Re: [AD] Status of A5 on OS X (ppc)

[Evert Glebbeek <eglebbk@xxxxxxxxxx>]

> I'll do the same for intel later. I think the situation is similar but
> maybe a little better.

Things are actually a lot better now than when I send that message.  
Maybe next time I should do an evening of debugging first and then  
write a report on what does and doesn't work. ;)
One of the advantages of just having moved continents: my old social  
life is on hold (all my friends are asleep in the evenings so I have  
no one to chat to) and I don't have a new local one yet, so my  
evenings are more or less free to work on Allegro!
I'm sure it'll change.

> Some things don't work because they aren't
> implemented - fullscreen resize is one, and responding to events on a
> frameless window is another (can be done, just need to subclass
> NSWindow and implement a function there).

Ok. Fullscreen resize is probably something I can implement when I  
look into it. I feel less sure about frameless windows, mainly  
because I'm also still getting used to the NSView/NSWindow interface.  
I don't know, maybe it's easier than I think it will be.

> On the cmake/demo thing - can someone else volunteer for that?

Already done.

> In the past I'm sure that mouse cursors were OK on multiple windows
> (not changing them, just the fact that they were visible or not) but I
> noticed it didn't seem to be working in ex_mouse_cursor yesterday.

I added a dpy->show_cursor = YES; in the function that sets up the  
window and it's fine now.
On a side note, it is a bit confusing that the OS X port stores  
whether the mouse should be shown, whereas the X11 port stores  
whether the mouse should be hidden. It doesn't really matter of  
course, but it's mildly annoying when looking at the two ports side  
by side.

> For all platforms, could we start putting together a new 'test'
> program that can be used to test all the features are running properly
> (and benchmarks in the future)?

That's a good idea, actually.

Evert