Are you complaining about Arma 3? You need to watch this video.

[brightcandle 114]

I can get the game down to 15 fps with 6 players and 20 AI. Part of our ongoing efforts is with very limited tools to try and determine what is the cause of the poor performance and if there is anything we can do about it without asking BIS to fix the underlying game it. We run 28 mods and any one or a combination of them could be impacting performance very severely, since the game is intrinsically CPU limited anything that adds complexity to the game will slow the frame rate. But without good tools for testing other peoples mods and being able to capture during a game what it is that is taking the time in the frame its nearly impossible to determine which scripts are causing the problem. Its also the reason why a lot of missions perform poorly but no one can really explain why, the mission makers can't themselves see the frame time impact of their scripts, they likely don't see the problems in their own testing and so without information they have to look towards BI. We all know that some missions run well and others don't, we know that some servers are worse than others because they are overloaded and despite that information the performance complaints persist. The reason is because we can't just hit a button and see the missions scripts are taking 16ms every frame and that is what caused performance to tank.

One of the problems with building a sandbox is that you need to provide some tools to make programming in that sandbox better. If you want to blame the users for the performance of their scripts then you need a way to measure the performance of those scripts in real world conditions. BI script is really backwards in what you can do with it. Most modern programming languages come with detailed debuggers and profilers that can really help get them working, its hard enough to get warnings or the right things to happen with the scripts for Arma, its hard to organise the code well and there is a large part of just nastiness to the way a lot of mods necessarily do things. In essence they get it done but they can't test it properly and it is part of the reason why performance can't be checked very well.

As I have maintained from the early days of Arma I and my community want to help, and indeed when dwarden and a BI developer asked us to do diag_captureFrame's of one of our games we got 35 players to install the profiler build, got all the mods working on it and went ahead and ran several missions on it, even though it ran worse than usual so we could capture what they asked for. We provided that information a little over a year ago and we never got any response, our posts with the links to the profiler output and our analysis of what climbed through the mission seems to have been consigned to the wind. Indeed those same missions with similar parameters and mods will run just as poorly today as it did 15 months ago. So its not like we didn't do what they wanted us to do to help fix performance, I am not complaining because it sucks, my goal right now is to call them out on their actions when we have tried to help. I feel like the call for this information was to try and stall us and give us a near impossible task not to actually improve performance in any meaningful way. I come to that conclusion after trying to determine the performance problem over a long period of time.

While I agree with the sentiments of the video my own personal experience hasn't been of BI being helpful. I remember reporting the cause of bugs for example in the Alpha (that damn seagull bug) being enormously harder than it should of been and the fix breaking a lot of things with versioning. We pleaded for them to do it a different way but they didn't and its still as it was set back then. Its always been an antagonistic relationship and my experience is that its always required pitchforks and torches from the community out to get BI to do anything, being helpful and voting up bugs in their tracker for example has been random at best. Sometimes they just close things with wont fix and other times a few people vote something up and it gets done immediately. In essence it appears they do what they want to do and sometimes it aligns with the community and often it doesn't. I don't think sitting in silence will help get anyones personal faults fixed. Yes the game will get better with time, that is one of the good things about BI, but if we want it to get better in the ways we want we need to talk about our issues. Its not really complaints or whining, its direction.

I don't care too much about the other issues, DLC is either something I want and pay for or its not, if its too expensive or splitting or whatever its not an issue I have with BI. I expect to pay for expansions and extra content to keep the development going and its something I want to happen. My primary concern is with performance right now and its this I want to fix. I personally think the next step is a diag_captureFrame like tool but which reports the timings of scripts for a frame at that moment. So this way if you are on a server and FPS is terrible you can type in the command and see the breakdown and see which scripts are slow. That way at least you can blame the mission and its implementation and we can start to focus on the parts that are going slow and improve the communities contributions so they perform better. That would enable the sandbox immensely, we need better debugging tools.

PS Still willing to do whatever it takes to fix the performance issues, I am a programmer myself with a lot of experience in profiling code so you know if BI wants to change this relationship and either help us help ourselves with determining the cause of our performance problems or get a capture to help them then more than willing. But its about the 20th time I have made that offer so not expecting to go any better this time around.

