.

2005-07-26  Artur Flinta  <aflinta@cvs.gnome.org>

* pl.po: Updated Polish translation by GNOME PL Team.

19 years agoAdd details
Matthias Clasen [Tue, 26 Jul 2005 02:38:26 +0000 (02:38 +0000)]
Add details

19 years agoCall msgfmt with -c to catch format errors.
Matthias Clasen [Mon, 25 Jul 2005 14:17:30 +0000 (14:17 +0000)]
Call msgfmt with -c to catch format errors.

2005-07-25  Matthias Clasen  <mclasen@redhat.com>

* Makefile.in.in: Call msgfmt with -c to catch
format errors.

19 years agoUpdated Translation
Ankitkumar Rameshchandra Patel [Mon, 25 Jul 2005 08:44:19 +0000 (08:44 +0000)]
Updated Translation

19 years agoUpdated tr.po from Onur Can Cakmak
Baris Cicek [Sun, 24 Jul 2005 20:58:50 +0000 (20:58 +0000)]
Updated tr.po from Onur Can Cakmak

19 years agozh_TW.po Updated Traditional Chinese translation.
Chao-Hsiung Liao [Sat, 23 Jul 2005 05:33:50 +0000 (05:33 +0000)]
zh_TW.po Updated Traditional Chinese translation.

2005-07-22  Chao-Hsiung Liao  <j_h_liau@yahoo.com.tw>

* zh_TW.po Updated Traditional Chinese translation.

19 years agoUpdated Portuguese translation.
Duarte Loreto [Fri, 22 Jul 2005 20:48:28 +0000 (20:48 +0000)]
Updated Portuguese translation.

2005-07-22  Duarte Loreto <happyguy_pt@hotmail.com>

* pt.po: Updated Portuguese translation.

19 years agoBump version
Matthias Clasen [Fri, 22 Jul 2005 17:21:52 +0000 (17:21 +0000)]
Bump version

19 years ago2.7.4 GLIB_2_7_4
Matthias Clasen [Fri, 22 Jul 2005 17:02:04 +0000 (17:02 +0000)]
2.7.4

19 years agoUpdates GTK_2_7_4
Matthias Clasen [Thu, 21 Jul 2005 19:27:00 +0000 (19:27 +0000)]
Updates

19 years agoMerged from gnome-2.10
Nguyen Thai Ngoc Duy [Thu, 21 Jul 2005 11:38:47 +0000 (11:38 +0000)]
Merged from gnome-2.10

19 years agouse G_CALLBACK for signal connections.
Manish Singh [Wed, 20 Jul 2005 20:46:58 +0000 (20:46 +0000)]
use G_CALLBACK for signal connections.

2005-07-20  Manish Singh  <yosh@gimp.org>

        * tests/refcount/signals.c: use G_CALLBACK for signal connections.

19 years agoShut up CVS
Manish Singh [Wed, 20 Jul 2005 20:44:28 +0000 (20:44 +0000)]
Shut up CVS

19 years agoRemove debugging spew
Matthias Clasen [Wed, 20 Jul 2005 17:12:45 +0000 (17:12 +0000)]
Remove debugging spew

19 years agoDon't get stuck in here if immediate is TRUE. (#310954, Hong Jen Yee)
Matthias Clasen [Wed, 20 Jul 2005 17:06:02 +0000 (17:06 +0000)]
Don't get stuck in here if immediate is TRUE. (#310954, Hong Jen Yee)

2005-07-20  Matthias Clasen  <mclasen@redhat.com>

* glib/gthreadpool.c (g_thread_pool_free): Don't get
stuck in here if immediate is TRUE.  (#310954,
Hong Jen Yee)

* tests/threadpool-test.c (main): Test immediate == TRUE.

19 years agoAdd some discouraging comments to the G_INLINE_FUNC macros. (#310836,
Matthias Clasen [Wed, 20 Jul 2005 15:41:19 +0000 (15:41 +0000)]
Add some discouraging comments to the G_INLINE_FUNC macros. (#310836,

2005-07-20  Matthias Clasen  <mclasen@redhat.com>

* glib/tmpl/macros_misc.sgml: Add some discouraging
comments to the G_INLINE_FUNC macros.  (#310836,
Callum McKenzie)

19 years agoMake this an inline function. Define it only if G_CAN_INLINE and not C++.
Tor Lillqvist [Wed, 20 Jul 2005 08:37:13 +0000 (08:37 +0000)]
Make this an inline function. Define it only if G_CAN_INLINE and not C++.

