Board index Games... Mods... Multiplayer! Deus Ex Questions about DeusEx ]

Framerate based action times; discussion


Author
Message
synthetic
Seraphic/8X
Seraphic/8X

Posts: 1321
Joined:
 27 Apr 2007, 15:29


Framerate based action times; discussion - Postby synthetic » 02 Jan 2013, 06:00


So I've been arguing, discussing, pondering, wondering about this topic for a while now, since after playing it so many years my favourite online game's increasing fire rate issue raises some emotions.

This thread is intended for those who have knowledge and experience to contribute, for those smarter than me and for those as passionate about it.

Nobody I've talked with has a clear idea of what is wrong, how this fire rate issue came to be, and how to stop it.

Given that we are talking about score and results based online shooter game, I believe I am not overly dramatic when I state that the current situation is the worst possible thing that could happen before all master servers being taken down.

This game is no longer about skill, but about who can configure, purchase, overclock, and otherwise optimise their hardware better, with the game and scores becoming an arbitrary feedback on your computers performance.

1. First big question is, why is it happening? What is it?

And technically, scientifically, I am looking at the more computer-savvy people here, coders and hardware fans alike, because I myself only have my uneducated theory of what is happening and why.

I believed that dxmp had stable base values for action times, event times. What we are witnessing here is contradicting that to a point where the former assumption cannot be true. I read that certain games (Quake3A for example at 125 FPS) base their action times and events on certain intended FPS cap, using it as time reference I presume, and the way DXMP gets accelerated seemingly in sync with FPS makes me think this is the case here as well.

This makes one wonder where did the cap go. Currently, this game accelerates with framerate, and without any enforcable cap. I recall that between some patches or mods (maybe MTL, actually) the choices for NETSPEED settings changed, reflecting better the contemporary needs. I suspect maybe it is that, seeing how lower netspeed values nicely cap FPS as well. Maybe, somehow, it also happens to be that this online patch simply doesn't have any sort of fps cap beyond the hardware capabilities of year 2000. Seemed fine on PCs back then and nobody bothered. Maybe.

2. Second big question is: are the action times absolutely synced to FPS, or are they dependent on additional/other elements beyond FPS which would be easiest to tackle?

There are so many myths, doubts, ideas and opinions about it. I myself have had conflicting thoughts on how it might be working. Right now it somehow makes sense to me that the acceleration is indeed absolutely bound to FPS, and individual hardware only contribute to the length (max possible FPS) and speed (how fast FPS value increases between rendered situations) of this acceleration.
If this was true, if someone here can confirm it with proper, technical opinion and maybe facts, then shitload of myths would fly right out of the window. Additionally, the nature of the problem would be easier to comprehend and fix.

From the other viewpoint, if it WASN'T linear acceleration synced to FPS, but instead exists as some kind of performance phenomenon, allowing someone to reach 2x fire rate at 200 FPS where the next guy reaches 2x fire rate at 150, then.. we need other, more thorough and time-consuming solutions to the issue.

The second idea has been the main suspicion in the community, one that I too used to wonder about, but now if you compare it to the first idea there, then it seems like a big fat fairy tale for the dark ages that most of us still live in. But, it could never the less be true, for some technical reason.

3. Third big question is: how do we find and enforce the default action and event times?

My experience, live-action gamer-style testing (ridiculously inferior to the methods used by good coder or engineer), and the idea of action times decreasing in sync with the increase in framerate value, all makes me think that the base action times might be capped to some very small FPS value, one that may even not function as intended due to the processing power of contemporary, relevant hardware.
This makes me suspect that trying to enforce default values through FPS cap may not work with satisfactory results; ie we all cap to 100 and play accelerated game (faster than default), or are forced to play at maybe 30 fps, unthinkable for most of us.

On my PC, vsync ON and @ 60 fps, lot of the timers felt "like they should", but as Blade pointed out, the mouse is experiencing some lag. Additionally, Blade was able to test vsync ON @ 75 FPS, and reported faster scoping times. Our opinions part when it comes to the value of the capped fps, however.
Interestingly, with FPS capped @ 60 fps, vsync OFF, the game (and IMHO the fire rate) was significantly faster. Faster enough to turn me from struggling and slowed veteran (60 fps+vsync) into ridiculous killing machine (60 fps, no vsync).

I don't understand fully why vsync has the effect that it has over similar fps value with vsync disabled. I fear that it might show the impossibility of fixing the fire rates (and all other times) through fps cap alone.

I would proceed to claim that every higher FPS margin results in faster everything (making everybody cheaters before everybody with lower fps value of any kind, in any range), including fire speed, but it is you, guys, who I turn to in hopes of finding some confirmation to any of these doubts, questions, and maybe even allowing the community gain better insight into possible solutions and the games future.

