tuxnes-devel

[Tuxnes-devel] A fue minor patches too 0.75.

[Mike M. <che...@ya...>]

Today when I built 0.75 I had to add AM_PROG_AS to configure.in.  There
was several AC_PROG lines and I was not sure witch one I should use AC or
AM.

There was also a race condition when the emulator was paused, I'l have to
look into adding a select for this.  It shoulden't be a big deal to just
halt execution inside the input handeler should it?
For CVS HEAD I'd also like to see if calling a generic input(alarm signal)
handeler(or a timer) for emulator type
events(pause,save,snapshoot,quit,ect) so were not relying on the nes rom
to pool these for us.



		
__________________________________
Do you Yahoo!?
Yahoo! Mail - Helps protect you from nasty viruses.
http://promotions.yahoo.com/new_mail

Re: [Tuxnes-devel] A fue minor patches too 0.75.

[Jason D. S. <jd...@us...>]

Mike Mestnik wrote:
> Today when I built 0.75 I had to add AM_PROG_AS to configure.in.  There
> was several AC_PROG lines and I was not sure witch one I should use AC or
> AM.
> 
> There was also a race condition when the emulator was paused, I'l have to
> look into adding a select for this.  It shoulden't be a big deal to just
> halt execution inside the input handeler should it?
> For CVS HEAD I'd also like to see if calling a generic input(alarm signal)
> handeler(or a timer) for emulator type
> events(pause,save,snapshoot,quit,ect) so were not relying on the nes rom
> to pool these for us.

It needs to pause while waiting for an "event".

An "event" may be input from keyboard or mouse or any arbitrary X (or 
whatever GUI) event.

Using a select loop _may_ work here since select will exit when there's 
an interrupt/signal received (but I don't know if the above events 
generate interrupts).  In that case all FD sets should be NULL.

jason

Re: [Tuxnes-devel] A fue minor patches too 0.75.

[Mike M. <che...@ya...>]

That's correct, I just used usleep for finding when the event might end. 
This way where not constintaly looking for the event, just once every time
frame.

While forward porting this too CVS HEAD I saw that non of the working
renderes still had this code.  This could be the clue I needed to get GGI
working, since it still has this code.

--- Jason Dorje Short <jd...@us...> wrote:

> Mike Mestnik wrote:
> > Today when I built 0.75 I had to add AM_PROG_AS to configure.in. 
> There
> > was several AC_PROG lines and I was not sure witch one I should use AC
> or
> > AM.
> > 
> > There was also a race condition when the emulator was paused, I'l have
> to
> > look into adding a select for this.  It shoulden't be a big deal to
> just
> > halt execution inside the input handeler should it?
> > For CVS HEAD I'd also like to see if calling a generic input(alarm
> signal)
> > handeler(or a timer) for emulator type
> > events(pause,save,snapshoot,quit,ect) so were not relying on the nes
> rom
> > to pool these for us.
> 
> It needs to pause while waiting for an "event".
> 
> An "event" may be input from keyboard or mouse or any arbitrary X (or 
> whatever GUI) event.
> 
> Using a select loop _may_ work here since select will exit when there's 
> an interrupt/signal received (but I don't know if the above events 
> generate interrupts).  In that case all FD sets should be NULL.
> 
> jason
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click
> _______________________________________________
> Tuxnes-devel mailing list
> Tux...@li...
> https://lists.sourceforge.net/lists/listinfo/tuxnes-devel
> 



		
__________________________________
Do you Yahoo!?
New and Improved Yahoo! Mail - Send 10MB messages!
http://promotions.yahoo.com/new_mail 

Re: [Tuxnes-devel] A fue minor patches too 0.75.

[Jason D. S. <jd...@us...>]

Mike Mestnik wrote:
> That's correct, I just used usleep for finding when the event might end. 
> This way where not constintaly looking for the event, just once every time
> frame.
> 
> While forward porting this too CVS HEAD I saw that non of the working
> renderes still had this code.  This could be the clue I needed to get GGI
> working, since it still has this code.

Well, you could replace it with a select(NULL, NULL, NULL, &tv), where 
tv holds the polling time.  The poll time doesn't have to correspond to 
a frame, though.  Using select _might_ give better results (lower lag) 
since it will exit if there's an interrupt.  But in most cases it 
probably doesn't matter.

Looking at the internals of a function like gtk_main() may show a better 
way to do this.  Most GUI libraries have code like this already.

jason

[Tuxnes-devel] How to pause with minimal effect on seti or distributed.net?

[Mike M. <che...@ya...>]

After looking into this I wonder how the current pause code may even work
at all, it's if'ed out in places.  As for select vs useep this is
imposible as we need to call XPending to get events and it's vary unlikely
we will receve any signals(interupts) since among other things we don't
have any children.

I'm going to run some ddd/gdb tests.

--- Jason Dorje Short <jd...@us...> wrote:

> Mike Mestnik wrote:
> > That's correct, I just used usleep for finding when the event might
> end. 
> > This way where not constintaly looking for the event, just once every
> time
> > frame.
> > 
> > While forward porting this too CVS HEAD I saw that non of the working
> > renderes still had this code.  This could be the clue I needed to get
> GGI
> > working, since it still has this code.
> 
> Well, you could replace it with a select(NULL, NULL, NULL, &tv), where 
> tv holds the polling time.  The poll time doesn't have to correspond to 
> a frame, though.  Using select _might_ give better results (lower lag) 
> since it will exit if there's an interrupt.  But in most cases it 
> probably doesn't matter.
> 
> Looking at the internals of a function like gtk_main() may show a better
> 
> way to do this.  Most GUI libraries have code like this already.
> 
> jason
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=4721&alloc_id=10040&op=click
> _______________________________________________
> Tuxnes-devel mailing list
> Tux...@li...
> https://lists.sourceforge.net/lists/listinfo/tuxnes-devel
> 



		
__________________________________
Do you Yahoo!?
Yahoo! Mail - You care about security. So do we.
http://promotions.yahoo.com/new_mail

Re: [Tuxnes-devel] How to pause with minimal effect on seti or distributed.net?

[Jason D. S. <jd...@us...>]

Mike Mestnik wrote:
> After looking into this I wonder how the current pause code may even work
> at all, it's if'ed out in places.  As for select vs useep this is
> imposible as we need to call XPending to get events and it's vary unlikely
> we will receve any signals(interupts) since among other things we don't
> have any children.
> 
> I'm going to run some ddd/gdb tests.

Last time I looked it definitely didn't work.

jason