diff options
| author | Pietro Gagliardi <[email protected]> | 2014-04-27 21:33:48 -0400 |
|---|---|---|
| committer | Pietro Gagliardi <[email protected]> | 2014-04-27 21:34:42 -0400 |
| commit | 34c84ce659d2c13bd603a8d77bbb3a98ee49128a (patch) | |
| tree | 8ec87b42382b6d2ab630c8062ee11716369d95a2 /uitask_windows.go | |
| parent | 41631be076545e3478e2154c1257d252e0f9c360 (diff) | |
More future plans and a clarification in the block comment at the top of uitask_windows.go.
Diffstat (limited to 'uitask_windows.go')
| -rw-r--r-- | uitask_windows.go | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/uitask_windows.go b/uitask_windows.go index ad8404f..f7a6ad5 100644 --- a/uitask_windows.go +++ b/uitask_windows.go @@ -14,8 +14,8 @@ problem: messages have to be dispatched on the same thread as system calls, and possible: solution: use PostThreadMessage() to send uimsgs out to the message loop, which runs on its own goroutine (I had come up with this first but wanted to try other things before doing it (and wasn't really sure if user-defined messages were safe, not quite understanding the system); nsf came up with it independently and explained that this was really the only right way to do it, so thanks to him) -problem: if the thread isn't in its main message pump, the thread message is simply lost -this happened when scrolling Area +problem: if the thread isn't in its main message pump, the thread message is simply lost (see, for example, http://blogs.msdn.com/b/oldnewthing/archive/2005/04/26/412116.aspx) +this happened when scrolling Areas the only recourse, and the one both Microsoft (http://support.microsoft.com/kb/183116) and Raymond Chen (http://blogs.msdn.com/b/oldnewthing/archive/2008/12/23/9248851.aspx) suggest (and Treeki/Ninjifox confirmed), is to create an invisible window to dispatch messages instead. |