Apologies for big post, but it is a big topic and I did my best to keep it as clear and straight forward as possible. I want skill back in deus ex, internet has enough benchmark software available already.


ShadowRunner
It's over 5000!
It's over 5000!
User avatar

Posts: 6970
Joined:
 28 Feb 2006, 21:07
Nationality:
Japan


Re: Framerate based action times; discussion - Postby ShadowRunner » 02 Jan 2013, 11:58


If I can touch on the bigger picture here. Yes it's a problem, but this problem is really mostly a problem because of the way people have changed DXMP from the original.

If people were actually making 0-aug maps that are Deus Ex maps and actually making it a viable proposition for a beginner, I'd have more sympathy for the cause. Currently the 0-aug game/server hosts maps where there is nowhere to hide at all, no shadows, no explosives, no devices, no codes, rarely any hacking/tooling, no water, no elevators, no NOTHING. It's not interesting for anyone, except ppl concerned with ping or fps or rifle reload speed. With a map like DXMP_CMD, the issue is far less than on a map like Pool.

Many elements of Deus Ex are totally missing in the 0-aug servers. To me, that is the far more serious problem than fps abuse. The fps is amplified, because we're playing on maps so small, only an assault gunner could survive.

There's no hope for explosive guys, stealth guys, low tech guys, every single map is tiny, dominated by rifles and guns, low ping, fps etc... now you guys are trying to fix this, when really, if you just hosted actual Deus Ex, it would be LESS of a problem.

I can't fix fps, but I can suggest rotating weps on server or having co-op servers.


synthetic
Seraphic/8X
Seraphic/8X

Posts: 1321
Joined:
 27 Apr 2007, 15:29


Re: Framerate based action times; discussion - Postby synthetic » 02 Jan 2013, 14:41


Look at Original DXMP maps, Shadowrunner, and then you know who to point finger to. I agree with you fully, but it has nothing to do with the topic other than you saying "not worth fixing". Also, Deus Ex is is made special by variety of significant augmentations, whereas skills and all the deco is generic rpg elements. 0augs doesn't even have that. You saying about real Deus Ex being hosted has nothing to do with the reality that the developers never made this game into a proper online shooter. Your words of real deus ex are likely just a reflection of your personal, subjective and biased opinion. If you code your little RPG city and change augmentations so they dont speed you as fast, protection as much, or cloak you as well, you've already deviated from the original vision, for example.

I posted the thread to learn from people who know the gaming mechanics, game engines, code, or have in-depth knowledge of how hardware affects a game this way, also how to fix it. Let us keep it on topic, please.

Also, your suggestion of rotating weapons still makes the game unfair for all players besides the few that coincidentally prefer same fps range or get similar fps. If your rotation puts pistol in, there are still those who shoot pistol faster, reload it faster, swap and activate things faster.


ShadowRunner
It's over 5000!
It's over 5000!
User avatar

Posts: 6970
Joined:
 28 Feb 2006, 21:07
Nationality:
Japan


Re: Framerate based action times; discussion - Postby ShadowRunner » 02 Jan 2013, 17:40


I'm not saying don't fix it, in fact it could do a lot to increase player numbers, just I'm saying it fixes the game for some people, but for others it is not a major concern or the thing that is ruining their game. Right you are though, a coding discussion.


~][FGS][Nobody~
Illuminated
Illuminated
User avatar

Posts: 5359
Joined:
 26 Oct 2005, 16:59
Location:
 Schland!
Nationality:
Germany


Re: Framerate based action times; discussion - Postby ~][FGS][Nobody~ » 02 Jan 2013, 20:22


Hidden's right. Your topic is definitely relevant, Rollo, but it's beyond the scope of this topic.


Nobody is perfect...
-----------------------
ô¿ô¥[GODZ]¥NOCHANC wrote:I can ban any one I want ANY time I want. You have no rights here.

Magus wrote:Maybe one day I will understand your arcane rituals of voting :)

chin.democ. wrote:You can use light bulbs that emit light, and when shot, do not.

synthetic wrote:and while every person is interesting in their own unique ways, there is some degree of uniqueness that a doctor can help with.

~ô¿ô~][FGS][Nobody~ said: THERE IS NO SPOON!
~¤¥ÐJ¥¤~ said: THERE IS NO CAKE!
TheWolf
Principality/2C
Principality/2C
User avatar

Posts: 226
Joined:
 30 Nov 2011, 03:21
Nationality:
Canada


Re: Framerate based action times; discussion - Postby TheWolf » 03 Jan 2013, 05:11


I play less, play other games more, play RPG and fun metabolic end product more, and map more because of fps cheats. Since I started multiplayer back in about 2007 I played with 30FPS. About a year ago I got a new pc and it was then at 60fps. A week ago I turned the fps up to 90. 90 is the highest i've ever seen in a game, which is why I decided to put it there. Even the jumped from 60 to 90 made a noticeable difference. 150 would be too far in my opinion, although I have never tried it. I think we should limit fps, and do it as soon as possible. Removing such advantages like that makes the game more fair.


