just installed and crashed after displaying the online server list

EDIT: seems random, sometimes does, sometimes not

Q: will i have incompatibilities if a stay with patch S?

Last edited by Honey, Sat, 29 Apr 2006 14:29.

Quote from DEVIL 007 :

With your experiencies probably you should know that all ISP sometimes shaping UPD packets very lot.If there is too much on network of UDP they can automaticly shape it down.Its really easy.
And you should know that if UPD packet is lost you never get the same one again as its possible with TCP but just the next one.

There are always some priorities on ISP`s network and to be honest its not the home user.

i understand, and if that maybe true for my home isp, but for my company isp the internet connection has a minimum bw guarranteed and i mad sure the first test was on our "high priority" gateway so its unlikely that bw saturated at anytime (also holiday time) because on very first test i didnt reserved bw only for lfs.
let's assume some udp loss occurs, for any reason, as soon any lfs packet succeed to reach destination, then car position is updated and this happens every day when you see those guys that lag as a hell...there is huge packet loss, but a new packet updates car position.

the bug i'm talking about make lfs simply stop sending packet "forever"...it is "similar" to the bug that surely happened to any of us, when a server suddenlly lags every car forever until server restart...whilst this now rarely happens, it still happens and it encourages me to believe that is a very subtle problem that still is present either in server and client

Quote from DEVIL 007 :

You mean LFS client didnt send UDP packets or server part?
If I understood that was test made on LAN?

the very first on internet, the other on lan
about server or client, you arise a dubt in me, i always assumed it was the client, but i don't remeber my collegue specifically said it was the client, next week i will return to work and ask him if he remembers

Quote from afastest :

That is very good you tested it thoroughly. Often people report bugs before proper testing, OC'ed, faulty hardware, bad connection and what not. This is not the case with you here, but one has to state these things from the beginning so Scawen knows what kind of fault he is dealing with. Or maybe you did it already and I missed it.

i'm in the "ict" area of my company and you cannot believe the absurd stories and the most subtle problems that come out everyday.
my home isp has not the best performance, but is one of the most reliable (in the past used to work at home and had embedded net prototypes working 24/7 flawlessly with hundreds of connections), moreover my connection is "interleaved" and not "fast" so delivery of packets is supposed to be "guarranteed".

i remember i've reported this bug on an already started thread, but it died without conlusions iirc, i always assumed scawen put that on his "to-fix list"

Quote from GP4Flo :

@ Honey: What about CTRL-T, does this help?

never tried: what's that supposed to do?
(i will try it anyway the next time it happens)

Quote from afastest :

The only time I had similiar problems was when I had uploads running or heavy packet loss.

Try to ping the host IP when you have that kind of bug. Don't leave the server or track, minimize LFS and run command line. Then ping the host for example one hundred times. On windows it is ping -n 100 IP. Then you can go back to LFS and play and let the pinging run. Later see what ping is and how much there is packet loss, if any.

I know you said restarting helps, but you can still try that just to see what kind of ping you will be getting.

I use netlimiter to get host IP, and I can also see traffic graphs in it from LFS.

last summer when most was on holidays (and nothing on my schedule) at work i got a clean pc -> it happened again, then i programmed the company routers to guarrantee such pc 100Mb bandwidth ...-> it happened again and from router logs i didnt see any packet from such pc, then -> i installed ethereal on such pc and the next time the bug showed, ethereal logs (according to a collegue of mine who is a geek for such application) revealed that no udp packets where sent by lfs.

so the problem is either lfs alone or some sort of stack overflow lfs causes to windows driver.

tests at work were done with windows xp sp2, but at home i experienced it also with windows xp sp1, never tried other windows flavours

Quote from Scawen :

LFS hasn't changed in this area.

As I've said before, every time there is a new patch, there are a few people claiming that the patch has caused them some new network problems. Usually the problem goes away miraculously a few weeks later.

