platform/upstream/glib.git
23 years agoFix GNU sedism (#55430)
Owen Taylor [Tue, 29 May 2001 22:19:46 +0000 (22:19 +0000)]
Fix GNU sedism (#55430)

Tue May 29 18:17:11 2001  Owen Taylor  <otaylor@redhat.com>

* autogen.sh (have_libtool): Fix GNU sedism (#55430)

23 years agoAdd test case for the g_qsort_with_data func. It works. This fixes bug
Sebastian Wilhelmi [Tue, 29 May 2001 14:55:39 +0000 (14:55 +0000)]
Add test case for the g_qsort_with_data func. It works. This fixes bug

2001-05-29  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* tests/Makefile.am, tests/qsort-test.c: Add test case for the
g_qsort_with_data func. It works. This fixes bug #52605.

23 years agoNow that we check for a C++ compiler in configure.in anyway, I added a
Sebastian Wilhelmi [Tue, 29 May 2001 14:17:23 +0000 (14:17 +0000)]
Now that we check for a C++ compiler in configure.in anyway, I added a

2001-05-29  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* tests/Makefile.am, tests/cxx-test.C: Now that we check for a C++
compiler in configure.in anyway, I added a test, that the GLib
headers can be inclued into a C++ program. This fixes bug #52605.

23 years agoDo not g_strdup, as promised in the function name.
Sebastian Wilhelmi [Tue, 29 May 2001 13:48:18 +0000 (13:48 +0000)]
Do not g_strdup, as promised in the function name.

2001-05-29  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gmodule.c (g_module_set_error_unduped): Do not g_strdup, as
promised in the function name.

23 years agoDon't bail out, if no C++ compiler is found. Define G_HAVE_ISO_VARARGS
Sebastian Wilhelmi [Tue, 29 May 2001 13:47:05 +0000 (13:47 +0000)]
Don't bail out, if no C++ compiler is found. Define G_HAVE_ISO_VARARGS

2001-05-29  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* configure.in: Don't bail out, if no C++ compiler is
found. Define G_HAVE_ISO_VARARGS only if appropriate.

* gmessages.h: Thus we can revert the previous patch.

23 years ago"#ifdef G_HAVE_ISO_VARARGS" will always succeed even when we don't have
James Henstridge [Tue, 29 May 2001 11:12:42 +0000 (11:12 +0000)]
"#ifdef G_HAVE_ISO_VARARGS" will always succeed even when we don't have

2001-05-29  James Henstridge  <james@daa.com.au>

        * gmessages.h: "#ifdef G_HAVE_ISO_VARARGS" will always succeed
        even when we don't have ISO vararg support.  Changed to
        "#if G_HAVE_ISO_VARARGS".

23 years agoerk, strconcat needs NULL as last arg. make sure we pass module name with
Tim Janik [Tue, 29 May 2001 11:05:53 +0000 (11:05 +0000)]
erk, strconcat needs NULL as last arg. make sure we pass module name with

Tue May 29 12:57:51 2001  Tim Janik  <timj@gtk.org>

        * gmodule.c (g_module_symbol): erk, strconcat needs NULL as last arg.
        make sure we pass module name with suffix into backend open.

23 years agoas a last resort, if we can't access() the module, leave it up to the
Tim Janik [Mon, 28 May 2001 18:23:31 +0000 (18:23 +0000)]
as a last resort, if we can't access() the module, leave it up to the

Mon May 28 20:19:25 2001  Tim Janik  <timj@gtk.org>

        * gmodule.c (g_module_open): as a last resort, if we can't access()
        the module, leave it up to the platform backends to find it.
        (fixes implicit search patch loading i broke with my last
        commit, spotted by Padraig O'Briain).

23 years agoCheck for __VARARGS__ separately for C and C++.
Owen Taylor [Mon, 28 May 2001 15:41:24 +0000 (15:41 +0000)]
Check for __VARARGS__ separately for C and C++.

Mon May 28 11:40:34 2001  Owen Taylor  <otaylor@redhat.com>

* configure.in: Check for __VARARGS__ separately for C and C++.

23 years agoFix typo.
Sebastian Wilhelmi [Mon, 28 May 2001 11:16:33 +0000 (11:16 +0000)]
Fix typo.

2001-05-28  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* configure.in: Fix typo.

* gobject/Makefile.am (EXTRA_DIST): Really distribute
        glib-mkenums, now that bin_SCRIPTS doesn't seem to care about
        doing so anymore.

23 years agofixed bad faux-pass, forgot to release signal lock when restarting
Tim Janik [Mon, 28 May 2001 04:57:07 +0000 (04:57 +0000)]
fixed bad faux-pass, forgot to release signal lock when restarting

Mon May 28 06:51:24 2001  Tim Janik  <timj@gtk.org>

        * gsignal.c (signal_emit_unlocked_R): fixed bad faux-pass, forgot
        to release signal lock when restarting emissions.

23 years agounlock context when bailing out with a warning. (g_main_context_check):
Tim Janik [Sun, 27 May 2001 18:28:58 +0000 (18:28 +0000)]
unlock context when bailing out with a warning. (g_main_context_check):

Sun May 27 05:09:18 2001  Tim Janik  <timj@gtk.org>

        * gmain.c (g_main_context_prepare): unlock context when bailing
        out with a warning.
        (g_main_context_check): same here.

        * gmain.c (g_main_context_check): before returning due to
        changed pollfds, unlock context.

Sun May 27 04:52:28 2001  Tim Janik  <timj@gtk.org>

        * gsignal.[hc] (g_signal_stop_emission_by_name): added variant
        to stop signal emissions through a detailed_signal string.

        * gsignal.c (signal_emit_R) (g_signal_emit_valist): account for
        the fact that g_value_* functions may cause signal emissons by
        unlocking the global signal system lock around g_value_* functions.
        (signal_emit_unlocked_R): renamed this from signal_emit_R() to reflect
        that this functions acquires the lock on its own now.