Ingame: Sev
Poor
Power/3D
Power/3D

Posts: 288
Joined:
 02 Jan 2011, 22:39
Nationality:
United States of America


Re: Framerate based action times; discussion - Postby Poor » 03 Jan 2013, 07:54


So I made a weapon that measures the client's fire rate and animation length. If you go to my cagematch server, you'll find an assault rifle near the shop. I would appreciate some numbers from players, especially from ones that "cheat".


synthetic
Seraphic/8X
Seraphic/8X

Posts: 1321
Joined:
 27 Apr 2007, 15:29


Re: Framerate based action times; discussion - Postby synthetic » 03 Jan 2013, 13:32


Poor and Nobody, you two are some of the main people I am hoping to get some comments from regarding this issue, is there any chance you could provide some thoughts about the problems mentioned?

Fundamentally, competitive sports and games lay down rules for action but also technical standards, so that it would be mainly one's ability that provides interesting results. Cheating in this game at this point is so significant that it does not even fit into the concept of "cheating in online games" rather than describes the general definition of "Cheating", because you can take an automatic weapon to revolver shooting contest, and fly a jet plane in a marathon, in dxmp.

There are no longer any rules or restrictions. When I run marathon I don't want to compete on my feet against someone driving an SUV all 50 kilometers.

Is this significant enough of an issue? I'll leave that question for you to answer quietly to yourself.

Apologies towards anyone who sees it as clearly as me and just considers the body of this post an unnecessary show by captain obvious, but sometimes it appears surprisingly common how often one needs to point out the obvious to get people's focus.

I also have to say that the discussion I wanted to spark with the opening post is necessary for people to actually understand and reliably use any mutators or gametype mods released. There are people who currently do not see any problem playing with FPS around 150, 230, etc, and have taken preference to these subjective and personal fps caps.
There are also wide doubts whether fps cap will equally solve the problem for the variety of hardware builds and configurations today. Some 4 years ago I heard my friends fire gets faster (noticeably for him) at 125 FPS, today I hear from another person that there is no noticeable difference at 150 FPS. These inconsistencies are extremely worrying, especially if I add in my own subjective opinion that the game along with its fire rate becomes faster from.. 40-60FPS onwards on multicore CPUs and powerful graphics cards.

On another note, to play DXMP in a playable manner many (if not all?) players playing on multi-core/multi-processor computers have to assign the program to a single core, or it'll be too fast. Still, the game accelerates. Old DOS games experience same issue, but to play them a lower cpu speed is emulated (correct me if I'm wrong) on something like DOSBox. What makes me curious is, what happens if you do something similar with deus ex, or even try it on dosbox itself? Wouldn't it be possible to achieve faster rates on something that takes advantage of such a configuration? If so, would FPS rate still be a tangible indicator of acceleration, or would you be able to reach 2x the speed at, say, 40-60 fps?

These suspicions, if they have any merit, would render Cozmos fix insufficient for 5-20% of active playerbase, whereas Poor obviously is taking more fundamental approach to the very nature of the problem. I suppose the most reliable way to fix the times would be to have the server tell the client the time reference, either by enforcing it or kicking for exceeding it. Fixing it clientside likely just calls for someone like Phantom to find a quick way around it, but for most of us anything is a god's gift at this stage.

I've written a lot on this topic, written passionately about it, but I want to make clear that my interest in this stems strictly from inherit disgust towards cheating. I used to have long discussions with Mole about the measures against aimbotting and cheating in the community even as I was a beginner. In spite of being called a cheater, aimbotter, etc, by a surprising number of players, I cannot stand any sign of it, and in retrospective my pings (60-90 for most years, 50-70 in recent 1-2 years) and framerates (10-40 for 3 years, 30-100 4 years, 30-130 for 1-2 years) have not granted me a significant advantage over the fellow players, often the opposite.


Poor
Power/3D
Power/3D

Posts: 288
Joined:
 02 Jan 2011, 22:39
Nationality:
United States of America


Re: Framerate based action times; discussion - Postby Poor » 03 Jan 2013, 22:36


From the data I collected (thanks to the testers and big thanks to Blade), the fire rate starts increasing beyond normal at around 250 FPS. At 350 FPS, the time between shots decreased to 3/5 of the normal time. When Blade got his FPS up to 500, his fire rate was triple (!) the normal rate.

Players who didn't change their FPS settings for the test got normal fire rates. I don't have data from enough players to know for sure, but I think most players have normal fire rates.

miniMTL attempts to limit fire rate to the correct value. It works on some weapons such as pistols and shotguns but it does not work on the assault rifle, flamethrower, or hand to hand weapons.