[clifdenhill 1]

I think the reason why ppl have something to complain about is because BI and the community don't usually agree on what features are most important and should be worked on right now.

BI has their road map of what to fix and the community has what they want fixed to mod and play the game.

I think that's where the frustration comes from and dyslexi is vending on behalf of BI because surely the are frustrated from the harsh critiques of their game they worked years on.

:)

[Gunter Severloh 3948]

Then, whats the point of Altis? Whats the point of allowing 100 players?

To be played with 20 players vs 20 AI?

Its just a map, its not designed for maximum players against maximum AI, Stratis is vary big too, again its just a map, BIS doesn't design maps for specific pvp,

or pvAI or whathave you, this aint call of duty or battlefield where the maps are set for specific game type this is just a map that you can play on whether you

use it for sp or mp or just to do whatever on through scripting or fooling around, its a place is all.

Most of the map makers in the community build maps that do the same thing, they dont build maps for specific MP gametypes, not everything

in this game is designed for mp in mind, theres alot of people that aren't interested in mp as well as are, so it goes both ways.

If theres such and issue with performance then use a caching script, last night i hosted my own mission i built on Altis that had over

100 AI on it including tanks, and planes flying around, and had no issues with performance on it, on top of using mods with it.

[I give up 152]

Its just a map, its not designed for maximum players against maximum AI, Stratis is vary big too, again its just a map, BIS doesn't design maps for specific pvp,

or pvAI or whathave you, this aint call of duty or battlefield where the maps are set for specific game type this is just a map that you can play on whether you

use it for sp or mp or just to do whatever on through scripting or fooling around, its a place is all.

Most of the map makers in the community build maps that do the same thing, they dont build maps for specific MP gametypes, not everything

in this game is designed for mp in mind, theres alot of people that aren't interested in mp as well as are, so it goes both ways.

If theres such and issue with performance then use a caching script, last night i hosted my own mission i built on Altis that had over

100 AI on it including tanks, and planes flying around, and had no issues with performance on it, on top of using mods with it.

Can we have a video of such great gameplay? (this thread is about videos)

[brightcandle 114]

If theres such and issue with performance then use a caching script, last night i hosted my own mission i built on Altis that had over

100 AI on it including tanks, and planes flying around, and had no issues with performance on it, on top of using mods with it.

This is the nugget of the issue the community has. There doesn't seem to be any obvious reason why some games run well and others don't. There is a shroud over where the problem comes from and if its scripts/mods there is no way for a community to determine which ones so they can get the mod makers to fix it. Some communities are having no issues, some are having enormous issues (and with perfect server FPS). The lack of information as to the cause is the problem, we aren't empowered to track the problems down and fix it very well especially when it only really manifests with multiple players in the game and in real scenarios mid game. Whatever we have needs to work within the game and it needs to point us in a particular direction so we can work it out.

Your performance is good (but didn't define good, some people are happy at 25, others want minimum 30, others prefer mostly 60 and some want 120+ so you need to use the absolute numbers) but that isn't everyones experience. High end rigs regularly have problems with the game and I haven't really seen why that is. Its not always the case its a poor performing server, or a heavily scripted mission. We see plenty of complaints about wasteland and altis life (and they do both in my experience perform awfully) but then the king of the hill stuff sometimes, but not always, runs great.

All this says to me is its a complex topic and those that have issues need more help to determine the problem, its not fair to dismiss them as whiners and complainers, the logical fallacy of these aren't true scotsman (arma 3 communities members or a minority) is simply an attempt to build prejudice and dismiss their issues.

[Jona33 51]

introduce a proper server browser, server lobby, squad management and loadout

management PLUS a proper official conventional PVP gamemode!

Although it sounds like we might be getting some of that in the next DLC, I fail to see what that has to do with the topic?

[katipo66 94]

Yeah that was a great video from Dslyecxi, i appreciated that!

[Gunter Severloh 3948]

Can we have a video of such great gameplay? (this thread is about videos)

Thread is about the OP's posted video and its pertaining subject not other videos to demonstrate perfect gameplay which you'll never have given certain circumstances.

What you expect from an open ended world milsim sandbox, nothing in the game is closed in where you have a 100x100 map to play in and get perfect fps.

