New beta patch 1.16

[BLSmith2112 0]

Glad to see they are managing to still continue to support Arma while a new game is nearing its end of development.

You guys rock :)

[ModeZt 0]

The non beta version (1.14) has a Linux server. Nobody is forcing you to test the betas :j:

i'm saying that server needs testing too. and quality of the server is usually more imortant then clinet's.

plus that statement that other developers put linux servers to the ends of their lists seems very false to me.

[trini scourge 1]

BIS, this is why I will buy Arma 2. Even though there were early missteps, the continued support of the 'old' game while still working on finalizing the new game, is worthy of my support. Runs flawlessly on Vista Ultimate 64 with an ATI 4870 512mb video card.

[Maddmatt 1]

i'm saying that server needs testing too. and quality of the server is usually more imortant then clinet's.

plus that statement that other developers put linux servers to the ends of their lists seems very false to me.

Windows server is available which allows for testing.

It's not like they can just pull a Linux version out of a hat right away, development takes time. It makes sense to make sure the Windows version works fully before porting to Linux.

Remember that they are trying to finish ArmA 2 as well.

[CarlGustaffa 4]

Yes I got the bad blocky stretchy shadow bug after upgrading to ACE107, prior to ArmA1.16 beta. So that bug is not related to ArmA patch.

[M.Andersson(SWE) 4]

Sooooo, how do i unisntall the 1.15 and install 1.16............

[evilfury 0]

Just manualy delete beta(1.5) in your ArmA folder. After that install 1.6 beta.

[.kju 3245]

It goes without saying that any improvement is very welcome and

your hard work on any patch is very much appreciated. There is no need

to discuss that.

Yet there are some points valid to discuss, so for further discussion head to Beta patches policy. :bounce3:

So again, sincere thanks Marek for your insights and thoughts!

Edited April 25, 2009 by kju

[gonk 0]

thx... I can now play the game without the -window parameter. good job..

[SaBrE_UK 0]

Thanks for the patch. Know that most community members are happy with the support you give patch-wise.

[rundll.exe 12]

I cant test right now, but does someone know if the no-autocenter steering sensitvity (introduced in 1.15) has been decreased? Hope BI didnt forget about that one.

[mr.g-c 6]

Ok Windows Server crashes still with "Too Many Virtual Memory Blocks requested".

Crashing of Dedicated Server is NOT FIXED.

I try it now with maxmem 256, lets see if that helps.

(I have sent the normal Windows error-report), i hope it reaches you!

0000: 41 70 70 6c 69 63 61 74   Applicat
0008: 69 6f 6e 20 46 61 69 6c   ion Fail
0010: 75 72 65 20 20 61 72 6d   ure  arm
0018: 61 5f 73 65 72 76 65 72   a_server
0020: 2e 65 78 65 20 31 2e 31   .exe 1.1
0028: 36 2e 30 2e 35 32 37 37   6.0.5277
0030: 20 69 6e 20 61 72 6d 61    in arma
0038: 5f 73 65 72 76 65 72 2e   _server.
0040: 65 78 65 20 31 2e 31 36   exe 1.16
0048: 2e 30 2e 35 32 37 37 20   .0.5277 
0050: 61 74 20 6f 66 66 73 65   at offse
0058: 74 20 30 30 33 63 61 35   t 003ca5
0060: 61 39 0d 0a               a9..    

1 Second later the event log get a entry about a false memory area:

0000: 42 75 63 6b 65 74 3a 20   Bucket: 
0008: 31 32 34 36 31 31 30 31   12461101
0010: 36 34 0d 0a               64..    

And "Dr. Watson" Reported that Armaserver had a access violation(in german sorry):

Die Anwendung "C:\games\ArmA\beta\arma_server.exe" hat einen Programmfehler verursacht. Datum und Zeit des Fehlers: 25.04.2009 um 13:37:54.312 Ausnahme: c0000005 an Adresse 007E7B32 (arma_server)

:(:(:(:(:(

[Rocco 0]

My critism wasn't against beta patches in general, more like that you release a second beta patch, before releasing a full patch before. We tested 115 long enough and there was basically no feedback from BIS.

My biggest concern was, that we get 3 communities, but it seems like that most of the 115 servers updated to 116 already thankfully.

