Posts: 2,020
Threads: 24
Joined: Jun 2002
hardly. you can always read the timer, and believe it or not, yes, you can clog memory and slow down your program with memory allocation, stack calls, string memory, etc. Again, it's minor but it's there. The system timer was put there for a reason.
i]"I know what you're thinking. Did he fire six shots or only five? Well, to tell you the truth, in all this excitement, I've kinda lost track myself. But being as this is a .44 Magnum ... you've got to ask yourself one question: 'Do I feel lucky?' Well, do ya punk?"[/i] - Dirty Harry
Posts: 1,752
Threads: 21
Joined: Jun 2002
You cannot always reprogram the timer to the resolution needed for sub-ms delays, because other programs need the timer. (A reliable TSR cannot reprogram the timer!)
Are you speaking from experience? It doesn't sound like it. *None* of the things you mentioned affect program speed in any way. The CPU cache and pipes are the only thing that affect the speed under pure DOS.
Posts: 2,020
Threads: 24
Joined: Jun 2002
You dont *need* to reprogram the timer at all. You can just refresh it to find the tick countdown to the next programmed cycle. add em' together and you have a reliable hi-res timer. You can do that with pure qb, TSRs or interrupt vectors or nothing. *forget* setting TSRs, I really dont care about that. I programmed using for loops for a very long time, and besides the fact that you can't reliably hold a frame rate with a delay, yes, it isnt a realiable delay at that.
sheeesh.
*looks up at thread* you know, this has kind of stretched into a two person rally. Let's just present some closing arguments and be done with it, 'k?
i]"I know what you're thinking. Did he fire six shots or only five? Well, to tell you the truth, in all this excitement, I've kinda lost track myself. But being as this is a .44 Magnum ... you've got to ask yourself one question: 'Do I feel lucky?' Well, do ya punk?"[/i] - Dirty Harry
Posts: 1,752
Threads: 21
Joined: Jun 2002
Again, you're making assumptions based on things you've never done. A TSR refreshing the timer will cause certain programs to lock up. I know because I've tried it. Have you?
The loop isn't to hold a "frame rate", it's to cause a delay. The CPU doesn't dynamically change speeds. Memory allocation has no effect on the speed of the loop. (Maybe you'd like to clarify what exactly a "stack call" is?)
The only reason this stretched into a rally is because you refuse to accept my statement that, in certain conditions, using a calibrated loop to delay a program is more accurate and more reliable than using the PIT. I know this is true from the experience of writing LCD drivers in C and assembly.
I agree that there are times to use the timer, and there are times to use loops. Since you've obviously never written any programs that required loops for a delay, maybe you can just take my word for it, instead of making up reasons why you think it doesn't work.
Posts: 2,020
Threads: 24
Joined: Jun 2002
Like I said before, foget' TSRs, forget all that. you *dont* need to use one. You can get the tick countdown until the timer refreshes, and that's enough to measure time at full 1,000,000 tick per second accuracy. Reprogramming the PIT is silly and unnecessary. Read it over. Trust me, it makes sense. Whatever. I'm tired. It's Friday. I'm done. I'm going out now. *whoosh*
i]"I know what you're thinking. Did he fire six shots or only five? Well, to tell you the truth, in all this excitement, I've kinda lost track myself. But being as this is a .44 Magnum ... you've got to ask yourself one question: 'Do I feel lucky?' Well, do ya punk?"[/i] - Dirty Harry
Posts: 3,279
Threads: 170
Joined: Nov 2003
Both have their purposes, both have their times to be used.
Stop bickering, you weirdos 
I'd knock on wood, but my desk is particle board.
Posts: 6,419
Threads: 74
Joined: Mar 2002
Posts: 1,752
Threads: 21
Joined: Jun 2002
Quote:Like I said before, foget' TSRs, forget all that. you *dont* need to use one. You can get the tick countdown until the timer refreshes, and that's enough to measure time at full 1,000,000 tick per second accuracy. Reprogramming the PIT is silly and unnecessary. Read it over. Trust me, it makes sense. Whatever. I'm tired. It's Friday. I'm done. I'm going out now. *whoosh*
I think you really are blind, man. I clearly stated when I needed to use loops to delay a program (a TSR, for that matter). And yet you still fail to listen...you don't think I tried getting the tick countdown? Guess what? It locked up with XTC-Play in the background. I wasn't reprogramming the PIT.
I also like how you sidestep my questions, make up theories, post that I'm wrong, and run off. I feel like I'm arguing with Ags.
Posts: 1
Threads: 0
Joined: Nov 2003
My intent was not to create the best or most accurate speed init, it was to give a simple bit of code to make some programs a little better. Some programers dont know how to use more complicated systems. This I belive will be a great way to share games so that users of all speeds of computers can play well.
lt;IRON>
Posts: 2,020
Threads: 24
Joined: Jun 2002
I never said you were wrong, buddy, chill. Have a drink, and relax. I say loops are inconsistant. You say not. I say loops are a silly way to go about things. You say not. *shrug* not worth continuing.
i]"I know what you're thinking. Did he fire six shots or only five? Well, to tell you the truth, in all this excitement, I've kinda lost track myself. But being as this is a .44 Magnum ... you've got to ask yourself one question: 'Do I feel lucky?' Well, do ya punk?"[/i] - Dirty Harry
|
