Timing improvement

[HASTA]

1/1000 of a second must be calculated in the game.

Times like 1.31.984 must be visible.

[R3DMAN]

they are arnt they?

[tristancliffe]

Nope. Currently they are only the hundredth - you must have seen Chris!

Do you know why Hasta? It's because the physics engine runs at 100Hz. But your avatar leads me to think that you wouldn't have realised that.

And why MUST they been calculated? Will it make any difference to the racing?

[StewartFisher]

Do you know why Hasta? It's because the physics engine runs at 100Hz. But your avatar leads me to think that you wouldn't have realised that.
Your point makes logical sense, but the physics engine in GPL runs at 288Hz and it records times down to 1000th of a second...does that mean the 0.00x figure in GPL is a fudge?
BTW, in-game GPL only displays 100ths of a second, you have to look at the stored lap times to find 1000ths.

I'd quite like to see 1000ths of a second introduced, if only because it's fairly standard across high-level motorsport. Incidentally, the timing system in F1 is accurate to at least 10,000ths of a second because that's what they use for the speed trap figures...they choose to display only 3 decimal places for timing.

[MyBoss]

I would welcome this, but its not the big issue for meh.

[HASTA]

But your avatar leads me to think that you wouldn't have realised that.


I didn't understand this, what is the connection of this thread with my avatar???

[Bob Smith]

You probably need to be here a while to get that one. :)

[HASTA]

You probably need to be here a while to get that one. :)
I'm here for one and a half year. I got what he's trying to say with this simple joke, but i didnt understand why such a meanless connection between NOS and Live for Speed. As a "SIM RACING" team, NOS Racing Team is waiting for an apologize from Tristan.

[tristancliffe]

You'll be waiting a loooooooooooooong time for an apology from me!!!! For starters I have done nothing wrong. And as you didn't understand me, you can't prove that I did do anything wrong either.

As it is Nitrous Oxide (NOS is a trademark of one particular firm, and by using their logo you might run into difficulties, especially if you ever managed to get sponsership, or make money from it) is solely used by poor drag racers (poor in terms of money, who can't afford to go Top Fuelling :P), and spotty 17 years olds with Vauxhall Novas, who have a (combined, i.e. all of them in the world ever) IQ of 17.3.

Whether or not you fall into that category I cannot say. But your avatar, and the impressions one gains from it (e.g. you think I'm an effeminate softy from my avatar) led me to the conclusion that you didn't have the gumption to link 100Hz to 0.01 timing intervals (although I may have been proven wrong on that point, case pending).

I would apologise, but by NOT understanding, and then demanding an apology despite not understanding you have confirmed by initial opinion. If you're lucky I'll get Ian H to come and have his say as well, as he can be, sometimes, even more blunt than me.

Welcome to LFS. Where everyone is appreciated. Unless they rice, cheat or swear online. Or use Nos. Or complain about tyre wear while drifting. Or suggestion blantently obvious features that they MUST realise someone will have suggested 3 years ago, and every week since. And yet, despite this, I still have good relations with many many LFSers.

So there... *sticks tongue out, and walks off*

[the_angry_angel]

Thousandths are measured (see the InSim thousandth's variable), but not displayed.

Edit:
:lovies: Tristan :D

[HASTA]

1- Vmax is a motorcycle brand owned by Yamaha.
2- I never thought about this hz thing, i forgot about it, thanks for your help.
3- Come to see our demo server sometimes to learn NOS Racing Team, or check our website and our forum.
4- We never ever ever drift!!! NOS Racing Team is a BL1 XFG nutters team.
5- I just asked why LFS are not using 1/1000 of a second. You gave the answer by "It's because the physics engine runs at 100Hz."
6- Please speak wisely and to the point, thanks.

[tristancliffe]

Yes Vmax is a brand owned by Yamaha. But vMax (or more accurately V<subscript>max, but we can't do this on forums/LFSW/ingame) is a generic physics term for Maximum Velocity (although we are called Velocity Maximum Simulation Racing, but I'm not 100% sure why). If your going to play smart, you'll have to research vMax properly (the term, not the team) first :P :D

I never said YOU drift, it's just the impression NOS gives.
I have no desire to know about demo teams. As far as I'm concerned LFS is very very cheap, and as such anyone committed enough to form/join a team should just pay for it.
I do speak wisely(ish, and of course imho), and I got to the point but ensured accuracy at all times.

[the_angry_angel]

To clarify, as far as we are aware, the external engine loop runs at 100Hz, but the internal loop runs at 2KHz. I'd be happy for Scawen to correct me, but thats what hes been quoted as saying at RSC. And yes, again, LFS does record thousandth's but its not displayed to the user. The thousandth's information is obtainable through InSim, so it cant just be making it up.

