Improvements to the Speed of Calculations?

[RyanCrierie]

Since from what I've been hearing, the AI has become really really good, perhaps next up should be looking into the code in an attempt to speed up the supply calculations, et al for really big scenarios when you first load them up.

I find that with COW, it takes a while for big scenarios to do supply calculations for the first turn, etc.

[Bloodybucket28th]

I recall reading somewhere that while Elmer takes his turn in some of the larger scenarios, you'd better find something to do while he ponders...While this might not be all bad depending on what that something else might be[:o], I'm curious to know what TOAWIII users are seeing in this regard. Does the speed of more modern computers pretty much make the wait a thing of the past?

[Captain Cruft]

It's not Elmer that takes the time, it's "inter-turn Housekeeping". Namely, calculating supply values for every hex on the map, amongst other things.

I have a very un-modern PC, viz. a PIII-500. Even with this steam-powered equipment the inter-turn wait has not exceeded c.1.5 hours, which was for Fire in the East and Europe Aflame.

The thing is, because TOAW III is a well-behaved Windows application, there is nothing to stop you going off and surfing the web or doing whatever else while you're waiting for the calcs to finish. At least that's been my experience.

[Bloodybucket28th]

Wow, an hour and a half? That sounds like a loooong time, but it sure must help to run it on a faster machine or with the non Supersized ("Welcome to Bucket O' Battalions! Would you like fries with your Eastern Front Blizzard Biggie Burger?") scenarios.

Besides, I might get some stuff done around the house if it does take that long.

[Captain Cruft]

LOL. I'm up for some of those burgers

I suppose I was merely trying to illustrate the worst case, and illustrate that this game is playable even with equipment well below the official minimum requirements.

If your PC was bought within the last 5 years it will be at least twice as fast as mine, and I would imagine any recent PC will zip through the same calcs in short order.

[ralphtricky]

They're still slower than they should be. I've identified two calculations that can take a long time, the supply calc, and the reconstitution check.

That's all I'm going to say about it. [;)]

Ralph

[Bloodybucket28th]

A Gumpism from Dr. Frankentrickey, who gave Elmer life!

Or perhaps he is more in the role of Dr. Jones:

"Old programming language, very dangerous. You go first!"

[geozero]

ORIGINAL: ralphtrick

They're still slower than they should be. I've identified two calculations that can take a long time, the supply calc, and the reconstitution check.

That's all I'm going to say about it. [;)]

Ralph

It sound like you have a firm grip on the two major time consuming calculation issues. Look forward to seeing these improved.

[Erik2]

ORIGINAL: Captain Cruft

It's not Elmer that takes the time, it's "inter-turn Housekeeping". Namely, calculating supply values for every hex on the map, amongst other things.
.....
I

I have one 1km scenario of Crete (300x300 hexes) wich I'm testing for play against the PO.
Elmer does take a very long time pondering his moves as the Allies, this may be related to that there are a lot of individual ships to keep track on. Moving as the Germans is fairly quick.

Erik

[Szilard]

Not the most important point, but anyway: I think Elmer might just be good enough now to make testing of scenarios via multiple AI vs AI runs something worth doing, before human vs human etc testing. I'm thinking in terms of *lots* of test runs - 50, 100? - depending on the scenario. Obviously, the faster you can make things, the more runs & the better the testing (hopefully).

[glvaca]

on a AMD 3600Mhz with 1GB Ram, the pre-calcs for the FitE scenario takes about 15-20m.
You can do something else while the check is in progress but you will probably lose the "in progress" screen so you have no idea where he's at.
While on the subject, TOAW III is very robust and I have yet to experience a crash even with several instances of the game open at the same time. Good job!
Best,
Glenn

[ralphtricky]

ORIGINAL: Erik Nygaard

ORIGINAL: Captain Cruft

It's not Elmer that takes the time, it's "inter-turn Housekeeping". Namely, calculating supply values for every hex on the map, amongst other things.
.....
I

I have one 1km scenario of Crete (300x300 hexes) wich I'm testing for play against the PO.
Elmer does take a very long time pondering his moves as the Allies, this may be related to that there are a lot of individual ships to keep track on. Moving as the Germans is fairly quick.

Erik

If you care to email me a copy, I'll try to take a look at it later.

Ralph

[ralphtricky]

ORIGINAL: geozero
It sound like you have a firm grip on the two major time consuming calculation issues. Look forward to seeing these improved.

Well, some people were complaining that the computer had crashed. That's getting a bit too slow, so I had to look into it.

Ralph

[CommC]

15 or 20 min for any calculations for a game as simple as this one is simply unacceptable. With the speed of modern computers, the slow down is not due to computational capacity of the computer, but simply poor programming. I hope Matrix will put a priority on improving this.

[Erik2]

So you think the game is simple?
Have you looked at a good researched scenario in the editor?
There are a lot of details there that the system needs to calculate for each side each turn,
especially if you play against the PO.
Believe me, there's nothing simplistic about TOAW.

[a white rabbit]

ORIGINAL: CommC

15 or 20 min for any calculations for a game as simple as this one is simply unacceptable. With the speed of modern computers, the slow down is not due to computational capacity of the computer, but simply poor programming. I hope Matrix will put a priority on improving this.

..simple ?!?!?..

..roflmao...

[geozero]

Speed of the game turn is relative to the Computer you use, specifically RAM and Processor type and speed. This is true for any game, there's no mystery here.

If you have a Pentium 2 with 256MB Ram it will be immensely slower than if you are using a new P4 with 2GB ram (which is what I have).

I think most people that have played the series will agree that the game is not simple. But I think the statement is valid in terms that the game is not graphical intensive, there is no 3D rendering and things like that which chew up a lot of processing power.

I'm not a programmer, bu I think it is probably written in older computer language with long and complex math calculations, etc.  The code might be messy.  It could likely be tightened up, but would require a new re-write entirely, thus a true NEW game. 

Until that  happens, all I can suggest is upgrade your PC, get as much RAM as possible, run it under a fast processor and WinXP (God forbid you are still using Win98 or Win95).

In any event there is always the other choice while waiting for turn completion....

Heat up a pizza and pop open another beer.[:D]

[golden delicious]

ORIGINAL: geozero

If you have a Pentium 2 with 256k Ram it will be immensely slower than if you are using a new P4 with 2MB ram (which is what I have).

Presumably you mean 256 MB and 2 GB respectively. My old Atari ST had 1 MB of RAM, and that was 18 years ago.

[geozero]

ORIGINAL: golden delicious

ORIGINAL: geozero

If you have a Pentium 2 with 256k Ram it will be immensely slower than if you are using a new P4 with 2MB ram (which is what I have).

Presumably you mean 256 MB and 2 GB respectively. My old Atari ST had 1 MB of RAM, and that was 18 years ago.

My bad. Post edited.

BTW - I used a Commodore which only booted up from a 5.25" floppy.

[golden delicious]

ORIGINAL: geozero

BTW - I used a Commodore which only booted up from a 5.25" floppy.

That's nothing. Before the ST we had a Commodore which loaded of cassette tapes. Those were the days.