[ Thread Index |
Date Index
| More lists.liballeg.org/allegro-developers Archives
]
On 2011-01-24, Matthew Leverton <meffer@xxxxxxxxxx> wrote:
> On Sun, Jan 23, 2011 at 6:37 PM, Peter Wang <novalazy@xxxxxxxxxx> wrote:
> >
> > Anything else?
> >
> I have a few things I've jotted down while using Allegro:
>
> * al_set_new_display_adapter() and al_get_monitor_info() could return
> false on out of range instead of (I assume) crashing. These aren't
> going to be performance sensitive functions, so I don't see any harm
> in Allegro doing the check.
Okay.
> * al_set_new_display_adapter() could use something like
> ALLEGRO_DEFAULT_DISPLAY_ADAPTER instead of the magic -1. (There
> doesn't seem to be many other hard coded magic numbers.)
Okay.
> * seems like const is missing in a lot of places. [e.g., the display
> parameter in al_get_display_height()] Probably could be updated in
> 5.2.
Not really important.
> * native image loader is buggy (unusable) in OS X due to 10.4 or PPC
> (or the combination). Allegro's routines, libpng, and libjpeg work
> though, so this could simply be a documentation thing.
Can you try rounding up the stride of the pixel array in the call to
CGBitmapContextCreate, to a multiple of 16 bytes, or even power-of-two?
The last image you posted on a.cc certainly looks like some sort of
alignment problem.
Peter