[Bob Smith]

While I believe that is accurate TAA, I thought the car is only moved forward on the external loop, while the suspension, forces and tyre physics and done on the inner loop? That's the impression I got.

[the_angry_angel]

The way I think it happens is that the inner loop also moves the vehicle, as well as doing sampling. But the bounding box sanity checking (i.e. intersection checking) only happens in the outer loop

The evidence for this is that we can get "inside" things and then pop back out as the game realises you shouldnt be there. Assuming Scawen uses this method, you get the accuracy of thousandths!

[dUmAsS]

The way I think it happens is that the inner loop also moves the vehicle, as well as doing sampling. But the bounding box sanity checking (i.e. intersection checking) only happens in the outer loop

thats plausable, because collision detection eats more cpu time then rigid body simulation

[the_angry_angel]

Exactly Tux :up:

[dUmAsS]

where is my cookie! :p

[the_angry_angel]

You can have a cookie on Oct 22nd :p

[tristancliffe]

Over there -------->

Edit: Damn Mr Angel keeps beating me...

[the_angry_angel]

<laugh type="evil">hohoho</laugh>

Cookies for all on Oct 22nd!

[dUmAsS]

lol

[GeForz]

And yes, again, LFS does record thousandth's but its not displayed to the user. The thousandth's information is obtainable through InSim, so it cant just be making it up.
So why is the thousandth's byte always zero?

[Bob Smith]

Yay. I'll buy a box of cookies to bring with me just incase, for I am the cookie monster in disguise. :D

What was the topic again? :p

[ayrton senna 87]

bob, crumble them up and throw them on track, then u will see some crashing action :D

[the_angry_angel]

So why is the thousandth's byte always zero?
Feel free to correct me if I'm wrong, but I'm pretty sure that its not always 0. I must admit I dont use InSim very often.

[Bob Smith]

bob, crumble them up
Blasphemy!

Nooooo, my precious! :p

[tristancliffe]

bob, crumble them up and throw them on track, then u will see some crashing action :D

WRONG

Crumble them up, put them on the track, and watch me stop to gather all the bits up and eat them... :P <---licking lips

[Dygear]

I would love to see this. There have been to may races where LFS had a 'tie' thanks to this.

[kwijibo]

+1 for this! :)

[duke_toaster]

Be great for drag racing, and leagues that do short races (like the league I am trying to organise (that no-one has signed up for)).

+1

[wheel4hummer]

Why? Isn't .xxx enough? What would be the point of the extra decimal place? The higher the frequency of the physics, the more CPU intensive it is. If we had a 300Hz physics engine, then you would need a proccessor 3x as fast just to get the same performance. This would mean that if your CPU was being used more, your framerate would lower, and more RAM would be eaten up.

[Vain]

I don't like doing my pb more than twice and that happens to me quite often. So I'd really like LFS to measure and show thousands. Though I won't scream in terror when the patch doesn't bring better time measurement. :shrug:

Vain

[Ian Wright]

For the love of all that does not suck, implement thousandths

It's all I want really. Is it difficult?

[keithano]

I wanna suggest this too. Thanks for bringing up the old thread.
Viewing the Hotlap time and some other high-standard league trackers (like OWRL), you can probably find that the time is equal thro.

[Hyperactive]

As said so colorfully earlier by our lovely community heroes it is probably dependant on... ehm... many things. Anyways, for qualifiy events, hotlapping and last laps on racing this would be good. It is possible to finish at the same 0.0x time but quite hard at 0.00x :)

+1, low priority warning and imho LX8 is more important ;)

[Sketchyrollin564]

+1, low priority warning and imho LX8 is more important ;)

More rally cars and tracks would be more important!:razz:

[Ian Wright]

I would like to add that some people I've chatted with in IRC see the lack of .00x's as an excuse to call LFS "not a proper simulation". I disagree with this but it's something to think about.

[boosterfire]

Altrought I do see an unusually high number of "same best lap" in races, I have the feeling that a new colision and impact absortion system is more important than this decimal.

[SpaceMarineITA]

Ok, there are lot of more things that can be considered more important of this, but i really would like to see .xxx measurements on split and lap times implemented.

[wheel4hummer]

I just went into LFS and tested. If you crash into a wall at a speed higher then 200kph, then you just go right through. I think that the tire collisions are done at a higher frequency then body collisions. What is the relation between speed and detection of collisons?

[bbman]

Tbh, I don't really miss that third decimal... Hardly ever I see identical times, and as other truthfully pointed out, there are things way more important to improve...

[ajp71]

LFS seems to do the results list in thousandths/absolute frequency terms anyway, I've had photofinishes where the time sheet reads a +0.00 gap but LFS still puts the correct car as finishing first.