Heres another thread for you talking about performance again and here us vets got to come and say the lag, or even other issues such as CTD's can be

cause by numerous things be it crappy machine, tons of AI, mods, server configuration, not optimized computer, eg defrag, registry errors, shit running

in the background, and of course we know that the game isn't completely optimized to utilize how ever meany cores you have for your cpu we can go on and on about it.

Only suggest one can make is to become more aware and observant on your end of such issues that do cause poor performance, if you have good

performance in sp or when you host your own missions and then go to a server and get shit performance then consider that its not completely

the games fault, surely it adds to it but server issues are there too.

[I give up 152]

Anyway, this whole debate is a bit pointless, whatever the video that shows a military gameplay does not reflect the current Arma 3 gameplay.

To realize that we just need to look at server browser, we only see servers running MMORPG game modes.

Arma is not a Military SIM anymore.

[brightcandle 114]

Anyway, this whole debate is a bit pointless, whatever the video that shows a military gameplay does not reflect the current Arma 3 gameplay.

To realize that we just need to look at server browser, we only see servers running MMORPG game modes.

Arma is not a Military SIM anymore.

It is, but those communities don't run public servers because the base game isn't very good for milsim. Those communities advertise for members and run lots of client size mods which really doesn't work on a public server (its empty because you can't just download and run the mods from within the game so it requires external tools and instruction guides and yeah pain the backside). You have applied the not a true scotsman logical fallacy there to those gaming communities as if they didn't exist whereas they do they just don't advertise in the same way. I don't dismiss the performance problems of the MMORPGers but you can't dismiss my community exists at all.

[Shagulon 1]

Can we have a video of such great gameplay? (this thread is about videos)

Actually this thread isn't about framerates and gameplay, it is about a bit of courtesy and manners when interacting with people on the internet. It is also about hanging in there while the performance issues are resolved, and words of encouragement that the BIS team are good people and doing a good job creating a unique product which we all love.

If you can't handle that, put the game down and come back in 6 months.

[naizarak 4]

funny video coming from a guy who works for BI and whose youtube channel is basically PR for arma, even though public arma is absolutely nothing like what he depicts in his videos. so yeah, people do have a right to be pissed at BI if arma doesn't meet their expectation especially based on his videos

[I give up 152]

Actually this thread isn't about framerates and gameplay, it is about a bit of courtesy and manners when interacting with people on the internet. It is also about hanging in there while the performance issues are resolved, and words of encouragement that the BIS team are good people and doing a good job creating a unique product which we all love.

If you can't handle that, put the game down and come back in 6 months.

You are correct.

[subunit 31]

I think the video's arguments have some merit in the sense that it's true that we often don't know why performance is suffering in any given situation, and that BIS is indeed less shitty than the majority of other developers, but I don't really see how either is exculpatory here. BrightCandle has a succinct history of the issue upthread- BIS was engaging with the community much more extensively in the past on the performance issue, but has over the last year or more failed to do so. Is it my fault that my missions are unoptimised when the platform provides literally no feedback on how I ought to be optimising the mission in the way dslyecxi suggests is absolutely critical to maintaining good performance?

Frankly it's telling that it's dslyexci defending BIS on this score and not someone from BIS. There's clearly been a communications failure here. The fact that both Rydygier, who probably knows about as much as anyone else does about mission design and optimisation, and BrightCandle, who seems to know about as much about the performance issue as anyone outside of BIS, feel that there's a problem with communciation, should be a wake-up call. If communication doesn't improve, it's hard to avoid BrightCandle's conclusion that these issues are being deliberately downplayed or slowrolled because BIS has decided internally that the issue isn't cost-effective to address.

[brightcandle 114]

I do feel if this message about performance (its complicated, we are fed up of the complaints, we are working on it, please be nicer) had come from Dwarden I doubt the community would have been so welcoming. I disagree that we should all put up with the status quo and trust that BI will solve it, because there is quite a lot of evidence to suggest they wont.