23 years agoback out change by mitch@convergence.de and apply patch in bug id #52067
CEST 2001 Paolo Molaro [Thu, 24 May 2001 19:30:40 +0000 (19:30 +0000)]
back out change by mitch@convergence.de and apply patch in bug id #52067

Thu May 24 21:24:16 CEST 2001 Paolo Molaro <lupus@ximian.com>

* gmarkup.c: back out change by mitch@convergence.de and apply patch in
bug id #52067 that fixes the same problem in a more complete manner.
This fixes also a segfault for a malformed XML file and adds a new
test case.

23 years agoreflect renaming of g_string_sprintfa to g_string_printfa
Hans Breuer [Thu, 24 May 2001 16:38:24 +0000 (16:38 +0000)]
reflect renaming of g_string_sprintfa to g_string_printfa

2001-05-24  Hans Breuer  <hans@breuer.org>

* grel.c : reflect renaming of g_string_sprintfa to g_string_printfa

* makefile.msc.in : create an additional static lib to be used from
(at least) glib-genmarshal

23 years agopthread isn't required anymore
Hans Breuer [Thu, 24 May 2001 16:36:47 +0000 (16:36 +0000)]
pthread isn't required anymore

2001-05-24  Hans Breuer  <hans@breuer.org>

* makefile.msc.in : pthread isn't required anymore

23 years agochanged depndencies to build glib-genmarshal first and statically linked
Hans Breuer [Thu, 24 May 2001 16:36:02 +0000 (16:36 +0000)]
changed depndencies to build glib-genmarshal first and statically linked

2001-05-24  Hans Breuer  <hans@breuer.org>

* makefile.msc.in : changed depndencies to build glib-genmarshal
first and statically linked with glib, which makes it independent
from the installed glib version. Added new object files to build.

* gobject.def : updated

23 years agogpointer, not gpointer *.
Owen Taylor [Thu, 24 May 2001 12:52:39 +0000 (12:52 +0000)]
gpointer, not gpointer *.

Thu May 24 08:52:02 2001  Owen Taylor  <otaylor@redhat.com>

* gobject.[ch] (g_value_set_object): gpointer, not
        gpointer *.

23 years agoTiny speed improvement suggested by noon@users.sourceforge.net.
Sebastian Wilhelmi [Wed, 23 May 2001 10:20:56 +0000 (10:20 +0000)]
Tiny speed improvement suggested by noon@users.sourceforge.net.

2001-05-23  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* garray.c (g_array_remove_index_fast): Tiny speed improvement
suggested by noon@users.sourceforge.net.

* glib/tmpl/arrays.sgml, glib/tmpl/arrays_pointers.sgml,
glib/tmpl/arrays_byte.sgml: Corrected documentation for the
..._sized_new functions. Discovered by noon@users.sourceforge.net.

23 years agomade 'make distcheck' happy.
Sebastian Wilhelmi [Tue, 22 May 2001 14:31:26 +0000 (14:31 +0000)]
made 'make distcheck' happy.

2001-05-22  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

        * Makefile.am (bin_SCRIPTS): made 'make distcheck' happy.

23 years agoAlso return TRUE for late arrived signals. Thanks to Steven Brooks
Sebastian Wilhelmi [Tue, 22 May 2001 14:20:58 +0000 (14:20 +0000)]
Also return TRUE for late arrived signals. Thanks to Steven Brooks

2001-05-22  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gthread-win32.c (g_cond_wait_internal): Also return TRUE for
late arrived signals. Thanks to Steven Brooks
<umbrook0@cs.umanitoba.ca> for pointing out.

23 years agoRemoved POSIX_*. Defined G_THREAD_SOURCE to "gthread-win32.c".
Sebastian Wilhelmi [Tue, 22 May 2001 12:28:06 +0000 (12:28 +0000)]
Removed POSIX_*. Defined G_THREAD_SOURCE to "gthread-win32.c".

2001-05-22  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

        * config.h.win32.in: Removed POSIX_*. Defined G_THREAD_SOURCE to
        "gthread-win32.c".

        * glibconfig.h.win32.in: Define G_HAVE_ISO_VARARGS for gcc, don't
        know about MSC. Define G_THREADS_IMPL_WIN32 instead of
        G_THREADS_IMPL_POSIX and define the right static mutex macros and
        types.

        * build/win32/make.mingw (CXX): Removed PTHREAD defs. Added
-O2 -Wall to compile flags.

        * gthread/gthread-impl.c (g_thread_init): Move the thread
implementation initialization to before assigning
   GThreadFuncs, which now is just struct assigned and not
memcpy'ed. Completed check for zero members of GThreadFuncs.

        * gthread/makefile.mingw: Don't link to pthread anymore.

        * gthread/gthread-win32.c: New file for native thread support for
        win32. Thanks to Hans Breuer <hans@breuer.org> to got me
        kickstarted.

        * gthread/Makefile.am: Also distribute gthread-win32.c.

23 years agoRemoved g_param_spec_stringc.
Sebastian Wilhelmi [Tue, 22 May 2001 12:12:06 +0000 (12:12 +0000)]
Removed g_param_spec_stringc.

2001-05-22  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gobject.def: Removed g_param_spec_stringc.

23 years agog_thread_create renamed to g_thread_create_full.
Sebastian Wilhelmi [Tue, 22 May 2001 12:06:59 +0000 (12:06 +0000)]
g_thread_create renamed to g_thread_create_full.

2001-05-22  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* glib.def: g_thread_create renamed to g_thread_create_full.

* gthread.c: memcpy is not necessary here. We can simply use
struct assignment.

* gmessages.c: Fix compilation error on win32. Added
GFileDescriptor, wich is gint on Unix and FILE* on win32.

