In-game speed

[Koriban]

So, what determines this currently? Before it was if you were on a Single Core PC with ACPI on Windows XP.

Now I know you've limited the frames, but weirdly certain people skip (in-game running faster than others) whilst certain people don't and it appears differently to other people.

It's 100% not down to specs of computer, as I'm by far on the best specced computer and don't notice myself skipping ahead of anyone. Is it still better to go single core? What optimisations have you made to the Code that's different to 2.3 standard?

Sadly this is an important aspect of Mir that could do with answering, as Speed in-game is often what makes someone win a fight. I was hoping this didn't exist and everyone was the same speed but I guess that'll never happen lol

 

[xXxX]

Yes this is a good question as i have been wondering if to disable like 3 of my cores out of the 6 or maybe even 4 as mir is single threaded.

/KENPACHI

 

[Insinia123]

What spec is your comp then interested to know why you think yours is " by far the best specced" ?

 

[FROZEN5]

What spec is your comp then interested to know why you think yours is " by far the best specced" ?

Well he must have something better than a dual 780 ti

 

[jez]

its all about ping, 8ms and below you will fly

 

[ILovePie:D]

its all about ping, 8ms and below you will fly

Wouldn't mind server IP to check my ping :P but guess it's off limits to everyone who aint testing :'(

 

[LionMir]

Lol best specced pc haha, im sure there is someone on here with better

 

[Carribean]

weirdly people skip past me on win 7, if i use xp im faster but when in large groups i get lag so god knows

also any idea why it takes considerably longer to log in using windows Xp ?

 

[Samuel]

PC Specification will mainly determine how quickly/well your computer handles drawing the images etc (which is limited at 60 frames per second as this is 1: the maximum the human brain can compute and 2: what UK based monitors are limited to).

Network ping will determine how fast the messages get from the client, to the server and back to the client...

e.g. if your ping is 8ms then the following would happen:

- Click to walk
- Client sends walk message to server (which takes 8 ms).
- Server computes message and responds with new cords (takes another 8ms).
- Client receives message and draws character moving (Pc spec determines how smooth this looks).

There are other contributing factors, but this is an overview of how it works.

The best person to actually provide a detailed explanation of how this works on the AM2 files is thedeath as he spent some time looking into it.

Sam

 

[Koriban]

What spec is your comp then interested to know why you think yours is " by far the best specced" ?

http://uk.pcpartpicker.com/p/1FHZU

I'd be very impressed if so, is all I'm saying. What bothers me, is my much much much less powered Laptop runs smoother/faster (in-game) than this for Mir, which leads me to believe the code is very unoptimized for newer tech.

 

[blah1233]

All that junk in a poor looking case.

 

[Koriban]

All that junk in a poor looking case.

Work related, has to look "professional" and all that. I'd of much loved to get a nicer looking case but, can't complain

 

[Tai]

That case though. It's not even cheap! Lols

Why are you playing Mir at work with your MS Office Student pack?

 

[Koriban]

That case though. It's not even cheap! Lols

Why are you playing Mir at work with your MS Office Student pack?

MS Office Student pack at the time was a cheaper deal than the standard pack, and I don't play Mir at work.

Anyways, e-peen proven, is TD around to explain what's going on with these files?

 

[thedeath]

fine i'll do my best to copy paste a pm i just send:

Sam says your the go to guy for how to be fast so what do I need to do master?

quote from sam
"The best person to actually provide a detailed explanation of how this works on the AM2 files is thedeath as he spent some time looking into it."

lol he's right, i did spend alot of time on it.
thing is: it's not that straight forward
the files have quite a few checks to identify ppl being faster then others.
in general the whole idea that ppl who 'jump' accros your screen are faster, is a semi myth.
the system works by sending a walk/run every xxx ms, i admit some pc's where doing this faster then others .generaly it wasnt so much that they did it faster, more that they had 'lag' like behavior where the one moment you'd do a walk every 700ms, and then you had a spike of 200ms and back to normal.

the am2 server files are perfectly capable of detecting those spilkes, however since i never managed to make the client 100% accurate, it was decided not to deal with them (aka ignore cheating, simply cause there are others who dont do it on purpose).