2005-07-20  Tor Lillqvist  <tml@novell.com>

* glib/gutils.h (g_win32_get_system_data_dirs): Make this an
inline function. Define it only if G_CAN_INLINE and not
C++. (#173098)

19 years agoAvoid an unnecessary strlen if len is -1. (#169692, Benoit Dejean)
Matthias Clasen [Tue, 19 Jul 2005 19:05:19 +0000 (19:05 +0000)]
Avoid an unnecessary strlen if len is -1. (#169692, Benoit Dejean)

2005-07-19  Matthias Clasen  <mclasen@redhat.com>

        * glib/gstring.c (g_string_chunk_insert_len): Avoid
        an unnecessary strlen if len is -1.  (#169692,
        Benoit Dejean)

19 years agoAdd docs for #307047, Bryan Silverthorn.
Matthias Clasen [Tue, 19 Jul 2005 18:36:39 +0000 (18:36 +0000)]
Add docs for #307047, Bryan Silverthorn.

2005-07-19  Matthias Clasen  <mclasen@redhat.com>

        * glib/tmpl/byte_order.sgml: Add docs for #307047,
        Bryan Silverthorn.

19 years agoFix g_atomic_pointer_compare_and_exchange on sparc64. (#167572, Gert
Matthias Clasen [Tue, 19 Jul 2005 15:30:23 +0000 (15:30 +0000)]
Fix g_atomic_pointer_compare_and_exchange on sparc64. (#167572, Gert

2005-07-19  Matthias Clasen  <mclasen@redhat.com>

        * glib/gatomic.c (g_atomic_pointer_compare_and_exchange):
        Fix g_atomic_pointer_compare_and_exchange on sparc64.
        (#167572, Gert Doering)

19 years agoFix some printf formatting in the German translation. (#310731, Max Horn)
Matthias Clasen [Mon, 18 Jul 2005 13:27:57 +0000 (13:27 +0000)]
Fix some printf formatting in the German translation. (#310731, Max Horn)

2005-07-18  Matthias Clasen  <mclasen@redhat.com>

* de.po: Fix some printf formatting in the German
translation.  (#310731, Max Horn)

19 years agoUpdated Simplified Chinese translation
Funda Wang [Mon, 18 Jul 2005 05:15:32 +0000 (05:15 +0000)]
Updated Simplified Chinese translation

19 years agoUpdated Lithuanian translation.
Žygimantas Beručka [Sun, 17 Jul 2005 18:29:26 +0000 (18:29 +0000)]
Updated Lithuanian translation.

2005-07-17  Žygimantas Beručka  <zygis@gnome.org>

        * lt.po: Updated Lithuanian translation.

19 years agoUpdated Albanian translation.
Laurent Dhima [Sat, 16 Jul 2005 16:16:41 +0000 (16:16 +0000)]
Updated Albanian translation.

2005-07-16  Laurent Dhima  <laurenti@alblinux.net>

* sq.po: Updated Albanian translation.

19 years agoBump version
Matthias Clasen [Fri, 15 Jul 2005 17:41:28 +0000 (17:41 +0000)]
Bump version

19 years ago2.7.3 GLIB_2_7_3
Matthias Clasen [Fri, 15 Jul 2005 17:28:17 +0000 (17:28 +0000)]
2.7.3

19 years agoUpdates
Matthias Clasen [Fri, 15 Jul 2005 16:52:58 +0000 (16:52 +0000)]
Updates

19 years agoMake refcounting threadsafe by using atomic operations. (#166020, Wim
Matthias Clasen [Fri, 15 Jul 2005 16:51:10 +0000 (16:51 +0000)]
Make refcounting threadsafe by using atomic operations. (#166020, Wim

2005-07-15  Matthias Clasen  <mclasen@redhat.com>

Make refcounting threadsafe by using atomic
operations.  (#166020, Wim Taymans)

* gobject.c: Use a recursive lock to protect the
notify queue.
(g_object_unref): Get rid of g_object_last_unref and
do the last unref handling in g_object_unref.
(g_object_ref, g_object_unref): Use atomic operations.

* gsignal.c (struct _HandlerMatch): Use a full integer
for the ref_count field.
(handler_ref, handler_unref_R): Use atomic operations.

* gparam.c (g_param_spec_ref, g_param_spec_unref):
Use atomic operations instead of a lock to make the
refcounting threadsafe.

* gclosure.c (g_closure_ref, g_closure_unref): Use atomic
operations. This is more complicated here, since the
refcount is stored in a bitfield, so we also have
to access all other bitfield members atomically.

* gsignal.c (handlers_find): Read the meta_marshal flag
of the closure atomically.