23 years agono message
Andy Lanoix [Tue, 22 May 2001 01:11:28 +0000 (01:11 +0000)]
no message

23 years ago*giowin32.c: G_IO_WIN32_WINDOWS_MESSAGES channels
Andy Lanoix [Tue, 22 May 2001 01:10:19 +0000 (01:10 +0000)]
*giowin32.c: G_IO_WIN32_WINDOWS_MESSAGES channels
not handled correctly in g_io_win32_check()

23 years agomake g_value_set_object() take a gpointer v_object so people don't need to
Tim Janik [Tue, 22 May 2001 00:48:06 +0000 (00:48 +0000)]
make g_value_set_object() take a gpointer v_object so people don't need to

Tue May 22 02:46:13 2001  Tim Janik  <timj@gtk.org>

        * gobject.c (g_value_set_object): make g_value_set_object()
        take a gpointer v_object so people don't need to G_OBJECT()
        cast NULL pointers (C code convenience).

23 years agoFix some problems in writing out varargs tests in configure.in
Owen Taylor [Sun, 20 May 2001 14:49:44 +0000 (14:49 +0000)]
Fix some problems in writing out varargs tests in configure.in

Sun May 20 10:47:47 2001  Owen Taylor  <otaylor@redhat.com>

        * configure.in: Fix some problems in writing out varargs
tests in configure.in

23 years agoadd glibintl.h
Havoc Pennington [Sat, 19 May 2001 05:32:50 +0000 (05:32 +0000)]
add glibintl.h

2001-05-19  Havoc Pennington  <hp@pobox.com>

* glib/Makefile.am (IGNORE_HFILES): add glibintl.h

* glib/tmpl/*.sgml: fix various missing docs

23 years agoUse compile checks for ISO C99 and GNU extension varargs macros syntax and
Owen Taylor [Fri, 18 May 2001 16:00:54 +0000 (16:00 +0000)]
Use compile checks for ISO C99 and GNU extension varargs macros syntax and

Fri May 18 10:52:23 2001  Owen Taylor  <otaylor@redhat.com>

* configure.in gmessages.h: Use compile checks for ISO C99 and GNU
extension varargs macros syntax and store result in glibconfig.h
(G_HAVE_ISO_VARARGS, G_HAVE_GNUC_VARARGS) rather than relying
on checking predefined macros.

23 years agodon't g_strdup()/g_free() all parsed attributes and their values twice but
Michael Natterer [Fri, 18 May 2001 12:38:32 +0000 (12:38 +0000)]
don't g_strdup()/g_free() all parsed attributes and their values twice but

2001-05-18  Michael Natterer  <mitch@convergence.de>

* gmarkup.c: don't g_strdup()/g_free() all parsed attributes and
their values twice but simply copy the string pointers from the
GMarkupAttribute struct to the string arrays before passing them
to start_element().

23 years agoUpdated.
Sebastian Wilhelmi [Fri, 18 May 2001 08:49:29 +0000 (08:49 +0000)]
Updated.

2001-05-18  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* glib/glib-overrides.txt, glib/glib-sections.txt,
glib/tmpl/thread_pools.sgml, glib/tmpl/threads.sgml: Updated.

23 years agoRenamed g_thread_create to g_thread_create_full and added macro
Sebastian Wilhelmi [Fri, 18 May 2001 08:44:57 +0000 (08:44 +0000)]
Renamed g_thread_create to g_thread_create_full and added macro

2001-05-18  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gthread.c, gthread.h: Renamed g_thread_create to
g_thread_create_full and added macro g_thread_create, which omits
'stack_size', 'bound' and 'priority' parameters. Also removed
'bound' from GThread struct.

* gthreadpool.h, gthreadpool.c: Adapted GThreadPool to the above
changes. GThreadPool lost the 'priority' and 'bound'
members. g_thread_pool_new the 'stack_size', 'bound' and
'priority' parameters.

* tests/mainloop-test.c, tests/thread-test.c,
tests/threadpool-test.c: Adapted to the above changes.

23 years agoFixed mutex deadlock.
Sebastian Wilhelmi [Fri, 18 May 2001 08:36:44 +0000 (08:36 +0000)]
Fixed mutex deadlock.

2001-05-18  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gmem.c (g_mem_profile): Fixed mutex deadlock.

23 years agoAdd gpattern.o to objects.
Sebastian Wilhelmi [Thu, 17 May 2001 15:15:00 +0000 (15:15 +0000)]
Add gpattern.o to objects.

2001-05-17  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* makefile.mingw.in: Add gpattern.o to objects.

* glib.def: Export g_thread_exit as well.

* Makefile.am: Fix BUILT_EXTRA_DIST exporting.

23 years agoAdded missing .cvsignore file.
Michael Natterer [Thu, 17 May 2001 12:19:09 +0000 (12:19 +0000)]
Added missing .cvsignore file.

23 years agoAdd inter-library dependencies.
Owen Taylor [Wed, 16 May 2001 18:35:39 +0000 (18:35 +0000)]
Add inter-library dependencies.

Sun May 13 10:31:17 2001  Owen Taylor  <otaylor@redhat.com>

* glib/Makefile.am gobject/Makefile.am gmodule/Makefile.am:
Add inter-library dependencies.

* acinclude.m4: Remove libtool macros.

* autogen.sh: Require libtool-1.4, automake-1.4p1.

23 years agom4macros/Makefile
Havoc Pennington [Tue, 15 May 2001 22:07:21 +0000 (22:07 +0000)]
m4macros/Makefile

2001-05-15  Havoc Pennington  <hp@redhat.com>

* configure.in (AC_OUTPUT): m4macros/Makefile

* m4macros/Makefile.am, m4macros/glib-2.0.m4,
m4macros/glib-gettext.m4:
m4 files moved here on server, Makefile.am added

This is so you can aclocal -I m4macros while avoiding acinclude.m4

* Makefile.am: add m4macros subdir, remove references to glib-2.0.m4

23 years agofix docs; they said we validated the UTF-8, but we can't possibly detect
Havoc Pennington [Tue, 15 May 2001 19:18:08 +0000 (19:18 +0000)]
fix docs; they said we validated the UTF-8, but we can't possibly detect

2001-05-14  Havoc Pennington  <hp@redhat.com>

* gutf8.c (g_utf8_get_char): fix docs; they said we validated
the UTF-8, but we can't possibly detect partial chars since
there's no length arg here, so trying to use this function
on invalid UTF-8 is a bad idea.

