Pietro Gagliardi
|
40dd4f51be
|
Removed constants from events_windows.go. This file can go away now...
|
2014-05-25 14:07:34 -04:00 |
|
Pietro Gagliardi
|
ad1dbc7b39
|
Removed extra newline generated in the zconstants_windows_GOARCH.go file from the Windows constants generator.
|
2014-05-25 14:05:23 -04:00 |
|
Pietro Gagliardi
|
1f94a68432
|
Fixed issues handling INVALID_HANDLE_VALUE in the Windows constant generator, and built the first build with generated constants!
|
2014-05-25 14:04:03 -04:00 |
|
Pietro Gagliardi
|
d2746b7d2c
|
Added a comment to the previous commit.
|
2014-05-25 13:20:19 -04:00 |
|
Pietro Gagliardi
|
a0f10000de
|
Added the ability to pass arguments to the go tool from the Windows constant generator's command line; this will be for debugging.
|
2014-05-25 13:18:45 -04:00 |
|
Pietro Gagliardi
|
9893387848
|
Changed the logic of the Windows constants generator so that it only generates unknown constants and complains on known constants (though that's disabled during the transitional period) as that's the situation that will happen now.
|
2014-05-25 13:02:56 -04:00 |
|
Pietro Gagliardi
|
339a77b5d6
|
Removed constants from windows_windows.go, controls_windows.go, and comctl_windows.go, as we can now use the constants generator.
|
2014-05-25 12:39:53 -04:00 |
|
Pietro Gagliardi
|
2ce95f3937
|
Removed CW_USEDEFAULT variable definition. It won't compile right now, but our cgo-based generator will make sure that isn't a problem.
|
2014-05-25 12:31:38 -04:00 |
|
Pietro Gagliardi
|
5583dee743
|
Updated the Windows constants generator to warn against variables that haven't been converted to constants (in case I forget I ever did this).
|
2014-05-25 11:35:55 -04:00 |
|
Pietro Gagliardi
|
eb0188a099
|
Changed most instances of var to const in the Windows code. Only one left...
|
2014-05-25 11:31:57 -04:00 |
|
Pietro Gagliardi
|
0d23bda925
|
Added quick credit thanks left out of previous commit by accident.
|
2014-05-25 03:15:49 -04:00 |
|
Pietro Gagliardi
|
36e6adcee9
|
Updated the Windows constant generator again; this time it now works fully, and I just have to nail the variables and nuke the static constants...
|
2014-05-25 03:13:49 -04:00 |
|
Pietro Gagliardi
|
3929dea818
|
Even more work on the Windows constant generator. There are still some edge cases involving variables that I need to resolve, and I still need to write the code that actually runs the generated program.
|
2014-05-25 01:28:59 -04:00 |
|
Pietro Gagliardi
|
8de17f0ade
|
Refactored the constant gathering program a bit.
|
2014-05-25 01:13:16 -04:00 |
|
Pietro Gagliardi
|
b27caae7c3
|
Renamed _LPARAM._X()/_LPARAM._Y() to _LPARAM.X()/_LPARAM.Y(): it doesn't matter since _LPARAM is not exported anyway, _WPARAM already had this change, and the constant checking tool complains that _X()/_Y() remain undefined (too early? do I need to merge the package files?).
|
2014-05-25 01:03:45 -04:00 |
|
Pietro Gagliardi
|
4408c1bf42
|
More work on the constant generation tool. We can almost actually write the generator portion now...
|
2014-05-25 00:40:45 -04:00 |
|
Pietro Gagliardi
|
4632d80576
|
More work on the constant generation tool.
|
2014-05-25 00:31:40 -04:00 |
|
Pietro Gagliardi
|
20306d881c
|
Added incomplete code for a program that generates a set of Windows constants using cgo.
|
2014-05-24 23:44:28 -04:00 |
|
Pietro Gagliardi
|
2c2d2dc0c8
|
More TODOs.
|
2014-05-24 21:30:21 -04:00 |
|
Pietro Gagliardi
|
32480a65c2
|
Resolved GtkEntry and GtkProgressBar resizing issues; decided from observing how they do padding that worrying about padding is counterproductive for now.
|
2014-05-24 21:28:28 -04:00 |
|
Pietro Gagliardi
|
511f774649
|
Finished the Area clicking stuff: formalized the behavior in documentation in area.go, pointed out that Mac OS X already behaves as expected, updated some comments, and removed the relevant TODO.
|
2014-05-23 23:48:17 -04:00 |
|
Pietro Gagliardi
|
f4afa2a951
|
Attempted to add focus change support to the new GTK+ high-order click code like we do on Windows; it isn't working quite yet but I'll mark it as a TODO for now.
|
2014-05-23 23:38:14 -04:00 |
|
Pietro Gagliardi
|
dbeccfa922
|
Added custom double-click/triple-click/higher-order click handling to GTK+ Areas.
|
2014-05-23 18:30:08 -04:00 |
|
Pietro Gagliardi
|
b359f995f5
|
Added click counter reset on window activate/deactivate for Windows Areas.
|
2014-05-23 17:11:59 -04:00 |
|
Pietro Gagliardi
|
582e3b4644
|
Implemented triple-clicks and up on Windows.
|
2014-05-23 17:01:56 -04:00 |
|
Pietro Gagliardi
|
92e4d32284
|
Added code that will replace the toolkit-internal double-click/triple-click handling in an effort to normalize click count handling.
|
2014-05-23 14:03:57 -04:00 |
|
Pietro Gagliardi
|
63ae00622a
|
More TODOs about mouse button click counts. The GTK+ implementation might be the killer here...