* tests/Makefile.am (SUBDIRS): Add tests/refcount

* configure.in: Add tests/refcount

* tests/refcount/properties.c: Test property changes
from multiple threads.

* tests/refcount/signals.c: Test signal emission from
multiple threads.

* tests/refcount/objects.c: Test refcounting from
multiple threads.

* tests/refcount/objects2.c:
* tests/refcount/properties2.c: Tests to measure the
overhead of threadsafe refcounting.

* glib/giochannel.c (g_io_channel_ref, g_io_channel_unref):
Use atomic operations to make refcounting
threadsafe.  (#166020, Wim Taymans)

19 years agoUpdates
Matthias Clasen [Fri, 15 Jul 2005 15:43:59 +0000 (15:43 +0000)]
Updates

19 years agoMake the tests silent on success.
Matthias Clasen [Thu, 14 Jul 2005 05:20:14 +0000 (05:20 +0000)]
Make the tests silent on success.

2005-07-14  Matthias Clasen  <mclasen@redhat.com>

* tests/spawn-test.c:
* tests/uri-test.c:
* tests/thread-test.c:
* tests/queue-test.c:
* tests/mainloop-test.c:
* tests/iochannel-test.c:
* tests/gio-test.c:
* tests/child-test.c: Make the tests silent on success.

19 years agoFix a typo. (#310243, Richard Laager)
Matthias Clasen [Wed, 13 Jul 2005 18:59:46 +0000 (18:59 +0000)]
Fix a typo. (#310243, Richard Laager)

2005-07-13  Matthias Clasen  <mclasen@redhat.com>

* glib/gfileutils.c (g_mkdir_with_parents): Fix a
typo.  (#310243, Richard Laager)

19 years agoAdd G_VA_COPY
Matthias Clasen [Wed, 13 Jul 2005 18:57:18 +0000 (18:57 +0000)]
Add G_VA_COPY

2005-07-13  Matthias Clasen  <mclasen@redhat.com>

* glib/glib-overrides.txt: Add G_VA_COPY

19 years agoWarn if a short name is not acceptable.
Matthias Clasen [Wed, 13 Jul 2005 13:31:30 +0000 (13:31 +0000)]
Warn if a short name is not acceptable.

2005-07-13  Matthias Clasen  <mclasen@redhat.com>

* glib/goption.c (g_option_group_add_entries): Warn if a
short name is not acceptable.

19 years agoUpdate some Update
Kjartan Maraas [Wed, 13 Jul 2005 08:53:42 +0000 (08:53 +0000)]
Update some Update

2005-07-13  Kjartan Maraas  <kmaraas@gnome.org>

* nb.po: Update some
* no.po: Update

19 years agodone
Ivan Stojmirov [Tue, 12 Jul 2005 21:15:03 +0000 (21:15 +0000)]
done

19 years agoAdd and implement a new flag to turn off the automatic <groupname>-
Matthias Clasen [Tue, 12 Jul 2005 18:56:25 +0000 (18:56 +0000)]
Add and implement a new flag to turn off the automatic <groupname>-

2005-07-12  Matthias Clasen  <mclasen@redhat.com>

* glib/goption.h (G_OPTION_FLAG_NOALIAS):
* glib/goption.c: Add and implement a new flag
to turn off the automatic <groupname>- prefixing
for conflict resolution of long option names. (#171840,
Adam McLaurin)

All optional callback arguments  (#308886, Pawel
Sliwowski)

* glib/goption.h (G_OPTION_FLAG_OPTIONAL_ARG):
* glib/goption.c: Add and implement a new flag
to indicate that a callback *optionally* takes another
argument.

* tests/option-test.c: Add tests for optional arguments.

19 years agoDon't lock if depth is zero. (#310148, Wim Taymans)
Matthias Clasen [Tue, 12 Jul 2005 16:56:51 +0000 (16:56 +0000)]
Don't lock if depth is zero. (#310148, Wim Taymans)

2005-07-12  Matthias Clasen  <mclasen@redhat.com>

* glib/gthread.c (g_static_rec_mutex_lock_full): Don't lock
if depth is zero.  (#310148, Wim Taymans)

19 years agoHungarian translation updated.
Gabor Kelemen [Tue, 12 Jul 2005 16:49:38 +0000 (16:49 +0000)]
Hungarian translation updated.

2005-07-12  Gabor Kelemen  <kelemeng@gnome.hu>

* hu.po: Hungarian translation updated.

