Mercurial > pidgin
view m4/gnome-x-checks.m4 @ 12115:e9790eb93216
[gaim-migrate @ 14415]
quoth charkins:
" This patch has a few small fixes for the visibility
stuff in gtkblist.c.
First, tracking of the ICONIFIED state of the blist was
removed. This was intended to allow the blist to
"remember" if it was minimized between restarts.
Unfortunately, this is not possible because the
ICONIFIED state gets set when the blist is on a
different desktop with many window managers.
Second, while talking about the ICONIFIED issue on
#gtk@GIMPNet, muntyan_ asked about a bug where the
blist would get shown on an account re-connect with
1.5.0. Luke mentioned something about this with cvs as
well. This patch introduces a check in
gaim_gtk_blist_show() to prevent the window from being
shown if it already exists and is visible.
Third, sadrul pointed me to a one-line fix for the
missing blist on startup. I added a second line to make
sure the blist restores its proper size as well.
Finally, when the last visibility manager is removed,
gaim will now minimize the blist if it was previously
hidden, rather than showing it. This could prevent a
race condition with out-of-process applets, preventing
gaim from maintaining the visibility state properly
between restarts.
This was 'cvs diff'ed against the last available anon
cvs from Friday. Hopefully it'll apply cleanly."
it did.
committer: Tailor Script <tailor@pidgin.im>
author | Luke Schierer <lschiere@pidgin.im> |
---|---|
date | Wed, 16 Nov 2005 17:55:26 +0000 |
parents | 68b230f8da5f |
children |
line wrap: on
line source
dnl GNOME_X_CHECKS dnl dnl Basic X11 related checks for X11. At the end, the following will be dnl defined/changed: dnl GTK_{CFLAGS,LIBS} From AM_PATH_GTK dnl CPPFLAGS Will include $X_CFLAGS dnl GNOME_HAVE_SM `true' or `false' depending on whether session dnl management is available. It is available if dnl both -lSM and X11/SM/SMlib.h exist. (Some dnl Solaris boxes have the library but not the header) dnl XPM_LIBS -lXpm if Xpm library is present, otherwise "" dnl dnl The following configure cache variables are defined (but not used): dnl gnome_cv_passdown_{x_libs,X_LIBS,X_CFLAGS} dnl AC_DEFUN([GNOME_X_CHECKS], [ AM_PATH_GTK(1.2.0,,AC_MSG_ERROR(GTK not installed, or gtk-config not in path)) dnl Hope that GTK_CFLAGS have only -I and -D. Otherwise, we could dnl test -z "$x_includes" || CPPFLAGS="$CPPFLAGS -I$x_includes" dnl dnl Use CPPFLAGS instead of CFLAGS because AC_CHECK_HEADERS uses dnl CPPFLAGS, not CFLAGS CPPFLAGS="$CPPFLAGS $GTK_CFLAGS" saved_ldflags="$LDFLAGS" LDFLAGS="$LDFLAGS $GTK_LIBS" gnome_cv_passdown_x_libs="$GTK_LIBS" gnome_cv_passdown_X_LIBS="$GTK_LIBS" gnome_cv_passdown_X_CFLAGS="$GTK_CFLAGS" gnome_cv_passdown_GTK_LIBS="$GTK_LIBS" LDFLAGS="$saved_ldflags $GTK_LIBS" dnl We are requiring GTK >= 1.1.1, which means this will be fine anyhow. USE_DEVGTK=true dnl AC_MSG_CHECKING([whether to use features from (unstable) GTK+ 1.1.x]) dnl AC_EGREP_CPP(answer_affirmatively, dnl [#include <gtk/gtkfeatures.h> dnl #ifdef GTK_HAVE_FEATURES_1_1_0 dnl answer_affirmatively dnl #endif dnl ], dev_gtk=yes, dev_gtk=no) dnl if test "$dev_gtk" = "yes"; then dnl USE_DEVGTK=true dnl fi dnl AC_MSG_RESULT("$dev_gtk") GNOME_HAVE_SM=true case "$GTK_LIBS" in *-lSM*) dnl Already found it. ;; *) dnl Assume that if we have -lSM then we also have -lICE. AC_CHECK_LIB(SM, SmcSaveYourselfDone, [GTK_LIBS="-lSM -lICE $GTK_LIBS"],GNOME_HAVE_SM=false, $x_libs -lICE) ;; esac if test "$GNOME_HAVE_SM" = true; then AC_CHECK_HEADERS(X11/SM/SMlib.h,,GNOME_HAVE_SM=false) fi if test "$GNOME_HAVE_SM" = true; then AC_DEFINE(HAVE_LIBSM) fi XPM_LIBS="" AC_CHECK_LIB(Xpm, XpmFreeXpmImage, [XPM_LIBS="-lXpm"], , $x_libs) AC_SUBST(XPM_LIBS) AC_REQUIRE([GNOME_PTHREAD_CHECK]) LDFLAGS="$saved_ldflags" AC_PROVIDE([GNOME_X_CHECKS]) ])