|
2014-05-23 13:00:19 -04:00 |
|
Pietro Gagliardi
|
ec2749d46c
|
Updated some TODOs (part 2 of yesterday's The Old New Thing post).
|
2014-05-22 23:23:10 -04:00 |
|
Pietro Gagliardi
|
0b609178ff
|
More of the previous commit.
|
2014-05-22 04:01:35 -04:00 |
|
Pietro Gagliardi
|
77d861354e
|
More KeyEvent Modifiers designing in the TODO file.
|
2014-05-22 03:59:19 -04:00 |
|
Pietro Gagliardi
|
6a07a06cb8
|
Updated TODO on modifier keys handling in KeyEvent.
|
2014-05-22 03:44:25 -04:00 |
|
Pietro Gagliardi
|
e2ab537474
|
Updated a TODO with a The Old New Thing blog post link posted today (!) that answers the question the TODO asks.
|
2014-05-21 23:19:55 -04:00 |
|
Pietro Gagliardi
|
8c386274e9
|
Removed TODO about pixel alignment of scrollviews on Mac OS X; mikeash in irc.freenode.net/#macdev told me this only happens when using fractional pixels (because Mac OS X lets you use fractional coordinates and sizes), so we don't have an issue since all our coordinates are integers.
|
2014-05-21 18:44:20 -04:00 |
|
Pietro Gagliardi
|
5b0f8e8cc7
|
Updated TODOs after confirming that Windows and Mac do have MouseEvent.Held[] sorted.
|
2014-05-20 12:32:05 -04:00 |
|
Pietro Gagliardi
|
c01549420d
|
Removed 10.6 build issues from TODO now that that's done.
|
2014-05-20 09:03:44 -04:00 |
|
Pietro Gagliardi
|
989ac438fc
|
Fixed the whole NSApplication.h thing so now the project builds with both 10.6 and 10.8. Ugh.
|
2014-05-20 08:35:39 -04:00 |
|
Pietro Gagliardi
|
c6d24f0e32
|
Changed #include to #import on the Objective-C files to settle the whole include guards thing. Now to continue fixing Mac OS X 10.6 building...
|
2014-05-20 08:08:55 -04:00 |
|
Pietro Gagliardi
|
3e9e767d79
|
Fixed the Mac OS X 10.6 building using a rather hackish approach... Now it won't build on 10.6 itself without problems, so...
|
2014-05-20 07:46:57 -04:00 |
|
Pietro Gagliardi
|
5f5f4f6c2a
|
Set up the Mac OS X build to always target Mac OS X 10.6 regardless of the host system; this is what we want and by default building on 10.8 produces binaries which won't run on 10.6. This isn't working yet, though: one enum in NSApplication.h becomes empty on 10.6...
|
2014-05-19 23:53:10 -04:00 |
|
Pietro Gagliardi
|
946e007e5d
|
Same as previous commit, but for Mac OS X now. The #cgo directives are in uitask_darwin.go.
|
2014-05-19 22:52:59 -04:00 |
|
Pietro Gagliardi
|
a48dd80add
|
Removed the extra #cgo directives from the Unix files. cgo collects directives from ALL files, rather than using each file's directive individually, so having them combined like this is pointless. Instead, the #cgo directives are (or is, since there's only one) in uitask_unix.go. Will do the same for Mac next; this will help since we're going to add a CFLAGS value afterward.
|
2014-05-19 22:32:38 -04:00 |
|
Pietro Gagliardi
|
4e820e86fc
|
Moved gtk_main() such that uitask_unix.go calls it directly via cgo. This is in preparation for the next two commits, which will remove the #cgo directives from all files except the uitask_*.go files, since they're being concatenated across files.
|
2014-05-19 22:18:53 -04:00 |
|
Pietro Gagliardi
|
7b7005ec19
|
TODO file update.
|
2014-05-19 00:11:50 -04:00 |
|
Pietro Gagliardi
|
cadbc174e3
|
Removed TODO about buttons 1+3 causing a button 2 click on Mac OS X Areas... it was actually on my host's end: X11/KDE is doing it, and I never noticed that before, nor do I remember that happening on the Windows/wine or GTK+ versions :S That's it for the TODOs in the Mac OS X Go files! :D
|
2014-05-18 12:53:15 -04:00 |
|
Pietro Gagliardi
|
216b5c183c
|
Made drawing of NSScrollView backgrounds explicit on Mac OS X. This is what Interface Builder sets for NSTableViews, so we do that too for Listbox. This also means NSScrollView draws the background for our Areas, settling the clear clip rect TODO.
|
2014-05-17 20:21:48 -04:00 |
|
Pietro Gagliardi
|
b9c12da838
|
More TODOs (see previous commit).
|
2014-05-17 17:10:23 -04:00 |
|
Pietro Gagliardi
|
24cac41de6
|
Decided to unconditionally redraw everything on a window resize on Mac OS X. Now to add TODOs for the other platforms to see if we should do the same.
|
2014-05-17 17:07:53 -04:00 |
|
Pietro Gagliardi
|
ce14390eff
|
Fixed variable names in Area drawing code on Unix to reflect that image.Rectangle is point-to-point, not origin/size like on Mac OS X. Windows was already fine; it just uses "left, top, right, and bottom" instead of "x0, y0, x1, and y1".
|
2014-05-17 17:01:52 -04:00 |
|
Pietro Gagliardi
|
352c60bf68
|
More clarification on the previous commit.
|
2014-05-17 16:21:25 -04:00 |
|
Pietro Gagliardi
|
eca42bc152
|
Removed TODO about not using bad-for-tracking features of NSEvent in Areas on Mac OS X since that's not relevant to our use of tracking areas.
|
2014-05-17 16:19:42 -04:00 |