i did notice my virtual machine client would have much much more spikes then my normal pc (both near identical windows 7 builds).
i think mostly cause the virtual machine was not an actual true processor (basicaly i think the virtual machine software was not getting full priority on the pc, so by 'lagging' the virtual machine it would spike the cpu for a few ms and cause invisbile speed increases)

as for all those morons saying someones faster just cause he's looking faster on someones client or whatever.
this is all an illusion.

am2 files dont have a fixed movement timer for every person on your screen, instead it's a timer used on your own char. basicaly if you happen to make a move halfway through the animation of someone elses: your client will '**** up' the timer and it'll either make that person's movement delayed or skip frames.
basicaly when you see ppl jumping spots, look like they are moving super fast: this is only possible cause of a bug in how mir (and am2 in particular) deals with the animations of any character on your screen, doesnt matter how fast someone moves: it's your own client that displays it wrong to begin with.

there's a few things to note (especialy on these files i guess):
- doesnt matter if you have a single threaded cpu or not (at least it shouldnt)
- how fast your ping is matters ALOT in mir (basicaly mir uses a rather outdated network system where you cant make your next action till after your previous action is confirmed by server), faster reply = your next action can be done sooner
- the fact that other ppl 'LOOK' like they run faster means nothing: this is all down to mir using an antiquated game engine that rely's on a single movement timer for every 'char' on your screen, including your own
- vmwares seem to have the ability to lag/speed spike the client (i think the speed spikes are basicaly vmware noticing the timing is off and allowing the emulated timer to catch up, this would mean that in theory you're slower overall but get those lucky times when you can do 2 actions almost simultanious during catchup moments

hope this helps

 

[Jamie]

Ping has a very small impact on peoples speeds, anything less then 200 shouldn't be noticable when running..

Example:

1. You click to Move.
2. You begin your animation.
3. You send the movement information to server.

IF you have not received a response from the server BEFORE your walk animation finish (Aproximately 500-600 MS) you will wait on the last frame OTHERWISE you will repeat 2 and 3.


At what point will your ping be taken into account when it is less then the movement animation?

 

[thedeath]

Ping has a very small impact on peoples speeds, anything less then 200 shouldn't be noticable when running..

Example:

1. You click to Move.
2. You begin your animation.
3. You send the movement information to server.

IF you have not received a response from the server BEFORE your walk animation finish (Aproximately 500-600 MS) you will wait on the last frame OTHERWISE you will repeat 2 and 3.


At what point will your ping be taken into account when it is less then the movement animation?

you might be mistaken there
i think i 'fixed' it in the ace files but: original design of the client, it wouldnt start the move animation till AFTER the server confirmed the move is ok (probably to sync your movements on everyones screen), it would also not start 'counting down' the time needed for your next move till after the confirmation.

dont kill me if i'm confusing details on this code but, think i changed it to how you said (although made it optional in code)

 

[Jamie]

you might be mistaken there
i think i 'fixed' it in the ace files but: original design of the client, it wouldnt start the move animation till AFTER the server confirmed the move is ok (probably to sync your movements on everyones screen), it would also not start 'counting down' the time needed for your next move till after the confirmation.

dont kill me if i'm confusing details on this code but, think i changed it to how you said (although made it optional in code)

If that was the case, no one would be able to walk through, there would be no repel/dash bug and there would be no push back because they wouldn't start to walk and get canceled. (if the server confirms before movement).


Also why do you attempt to walk when disconnected?

 

[LionMir]

Work related, has to look "professional" and all that. I'd of much loved to get a nicer looking case but, can't complain

You definatly dont have best specced pc out of chronicles I know atleast 3 people who have better and as for professional case, should of gone with a corsair 650d there very nice.

Please dont tell me you payed that for your pc aswell?

 

[Koriban]

If that was the case, no one would be able to walk through, there would be no repel/dash bug and there would be no push back because they wouldn't start to walk and get canceled. (if the server confirms before movement).


Also why do you attempt to walk when disconnected?

All of the above seems to be the case on this server except for "push back". This still happens. I've not had the displacement bug on this server yet.