23 years agoUpdated German translations for Glib and GTK+ HEAD. Phew.
Matthias Warkus [Tue, 15 May 2001 17:47:49 +0000 (17:47 +0000)]
Updated German translations for Glib and GTK+ HEAD. Phew.

23 years agoAvoid using stderr to be as robust as possible in out-of-memory.
Owen Taylor [Mon, 14 May 2001 14:53:59 +0000 (14:53 +0000)]
Avoid using stderr to be as robust as possible in out-of-memory.

Thu May 10 23:21:30 2001  Owen Taylor  <otaylor@redhat.com>

* gmessages.c (g_log_write_prefix): Avoid using stderr
to be as robust as possible in out-of-memory.

* gmessages.c (g_log_default_handler): Remove some dead
code.

* gutils.c (g_parse_debug_string): Fix to avoid mallocs.

23 years agonew function to check whether a handler is still connected to an instance
Tim Janik [Mon, 14 May 2001 13:55:51 +0000 (13:55 +0000)]
new function to check whether a handler is still connected to an instance

Mon May 14 01:49:01 2001  Tim Janik  <timj@gtk.org>

        * gsignal.[hc] (g_signal_handler_is_connected): new function
        to check whether a handler is still connected to an instance
        (by handler id).

23 years agomake these safe against removal of the current element.
Tim Janik [Fri, 11 May 2001 18:05:16 +0000 (18:05 +0000)]
make these safe against removal of the current element.

Fri May 11 18:25:23 2001  Tim Janik  <timj@gtk.org>

        * gdataset.c:
        (g_dataset_foreach):
        (g_datalist_foreach): make these safe against removal of the
        current element.

23 years agoUpdated Ukrainian transaltion
Yuri Syrota [Fri, 11 May 2001 12:09:21 +0000 (12:09 +0000)]
Updated Ukrainian transaltion

23 years agoUpdated Ukrainian transaltion
Yuri Syrota [Fri, 11 May 2001 12:02:34 +0000 (12:02 +0000)]
Updated Ukrainian transaltion

23 years agochange env variable to G_MESSAGES_PREFIXED, suggested by Tim
Havoc Pennington [Fri, 11 May 2001 03:07:53 +0000 (03:07 +0000)]
change env variable to G_MESSAGES_PREFIXED, suggested by Tim

2001-05-10  Havoc Pennington  <hp@pobox.com>

* gmessages.c (g_log_write_prefix): change env variable to
G_MESSAGES_PREFIXED, suggested by Tim

23 years agouse g_str_hash() algorithm to generate hashes.
Tim Janik [Thu, 10 May 2001 13:58:40 +0000 (13:58 +0000)]
use g_str_hash() algorithm to generate hashes.

Thu May 10 15:19:01 2001  Tim Janik  <timj@gtk.org>

        * gscanner.c (g_scanner_key_hash): use g_str_hash() algorithm
        to generate hashes.

        * gmem.c (standard_calloc): free() doesn't return a value
        (Mark Murnane).

Thu May 10 14:00:48 2001  Tim Janik  <timj@gtk.org>

        * gparamspecs.[hc]: removed g_param_spec_stringc() in lack of
        evidence of public need.

        * gsignal.h: added g_signal_disconnect_by_func(),
        g_signal_block_by_func() and g_signal_unblock_by_func() convenience
        macros as per owen's request.

        * gtype.c (SIZEOF_FUNDAMENTAL_INFO): align sizeof (GTypeFundamentalInfo)
        to size of longs and pointers.

23 years agoGet rid of --enable-msg-prefix
Havoc Pennington [Wed, 9 May 2001 16:36:17 +0000 (16:36 +0000)]
Get rid of --enable-msg-prefix

2001-04-26  Havoc Pennington  <hp@redhat.com>

* configure.in: Get rid of --enable-msg-prefix

* gmessages.c: make whether to prefix the messages with
appname/pid a runtime setting, not a compile-time setting.  Change
default to include prefix for debug/warning/error type messages.

23 years agoOps, forgot this.
Sebastian Wilhelmi [Wed, 9 May 2001 12:54:37 +0000 (12:54 +0000)]
Ops, forgot this.

Ops, forgot this.

23 years agoUpdates after some renaming took place.
Sebastian Wilhelmi [Wed, 9 May 2001 12:53:42 +0000 (12:53 +0000)]
Updates after some renaming took place.

2001-05-09  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* glib/tmpl/thread_pools.sgml, glib/tmpl/thread_pools.sgml:
Updates after some renaming took place.

23 years agoRenamed 'value' and 'arg' to 'data' and 'thread_func' to 'func' to make it
Sebastian Wilhelmi [Wed, 9 May 2001 12:51:21 +0000 (12:51 +0000)]
Renamed 'value' and 'arg' to 'data' and 'thread_func' to 'func' to make it

2001-05-09  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gthread.c, gthread.h: Renamed 'value' and 'arg' to 'data' and
'thread_func' to 'func' to make it more consistent with the rest
of GLib.

* gthreadpool.c, gthreadpool.h: Moved 'stack_size' from the public
members of GThreadPool to the private ones. Renamed 'thread_func'
to 'func' as above. Moved up 'user_data' in g_thead_pool_new
argument list and in GThreadPool struct.

23 years agoClarification for g_mutex_trylock.
Sebastian Wilhelmi [Tue, 8 May 2001 14:21:40 +0000 (14:21 +0000)]
Clarification for g_mutex_trylock.

2001-05-08  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* glib/tmpl/threads.sgml: Clarification for g_mutex_trylock.

23 years agoremoved #include <unistd.h> sneaked in by sopwith in november.
Tim Janik [Tue, 8 May 2001 13:50:38 +0000 (13:50 +0000)]
removed #include <unistd.h> sneaked in by sopwith in november.