The solution is to create a mutator that replaces all weapons with fixed versions that will ignore a player's request to fire if not enough time has passed since the last shot. The server won't register their shots but can warn the player that there is an issue. Fast reloading and weapon switching can also be fixed.


~Blade~
Power/3D
Power/3D

Posts: 252
Joined:
 17 Jan 2010, 20:56


Re: Framerate based action times; discussion - Postby ~Blade~ » 03 Jan 2013, 22:43


awesome! can't wait


TétrãHýdrøÇäññÂßîñø£

Mastakilla wrote:If you bring even a thousand players to DX I will come over and suck your male sex organ.
I really will.

~][FGS][Nobody~ wrote:...you're doing the best to decrease my rationality at the moment.

jager774 wrote:He starts a fight and blames you for starting it and bitching on it. OKAY KEN.
Cozmo
Power/3D
Power/3D

Posts: 276
Joined:
 08 Apr 2010, 22:00


Re: Framerate based action times; discussion - Postby Cozmo » 04 Jan 2013, 09:38


Nice. Will we get a chance to do more testing with the values? Mine were usually slightly lower than expected (eg. 4.8/5), even with default D3D 60fps cap and Kenties, etc. I'm putting that down to vsync or something


CDX - A huge Deus Ex Multiplayer patch
Cozmo_RPG (v1 & v2)
MPConversations - A great tool for creating co-op or rpg content. It will never be used.
synthetic
Seraphic/8X
Seraphic/8X

Posts: 1321
Joined:
 27 Apr 2007, 15:29


Re: Framerate based action times; discussion - Postby synthetic » 04 Jan 2013, 10:15


Did some testing there myself (as gulag-[ren]) and it was quite fascinating. I have a strange question, though: is it possible to host the checking mod/mutator on another server, to make sure nothing else messes with the times (server rates, latency, god knows what)?


~][FGS][Nobody~
Illuminated
Illuminated
User avatar

Posts: 5359
Joined:
 26 Oct 2005, 16:59
Location:
 Schland!
Nationality:
Germany


Re: Framerate based action times; discussion - Postby ~][FGS][Nobody~ » 04 Jan 2013, 13:06


To give my two cents to this topic.
The best solution to solve all that timing issues would be client-independent timing checks on the server.
For instance, just executing the serverside called functions for shooting bullets or projectiles, if the last call was at least ShotTime seconds ago.
The same could be done for reloading. Independent of the reload animation (or whatever the client does meanwhile), client instigated shots are just accepted by the server when the ReloadTime has passed.
However, as Poor stated, all weapons had to be modified.

According to a second server. I'd add the mutator to the FGS Altfire server, if that helps.


Nobody is perfect...
-----------------------
ô¿ô¥[GODZ]¥NOCHANC wrote:I can ban any one I want ANY time I want. You have no rights here.

Magus wrote:Maybe one day I will understand your arcane rituals of voting :)

chin.democ. wrote:You can use light bulbs that emit light, and when shot, do not.

synthetic wrote:and while every person is interesting in their own unique ways, there is some degree of uniqueness that a doctor can help with.

~ô¿ô~][FGS][Nobody~ said: THERE IS NO SPOON!
~¤¥ÐJ¥¤~ said: THERE IS NO CAKE!
synthetic
Seraphic/8X
Seraphic/8X

Posts: 1321
Joined:
 27 Apr 2007, 15:29


Re: Framerate based action times; discussion - Postby synthetic » 04 Jan 2013, 15:00


I am quite puzzled by the results we got from Poor's tests, though. I have more questions than answers right now. Although I'd like to chat with Poor about couple things to check, right now it looks like playing with FPS from 60 (100) to 200 (220-250) actually gives a handicap, because if the enemy is not shooting faster (or has some other advantage) then I sure as hell DIE faster. :S Really confused about what is going on here, I figured fire rate increase would be gradual from early frame rates.


chin.democ.
chinny!
chinny!
User avatar

Posts: 2859
Joined:
 10 Aug 2006, 17:19
Location:
 London
Nationality:
Great Britain


Re: Framerate based action times; discussion - Postby chin.democ. » 04 Jan 2013, 15:16


A mutator sounds great, like for like changes no fuss (well not for me anyway! :P). I tested ~30 fps vs ~100 fps (me looking at a wall rendering about 1-3 nodes lol) and there was no change in speed, so in my limited short tests I concluded that the results only change after well above 100 fps to support the general theory. So looking at Poor's data you could conclude that someone will have the same shooting ability with 30 fps as someone with 200 which is interesting.




Return to Questions about DeusEx

Who is online

Users browsing this forum: No registered users and 5 guests



cron