New Beta build 70100 is up!

[SASrecon 0]

http://www.arma2.com/beta-patch.php

No news regarding what they updated yet.

Changelog still says the same

[69714] "enablesimulation" fixed on dedicated server

[69683] VoN volume slider

[69599] Optimized: Texture loading done on background thread.

[68990] Improved: Reduced stutter when looking around, esp. with visibility 5 km or more.

Gonna give it a try a.s.a.p

Yea, as Snake said it seems like performance has gone down very slightly.

But it also seems like the AI is becoming much more responsive as far as I can tell :)

Edited May 4, 2010 by SASrecon

[jw custom 56]

Scriptlag still not fixed!

[otrebla_snake_ita 2]

The changelog is the same: probably there are minor changes or performance improvements

[ACPL Jon 68]

Dwarden said it's actually an experiment

[dmarkwick 261]

Scriptlag still not fixed!

I opened an issue for this several months back. In what situations are you experiencing, or rather, noticing, script lag? It seems to be difficult to get people to believe it's an issue.

[otrebla_snake_ita 2]

Dwarden said it's actually an experiment

Hm?

However, I've tried Benchmark 1 and Benchmark 2.

With the last betapatch I had

Bench1: 24fps average

Bench2: 9fps average

Now I get

Bench1: 23fps average

Bench2: 8fps average

Edited May 4, 2010 by otreblA_SNAKE_[ITA]

[jw custom 56]

;1621801']Hm?

However' date=' I've tried Benchmark 1 and Benchmark 2.

With the last betapatch I had

Bench1: 24fps average

Bench2: 9fps average

Now I get

Bench1: 23fps average

Bench2: 8fps average[/quote']

Losing 1 or 2 fps could be caused by a million things in your computer!

I opened an issue for this several months back. In what situations are you experiencing, or rather, noticing, script lag? It seems to be difficult to get people to believe it's an issue.

Read here:

http://dev-heaven.net/issues/10383

That was what I was talking about.

And it causes me a lot of trouble as some things simply don't work as expected anymore.

Xeno

[malick 0]

Mmmh, does NOT start for me :

"Error compiling pixel shader PSSpecularAlpha:0"

Any idea ?

Thanks

[thr0tt 12]

Mmmh, does NOT start for me :

"Error compiling pixel shader PSSpecularAlpha:0"

Any idea ?

Thanks

Delete the beta folder and reinstall the beta.

[Inkompetent 0]

Mmmh, does NOT start for me :

"Error compiling pixel shader PSSpecularAlpha:0"

Any idea ?

Thanks

If you are using okt_noblur, disable it. It doesn't work for the last two betas. If you aren't using it, make sure that the shortcut is correct

target: "/WhateverPathToYourInstall/ArmA 2/beta/arma2.exe" -mod=beta

start in: "/WhateverPathToYourInstall/ArmA 2"

[otrebla_snake_ita 2]

Losing 1 or 2 fps could be caused by a million things in your computer!

Read here:

Sure, I wanted to say that there is no performance improvement at all

[manzilla 1]

They sure are adding some good stuff to the patches lately. Thanks BIS, the last few patches have improved the game for me a lot.(And it already ran well)

[xeno 234]

I opened an issue for this several months back. In what situations are you experiencing, or rather, noticing, script lag? It seems to be difficult to get people to believe it's an issue.

It was no problem till the last beta patches, you could easily avoid it.

Take ACE as an example, we have gazillions of scripts but you allmost don't notice it.

Now look at JTD Fire and Smoke as another example (sorry, no offense :)). Fire and Smoke is allready causing troubles with 63826 or 1.05 as many very expensive operations (loops with nearestObjects and string operations, running for each unit) can or will run at once without sleeps or waitUntils.

What happens is that during this time other scripts get no time from the scripting engine as those expensive operations eat everything up. Execution of the other scripts gets moved to the next frame, then the next frame, next frame, next fra...

But now, and that is the difference, even scripts that were fine with 63826 and 1.05 get in troubles and delayed.

Though I think it won't stay like it is now, BIS is using many scripts in ArmA 2 too, and I bet it will not change with OA :)

Xeno

[-=seany=- 5]

Hmm this "experiment" is just proving what we already know. Arma is a pain to benchmark. so many variations: Play online/offline, play with mods, different gfx settings, different servers with different missions and hardware. All the conflicting results are because of this.

If they want consistent results they must outline benchmarks they want us to run to compare.

I mean, people are testing with ACE, this gets updated all the time and who knows what perf differences their changes may make/cause. I don't blame them, but I would not post results based on Mods here.

Personally, for me to properly test the new beta's each time would take about 2 hours; off line with with controlled benchmarks. I, and many others, don't have the time to test like this. So we end up applying the new beta and running the game as we would normally (online with mods). Maybe the server is not running as good as it normally does so I get less FPS and stutter, does that mean the new beta is to blame? Probably not.

We need more guidelines for how they want us to test.

My advise if you want to report results here:

Don't use Mods to test, If you do and find anomalies then test with vanilla

Don't test online (too many variables)

Don't change your gfx driver while testing

Keep your gfx settings the same as much as possible.

