Pietro Gagliardi
|
b97e6b864f
|
Fixed 32-bit Mac OS X builds: Area crashed due to random ABI changes. Also more TODOs.
|
2014-04-04 22:51:30 -04:00 |
Pietro Gagliardi
|
8eee2a92b7
|
Changed the way drawRect: was being added to our Mac OS X Area such that we no longer need a separate function to add it to the class.
|
2014-04-04 22:21:53 -04:00 |
Pietro Gagliardi
|
8c109a0ae9
|
Updated the Area drawRect: adding error message to look like the ones that come out of the new class system.
|
2014-04-04 21:35:14 -04:00 |
Pietro Gagliardi
|
7ea493873c
|
Changed the Area class on Mac OS X to use the new class creation code. This gets rid of ALL the class creation code from delegate_darwin.go.
|
2014-04-04 21:32:10 -04:00 |
Pietro Gagliardi
|
dd20d56f25
|
Created a new, simpler, easier to work with, and consistent way to create the necessary Objective-C classes in our Go code, and converted the appDelegate class to use it. Now for goArea...
|
2014-04-04 21:06:01 -04:00 |
Pietro Gagliardi
|
9488824240
|
Wrapper functions are no longer used (see the past few commits), so that comment was removed from bleh_darwin.m. That comment also had the wrong filename all this time (bleh_darwin.go instead of objc_darwin.go).
|
2014-04-04 20:36:19 -04:00 |
Pietro Gagliardi
|
e35457b968
|
Removed objc_alloc(). This should be all of them. Now to redo how classes are handled...
|
2014-04-04 20:34:35 -04:00 |
Pietro Gagliardi
|
5a5b486b6a
|
Removed objc_msgSend_rect(). This also consolidates all initWithFrame: calls into a simple initWithDummyFrame().
|
2014-04-04 19:51:59 -04:00 |
Pietro Gagliardi
|
dd4a7b9cbe
|
Removed objc_msgSend_rect_uint_uint_bool().
|
2014-04-04 19:14:03 -04:00 |
Pietro Gagliardi
|
461b0cc023
|
Removed objc_msgSend_rect_bool().
|
2014-04-04 19:08:25 -04:00 |
Pietro Gagliardi
|
ba2c1d6d20
|
Removed objc_msgSend_uint().
|
2014-04-04 18:56:37 -04:00 |
Pietro Gagliardi
|
bf1e7aae42
|
More TODOs.
|
2014-04-04 18:39:06 -04:00 |
Pietro Gagliardi
|
ef43865923
|
Removed objc_release().
|
2014-04-04 18:08:23 -04:00 |
Pietro Gagliardi
|
fbde220ab1
|
Added a TODO related to the previous commit.
|
2014-04-04 18:00:22 -04:00 |
Pietro Gagliardi
|
1b05312da9
|
Mac OS X indeterminate progresbars were not animating under certain conditions; fix that.
|
2014-04-04 17:59:23 -04:00 |
Pietro Gagliardi
|
3a130b756a
|
Removed objc_new(). The idea is to remove all the objc_* functions that aren't prefixed with C. (except objc_getClass()).
|
2014-04-04 17:50:27 -04:00 |
Pietro Gagliardi
|
1a7bf4cd89
|
Got rid of objc_addDelegate(); it was only called in one place (Mac OS X Window creation). Also forgot to mention that the previous commit also disables mouse movements in Cocoa Areas for now; I need to replace the current code with one that uses tracking because mouse movements are system-global.
|
2014-04-04 15:51:40 -04:00 |
Pietro Gagliardi
|
be668a965a
|
Fixed Mac OS X Button appearance; turns out it was using the wrong font. Also updated the TODO file to mark this fix and remove other stale/already fixed details.
|
2014-04-04 14:52:38 -04:00 |
Pietro Gagliardi
|
2884d45f0f
|
Fixed mouse move events on Mac OS X not returning buttons right, I think: turns out I was sending the wrong message in. Now I have to figure out how to filter out mouse move messages; then I can really make sure this works right...
|
2014-04-03 23:52:26 -04:00 |
Pietro Gagliardi
|
2e617611c5
|
Adjusted Button and Combobox sizes on Windows to be better...
|
2014-04-03 21:00:38 -04:00 |
Pietro Gagliardi
|
89e3afc413
|
Added question about Windows proper sizes... for now...
|
2014-04-03 19:19:19 -04:00 |
Pietro Gagliardi
|
2e72ab689a
|
Updated TODO on dragging in an Area (decided to go with what GTK+ and Windows currently do; dragging was working on Windows all along; now dragging is weird on Mac).
|
2014-04-03 18:33:57 -04:00 |
Pietro Gagliardi
|
01d9e1b02b
|
Enabled mouse move events on Mac OS X. Mosue events now just got stranger and buggier there...
|
2014-04-03 18:33:37 -04:00 |
Pietro Gagliardi
|
773668c35b
|
Fixed the major scrolling corruption on Windows: had to commit the scroll changes before we can get to them in our painting routine =P This doesn't fix ALL the issues; check the (now-updated) TODO.
|
2014-04-02 23:40:38 -04:00 |
Pietro Gagliardi
|
a325e9e404
|
More TODOs.
|
2014-04-02 21:56:45 -04:00 |
Pietro Gagliardi
|
07c8aab1e6
|
More TODOs.
|
2014-04-02 19:59:39 -04:00 |
Pietro Gagliardi
|
4ac5290771
|
Fixed Listbox and LineEdit borders on Windows. Also more TODOs.
