Test for "LAG"

Ticonderoga60 · Post by **Ticonderoga60** » Tue Jan 09, 2018 10:14 am

I did a test. El Alamein scenario. After several turns, the lag increased exponentially each turn.
After 6 turns of play, the lag reached 30-35 seconds! Then I did a test:
I turned off the PC, I turned it back on and I uploaded TOTH again.
The lag now lowers to 5 seconds (instead 30-35)! It is as if the computer memory was filled round after
shift without emptying.
Peter, is it possible for the game to constantly allocate memory without empty the RAM?

Hailstone · Post by **Hailstone** » Tue Jan 09, 2018 1:36 pm

My experience is quite different. The scenarios with lag problems that I have saved are repeatable regardless of restarting my PC. Every time I reload a saved laggy game the delays just pick up from where they left off when I saved the game. Never found any difference with restarts.[&:]

Peter Fisla · Post by **Peter Fisla** » Tue Jan 09, 2018 2:32 pm

ORIGINAL: Ticonderoga60

I did a test. El Alamein scenario. After several turns, the lag increased exponentially each turn.
After 6 turns of play, the lag reached 30-35 seconds! Then I did a test:
I turned off the PC, I turned it back on and I uploaded TOTH again.
The lag now lowers to 5 seconds (instead 30-35)! It is as if the computer memory was filled round after
shift without emptying.
Peter, is it possible for the game to constantly allocate memory without empty the RAM?

It's the LOS and FOW processing that takes the hit, these two processes are very CPU demanding. I have optimized both functions to get about 45% performance improvement in the UPDATE4. Are you playing with UPDATE4 + Beta 1.0.82 patch ( posted here, tm.asp?m=4384334&mpage=3&key= Post 74 form me).

Ticonderoga60 · Post by **Ticonderoga60** » Tue Jan 09, 2018 8:20 pm

No Peter. I haved 1.0.81. Tomorrow try with 1.0.82. Thanks for the moment.

Peter Fisla · Post by **Peter Fisla** » Wed Jan 10, 2018 12:51 am

ORIGINAL: Ticonderoga60

No Peter. I haved 1.0.81. Tomorrow try with 1.0.82. Thanks for the moment.

The 1.0.82 addresses some old CPUs not being able to run 1.0.81. So if you are ok with the released UPDATE4 (meaning your CPU can run build 1.0.81) then you are up to date and don’t really need 1.0.82.

Ticonderoga60 · Post by **Ticonderoga60** » Wed Jan 10, 2018 9:38 am

So, after installing 1.0.82 the situation is as follows:
El alamein, (lOS quality low):
turn 7: 5-6 seconds (very acceptable)
turn 8: 12-14 seconds
turn 9: 16-20 seconds
The situation has improved compared to 1.0.81
However, I turned off the PC (Hailstone: switched off and not simply restarted) and switched on again
I uploaded the saved file and the lag is back to 6 seconds.
Evidently, something in TOTH progressively fills the RAM, worsening performance as the game continues.
However, if they were 6-7 seconds, it would still be excellent.
Peter, a bug: AFI managed by the AI can interdict player's broken units
Thanks for the attention. TOTH always remains a masterpiece! In game programming, one of the most difficult, if not impossible, things is AI management! Actually level of IA is not excellent but is good. I enjoy very well!

Ticonderoga60 · Post by **Ticonderoga60** » Wed Jan 10, 2018 9:47 am

Ops! IA AFV are BU!

Peter Fisla · Post by **Peter Fisla** » Wed Jan 10, 2018 10:23 am

Each time AI plays a segment, the AI processing will different as each time the AI thinks slightly differently. Also in Windows 8 and 10 the performance testing with the game will report different numbers as the windows os by itself does things on its own so the environment is never the same either.

