Re: [AD] Proposed changes for Allegro 5 (6?)

[Bob <ohannessian@xxxxxxxxxx>]

Stepan Roh wrote:

> On Tue, 27 Nov 2001, Bob wrote:
>
>
> > http://pages.infinit.net/voidstar/newalleg.txt
>
> Splitting packfiles handling to compression/encryption layers is good, but
> removing default implementation is bad, bad, bad :-) Why?


I just thought about lightening the lib. There's no reason really, and I'm 
not advocating it. I just thought I'd bring it up incase someone else would 
argument for it :)


> Adding networking is not good thing, I think. Too much new code without
> any connection to current Allegro.


The current Allegro has nearly no connection to Allegro 3.0 appart from most 
of the external API. Why is networking any different?


> Allegro 5 should have cleaner and consistent API with better decomposition
> to modules (unicode handling, gfx, sound, input, packfiles, etc.). This
> should be the main goal. And, of course, cleaning up the sources a bit.


Agreed.


> Generated API docs is very good idea, but forget Javadoc. It's too
> Java-centric and too slow.


Doxygen? They use the same syntax (AFAIK), so we could just document the 
code like this:

/** Draws a bitmap. (insert blit() docs)
 *
 *  \param source The source bitmap
 *  \param dest   The destination bitmap
 *    (and so on)
 *
 *  \sa draw_sprite masked_blit stretch_blit
 */
void blit(...) {
}


Then docs are in the sources, get updates with the sources, and get generated from the sources.
Docs get generated in html, latex, rich text, and some other formats I can't remember.



--
- Robert J Ohannessian
"Microsoft code is probably O(n^20)" (my CS prof)
http://pages.infinit.net/voidstar/