aboutsummaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
...
| * | | Add dbus-menu-clean.xml to EXTRA_DIST and remove previous builddir hack to ↵Ken VanDine2012-03-291-4/+5
| | | | | | | | | | | | | | | | fix distcheck
| * | | distcheck fixesKen VanDine2012-03-291-4/+4
|/ / /
* | | merge lp:~ted/dbusmenu/instruction-count-testing to support benchmarking ↵Charles Kerr2012-03-295-8/+42
|\ \ \ | | | | | | | | | | | | based on callgrind measurements h/t Chase
| * | | Adding in a test to print the number of instructions it takes to export a menuTed Gould2012-03-282-1/+24
| | | |
| * | | Making the json test signal based instead of timeout basedTed Gould2012-03-282-6/+8
| | | |
| * | | Insert callgrind annotations into the serverTed Gould2012-03-282-2/+11
| | | |
* | | | merge lp:~ted/dbusmenu/testing-fixes for improvements to the dbusmenu testsCharles Kerr2012-03-284-24/+77
|\| | |
| * | | Grabbing the firefox extreme caseTed Gould2012-03-271-0/+43629
| |\ \ \
| * | | | Making it so that the JSON handler can handle an array of bytes to specify ↵Ted Gould2012-03-271-3/+21
| | | | | | | | | | | | | | | | | | | | images
| * | | | Change the dbusmenu-dumper to be realization based instead of timeout basedTed Gould2012-03-271-1/+53
| | | | |
| * | | | Using the new --wait-for attribute of dbus-test-runnerTed Gould2012-03-272-20/+3
| | | | |
* | | | | ChrisCoulson's patch to add a lookup hash to avoid the overhead of calling ↵Charles Kerr2012-03-281-7/+55
|\ \ \ \ \ | | | | | | | | | | | | | | | | | | dbusmenu_menuitem_find_id() when possible. See lp bug #801699
| * | | | | this is ChrisCoulson's experimental patch to add a lookup hash to avoid the ↵Charles Kerr2012-03-271-7/+55
| |/ / / / | | | | | | | | | | | | | | | overhead of calling dbusmenu_menuitem_find_id() when possible.
* | | | | fold together client.c's two PROPERTY_CHANGED signal handlersCharles Kerr2012-03-281-12/+2
|\ \ \ \ \ | |_|_|_|/ |/| | | |
| * | | | fold client's two DBUSMENU_MENUITEM_SIGNAL_PROPERTY_CHANGED signal handlers ↵Charles Kerr2012-03-271-12/+2
| |/ / / | | | | | | | | | | | | together for clarity
* | | | Don't sent event replies that aren't requestedTed Gould2012-03-281-1/+3
|\ \ \ \
| * | | | server: don't send replies for events that don't request repliesRyan Lortie2012-03-271-1/+3
| |/ / /
* | | | merge lp:~ted/dbusmenu/firefox-extreme-case for tests/test-json-02.jsonCharles Kerr2012-03-271-0/+43629
|\ \ \ \ | |/ / / |/| / / | |/ /
| * / Adding a Firefox extreme menu case for testingTed Gould2012-03-271-0/+43629
|/ /
* | 0.5.95Ted Gould2012-03-211-3/+3
| |
* | merge lp:~charlesk/dbusmenu/lp-953509 to fix a crash in dbusmenu-gtk's ↵Charles Kerr2012-03-212-59/+11
|\ \ | | | | | | | | | handling of menu icon theme changes
| * | Don't listen for "changed" events from the screen's default GtkIconTheme. ↵Charles Kerr2012-03-211-51/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | Fixes lp bug #953509 The crash in #953509 was being caused by the last two lines of theme_changed_cb() (a) leaving a dangling handler id in priv.theme_changed_sig, and (b) not remembering the handler id of its own signal connection. However after testing I don't see any reason to keep any of the theme handling code at all. "But wait," you say. "How will our menu icons follow the theme changes?" It works in practice anyway because we always listen for property changes to our reference GtkImage, **and** we listen for property changes to its GtkImageMenuItem parent so that if the GtkImageMenuItem changes GtkImages we can stop listening to the old one and start listening to the new one.