Don't report changes of 2fps less or more and then vote "better/less performance".

[dmarkwick 261]

Execution of the other scripts gets moved to the next frame, then the next frame, next frame, next fra...

Yep :) but, this is the crux of the issue I raised: that the scheduling of the delayed scripts doesn't *seem* to prioritize correctly, so that scripts that were delayed, that should be getting a "priority" for the next cycle, seem to be treated as normal for the next cycle, no perceived priority.

It's the only way I can explain some scripts not kicking in for as many as 10 or 15 seconds in extreme examples. There's no way some scripts can build up that sort of delay with correct prioritization (and I'm not talking just JTD scripts, I see it in ACE2 scripts etc, for example resting weapons on wall function).

[ACPL Jon 68]

Hm?

it's social experiment there might be (or not) some fixes inside ...

truth will be revealed after end of this experiment ...

in meanwhile keep reporting abnormal findings, crashes and bugs ...

<too short>

[MavericK96 0]

I can only hope all these betas means an official patch coming soon. I'd really love to play online with some of these fixes.

I'd also like to see an actual changelog for the last two betas. :cool:

[otrebla_snake_ita 2]

<too short>

Understood. Thank you for the info

[malick 0]

If you are using okt_noblur, disable it. It doesn't work for the last two betas. If you aren't using it, make sure that the shortcut is correct

My bad, I accidently deleted the -mod=beta when removing all mods to test the patch ^^ Yeah, silly me !

I don't see much improvement otherwise, compared to 69782. Same stutter. Some visual artifacts here and there (white ground textures). Haven't had time to test much further.

At least, no crash or computer burning. Good start :)

Thanks anyway.

[=wfl= sgt bilko 10]

For me:

- The Error message about missing "CfgMarkers.Faction_INS" (or something) is gone now in this beta patch.

- The Arma2Mark gives slightly better overall result (vanilla 3626, this beta 3850 average). The real winner compared to vanilla is the last test where you zoom in from above (probably due to the fixes to grass).

- Benchmark1/2, about same as before 36/16 (which is surprising considering the difference for arma2mark).

I haven't played any longer sessions lately so I can't comment on stability.

Volume levels were reset (maybe it had something to so with the added VON slider).

No addons loaded during testing.

Edited May 4, 2010 by =WFL= Sgt Bilko

[rok 0]

No changes in changelog might be related to this:

Originally Posted by Suma

We never put fixes of the issues which were not present in the stable versions into the change-log.

Anyone else having issues with Fraps and latest betas? For me the framerate drops literally to 1 (one) when recording. With 1.05 the framerate stays over 25 at a same scene.

EDIT: Did a quick search and seems other people have this issue too. Also did test with beta patch 63826 and with it Fraps worked all ok. All newer betas seem to have this issue.

HD4850, Cat 10.4, Win7 RC 64bit

Edited May 4, 2010 by Rok

[Dwarden 1125]

keep reporting ... and if you get crash ...

remember to post it on CIT including crashdumps (especially bidmp)

[Inkompetent 0]

I never tried 70054, but I do have small graphical artifacts now.

When close to trees I get flashing white pixels a bit here and there on the tree branches/leafs, randomly flashing all over it.

System:

Windows XP Professional 32-bit SP3

Intel Core2 Duo E6750 @ 3.60GHz.

2x1GB Corsair Dominator DDR2 PC8500 @ 4-4-4-14

Gigabyte GA-P35-DS3 Bulk 1

ASUS Radeon HD4850 512MB, 10.4a beta drivers

Samsung 7200rpm 500GB SATA-II 16MB cache

ArmA2 settings:

Resolution: 1280x994 Windowed

3D resolution: 100%

View distance: 3000m

Texture detail: Normal

Video memory: Default

Anistropic filtering: Normal

Antialiasing: Disabled

Terrain detail: High

Objects detail: High

Shadow detail: High

Postprocessing: Disabled

[Alex72 1]

Tested Chernogorsk by running around there a bit. Felt good after minor testing. About 30-45FPS with below settings. Setting terrain detail to very low gives me 60-70FPS. Wish that in particular could get some attention as it seems its the grass itself that do the ~25FPS drop. But the beta feels pretty good after a quick test and i will continue tomorrow. Thanks BIS.

ArmA2 settings:

Resolution: 1280x1024 Full

3D resolution: 100%

View distance: 3000m

Texture detail: High

Video memory: High

Anistropic filtering: High

Antialiasing: High

Terrain detail: Normal

Objects detail: High

Shadow detail: High

Postprocessing: Low

Addons: No

i7 930 @ 2.9Ghz (no OC)

Nvidia GTX275 (no OC)

Asus P6X58D Premium

6GB OCZ 1600Mhz (no OC)

WD Black 7200RPM 32MB Cache

Edited May 5, 2010 by Alex72

[MavericK96 0]

About the same as the last beta here. Good performance and besides some fairly minor slowdown/stutter when flying extremely low and fast over Chernogorsk, no issues. Z-Fighting seems a little better (not gone, though) but that could be a placebo. Still no v-buffer crashes even with multiple (10+) low flybys of Chernogorsk. :D