Tue May  8 15:33:31 2001  Tim Janik  <timj@gtk.org>

        * gcompletion.h: removed #include <unistd.h> sneaked in by
        sopwith in november.

23 years agoUpdated.
Sebastian Wilhelmi [Tue, 8 May 2001 08:25:43 +0000 (08:25 +0000)]
Updated.

2001-05-08  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* glib/tmpl/threads.sgml, glib/glib-overrides.txt: Updated.

23 years agoMoved func and arg members from GRealThread to GThread, such that they can
Sebastian Wilhelmi [Tue, 8 May 2001 08:23:18 +0000 (08:23 +0000)]
Moved func and arg members from GRealThread to GThread, such that they can

2001-05-08  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gmain.c, gthread.c, gthread.h: Moved func and arg members from
GRealThread to GThread, such that they can be accessed by the
user.

* gthread.c, gthread.h: Due to popular demand (Tim being the
populus here ;-) threads now have a 'return value', which is
returned by g_thread_join and is either the return of the topmost
thread function or the value given to g_thread_exit.

* gthreadpool.c, tests/mainloop-test.c, tests/thread-test.c:
Adapted to the above change.

23 years agodon't g_assert() the user specified time, but g_return_val_if_fail() here.
Tim Janik [Tue, 8 May 2001 05:25:35 +0000 (05:25 +0000)]
don't g_assert() the user specified time, but g_return_val_if_fail() here.

Fri May  4 04:14:45 2001  Tim Janik  <timj@gtk.org>

        * gthread-posix.c (g_cond_timed_wait_posix_impl): don't g_assert()
        the user specified time, but g_return_val_if_fail() here.

23 years agodon't try to export g_strcpy, it is g_stpcpy; updated and added some
Hans Breuer [Sun, 6 May 2001 13:57:21 +0000 (13:57 +0000)]
don't try to export g_strcpy, it is g_stpcpy; updated and added some

2001-01-06  Hans Breuer  <hans@breuer.org>

* glib.def : don't try to export g_strcpy, it is g_stpcpy;
updated and added some tweaking for functions which got
renamed recently, to avoid the update hassle if it can
simply be done compatible. Should vanish if there is a
stable version for win32.

23 years agoglib-mkenums is not a compiled program, so shouldn't be in the
James Henstridge [Sat, 5 May 2001 08:38:35 +0000 (08:38 +0000)]
glib-mkenums is not a compiled program, so shouldn't be in the

2001-05-05  James Henstridge  <james@daa.com.au>

        * Makefile.am: glib-mkenums is not a compiled program, so shouldn't
        be in the bin_PROGRAMS primary.  Put it in bin_SCRIPTS instead, and
        removed the glib_mkenums_SOURCES var.

23 years agoReleased 1.3.5 GLIB_1_3_5
Owen Taylor [Fri, 4 May 2001 22:34:43 +0000 (22:34 +0000)]
Released 1.3.5

Fri May  4 11:49:18 2001  Owen Taylor  <otaylor@redhat.com>

* Released 1.3.5

* NEWS: Updated

* configure.in (GLIB_MICRO_VERSION): Up version to 1.3.5,
interface/binary age 0.

23 years agofixed a typo in a comment.
Sven Neumann [Fri, 4 May 2001 17:01:56 +0000 (17:01 +0000)]
fixed a typo in a comment.

2001-05-04  Sven Neumann  <sven@convergence.de>

        * ghash.c: fixed a typo in a comment.

        * gtree.[ch]: added new functions g_tree_new_full(), g_tree_replace(),
        g_tree_steal() and g_tree_foreach() to adapt GTree to the GHashTable
        API. Moved comments into the C file.

        * docs/reference/glib/glib-sections.txt
        * docs/reference/glib/tmpl/glib-unused.sgml
        * docs/reference/glib/tmpl/hash_tables.sgml
        * docs/reference/glib/tmpl/linked_lists_double.sgml
        * docs/reference/glib/tmpl/linked_lists_single.sgml
        * docs/reference/glib/tmpl/macros_misc.sgml
        * docs/reference/glib/tmpl/trees-binary.sgml: updated documentation

23 years agoMake foreach() safe against removal of the _current_ element. While this
Owen Taylor [Thu, 3 May 2001 10:47:32 +0000 (10:47 +0000)]
Make foreach() safe against removal of the _current_ element. While this

Thu May  3 06:38:28 2001  Owen Taylor  <otaylor@redhat.com>

* g[s]list.c (g_[s]list_foreach) docs/Changes-2.0.txt: Make
foreach() safe against removal of the _current_ element. While
this could break some code, the new behavior is consistent with
the rest of GLib/GTK+ and probably is what people expect in most
cases. (Suggested by Paul Kuykendall, #50071)

23 years agoMake types of ternary operator correspond. (Fixes compilation errors with
Owen Taylor [Thu, 3 May 2001 10:12:46 +0000 (10:12 +0000)]
Make types of ternary operator correspond. (Fixes compilation errors with

Thu May  3 06:10:23 2001  Owen Taylor  <otaylor@redhat.com>

* gobject.c (g_object_set_[q]data_full): Make types of ternary
        operator correspond. (Fixes compilation errors with Sun CC,
#52230)

23 years agoFix read past end of the string. (#50404, fix from Jonas Borgström)
Owen Taylor [Wed, 2 May 2001 15:19:55 +0000 (15:19 +0000)]
Fix read past end of the string. (#50404, fix from Jonas Borgström)

Wed May  2 11:10:22 2001  Owen Taylor  <otaylor@redhat.com>

* gutf8.c (g_utf8_to_ucs4_fast): Fix read past end of the string.
(#50404, fix from Jonas Borgström)

23 years agoAdded Catalan file
Pablo Saratxaga [Mon, 30 Apr 2001 20:40:51 +0000 (20:40 +0000)]
Added Catalan file

23 years agoremoved \v escaping, newer perl versions don't support this, and it'S
Tim Janik [Mon, 30 Apr 2001 17:59:47 +0000 (17:59 +0000)]
removed \v escaping, newer perl versions don't support this, and it'S

Mon Apr 30 20:03:56 2001  Tim Janik  <timj@gtk.org>

        * glib-mkenums (usage): removed \v escaping, newer perl versions don't
        support this, and it'S probably useless anyways.