|
2014-04-02 19:57:49 -04:00 |
Pietro Gagliardi
|
77b5afe824
|
More TODOs.
|
2014-04-02 13:48:49 -04:00 |
Pietro Gagliardi
|
150b999c0a
|
Forced Common Controls version 6 on Windows by embedding the needed manifest in.
|
2014-04-02 13:47:11 -04:00 |
Pietro Gagliardi
|
43311a668f
|
Switched to calling the Windows MulDiv() function instead of reimplementing it ourselves in prefsize_windows.go.
|
2014-04-02 10:43:28 -04:00 |
Pietro Gagliardi
|
fa880a71cc
|
Added a println() to the main test window to signal that window closing has happened. This is used for ensuring that closing programs from a task switcher works as expected (that is, that the appropriate w.Closing event is triggered if an individual window is asked to be closed, and something else if an application as a whole is...).
|
2014-04-02 10:27:41 -04:00 |
Pietro Gagliardi
|
d1c0b9b29d
|
Not using PostThreadMessage() (a few commits back) seems to fix the weird Invalid thread ID. error that happens when running package ui-based programs in wine for the first time (I assume that issue is related to the one that led to the PostThreadMessage() fix).
|
2014-04-02 10:17:24 -04:00 |
Pietro Gagliardi
|
973853c835
|
Fixed odd Listbox sizing on Windows. (Windows by default sizes List Box controls so that partial items are not visible.)
|
2014-04-02 10:15:04 -04:00 |
Pietro Gagliardi
|
3fe4c2f113
|
Control candidates document updates.
|
2014-04-01 21:34:53 -04:00 |
Pietro Gagliardi
|
cbf439223b
|
Changed the font loading on Windows to always load the control font unless a flag to not load any font is set; when menus and statusbars are added, they need special casing. This also means the preferred size calculations reads controlFont directly (it will need the same change).
|
2014-04-01 21:24:20 -04:00 |
Pietro Gagliardi
|
8cb3991ef9
|
Fixed Windows control sizing: turns out simply calling GetTextMetrics() was not enough, as the GetDC() functions don't load the control font into the DC; we have to do it ourselves with SelectObject() (according to GetTextMetrics()'s docs on MSDN). Upon re-evaluation, the only things that need custom fonts are menus and statusbars; I don't know if menus can be done with the standard contorls and statusbars change the font of all controls inside... so how fonts are handled in classData needs to change now.
|
2014-04-01 21:17:27 -04:00 |
Pietro Gagliardi
|
af95232589
|
More TODOs.
|
2014-04-01 20:01:38 -04:00 |
Pietro Gagliardi
|
76de7db2c0
|
Remove TODO about GTK+ and Cocoa errors since we're just assuming errors are mistakes I made in the implementation because at this point that's really what we're left with since on Cocoa doing manual exception handling is discouraged... or something like that.
|
2014-04-01 19:27:45 -04:00 |
Pietro Gagliardi
|
df087d0ea5
|
Fixed the Windows stalling when scrolling an Area problem: messages were getting lost because (I assume) scrolling is done in a modal event loop. Alas, we have to change to using a dummy window to handle events. Also altered the respective TODO to document bugs now exposed by this fix.
|
2014-04-01 19:10:30 -04:00 |
Pietro Gagliardi
|
2efda7bc39
|
Separated initial text from sysData.make() on the Mac OS X backend, completing that set of commits.
|
2014-04-01 17:05:16 -04:00 |
Pietro Gagliardi
|
7a079f1cbb
|
Made the sysData.make() change from the previous commit on the Windows backend.
|
2014-04-01 16:47:36 -04:00 |
Pietro Gagliardi
|
2647d28f2b
|
Removed the initText parameter from sysData.make() and changed all invocations to call sysData.setText() separately; this avoids the need to check if sysData.setText() is valid. Also implemented that on GTK+.
|
2014-04-01 16:43:56 -04:00 |
Pietro Gagliardi
|
81153617dc
|
Got rid of gtkWidget and changed to using *C.GtkWidget everywhere. This gets rid of some of the casting functions and invocations of said functions, and only adds sysdata_unix.go to the cgo list.
|
2014-04-01 16:30:38 -04:00 |
Pietro Gagliardi
|
a003ff62d5
|
Removed winerrors.md because I screwed it ages ago.
|
2014-04-01 16:04:23 -04:00 |
Pietro Gagliardi
|
14770e890a
|
Moved <stdlib.h> includes in the GTK+ backend to gtk_unix.h; organized that file a bit.
|
2014-04-01 16:01:49 -04:00 |
Pietro Gagliardi
|
854ec2d17c
|
More TODOs.
|
2014-04-01 15:54:55 -04:00 |
Pietro Gagliardi
|
dbbabf35c3
|
Changed the cSysData dummy functions to an interface that a dummy sysData instance is tested against to make sure that not only all functions exist, but also that they are all the correct type.
|
2014-04-01 15:34:51 -04:00 |
Pietro Gagliardi
|
593ded848b
|
Removed a stray TODO comment whose purpose is long forgotten.
|
2014-04-01 15:16:29 -04:00 |
Pietro Gagliardi
|
b2f0fe1956
|
Reordered each part of the Windows uitask so that it does things in the same order and with the same goroutine setup as the other platforms; this gets rid of a few channels. Also panics on more errors (gets rid of a few more channels) and removed some stray TODOs.
|
2014-04-01 15:14:57 -04:00 |
Pietro Gagliardi
|
acc0f72379
|
Added another row to the open/save dialog test table.
|
2014-04-01 12:52:13 -04:00 |