I feel like I partially started this recent trend of complaints and I did it on purpose. When arma 3 performance was a problem in the Alpha, BIS acknowledged it was. By the time release happened everyone had the same complaints, most hadn't seen any useful difference in performance but now BI were saying its not client side, that the issue was with server performance impacting the client. We provided some rudimentary data in the Autumn (player numbers, AI numbers, mission details, mods) and were asked to use the profiler. We did, we got quite a bit of data and we put it on these forums and provided the full profile files. In addition I wrote some software to sum up methods with the same name (to provide top lists) and then compared the growth throughout a game. As a developer personally on one of my projects I could have done quite a lot with that information assuming the profiling data was easily mapped to code including guessing where the problems came from. I have done a lot of performance improving work in my years as a programmer so I was pretty confident that good things would come from it. Then we went into hibernation awaiting the fixes just as this wonderful video has suggested everyone do today, be patient because its coming. Its been 15 months and it hasn't. Performance as far as I can see is still all about the server, this King of the hill tweaking and testing seems focussed on the server side performance mostly to tweak the binaries which is why only BI run it. I felt like we did everything BIS asked us to and they ignored it all. They haven't ever responded to that post, my analysis is the last thing said in it.

So I came back after a long wait and I called them out for not responding, not addressing the issue. I did so with some more rudimentary data and a link back to the original post and data and profiling information and asked how things were going and if there was anything more we could do. I got no response from BI again. But the community here has gained a second wind of complaints since I did it, real information about the underlying causes of the issues are really easy to see from my screen capture of one of the profiles I took and I think that really irritates people especially considering how long ago it was we did this.

Then we get this video in response, which I kind of view as how BI talks to us back, because where it comes from and who is pushing at the heart of this gets lost. How do you think that makes me feel given what we have done and what we are committed to doing in the future (helping fix performance anyway we can). I am not this toxic individual that is mean and horrid to everyone, I just want a genuine problem that I have had with the game since I first played it in Alpha addressed. I have the right to request that and if BIS wont fix it I would like to know. But I can't just sit on my hands for another 6 months and see if they choose to deal with it, mostly because I don't believe their existing strategy with the King of the Hill scenario is likely to help my community at all. Then I see they are adding a boolean to the game that says if its vanilla or not which the sole purpose of which seems to me to be to blame the mods and scripts for performance (very likely to be the cause). I just don't see this combination of actions addressing the issue that the game appears to have, my anaylsis suggests that its nothing to do with the server and entirely CPU dominated on the client and in a few key 100% single threaded activities.

We don't have to be jerks or even really complaining about performance. But being quiet doesn't get us anywhere either. I want them to change their strategy because I believe we have a community that can solve this faster if they will enable us to do so, or at least solve our part of it. We can't just help ourselves in this case, they hold all the code for profiling the script engine. Some of this problem is theirs, some is ours and we both need to get cracking on fixing it from data obtained in the games played every week by thousands of communities where the real problems manifest. One of the problems I have is that testing mods for performance is extremely tricky as its a lot of event driven code. Without the ability to do it in the moment its basically impossible to setup the scenario that will show the problem in a controlled condition and slowly eliminate all the possibilities without changing the scenario (how do you see if the medic script causes performance issues in a repeatable way, I just don't know).

So hate me if you like for inciting a lot of this complaining if you will but I do it out of love for my most played game, a game I simply want to run at 60fps and never below 30. I am here wanting to help because a problem a lot of people agrees exists has been present since the first day of alpha, nearly 2 years now and the efforts we have made were ignored and the existing strategy for the last year has failed to deliver. Be nice and pleasant about it, I don't believe in hate, but I don't want people to stop talking about their main issues with the game, without feedback BIS gets nowhere and they certainly wont change direction. Discussing the options we have, how we go about fixing this situation has helped me understand what I need and more data is necessary.

For example we found a few days ago that performance with mods was about 20% slower just on an empty map verses vanilla. In the past we hadn't found any appreciable difference and this was the first time we could obviously see the mods having an impact in a basic scenario. We tracked it down to AGM. We don't yet know which module, we certainly don't know which script and I don't even really know if we will be able to get beyond modules based on what we have since it will require a deep knowledge of the code. That one was relatively easy (if you call launching the game 28 times with different mods and testing performance over and over easy of course) but once the dead bodies are piling up and the firefights are on going and there is still a problem its going to be impossible to apply the same strategy of isolating the mods, we can't do that with real players or complicated scenarios. I want this community to start testing, make benchmarks that show bad situations, report single player missions that perform awfully etc etc. There is a positive productive way to report issues and "complain" that is extremely important. For example does the SP still perform awfully in the introduction? I can't say I have played it for ages but I remember fps being terrible in single player and at the time I was just saying "You don't need profiles from us, just run the single player and 30 seconds in its like 10 fps". So I hope that this video doesn't drive people awayfor another 6 months, because I fear it'll be another 6 months wasted waiting.

