| Age | Commit message (Collapse) | Author |
|---|
|
|
|
Remove unnecessary warnings
|
|
|
|
|
|
Fix error handling in gl as in wx.
|
|
Previously error from async functions made an exit when the next
sync call checked the message queue. This have been changed to an
error report instead since the errors where async there where really
hard to handle.
Also changed the error report format to make it easier to filter
them with a custom error_handler.
|
|
|
|
To avoid failing on slow machines
|
|
Needed functionality, to avoid abusing opauqe objects.
|
|
and avoid large logs when tests succed
|
|
Commands could be executed twice, if the command was dispatched
from a callback and caused a recursive invocation of command loop.
Solution is to mark op as -1 before calling wxWidgets.
Also commands could be missed when idle processing was done inside
while executing a recursive callback, solved be always resetting peak
index after idle processing is done.
|
|
|
|
Introduced when I optimized the binary handling.
|
|
Available since 3.0
|
|
Did not work on windows, on windows getItem uses wxLogWindow instead
of assert in case of out of range, so the test still caused a log window.
|
|
Avoids spawn but can deadlock
|
|
So they work on both wxWidgets-2.8 and 3.0
|
|
Instead of popping up an annoying msgbox
|
|
|
|
Remove erlang:now() and wait for windows to appear when neccesary,
which it often is when using wxWidgets-3.0 on gtk.
Also cleanup radioBox test.
|
|
To many floats in return value could cause a reallocation and
pointers to reallocated data. Fix it by ensuring that the temp
float array is large enough before add values to it.
|
|
Some events where missing useful information.
May require user code recompilation, since some records in wx.hrl
have chnaged it's definition.
|
|
Remove allocations
|
|
* maint:
debugger: Fix debugger save options on mac
wx: Do not crash server when going down
|
|
If a connect request was made during the takedown of wx it would crash.
|
|
|
|
|
|
Useful class though not documented in wxWidgets-2.8
|
|
* dgud/wx/fix-wx3-windows64:
wx: Send destroy message direct
wx: Add wxLocale class
wx: Added misc functions
wx: Fix memory garbage collector
wx: Fix possibility to fetch early open msgs on mac
wx: Fix listCtrl sort callback
wx: Fix configure for win64 with wxWidgets-3.0
|
|
|
|
|
|
Add the mentioned test suites for *all* library and touched
non-library applications.
|
|
and more dbg info
|
|
Remove the extra wxEvtListener that only caused confusion,
now everything is automatically cleaned up by the destructors
and event sent to erlang when it becomes delete for all cases.
|
|
Previously we could do a cleanup while we where recursed down
and thus delete the objects we where invoking.
|
|
It was missing.
|
|
|
|
wxStyledTextCtrl had no wrapper class so, because it was not detected
that it had virtual destructors, thus references to it was not cleaned
up on destruction. When a process dies and wx cleans up the event handlers
access was made to the deleted object and caused a seg fault.
Added a testcase which I thought could provoke the bug but didn't.
|
|
|
|
Do not use disconnect event listener when we are exiting the port, it may
interfere with window destructions and cause a crash.
|
|
Event handlers was not removed after objects/process had been
delete/died, which causes memory leakage and that fun's was
kept in the wx_server process. Code that might be purged and
the server died.
|
|
|
|
|
|
|
|
|
|
Fix bad hit test.
Fix timing issue in test.
|
|
Since I have replaced the wxTreeItemIdClass with integers, there
is no way to verify that it is a valid wxTreeItemId that is returned
from the wxTreeCtrl:getItem* functions.
Still the check is that is non null so beware don't construct your own id's.
|
|
Fixes wxTreeCtrl:getBoundingRect/2 and wxTreeCtrl:hitTest/1.
wxTreeCtrl:hitTest now returns a tuple, i.e. not bug compatible with previous
releases but needed to know if the returned wxTreeItemId is valid or not.
|
|
Added debugging and workarounds for wx-2.9 on Mac
|
|
Things have changed in 2.9 modify tests so they are still valid for
both 2.8 and 2.9 branches, so far only tested on linux.
|