23 years agomade hook ids a gulong.
Tim Janik [Sun, 29 Apr 2001 03:04:27 +0000 (03:04 +0000)]
made hook ids a gulong.

Sun Apr 29 00:37:34 2001  Tim Janik  <timj@gtk.org>

        * ghook.[hc]: made hook ids a gulong.

Sat Apr 28 23:39:42 2001  Tim Janik  <timj@gtk.org>

        * gsignal.[hc]: made signal handler and emission hook ids gulongs.
        (signal_handlers_foreach_matched_R): only invoke callback for handlers
        that are not disconnected (id>0).
        (signal_emit_R): prevent invocation of signal handlers during the
        emission they were connected within.

        * glib-mkenums: publically installed perl-script to parse C code
        enums and generate descriptions thereof.
        * glib-mkenums.1: assorted man page.

23 years agoAdd a check for the Darwin dynamic linker. Use AC_TRY_LINK when checking
Dan Winship [Fri, 20 Apr 2001 17:08:57 +0000 (17:08 +0000)]
Add a check for the Darwin dynamic linker. Use AC_TRY_LINK when checking

        * configure.in: Add a check for the Darwin dynamic linker. Use
        AC_TRY_LINK when checking for "nonposix getpwuid_r" so it notices
        "no getpwuid_r" correctly.

        * testglib.c (main): Make template[] bigger to prevent an overrun.
        Remove an unused variable. Initialize error to NULL.

        * tests/gio-test.c (main): Add a cast to prevent a warning when
        size_t is a long.

        * tests/type-test.c (main): Add an #ifdef to prevent a warning
        when G_HAVE_GINT64 is defined and G_GINT64_FORMAT isn't.

23 years agogmodule implementation for Darwin/Mac OS X
Dan Winship [Fri, 20 Apr 2001 17:05:38 +0000 (17:05 +0000)]
gmodule implementation for Darwin/Mac OS X

        * gmodule-dyld.c: gmodule implementation for Darwin/Mac OS X

        * gmodule.c:
        * gmoduleconf.h.in: Add gmodule-dyld support

        * testgmodule.c (main): Fix spelling

23 years agoWe were returning junk memory here, because we didn't copy the value
Havoc Pennington [Thu, 19 Apr 2001 21:48:21 +0000 (21:48 +0000)]
We were returning junk memory here, because we didn't copy the value

2001-04-19  Havoc Pennington  <hp@redhat.com>

* gobject.c (g_object_get_valist): We were returning junk memory
here, because we didn't copy the value (G_VALUE_NOCOPY_CONTENTS
passed to G_VALUE_LCOPY) and then we freed the GValue immediately
after. Removed G_VALUE_NOCOPY_CONTENTS from here; need a
G_VALUE_STEAL_CONTENTS or the like if we want this optimization.

23 years agoChanges due to deprecation update.
Sebastian Wilhelmi [Thu, 19 Apr 2001 14:50:44 +0000 (14:50 +0000)]
Changes due to deprecation update.

2001-04-19  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* glib/tmpl/hash_tables.sgml, glib/tmpl/misc_utils.sgml,
glib/tmpl/scanner.sgml: Changes due to deprecation update.

23 years agoRemove definition of g_hash_table_freeze and g_hash_table_thaw. Instead
Sebastian Wilhelmi [Thu, 19 Apr 2001 13:33:31 +0000 (13:33 +0000)]
Remove definition of g_hash_table_freeze and g_hash_table_thaw. Instead

2001-04-19  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* ghash.c, ghash.h: Remove definition of g_hash_table_freeze and
g_hash_table_thaw. Instead added G_DISABLE_DEPRECATED-guarded
macros to ghash.h to go along the lines of the standard.

* gscanner.c, gscanner.h: Dito for g_scanner_freeze_symbol_table
and g_scanner_thaw_symbol_table.

* gutils.c, gutils.h: Dito for g_dirname. g_basename is still
defined in gutils.c, but declared ing gutils.h only
G_DISABLE_DEPRECATED-guarded.

23 years agoRemoved bashism in test for the pkg-config version.
Sebastian Wilhelmi [Thu, 19 Apr 2001 08:26:21 +0000 (08:26 +0000)]
Removed bashism in test for the pkg-config version.

2001-04-19  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* configure.in: Removed bashism in test for the pkg-config
version.

* configure.in: Rewrote test for multithread flag. Now uses
localtime_r, which hopefully has a consistent prototype across
different platforms. Also it uses a clever double EGREP trick
instead of compiling, which could give false positives. Thanks to
Dan Winship <danw@ximian.com> for the hint.

23 years agoAdd missing static pointed out by Michael Meeks.
Owen Taylor [Wed, 18 Apr 2001 21:37:25 +0000 (21:37 +0000)]
Add missing static pointed out by Michael Meeks.

Wed Apr 18 17:35:38 2001  Owen Taylor  <otaylor@redhat.com>

* gutils.c (_glib_gettext): Add missing static pointed
out by Michael Meeks.

23 years agoChange C++ comments for FIXME's to C comments. (Patch from Andres Salomon)
Owen Taylor [Wed, 18 Apr 2001 13:47:59 +0000 (13:47 +0000)]
Change C++ comments for FIXME's to C comments. (Patch from Andres Salomon)

Wed Apr 18 09:46:56 2001  Owen Taylor  <otaylor@redhat.com>

* gobject.c gsignal.c: Change C++ comments for FIXME's to
        C comments. (Patch from Andres Salomon)

23 years agoRemoved. Keeping README, README.cvs-commits HACKING, and AUTHORS up to
Owen Taylor [Wed, 18 Apr 2001 13:41:45 +0000 (13:41 +0000)]
Removed. Keeping README, README.cvs-commits HACKING, and AUTHORS up to

Wed Apr 18 09:37:07 2001  Owen Taylor  <otaylor@redhat.com>

* MAINTAINERS: Removed. Keeping README, README.cvs-commits
HACKING, and AUTHORS up to date is plenty without extra
random files that someone thought a module should have.
(Actually, I believe this was used for debbugs in the past.)