Edited January 27, 2015 by BrightCandle

[pharoah 65]

I must've missed something - what is the 'next DLC' and future of A3 other than the sniper pack? funny to see people complaining about how big Altis is. Friggin 12 months ago (or whatever), everyone was bored with Stratis because it was so small!

No, BI...what we need is a jungle map - N'zwasisogo or whatever its called is a good example. We've done the open areas, now lets have a 'closed' area (jungle, swamps, rivers, creeks, snakes, crocs, mossies, etc) where I can be patrolling 10m away from the enemy and not know about it until it was too late ie. Vietnam!!

[the_demongod 31]

I must've missed something...

You did.

[Imperator[TFD] 444]

I must've missed something - what is the 'next DLC' and future of A3 other than the sniper pack? funny to see people complaining about how big Altis is. Friggin 12 months ago (or whatever), everyone was bored with Stratis because it was so small!

No, BI...what we need is a jungle map - N'zwasisogo or whatever its called is a good example. We've done the open areas, now lets have a 'closed' area (jungle, swamps, rivers, creeks, snakes, crocs, mossies, etc) where I can be patrolling 10m away from the enemy and not know about it until it was too late ie. Vietnam!!

I think you meant to say: "what you need is a jungle map...."

Frankly I couldn't give two hoots about a jungle map and find that terrain type immensely boring so please don't speak for "we".

I love dyslexci's video and I love how crazily passionate he is about this game and it's community. I too am well and truely tired of seeing all the various media feeds of BIS being spammed by moronic trolling and agree with almost all of the points raised in this video.

As for Bratwurst: I was playing a 27 player mission against some 160+ AI on Saturday night and experienced no performance loss whatsoever. So saying that the footage seen in these videos is not representative of what can be achieved in the game I say "try harder."

[lexx 1331]

even though public arma is absolutely nothing like what he depicts in his videos.

Well, most of the time public servers are nothing like that. Usually they never feature much of that, because they are full of people who do their own stuff. ArmA is hardly the type of game that can be played on a public server with random people in a huge kind of tactical way. But that's not a problem of ArmA... this happens with all such games on public servers. The hell, back in the days when Counter-Strike was still a thing it was usually a mess on public servers, because everyone wanted to be the lone-killer Rambo style.

Long story short... complaining about ArmA not being what it should be on a public server is a bit silly, imo.

/Edit: About that jungle map... Considering I sometimes find it already hard to spot an enemy on a "normal map", the jungle setting sounds horrible unfun to me.

[pharoah 65]

I think you meant to say: "what you need is a jungle map...."

Frankly I couldn't give two hoots about a jungle map and find that terrain type immensely boring so please don't speak for "we".

Jeez, some people are really touchy...okay, 'we' EXCLUDING 'you'. Better? :)

[gibonez 18]

Complaining about things is not a bad thing so long as it provides meaningful criticism that the devs can digest and work on.

[TheMightyKovacs 6]

Jeez, some people are really touchy...okay, 'we' EXCLUDING 'you'. Better? :)

Well , I for one wouldn't like a jungle map, however I can only speak for myself. Saying we is just including all of us (aka the community without specification), and you can't really do that, unless you know what we are all thinking. But I digress!

[pharoah 65]

OH MY GOSH. Guys...FOCUS. I'm throwing a suggestion out there that maybe the devs should consider doing a jungle map instead of a 'flat' one like what we have with Altis. Whether I use 'I' or 'we' or 'her' or 'them' or 'those blokes over there', it doesn't matter. Debate the merits of a jungle map, not my use of grammar.

[TheMightyKovacs 6]

Yeah sorry. Anyways. I don't think they'll move away from the whole Mediterranean themed idea. But that's just me. I won't complain either way they choose to go.

[fabio_chavez 103]

Although it sounds like we might be getting some of that in the next DLC, I fail to see what that has to do with the topic?

and i fail to see what your question has to do with the topic.