Really, really slow

[Agema]

Hi, I've got quite a high-spec system (AMD Athlon 3000+, 1Gb RAM) but I find COTA runs much, much more slowly than HTTR. On one level, this might not be a surprise due to the extra AI coding. Sometimes the double speed goes no faster than normal speed. Not always: sometimes it runs fine, other times it really drags. I've rarely noticed triple speed have any effect over double. Also occasionally everything stops for from about 5 seconds to as long as half a minute, then starts again. Any ideas?

[RedDevil]

I also have what I consider a high-spec system and have experienced the same.

It is a bit of a shame because I was planning to play the game on my laptop (of course after uninstalling it from the PC [:D]) but I am afraid that it won't be playable there.

Yesterday for example, while playing the tutorial a bridge was blown and it took quite a big amount of time to re-calculate the movement tables. Also loading a scenario takes quite some time, not longer than other games, but I think I was spoilt by the speed of HTTR.

Is there any way to speed up the game?

cheers,

RedDevil

[simovitch]

Check the read me file in the START->PROGRAMS->CotA menu, it includes tips for speeding up the game and reasons why the game will pause every once in a while.

Hope this helps.

[GoodGuy]

Well, I've experienced these same freezes......I'm playing COTA on a AMD 64 3200+ rig w/ 1 GB RAM.

Excerpts from the readme:

COTA readme file

....Occasionally, the simulation clock may come to a halt for a brief period of time while the map continues to scroll smoothly. This is not a bug or a performance issue. Behind the scenes the AI has initiated a replanning process which can be very CPU intensive. Effectively this activity has momentarily sucked CPU away from the simulation.......


.....(5) The CPU load is directly proportional to the number of units displayed on the map. Consider playing scenarios with fewer units if your machine has trouble with CPU load. Annex F of the Reference Manual, provides a Scenario Complexity Table which lists each scenario and the total number of units involved. Note: In some scenarios, the majority of units can arrive after scenario start as reinforcements........

Well, sounds logical.....but what about a game where 3 enemy units are left on the map, with only just 1 Bn doing anything (chasing these enemy units in move mode)? That just happened to me on the weekend. I can't imagine that the management of this massive number of enemy units (3) would steal much CPU time. Eh?

If this wouldn't have occurred in HttR as well, I'd say that it'd be a bit suspicious (i.e. memory leak).

[Arjuna]

GoodGuy,
 
There are no significant memory leaks during play. Yes things have slowed down from HTTR but then the game is doing a whole lot more. There is no getting around that. Yes the game does sometimes pause while plans are developed. The usual culprit being the calculating of routes. With mixed mode movement we now invariable do two to three times the processing for routes. If a bridge is blown the game will pause for some time while movement maps are recalculated. The latter we'll be addressing for BFTB when we tackle dynamic terrain for minefields and roadblocks. For this we'll be precalculating data to reduce the need to recalc during play.

[GoodGuy]

ORIGINAL: Arjuna
For this we'll be precalculating data to reduce the need to recalc during play.

Great news !

I'd love to get an answer regarding my 3-unit example, though [:)] [&o]
I mean....only one Bn moving, on the entire map, with 3 enemy units left?

Btw, I didn't claim that there's a leak, just to clarify things.....I said IF I wouldn't know HttR I'd be tempted to think there's one in COTA.

[Arjuna]

GoodGuy,
 
Send us a Saved Game plus a short description. BTW you will here me say this a lot in regards to bug fixes. For us to determine if there is anything wrong and if so what it is, we need a saved game where the behaviour can be repeated readily.

[Agema]

I doubt very much there's a memory leak: normally that causes a game to slow down more the longer it's going, together with a crunchingly long time when you exit as it clears everything up, neither of which I've observed.