23 years agoReleased 1.3.4 GLIB_1_3_4
Owen Taylor [Tue, 17 Apr 2001 22:39:00 +0000 (22:39 +0000)]
Released 1.3.4

Tue Apr 17 11:47:07 2001  Owen Taylor  <otaylor@redhat.com>

* Released 1.3.4

* NEWS: Updated

23 years agoDefine _GNU_SOURCE for stpcpy
Owen Taylor [Tue, 17 Apr 2001 14:48:11 +0000 (14:48 +0000)]
Define _GNU_SOURCE for stpcpy

Tue Apr 17 10:43:36 2001  Owen Taylor  <otaylor@redhat.com>

* gstrfuncs.c: Define _GNU_SOURCE for stpcpy

* tests/mainloop-test.c (main): Wait for all threads
to start before beginning tests.

23 years agoUntil now every thread pool always had at least one tread waiting to avoid
Sebastian Wilhelmi [Tue, 17 Apr 2001 11:48:45 +0000 (11:48 +0000)]
Until now every thread pool always had at least one tread waiting to avoid

2001-04-17  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gthreadpool.c (g_thread_pool_thread_proxy): Until now every
thread pool always had at least one tread waiting to avoid
switching overhead in case a new task would be added soon after
one finished. This however means a big waste of threads, if many
mostly inactive thread pools are involved. Now such a waiting
thread will only wait for half a second (This value is of course
very randomly picked) and go to the global threadpool afterwards.

MCVS: ----------------------------------------------------------------------

23 years agoRemove warnings about conflicts with the stable version. GLIB_1_3_3
Owen Taylor [Tue, 17 Apr 2001 00:55:34 +0000 (00:55 +0000)]
Remove warnings about conflicts with the stable version.

Mon Apr 16 12:04:52 2001  Owen Taylor  <otaylor@redhat.com>

        * configure.in: Remove warnings about conflicts with the
        stable version.

 * glib-2.0.m4: Fix some of the error text to be halfway
        up to date.

        * README.in INSTALL.in: Add these to generate README, INSTAL
        (as in the stable branch). Update.

  * HACKING: Update.

23 years agodocs
Havoc Pennington [Mon, 16 Apr 2001 20:05:25 +0000 (20:05 +0000)]
docs

2001-04-16  Havoc Pennington  <hp@redhat.com>

        * gqsort.c: docs

        * gfileutils.c: docs

        * gwin32.c: docs fixes

        * gconvert.c: docs

        * guniprop.c: docs

        * gutf8.c: docs

23 years ago.cvsignore fixes
Owen Taylor [Mon, 16 Apr 2001 16:35:11 +0000 (16:35 +0000)]
.cvsignore fixes

23 years agoMore tweaks from GTK+. Reenable $(HTML_IMAGES) with shell portability
Owen Taylor [Mon, 16 Apr 2001 16:34:15 +0000 (16:34 +0000)]
More tweaks from GTK+. Reenable $(HTML_IMAGES) with shell portability

Mon Apr 16 12:18:26 2001  Owen Taylor  <otaylor@redhat.com>

        * {glib,gobject}/Makefile.am (tmpl-build.stamp): More tweaks
        from GTK+. Reenable $(HTML_IMAGES) with shell portability fixes.

23 years agoput AC_PATH_PROG(pkg-config) before "Checking for glib" so the output
Havoc Pennington [Mon, 16 Apr 2001 15:21:19 +0000 (15:21 +0000)]
put AC_PATH_PROG(pkg-config) before "Checking for glib" so the output

2001-04-16  Havoc Pennington  <hp@redhat.com>

* glib-2.0.m4: put AC_PATH_PROG(pkg-config) before "Checking for
glib" so the output looks right

23 years agomake GDebugKeys argument const
Havoc Pennington [Mon, 16 Apr 2001 15:04:17 +0000 (15:04 +0000)]
make GDebugKeys argument const

2001-03-23  Havoc Pennington  <hp@redhat.com>

* gutils.c (g_parse_debug_string): make GDebugKeys argument
const

23 years agoupdated
Hans Breuer [Fri, 13 Apr 2001 23:41:53 +0000 (23:41 +0000)]
updated

2001-04-14  Hans Breuer  <hans@breuer.org>

* glib.def :
* makefile.msc.in : updated

* gpattern.c : include "gutils.h" to resolve the inline hassle

23 years agoupdated
Hans Breuer [Fri, 13 Apr 2001 23:40:39 +0000 (23:40 +0000)]
updated

2001-04-14  Hans Breuer  <hans@breuer.org>

* gobject.def : updated

23 years agoPass pkg-config options before the other args so it works even if
Alexander Larsson [Wed, 11 Apr 2001 14:58:29 +0000 (14:58 +0000)]
Pass pkg-config options before the other args so it works even if

2001-04-11  Alexander Larsson  <alexl@redhat.com>

* glib-2.0.m4: Pass pkg-config options
before the other args so it works even if
POSIXLY_CORRECT is set.

23 years agoincrement version to 1.3.4 (binary 0, interface 0).
Tim Janik [Mon, 9 Apr 2001 17:03:55 +0000 (17:03 +0000)]
increment version to 1.3.4 (binary 0, interface 0).

Mon Apr  9 18:57:44 2001  Tim Janik  <timj@gtk.org>

        * configure.in: increment version to 1.3.4 (binary 0, interface 0).

