Re: widget port

new topic     » topic index » view thread      » older message » newer message

Mike Sabal wrote:

>I've hit a major snag in porting the Widgets.  Converting all the
>machine_proc calls to their graphics.e equiv's has been no trouble.

There are only very few machine_proc calls in widgets, basically just
to remove one level of indirection in critical routines.

>And removing the assembler codes for the rom font only needed a few
>comment lines.

As far as I remember there is no assembly in the font module. So I am
not quite sure what you were commenting on or out.

>However, the "rat" logic is giving me nothing but headaches.  Is
there
>any reason you didn't use the mouse.e routines or other such
libraries,
>Jiri;

Just four:

    1. I like to roll my own, but only if I can do it smarter.
    2. I want my cursor transparent, no-one else does it.
    3. When I say 'show', I want my cursors displayed, and when I say
       'hide', I expect them to be hidden, irrespective of how many
times
       I said 'show' anywhere else in the murky depths of my code...
    4. No other set of mouse routines works reliably in SVGA
        modes, with possible exception of Pete's 'Mighty' or whatever
is
        the name he gave it, and from which I stole a trick.

>and is there any way to rework it without all the DOS interrupts?
I'm
>a bit stumped.

I don't know. Don't you have interrupts under Linux, that's a bit
rough?!

jiri

new topic     » topic index » view thread      » older message » newer message

Search



Quick Links

User menu

Not signed in.

Misc Menu