I really appreciate your work, Ofp+ArmA are the best games ever for me.

Still hoping for a Linux 116 patch though.

Very strange reading this discussion.

Does anyone care about the fixes made (as despite low number, all fixes made are of very essential nature: vista users with ATI should run the game at full screen in all configuration, BE shouldn't kick out players randomly, memory fragmentation dramatically reduced on ATI cards, most frequent crash more likely fixed).

Anyway, if you prefer talking about BI release policy, here you go:

1) 1.15 exhibited major issues preventing it from being final patch as main fixes we wanted there didn't fully work

2) 1.16 beta will stay beta only for some time, if nothing bad happens again, will be raised to final 1.17 pretty soon

Why do we bother creating another patch for Arma? Mostly to leave Arma 1 in a shape to run on Vista with X GB of memory in years to come. If you don't care, fair enough, but what annoys you about this free optional patch I simply don't get. If you don't want to try it, simply don't.

Small fixes... well, I better not to tell this to people implementing it. Low level D3D call changed based on year long communication with ATI about problems related to drivers, low level Vista fix, low level netcode protocol change to avoid BE kick off.

I am going back to Chernarus.

p.s.: as to the future: no betapatches = no patches later in life of the game. But I am pretty sure many people in the community will understand why beta patches are needed. The ways to play Arma are nearly endless, as are user modifications, hardware configurations etc. How can you expect us to test it all? No way. Beta patches are needed reality.

[KeyCat 131]

Nice to see another patch (personally I like the beta approach), thanks BIS!

Unfortunately my gaming rig is currently down so I can't give any feedback this time but will follow the threads. I understand that some crash/ATI/4GB+ fixes have high priority but still hope some AI fixes/tweaks (issues mentioned in the 1.15b feedback thread) will make it in before final patch.

@Mojo: Please keep us/me posted on how the IC test pan out, I'm currently on R&R...

/KC

Edited April 25, 2009 by KeyCat

[subtee 0]

This is great, big increase in FPS compared to 1.15 beta !

Cant wait for Arma 2. pre-ordered already :)

[Mojo 0]

Ok Windows Server crashes still with "Too Many Virtual Memory Blocks requested".

Crashing of Dedicated Server is NOT FIXED.

...

On 2008 server we are seeing a very improved CPU value and lower network usage with no crashes. We will know more in a couple hours.

--What are you running, 2008, 2003?

--What other applications are running?

--How many instances of the server are you running?

--Do you have affinity set?

Edited April 25, 2009 by Mojo
added questions

[mr.g-c 6]

Mojo, you are certainly not the person to help us here, even though i appreciate your help-attemot.

We are running a RPG mission which was running before 1.14 over 5000mins, sometimes 10.000 minutes without problems. Then came 1.14 and all went down the road.

So unless you have at least 20+ players constantly for over 1000 minutes on your server combined with a very very script-intensive mission, you can't compare it ever with our setup.

The Problem is known and the lead-dev had addressed it:

http://forums.bistudio.com/showthread.php?t=68360&highlight=Virtual+Memory+Blocks

Still neither maxmem=256 nor anything else helps.... Its a big shame since the serve runs so fluid and so awesome... even after 850 minutes we were still having above 40 FPS, then suddenly alsmost a hour latet -> crash.

With maxmem=256 it crashed just now after 300mins with less players, so even worse than without maxmem. Actual screenshot

[deady 0]

Thanks for the continued support of Arma long after other companies would stop releasing patches. This is what seperates you guys from a company such as...Codemasters ;)

[Zipper5 74]

Mojo, you are certainly not the person to help us here, even though i appreciate your help-attemot.

We are running a RPG mission which was running before 1.14 over 5000mins, sometimes 10.000 minutes without problems. Then came 1.14 and all went down the road.

You do know he's a tech admin at IC:ArmA? Where we have upwards of 100 players playing non-stop for 3 hours at a time?

[mr.g-c 6]

You do know he's a tech admin at IC:ArmA? Where we have upwards of 100 players playing non-stop for 3 hours at a time?

Yes i thought to remember that he had something to do with one of the bigger clans/teams, doesn't matter in this case as 3 hours are a joke compared to our game lengths of serveral days/weeks, which were like i said running great during pre-1.14 times.