Mon Apr  9 18:56:15 2001  Tim Janik  <timj@gtk.org>

        * gclosure.c (g_closure_invoke): only require marshal/meta_marshal if
        we're valid (about to actually do marshalling).

23 years agoAnd other fixes. Thanks to Göran Uddeborg <goeran@uddeborg.pp.se>.
Christian Rose [Thu, 5 Apr 2001 21:38:19 +0000 (21:38 +0000)]
And other fixes. Thanks to Göran Uddeborg <goeran@uddeborg.pp.se>.

23 years agoFixed another message in the Swedish translation. Thanks to Jörgen Tegnér
Christian Rose [Thu, 5 Apr 2001 20:27:46 +0000 (20:27 +0000)]
Fixed another message in the Swedish translation. Thanks to Jörgen Tegnér
<teg@post.utfors.se>.

23 years agoFixed Swedish translation, thanks to Veronica Loell <info@nakawe.se> and
Christian Rose [Thu, 5 Apr 2001 11:05:24 +0000 (11:05 +0000)]
Fixed Swedish translation, thanks to Veronica Loell <info@nakawe.se> and
Martin Norbäck <d95mback@dtek.chalmers.se>.

23 years agoAdded Swedish translation.
Christian Rose [Thu, 5 Apr 2001 00:51:52 +0000 (00:51 +0000)]
Added Swedish translation.

23 years agoReleased GLib-1.3.3.
Tim Janik [Wed, 4 Apr 2001 07:19:45 +0000 (07:19 +0000)]
Released GLib-1.3.3.

Wed Apr  4 09:18:55 2001  Tim Janik  <timj@gtk.org>

        * Released GLib-1.3.3.

23 years agodisabled "for i in $(HTML_IMAGES) ;" dist-hook because $(HTML_IMAGES) is
Tim Janik [Wed, 4 Apr 2001 02:31:40 +0000 (02:31 +0000)]
disabled "for i in $(HTML_IMAGES) ;" dist-hook because $(HTML_IMAGES) is

Wed Apr  4 00:56:00 2001  Tim Janik  <timj@gtk.org>

        * gobject/Makefile.am: disabled "for i in $(HTML_IMAGES) ;"
        dist-hook because $(HTML_IMAGES) is empty.

23 years agoupdates.
Tim Janik [Tue, 3 Apr 2001 19:22:44 +0000 (19:22 +0000)]
updates.

Tue Apr  3 20:22:59 2001  Tim Janik  <timj@gtk.org>

        * NEWS: updates.

        * NEWS.pre-1-3: take over old news.

Tue Apr  3 20:23:24 2001  Tim Janik  <timj@gtk.org>

        * NEWS: updates.

23 years agoExit with an error message that you should use pkg-config instead.
Owen Taylor [Tue, 3 Apr 2001 18:27:59 +0000 (18:27 +0000)]
Exit with an error message that you should use pkg-config instead.

Tue Apr  3 12:38:16 2001  Owen Taylor  <otaylor@redhat.com>

* glib-config-2.0.in: Exit with an error message that you
should use pkg-config instead.

* configure.in (PACKAGE): Require pkg-config.

* tests/Makefile.am (EXTRA_DIST): Add utf8.txt.

* configure.in (GLIB_MICRO_VERSION): Up MICRO to 3,
leave interface/binary at 0.

23 years agoadd missed (empty) file
Owen Taylor [Tue, 3 Apr 2001 18:27:29 +0000 (18:27 +0000)]
add missed (empty) file

23 years agoAdd explicit path to tmpl files in dependencies.
Owen Taylor [Tue, 3 Apr 2001 18:27:08 +0000 (18:27 +0000)]
Add explicit path to tmpl files in dependencies.

Tue Apr  3 11:51:02 2001  Owen Taylor  <otaylor@redhat.com>

* {glib,gobject}/Makefile.am (sgml-build.stamp): Add
explicit path to tmpl files in dependencies.

* glib/glib.types gobject/gobject.types: empty files
to keep generic makefiles happy.

23 years agoadded g_list_nth_prev() which walks ->prev instead of ->next.
Tim Janik [Tue, 3 Apr 2001 13:15:41 +0000 (13:15 +0000)]
added g_list_nth_prev() which walks ->prev instead of ->next.

Tue Apr  3 13:46:22 2001  Tim Janik  <timj@gtk.org>

        * glist.[hc]: added g_list_nth_prev() which walks ->prev instead
        of ->next.

        * gpattern.[hc]: added shell-style pattern matching code from beast,
        derived from the gtk_pattern_*() code, but with a couple of bug fixes
        and a number of optimizations.

Tue Apr  3 14:06:00 2001  Tim Janik  <timj@gtk.org>

        * gparam.[hc]: added g_param_spec_pool_list() to list pspecs per
        owner_type. the pspecs are not referenced, so the caller is
        supposed to have some idea about owner_type not randomly
        nuking his pspec's. if this is going to provide problems in
        the future, we can either auto-ref the pspecs, or add a
        _foreach variant, though the latter would have to invoke
        the callback while pspec's mutex is acquired, so i just
        went for the _list variant for now.

        * gclosure.h (G_CALLBACK): made GCallback a void (*) (void) fucntion.

23 years agoAdded special case for priorities on FreeBSD. Thanks to David Reid
Sebastian Wilhelmi [Tue, 3 Apr 2001 12:49:44 +0000 (12:49 +0000)]
Added special case for priorities on FreeBSD. Thanks to David Reid

2001-04-03  Sebastian Wilhelmi  <wilhelmi@ira.uka.de>

* gthread-posix.c: Added special case for priorities on
  FreeBSD. Thanks to David Reid <dreid@jetnet.co.uk> for the info.

* gthread-impl.c: Made two macros safe with ().

23 years agoOps, ChangeLog was incomplete.
Sebastian Wilhelmi [Tue, 3 Apr 2001 12:45:23 +0000 (12:45 +0000)]
Ops, ChangeLog was incomplete.

Ops, ChangeLog was incomplete.