kicad-developers team mailing list archive

[Chris Pavlina <pavlina.chris@xxxxxxxxx>]

Have any OS X devs had a chance to test this? I got a test from one, but
haven't been able to track down a followup to a potential regression.

I have a report that using <space> as a hotkey on OSX with this patch causes a
bell. Does someone have a chance to check that for me? Most importantly right
now, check whether that happens on an unpatched build - i.e., whether it is a
regression.

On Fri, Jan 22, 2016 at 04:04:44PM -0500, Chris Pavlina wrote:
> Updated patch - quite a few changes. First I noticed I was inadvertently 
> trapping some keys I shouldn't (try using Ctrl+S with the original 
> patch...), and then I decided to tackle the old "bug" that certain key 
> combinations, like Shift+Tab, cannot be used.
> 
> To make sure hotkeys are recognized equally everywhere, I took the code 
> I originally wrote for the hotkey editor widget to handle wxEVT_CHAR and 
> wxEVT_CHAR_HOOK keypress events "intelligently" and factored it out into 
> a separate class to be used in other places. I then pulled it into 
> EDA_DRAW_PANEL and EDA_DRAW_PANEL_GAL to replace the existing keycode 
> rewriting code (which originally was just "if hotkey is in WXK_CONTROL_A 
> through WXK_CONTROL_Z, remap to Ctrl + A..Z").
> 
> With this patch, all 'standard' combinations of keys on a US keyboard 
> should be usable, support for international keys is the same as before.
> Let me know if you find any combinations that don't work.
> 
> Please test! This is platform-dependent stuff - it's the same 
> platform-dependent stuff as before, but I need to make sure I haven't 
> added regressions. I've tested on Linux and Win10; more thorough testing 
> even on those same platforms is welcome.
> 
> 
> On Wed, Jan 20, 2016 at 10:44:36PM -0500, Chris Pavlina wrote:
> > There is an old bug that people turned up while testing my new hotkey 
> > editor, attached is a patch that fixes it. This patch pokes into the 
> > main hotkey handling code, so I really want as many people to test it as 
> > possible - it's a relatively minor bug, I don't want to go introducing 
> > twelve regressions to fix one small bug. Here's what I want to keep an 
> > eye out for:
> > 
> > - Hotkeys (Ctrl+Tab), (Tab), and (Ctrl+I) are all independent and 
> >   distinguished from each other, both in the hotkey editor and in actual 
> >   use. Make sure all of them work, and make sure none of them answers 
> >   for the others.
> > 
> > - Hotkeys that are *not* handled by the main hotkey code (for example, 
> >   menu keys like Alt+F to call up the File menu) are not affected.
> > 
> > - Behavior on OSX, with its weird Ctrl mapping, is not broken (or, at 
> >   least, is no more broken than usual ;)
> > 
> > I have tested on Linux and Windows 10, though more thorough testing on 
> > those platforms is welcome.
> > 
> > Patch/bug summary:
> > 
> > [PATCH] Fix (Ctrl)+(ASCII control key) hotkey handling
> > 
> > wxWidgets has quirks with how it handles these keys. For example, in
> > wxEVT_CHAR, Ctrl+Tab and Ctrl+I are indistinguishable.
> > 
> > - Modify the special wxEVT_CHAR_HOOK handler from WIDGET_HOTKEY_LIST to
> >   handle this case as well as the other funny cases it already handles.
> > 
> > - Factor this handler out into a function in hotkeys_basic.h for use
> >   elsewhere.
> > 
> > - Add this handler to the central hotkey handler, remove existing
> >   (buggy) ASCII control key handling.
> > 
> > Thanks for testing.
> > 
> > -- Chris