However it is still a known problem like you can see in the linked topic and apparently it was not fixed by BIS, nor does the by Suma adviced maxmem=256 command helps.

I dunno if its the large addons we use, but we used addons since ages and crashes tarting with 1.14 i'm almost 100% certain about that.

I am not sure as i have no real insight into any process but i think the limit of 1024MB for virtualblock was not raised like Suma suggested it in Early 2008:

Virtual memory used for higher level "object" caches is allocated in 64 MB blocks in ArmA. There are at most 16 such blocks allowed. I guess for the sake of dedicated server, which is not exhausting its memory space by allocating textures and other D3D resources, we should perhaps extend the limit.

Edited April 25, 2009 by mr.g-c
Added Suma quote

[Alex72 1]

Well, more of a f**, like a feature. Anyway, 1.16 focuses more on low level architecture and not any gameplay improvements over 1.15 and that's how it is going to stay.

But to Alex[Dev]72 - I can't see how could the 1.16 change behaviour on your Nvidia card. Are 100% sure that the patch is the only thing you changed and for example if you uninstall 1.16 and install 1.15 the problem really disapears?

Hi Maruk,

The texture blur was there in 1.15b as well even though now its a bit more agressive. I havent changed anything when i went from 1.15b - 1.16b apart from installing the 1.16b patch. What i noticed though with this unloading of textures is that when i am close to dead bodies - this is when the ground textures unload/change to lowres. Tested this so many times and its always when a dead body is close to me. When i move away from it - all fine again.

This is more agressive in 1.16b for me. Overall performance is better in the new patch and i really like it apart from this stupid texture thing i have - and it seems many others i ask. When it starts chopping/lagging during gameplay its a bit more twitchy now. It felt like usual lag in 1.15b and in 1.16b its a bit more, well different. Not that it bothers me though. Just saying what im noticing. But the ground texture change to low res when i am close to dead bodies is weird (in 1.15b also). Alive players i can have many close to me without the ground changing. Only dead bodies mess with the ground textures. Is it the damage textures that are messing with me? Wonder if a dead body normally take more performance than alive one that doesnt "think"?

Anyway, im playing with 1.16b and i notice that when the ground change to low res cause of dead bodies - the performance is worse than with 1.15b. And on Afghan Village map by Opteryx i have a mission, and that mission plays worse on 1.16b when the fight begins. This is the map i saw objects (units/my weapon) change shape everytime i went in and out from my scope. That never happened in 1.15b. Never (love 1.15b :) ).

Dont know what to do apart from looking if there is even newer NVIDIA drivers (mine are some days old). And i play ArmA everyday all the time. I notice when something changes, believe me. Did get my diploma for hardware, OS & networks just to let you know im not sitting here making things up. With a new system i hope this goes away cause it does take away a lot of the fun - even though i cant stay away from it. :)

EDIT: Forgot to say that i crash with this new patch much more often. Its that known 1.14 "bug" where the screen goes white/weird colours. This has happened to me 4 times since i installed 1.16b (2 times a day). It happened once a week on 1.15b maybe or even less. I saved to crash RPT's with DUMP files as well. I saved only the FROZEN START - FROZEN END. Dont know if the rest is needed.

Download: RPT's + DUMP's (7zipped)

Regards

Alex

Edited April 25, 2009 by Alex72

[ricoadf 0]

Well no crashes or obvious issues, so a successful patch on my end, the game seems alittle smoother so glad to see, thanks BIS :)

[MadMax28MEU 10]

What so good about this patch?

[Mr_Tea 0]

What so good about this patch?

For example, the better performance, bugfixes, support for an game other Dev`s would have dropped at least month ago.

Thank you BIS.:notworthy:

[SWT_Janowich 10]

Hi folk,

I've got a new bug. I call it the pigeons bug.

The Bug itself

The Bug itself is that the mouths remain moving after a spoken command

in an endless period.

Since 1.16

System:

ATI HD 2900 XT

8GB Ram

Windows Vista 64 Bit

> 100kb

I'm able to reproduce it.

No Mods enabled. Would be nice to hear a word from a dev.