My brain knows what it's doing, but i don't have a clue!
---------- Post added at 05:56 PM ---------- Previous post was at 04:51 PM ----------
So, about how laggy JSRS 1.4 will be.
I did some SP scale tests with around 220 AI soldiers.
Without 1.4 average fps: 36,5
With 1,4 average fps: 32,5
Without 1.4 average fps: 39
With 1.4 average fps: 33
Means: Without 1.4 37,75 fps, with 1.4 32,75 fps. Its a drop of of around 5 fps in total. Thats alot I guess. Well, we have CBA running with the 1.4 tests. I dont know if this is causing any lags or performance troubles.
But I dont know how much 1.33 had effect the performance, so maybe there will be no difference at all compared with 1.33.
At least, I can life with that. We do some dedi server testing maybe today!
We can see the AI soldiers as real players because of the way the scripts is build. So that means, we would have a drop of of 5 FPS with around 200 player. I dont know how much this will effect an dedi server, but we'll see this maybe today.
Last edited by LordJarhead; Dec 19 2011 at 16:03.
Can you send me your test mission? I want to test it with 1.3.
What was the spec of the PC hosting the test game LJ? Unless it is a very high end rig I agree that a game with 200+ units is acceptable to show a drop of ~5fps.
The British Army will never yield!
The drop is not increesed trough the ai! The difference arises trough the mod - both test conditions are with 200+ ai's! You missunderstand this Munger.
---------- Post added at 12:22 AM ---------- Previous post was Yesterday at 11:52 PM ----------
here are my performance test conditions and results:
- i tested with exactly the same map like LJ (+220 AI's in a large scale fight with tanks and inf)
- format: lowest frames / average over the time / highest
- all tests repeated 2 times
vanilla: 43 / 58 / 81
jsrs 1.3: 37 / 41 / 79
jsrs 1.4: 36 / 56 / 78
drop in avarage to vanilla:
1.3 --> 17 frames
1.4 --> 2 frames
In my eyes i didnt see a problem with jsrs and the new distant sound system.
However, I see a marked improvement in the average compared to the current version 1.3. --> passed!
btw: in all 3 tests are around 100 enemys in summary alive at the end of the test.
Last edited by [TcB]-Psycho-; Dec 19 2011 at 22:27.
I was also running tests with big and demanding missions, no performance problems so far. Now we need the MP test to tell if stability is really ok, but MP shouldn't be any different than SP since the stuff is running client side.
Wow that's quite a performance drop, 17fps
Shoothouse Mission http://forums.bistudio.com/showthread.php?t=126312
Yes, it is a big drop. But 220 men firing at once is not typical.
Could we get a test with two full squads unloading on each other? The drop here might be negligible.