On Wed, 29 May 2002, Regis Nicolas wrote:
> > ------------------------------
> > Accessing the screen directly may also cause some bpp issues.
> > The new Treo 90 is apparently is 4-4-4 12bpp mode.
> > ------------------------------
>
> my treo 90 is on its way over (fedex) and, i should be able to tell
> you guys how it works when i get the sucker :P
>
> --------------------------------
> I can't wait longer...
> --------------------------------
tell fedex to get on their horses then *g*
> > ------------------------------
> > As Aaron said the OS itself is slower if you call SysHandleEvent too
> > often. Did you try calling EvtEventAvail and skip SysHandleEvent if none
> > available or are you calling EvtGetEvent with a (short) timeout value and
> > always calling SysHandleEvent ?
> > ------------------------------
>
> SysHandleEvent is a switch-case statement internally right? at least,
> it better be :P if it is a switch-case, it'll not be worth calling
> EvtEventAvail() :)
>
> --------------------------------
> More something like:
> if (xxx1)
> yyy1;
> if (xxx2)
> yyy2;
> if (event == nilEvent)
> zzz; // Not really useful actually - I can't say much
> switch (event)
> {
> }
eek.. who wrote that crap? :P [/me looks at os sources] - surely
this code could have been cleaned up a bit :P quite a few if-then-else
statements there :P
not to mention that SysHandleEvent() is a wrapper function :P the
timer code and interrupt handling is a must of course - but, heck..
oh well :P
at least put an else between penUpEvent/penDownEvent checks :P minor
speed up there - might produce some better code.. could have done a
switch statement for the event type.. would give cleaner and more
optimized code (if compiler is good).
// az
[EMAIL PROTECTED]
http://www.ardiri.com/
--
For information on using the Palm Developer Forums, or to unsubscribe, please see
http://www.palmos.com/dev/support/forums/