Ticonderoga60 · Post by **Ticonderoga60** » Wed Jan 10, 2018 10:26 am

About bug Peter?Is really a bug? I think yes.

rico21 · Post by **rico21** » Wed Jan 10, 2018 11:17 am

The mainly lag is the distance between AI units and Victory points.
More than twenty hexes is too many.
And the other hand, If AI unit has to choice between several victory points, it's worse.
Howewer...[:D]

Peter Fisla · Post by **Peter Fisla** » Wed Jan 10, 2018 12:22 pm

ORIGINAL: Ticonderoga60

About bug Peter?Is really a bug? I think yes.

Sorry, what bug?

zakblood · Post by **zakblood** » Wed Jan 10, 2018 12:45 pm

might be better to share a report on your system spec tbh, as not had the issues myself and play on an old test system

Processor: Intel(R) Core(TM) i7-2700K CPU @ 3.50GHz (8 CPUs), ~3.5GHz
Memory: 16384MB RAM
User DPI Setting: 120 DPI (125 percent)
System DPI Setting: 144 DPI (150 percent)
Card name: NVIDIA GeForce GTX 960
Dedicated Memory: 4062 MB
Current Mode: 1920 x 1080 (32 bit) (60Hz) 920 x 1080(p)

How to get a DirectX Diagnostic file:

1. click on "START"
2. click "run"
3. type in "dxdiag" and click "OK"
4. check through the tabs and perform all tests
5. click "Save all Information..." and save the text file to a location of your choice

We prefer that DxDiag files be posted as message attachments rather than pasted into the body of the post.

once the report has been made, just attach it to your post with first adding the file to a zip, which is then permitted to attach to the forum

rico21 · Post by **rico21** » Wed Jan 10, 2018 1:16 pm

Is it serious doctor?[:D]

Ticonderoga60 · Post by **Ticonderoga60** » Wed Jan 10, 2018 2:39 pm

BU AFV managed by the AI can interdict player's broken units. This isn't correct

UP844 · Post by **UP844** » Wed Jan 10, 2018 4:20 pm

I think this is another feature (like radioless, drone-directed OBA) aimed at providing some help for the poor AI.

A human player can also make use of the advantages of being CE (better accuracy, AAMG fire, better road movement rate). Once the Panzers were killed, the CE Stuarts in Alamein slaughtered German infantry with ease

.

On the other hand, I think teaching the AI how to evaluate when it is better being CE rather than BU would be a major issue.

Peter Fisla · Post by **Peter Fisla** » Wed Jan 10, 2018 5:19 pm

ORIGINAL: Ticonderoga60

BU AFV managed by the AI can interdict player's broken units. This isn't correct

I will look into it, you have a saved game that could help me.

Ticonderoga60 · Post by **Ticonderoga60** » Thu Jan 11, 2018 7:23 am

No, I haven't in this moment.
UP, I agree with you. It is not a big deal. Also in El Alamein, the English Stuart has killed two broken teams ...
It is also true that the AI does not use the fast movement, which could help him win on several occasions.
Peter, I noticed one thing: an AI squad used Molotov against my PzIII. There is a shadow on the target hexagon. Is it maybe smoke?

Peter Fisla · Post by **Peter Fisla** » Thu Jan 11, 2018 9:37 am

ORIGINAL: Ticonderoga60

No, I haven't in this moment.
UP, I agree with you. It is not a big deal. Also in El Alamein, the English Stuart has killed two broken teams ...
It is also true that the AI does not use the fast movement, which could help him win on several occasions.
Peter, I noticed one thing: an AI squad used Molotov against my PzIII. There is a shadow on the target hexagon. Is it maybe smoke?

Do you know at least what buttoned up vehicle was firing vs the broken unit as it routed?

There is no Molotov in the game, only Molotov projector.

rico21 · Post by **rico21** » Thu Jan 11, 2018 9:42 am

I was reproduce the case...

rico21 · Post by **rico21** » Thu Jan 11, 2018 9:43 am

The file only for Peter eyes...[:D]

Test for "LAG"

Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"

RE: Test for "LAG"