[VIP]

I too would like to see the third decimal implemented in the game.

Does anybody knows if the developers ever talked about implementing it in the future, or the case is closed?

[Renku]

Does anybody knows if the developers ever talked about implementing it in the future...
They rarely talk about the future of LFS. Maybe you'll see it in the next big patch or maybe you don't.

PS: Yes, we know about AI change, SO and WE improvements + some other things, but we won't have any clue about these smallish things. You'll know after S2 final!

[nesrulz]

GO GO GO third decimal!

[NightShift]

Actually the bit about internal vs external loop was quite interesting :thumb:

[Zen321]

This thread is the most bumped thread ever :
2005-2006-2007-2008

[Rob_NL]

Well I just came back from a 3 year absence into LFS and I must say If this software wants to be called a SIMULATOR I think the devs need to get on the stick and improve the timing.

Its not just inconvienient to be consistant and have a lot of times the SAME lap time but its needed. Look at formula1 they are talking about going to .xxxx digits due to much closer performance between cars.

Comon DEV'S please show us you agree and update our favorite SIM to .xxx timing capability.


Rob

[SidiousX]

As it is Nitrous Oxide (NOS is a trademark of one particular firm, and by using their logo you might run into difficulties, especially if you ever managed to get sponsership, or make money from it) is solely used by poor drag racers (poor in terms of money, who can't afford to go Top Fuelling :P), and spotty 17 years olds with Vauxhall Novas, who have a (combined, i.e. all of them in the world ever) IQ of 17.3.




wow... that's not ignorant at all

[SidiousX]

wow...just realised how old this damn thread is lol

[Keling]

Better for drag / qual / hotlapping.　+1

[Dygear]

Thousandths are measured (see the InSim thousandth's variable), but not displayed.

Am I missing something, because I can't find this ... I did a string search for thousandths in the InSim.txt file, and could not find anything. Was it removed?

Incidentally, the timing system in F1 is accurate to at least 10,000ths of a second because that's what they use for the speed trap figures...they choose to display only 3 decimal places for timing.

Please cite your source for that, I find it very interesting.

While I believe that is accurate TAA, I thought the car is only moved forward on the external loop, while the suspension, forces and tyre physics and done on the inner loop? That's the impression I got.

To clarify, as far as we are aware, the external engine loop runs at 100Hz, but the internal loop runs at 2KHz. I'd be happy for Scawen to correct me, but thats what hes been quoted as saying at RSC. And yes, again, LFS does record thousandth's but its not displayed to the user. The thousandth's information is obtainable through InSim, so it cant just be making it up.

I remember him saying this as well and again, if there is a way to get access to the thousandths, I'd be very interested.

This thread is the most bumped thread ever :
2005-2006-2007-2008

And now 2010.

[amp88]

A bit more discussion on thousandths starting here (http://www.lfsforum.net/showthread.php?p=1408565#post1408565).

Thousandths are measured (see the InSim thousandth's variable), but not displayed.

Well, times reported in InSim have the scope for providing thousandths precision, but everything I've seen coming directly from InSim has a 0 value for the thousandths. It seems to me it's more of a placeholder for moving to thousandths in the future.

[NotAnIllusion]

Am I missing something, because I can't find this ... I did a string search for thousandths in the InSim.txt file, and could not find anything. Was it removed?
Yep, the timing has changed since 2005. http://en.lfsmanual.net/wiki/InSim_PHP5-Tutorial#MSHT

[amp88]

Yep, the timing has changed since 2005. http://en.lfsmanual.net/wiki/InSim_PHP5-Tutorial#MSHT

Were the values for the thousandths ever non-0?

[NotAnIllusion]

If I'd have to guess, it'd be a no. I linked that because it used to be in the insim.txt. Anyhow, since timing is now in ms, all that's left to be done is for the server & client to communicate the digit :shrug: On a tangent, ever driven a hotlap with the exact time as a benchmark? There I was thinking I'd got the time, but upon uploading I was still +0.003 away.. :(

[joshdifabio]

If I'd have to guess, it'd be a no. I linked that because it used to be in the insim.txt. Anyhow, since timing is now in ms, all that's left to be done is for the server & client to communicate the digit :shrug: On a tangent, ever driven a hotlap with the exact time as a benchmark? There I was thinking I'd got the time, but upon uploading I was still +0.003 away.. :(

I guess that's just a rounding error. 103% of x.x10 = y.yy3.

[amp88]

I guess that's just a rounding error. 103% of x.x10 = y.yy3.

Yep, that's how I've seen it explained before. I'm open to being told I'm wrong (as long as proof is provided) but I believe LFS has never provided thousandth precision.

[Dygear]

I believe LFS has never provided thousandth precision.

If that's right I won't feel like such a half whit.