Re: [AD] Using system mouse cursor

[Chris <kcat@xxxxxxxxxx>]

Evert Glebbeek wrote:
> This is just the current Allegro (WIP 4.1.15) behavior - in other words, 
> the patch changes nothing if the software cursor is in use.

Which is the problem. I thought we were trying to get away from having 
an implicit "mickey mode" for the mouse by making it explicit via 
enabling/disabling the hardware cursor.

> Redesigning
> that is IMO a seperate issue from adding the hardware cursor.

They're going to end up being connected, though.. since the hardware 
cursor can't coexist with proper mouse mickeys (if you have a hardware 
cursor, you can't have proper mouse mickeys, and vice versa.. in X at 
the very least).

> Then, if the software cursor is in use by default, you have an ill-behaved 
> windowed application. This would only be possible, I think, if the 
> software cursor is not the default cursor.

If the software cursor is default, you'll create an app that locks the 
mouse to the window by default (when using windowed modes.. fullscreen 
is completely unaffected) which can be undone by enabling the hardware 
cursor (which doesn't even have to show anything.. show_mouse(NULL)).

If the hardware cursor is default, you'll slightly break the API by 
making get_mouse_mickeys not do what the docs say, which it was supposed 
to do this whole time (pre-4.0), without resorting to potentially buggy 
hacks.

> Possibly - although I'm not sure if some GUI based applications doesn't 
> check mouse mickeys from a custom procedure.

There can be limited mickeys with the hardware mouse.. just leave it 
undefined at the screen's edge. Although, perhaps the entire GUI may be 
a bad idea.. the file selector/alert boxes/etc should be fine though 
since Allegro has total control over those.

> Ok. Limiting the amount of time between warps could still be a good idea 
> though.

How would you do this without a(nother) timer?

- Kitty Cat