Pietro Gagliardi
|
b9c12da838
|
More TODOs (see previous commit).
|
2014-05-17 17:10:23 -04:00 |
|
Pietro Gagliardi
|
9ed8d9c816
|
Consistency change: all Windows API structure field names now match case-wise (some were forcibly exported in the past for historical reasons: I originally started package ui by trying to write a Windows API wrapper, but decided to ditch that and just use the Windows API directly from package ui). Also more TODOs.
|
2014-05-11 11:11:57 -04:00 |
|
Pietro Gagliardi
|
719cab8742
|
Migrated cursors_windows.go and icons_windows.go.
|
2014-04-13 13:11:17 -04:00 |
|
Pietro Gagliardi
|
3bf215ae4e
|
Split out the common resizing code into its own function so that it's all in one place. This will also affect what happens if I switch to DeferWindowPos() on Windows.
|
2014-04-07 14:32:25 -04:00 |
|
Pietro Gagliardi
|
d1a41cd95a
|
Corrected a copy-paste fail in stdwndclass_windows.go that broke compilation (this is why you test builds on all platforms, folks!) and added a TODO about resizing on Windows in the meantime.
|
2014-03-24 11:36:15 -04:00 |
|
Pietro Gagliardi
|
64d5eb541e
|
Changed the new resizing code so that it uses the same allocated slice per window instead of making a new one to store all the resize requests each time.
|
2014-03-17 21:09:03 -04:00 |
|
Pietro Gagliardi
|
19227080da
|
Changed window resizes so that the actual Control.setRect() functions appended to an array of requests that the resize() function set all at once instead of having each done individually. This will be necessary for what I think will be a solution to the deadlocks. It doesn't work right now; I'm assuming it's allocating too much memory. I know how to fix this, but I'm committing what I have so far to be safe.
|
2014-03-17 20:42:36 -04:00 |
|
Pietro Gagliardi
|
34ce16d586
|
Separated file creation dates from the package comment.
|
2014-03-12 21:55:45 -04:00 |
|
Pietro Gagliardi
|
962a84e056
|
Fixed Mac OS X control placement flipping. This is done by passing the height of the window around setRect() calls to avoid polling the content view frame each time.
|
2014-03-03 17:44:03 -05:00 |
|
Pietro Gagliardi
|
e39a5d928c
|
Fixed the unregistered window class error on 64-bit Windows: turns out it was a difference in Go's int vs. what Windows expected (thanks Microsoft for not using your special types here). Fixed all instances of this, even in unmigrated/. Still does not work in wine...
|
2014-02-21 02:27:59 -05:00 |
|
Pietro Gagliardi
|
fb82e651a4
|
Moved to a proper package; main() is now a test and go test -c is used to build. Once I iron out a bug with Windows event handling, I'll add a README.
|
2014-02-19 11:41:10 -05:00 |
|
Pietro Gagliardi
|
1f08c874e0
|
Made event dispatch asynchronous to avoid deadlocks; right now events that cannot be dispatched get dropped.
|
2014-02-18 09:57:19 -05:00 |
|
Pietro Gagliardi
|
0595135d9a
|
Fixed the GTK+ crashes (I think) by making resizes synchronous. Since all control resizes happen on the UI thread anyway, we don't need to dispatch back; just call the resizing functions directly. Windows gets this fix too for consistency (and also because it gets rid of the only asynchronous oddity in the system).
|
2014-02-17 01:40:53 -05:00 |
|
Pietro Gagliardi
|
a400ec6666
|
Fixed the resizing deadlock.
|
2014-02-13 11:42:21 -05:00 |
|
Pietro Gagliardi
|
5626b9e35c
|
Added sizing of windows and the main window control. It presently deadlocks; I'll need to redo my mutexes...
|
2014-02-13 05:28:26 -05:00 |
|
Pietro Gagliardi
|
3d2df2334e
|
Fixed compiler errors. New code structure success!
|
2014-02-12 21:33:24 -05:00 |
|
Pietro Gagliardi
|
403ca88316
|
Whoops, class registration now has to go through the UI task.
|
2014-02-12 21:11:21 -05:00 |
|
Pietro Gagliardi
|
98d56767a3
|
Each control may only need one event, so combine them all into a single channel whose use depends on the sysData itself. Also we won't need to save the parentWindow anymore, as the change to the make() function will take care of that for us.
|
2014-02-12 20:55:05 -05:00 |
|
Pietro Gagliardi
|
4b16716e61
|
Restructuring the codebase to make things neater. First off: each window now has its own class; the sysData knows about children. This updates the window class to become a generator and rewrites the WndProc as well.
|
2014-02-12 20:51:07 -05:00 |
|
Pietro Gagliardi
|
5f587fbf47
|
Added button events.
|
2014-02-12 11:32:17 -05:00 |
|
Pietro Gagliardi
|
0f373195de
|
Changed the sysData cache so that if we didn't fill it in yet, getSysData() returns nil instead of panicking, and the standard window procedure jumps directly to the default handler. Not what I originally wanted to do, but whatever, it finally works!
|
2014-02-11 22:41:55 -05:00 |
|
Pietro Gagliardi
|
88363e131c
|
Got rid of the reverse HWND->sysData lookup for now, to see if there's a mistake elsewhere. Now we're up to a panic... with Success...
|
2014-02-11 20:14:19 -05:00 |
|
Pietro Gagliardi
|
cee78198e4
|
It appears the result of syscall.StringToUTF16Ptr() is only temporary, so removed all that. Now the program just hangs...
|
2014-02-11 19:52:39 -05:00 |
|
Pietro Gagliardi
|
a4182c606b
|
Whoops, copy-paste fail somehow changed init error checks. Now it starts... and hangs before creating a window...
|
2014-02-11 19:27:58 -05:00 |
|
Pietro Gagliardi
|
73fa611b37
|
All build errors resolved. Now to test...
|
2014-02-11 19:18:03 -05:00 |
|
Pietro Gagliardi
|
09a42e0c34
|
Added the init and testing main functions and started fixing errors. Let's fix the rest and hope it works...
|
2014-02-11 18:57:03 -05:00 |
|
Pietro Gagliardi
|
b918496ae4
|
Filled in the standard window procedure.
|
2014-02-11 18:22:39 -05:00 |
|
Pietro Gagliardi
|
c6a8a4d2f7
|
(see previous commit; part 2)
|
2014-02-11 17:39:41 -05:00 |