| * | merge lp:~ted/dbusmenu/regex_merge to fix a warning that comes up a lot with ↵Charles Kerr2012-03-201-8/+11
|/| | | | | | | | | | | unity-panel-service about not having a label.
| * | Fix string leakTed Gould2012-03-201-1/+1
| | |
| * | Handle the case of the label being NULL which can happen on custom itemsTed Gould2012-03-201-7/+10
|/ /
* | Don't ask for a reply if no one caresTed Gould2012-03-201-0/+12
|\ \
| * | glib client: don't request a reply on send_event() if nobody is listening for itRyan Lortie2012-03-161-0/+12
|/ /
* | 0.5.94Ted Gould2012-03-091-3/+3
| |
* | Removing underscores in the ATK nameTed Gould2012-03-091-7/+15
|\ \
| * | Restructuring slightly to have two cases, allocating memory and not allocatingTed Gould2012-03-091-7/+9
| | |
| * | Remove the underscores used for mnewmonics, as orca speaks them, which is notLuke Yelavich2012-03-091-1/+7
| | | | | | | | | | | | | | | what we want.
* | | Making sure to disconnect theme changed handler universallyTed Gould2012-03-091-4/+12
|\ \ \ | |/ / |/| |
| * | Attaching bugTed Gould2012-03-090-0/+0
| | |
| * | Missing a couple of cases of removing the signal handler and clearing the ↵Ted Gould2012-03-091-4/+12
|/ / | | | | | | stored value
* | Making sure the ATK object name always gets set to the label if nothing elseTed Gould2012-03-081-1/+5
|\ \
| * | Set the accessible name from the dbusmenu item label. THis is a work-around ↵Luke Yelavich2012-03-081-0/+8
| | | | | | | | | | | | until GTK follows atk docs.
| * | Merge from trunk.Luke Yelavich2012-03-085-64/+87
| |\ \
| * | | Do not set the accessible name to an empty string if the accessible_descLuke Yelavich2012-02-171-4/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | property = NULL. Yes this was recently changed so that dbusmenu conformed to Atk documentation, but the GTK menu item accessibility code does not yet conform to sed documentation. As a result, all dbusmenu menu items that do not have the accessible_desc property set end up getting an empty string for their accessible name. In the long term, GTK accessibility needs to be fixed.
* | | | 0.5.93Ted Gould2012-03-081-3/+3
| | | |
* | | | Fixing a trivial annotation bugTed Gould2012-03-081-2/+2
| | | |
* | | | Fixing resetting an icon numerous timesTed Gould2012-03-082-19/+49
|\ \ \ \ | |_|/ / |/| | |
| * | | Switching out logic, we're using replaced now because it gets set everytime ↵Ted Gould2012-03-081-2/+1
| | | | | | | | | | | | | | | | that the hashtable is modified, and if we weren't modifying the hash table in some way, we don't want to signal. And, conversely, no one cares if we didn't modify the hash table.
| * | | Attaching bugTed Gould2012-03-020-0/+0
| | | |
| * | | Don't need to setup this signal twiceTed Gould2012-03-021-3/+0
| | | |
| * | | Not sending the update signal if the value is being cleared and it was ↵Ted Gould2012-03-021-1/+2
| | | | | | | | | | | | | | | | already cleared
| * | | Moving the adding of signals to the update_icon function so that everyone ↵Ted Gould2012-03-021-12/+32
| | | | | | | | | | | | | | | | gets them
| * | | Changing 'update_icon' to take more variables so we can do more fun stuff ↵Ted Gould2012-03-021-6/+12
| | | | | | | | | | | | | | | | with it soon!
| * | | Track the theme changed signal so we ensure that we can drop it eventuallyTed Gould2012-03-021-1/+8
| | |/ | |/|
* | | Set the ATK role on menuitemsTed Gould2012-03-071-0/+24
|\ \ \
| * | | Fixing for GTK2Ted Gould2012-03-011-0/+9
| | | |