19 years agoRemove G_GNUC_NULL_TERMINATED from g_object_new again, since that makes
Matthias Clasen [Mon, 11 Jul 2005 17:58:29 +0000 (17:58 +0000)]
Remove G_GNUC_NULL_TERMINATED from g_object_new again, since that makes

2005-07-11  Matthias Clasen  <mclasen@redhat.com>

* gobject.h:
* gobject.symbols: Remove G_GNUC_NULL_TERMINATED from
g_object_new again, since that makes gcc warn on calls of
the form g_object_new (some_type, NULL)

19 years agoUpdated Czech translation.
Miloslav Trmac [Sun, 10 Jul 2005 17:22:37 +0000 (17:22 +0000)]
Updated Czech translation.

2005-07-10  Miloslav Trmac  <mitr@volny.cz>

* cs.po: Updated Czech translation.

19 years agoIgnore anomalous environment entries which are not of the form
Matthias Clasen [Sun, 10 Jul 2005 04:44:27 +0000 (04:44 +0000)]
Ignore anomalous environment entries which are not of the form

2005-07-10  Matthias Clasen  <mclasen@redhat.com>

* glib/gutils.c (g_listenv): Ignore anomalous environment
entries which are not of the form variable=value.  (#309859,
Morten Welinder)

19 years agoUpdated Bulgarian translation by Alexander Shopov <ash@contact.bg>
Alexander Shopov [Sat, 9 Jul 2005 12:48:39 +0000 (12:48 +0000)]
Updated Bulgarian translation by Alexander Shopov <ash@contact.bg>

2005-07-09  Alexander Shopov  <ash@contact.bg>

* bg.po: Updated Bulgarian translation by
Alexander Shopov <ash@contact.bg>

19 years agoTranslation updated by Tino Meinen.
Vincent van Adrighem [Sat, 9 Jul 2005 09:12:52 +0000 (09:12 +0000)]
Translation updated by Tino Meinen.

2005-07-09  Vincent van Adrighem  <adrighem@gnome.org>

* nl.po: Translation updated by Tino Meinen.

19 years agoTypo.
Tor Lillqvist [Sat, 9 Jul 2005 00:12:04 +0000 (00:12 +0000)]
Typo.

19 years agoTotally rewritten socket channel implementation. See discussion in bug
Tor Lillqvist [Fri, 8 Jul 2005 23:27:31 +0000 (23:27 +0000)]
Totally rewritten socket channel implementation. See discussion in bug

2005-07-09  Tor Lillqvist  <tml@novell.com>

* glib/giowin32.c: Totally rewritten socket channel
implementation. See discussion in bug #147392.

19 years agoDon't use autoconf variables for the resource object files on Win32 any
Tor Lillqvist [Fri, 8 Jul 2005 23:04:06 +0000 (23:04 +0000)]
Don't use autoconf variables for the resource object files on Win32 any

2005-07-09  Tor Lillqvist  <tml@novell.com>

* configure.in: Don't use autoconf variables for the resource
object files on Win32 any longer. Instead handle that in the
Makefile.am files. Check for windres.

* glibconfig.h.win32.in: Minor tuning to match the
configure-produced glibconfig.h closely.

* */Makefile.am: Don't use the scripts in build/win32 to
compile the .rc file into a resource object file. (This means
we loose the build number increment magic, but I doubt it was
that useful anyway.) Instead use windres directly. To pass the
normal .o file produced by windres through libtool, which
wants .lo files, pass it directly to the linker using a -Wl
option.

* */*.rc.in: Thus replace BUILDNUMBER with 0.

19 years agoBump version
Matthias Clasen [Fri, 8 Jul 2005 19:25:07 +0000 (19:25 +0000)]
Bump version

19 years ago2.7.2 GLIB_2_7_2
Matthias Clasen [Fri, 8 Jul 2005 15:48:32 +0000 (15:48 +0000)]
2.7.2

19 years agoAdd the "new in 2.8" index.
Matthias Clasen [Fri, 8 Jul 2005 15:04:29 +0000 (15:04 +0000)]
Add the "new in 2.8" index.

2005-07-08  Matthias Clasen  <mclasen@redhat.com>

* gobject/gobject-docs.sgml:
* glib/glib-docs.sgml: Add the "new in 2.8" index.

19 years agoUpdated Serbian translation.
Danilo Šegan [Fri, 8 Jul 2005 08:55:02 +0000 (08:55 +0000)]
Updated Serbian translation.

19 years agoupdate by amanpreetalam@yahoo.com
Amanpreet Singh Alam [Fri, 8 Jul 2005 04:24:01 +0000 (04:24 +0000)]
update by amanpreetalam@yahoo.com

19 years agoForgot one case
Matthias Clasen [Fri, 8 Jul 2005 03:30:24 +0000 (03:30 +0000)]
Forgot one case