🚀 go-pugleaf

In article <1992Nov20.182445.9500@jarvis.csri.toronto.edu>  
west@turing.toronto.edu (Tom West) writes:
>>  A more general implementation question from a tyro NeXT programmer.
>>
>>  I am porting over an interpreter to the NeXT.  When execution is
>>called for, the program calls the execution module.  It displays
>>output to its output window and occasionally asks for input from
>>the same window.  Or at least, it should.  
>>
>>  I am at a loss as to how I am supposed to:
>>	(1) Have the rest of the system keep responding
>>	(2) Have the execution wait for key input.
>>
>>  Rather kindly, the execution process does come up for air fairly regularly.
>>Looks like: 	do
>>	        {
>>		    running = Execute ();
>>		} while (running != FINISHED)
>>
>>  On other systems, I just insert a call to the main event loop in the
>>do loop and the application is able (with a little help) to process
>>events.  Is there any equivalent to this on the NeXT?
>>
>>  Is creating a MACH thread my only alternative?  That's going to require
>>much more programming overhead than I would like (to say nothing of the
>>fact that I would have to learn all about threads in about 2 days :-)).  
>>Is there a simple method of doing all of this that I've missed?
>>
>>  Many thanks in advance.
>>-- 
>>
>>	Tom West		
>>
>>   west@turing.toronto.edu  or			"Children of the Blood"
>>tomwest@gpu.utcs.utoronto.ca  or		  -Michelle Sagara
>>       west@hsa.on.ca

Just a shot in the dark but...What about using the Subprocess class ? (see 
/NeXTDeveloper/Examples/Unix/Subprocess)  I used it to talk to a Mathematica  
kernel via an external C program (supplied with Mma).  Super easy to use.

--
Ronald Pomeroy		[Objective-Cruntime: [Smalltalk runtime]];
Advanced Micro Devices	[self dream-on];
CAM Applications Group
rpomeroy@aunext1.amd.com (NeXTmail preferred)

west@turing.toronto.edu (Tom West) writes:
>   I am porting over an interpreter to the NeXT.  (...)
>   Rather kindly, the execution process does come up for air fairly regularly.
> Looks like: 	do
> 	        {
> 		    running = Execute ();
> 		} while (running != FINISHED)

In the past you have been simply polling the execution loop from the event  
loop.  The event loop is rather more hidden under NeXTstep, but you can get the  
same effect by installing a timed entry that runs the Execute() loop.

Timed entries happen "between" the dispatch of normal events (like keypresses).   
You must specify the interval at which you want to invoke the function  
(presumably a very small value in your case).  One thing to be aware of is that  
timed entries normally can go off any time you ask for an event (like "what is  
the next key?")  A timed entry will never directly reenter the handler function  
(that is, the function is locked out until the timed entry returns) but it  
could happen if you call the Execute() handler responding to a key-down event  
and you look for another event.  If you need to do this, a simple semaphore  
will keep you happy.  Oh yeah, the documentation is under DPSAddTimedEntry().

-sam

--
Opinions expressed herein are not those of my employer.  They're not even
mine.  They're probably wrong besides.  How did they get in here, anyway?