It's just a coincidence. Let's say 1 in 100 people with an internet connection, is having a problem with it, at any time. Now a new patch comes out and 10,000 people try it. 100 of them will experience connection problems at exactly that time, and they may blame it on the patch.

i know what you mean, but i'm not such kind of person and i try solve anything by myself first...as i said it's many month i'm experiencing this problem, there is also an old thread in the bug section related to this, as you can quickly verify here -> http://www.lfsforum.net/showthread.php?p=115379#post115379 (and the very next reply) an not the only one, it's an hard to recognize bug and it comes at least from patch L.

so fix the very urgent things and take the deserved rest, but please do not forget this bug, because people who experence this, simply don't figure it out and thinks that every s2 licensed racer is a stupid wrecker because everyone's ramming at him.

Quote from Scawen :

No, that is some strange problem with your ISP, not sending your UDP position packets to the host. Nothing I can do about that without some major changes.

i'm fine if you not plan to fix it for the near future (and thanks for answering me), but i cannot accept your statement that it's my ISP problem, it not happens only to me, also to people from other countries/ISP...if it's the isp, why exiting lfs and restarting solves the problem? and why it never happened till S2K and started since S2L? why happen also with other isp?, why other udp apps continue working fine without a glitch?
please do not jump at conclusions: if i'm losing time writing here, it's because i already verified it's an lfs problem...

Quote from axus :

Errr... I don't think noemfie was refering to your comment. Relax.

i apologise, but the question period of his post was a new line and seemed a statementent rather than a feature request, it completely changed the sense of the post, i pologise anyway

Quote from noemfie :

Probably not the right place to ask , but the 23 client limit , is it something the devs are working on to try and increase or is it just not possible atm ?

CAN YOU UNDERSTAND ANY OF THE WORD I'VE WRITTEN???????!!!!!
what this bug -> http://www.lfsforum.net/showthread.php?p=115379#post115379 has to do with cilent limit????????!!!!

Scawen, i'm sorry to bother but may i ask if fixing the "invisible car" bug is planned for patch U?
is just to know, i understand you have much pressure right now

Quote from Micha :

I made a small movie about the new tyre deformation:

http://www.pimba-racing.de/data/tyre_deformation.wmv

as i said i think the lfs deformation for road tires is a bit too much to me, but the real point is not the amount as much HOW...it seems that as soon as the deformation starts, the contact part of the tire seem to fold innerwise making cornering too weird and innatural

Quote from Ball Bearing Turbo :

Shut up

don't give him so much importance: he's slow! ...he's not even worth a smalltalk

Quote from GAS-Hugo :

This is exactly my problem - besides, that i can fix it by just leaving and rejoining the server.

... i´m wondering on how many others would have the problem, but won´t recogniced it, because everything seems to be ok - just that a lot of crashers are on track...

yeah you're right! it took several weeks to me to figure it out, actually i didn't, but was another racer that told me he cannot see my car because i just won the race few tenths of a second before him, but he tought he was the first because he couldn't see my car (i was thinking he was a dirty racer because in the last corners he was ramming me)

since this thread seems used to fix the current release from bugs, i post here:
tonight i experienced again (first time since patch S) the "invisible car" bug that was already addressed in the bug section a long ago.
i have patch S still.
just to remind what this bug is about, i summarize:
sometimes at some point when connected to a server my lfs stops sending packets and everyone else no longer see my car (just the "lag" writing and my dot not moving) to me is everything normal (except everyone crash into me because they don't see me) i can win races but if i'm too far from the first racer the server forces me into spectate mode (because it sees me not moving), leaving a server and rejoining (or joining another server) dosent solve, i MUST exit lfs and restart it to make it work again.

i know to happen to other people, even if i don't think is a very popular bug.
i hope it could be fixed once for all, thanks

Quote from Ball Bearing Turbo :

:doh:
Evidently...

But why?

maybe a thing called "privacy", someone may prefer to not show with his license name (at start i used another nick for that, but then i changed my mind) also because there are some people that watch your hotlaps and ranking to decide if you are entitled to talk

Quote from Glenn67 :

Forget for a moment the force the road is exerting on the tyre and instead think for a moment of the wieght (force) of the car acting through the tyre...

Now as the tyre (a standard steel belted radial on the outside of the turning circle which is under the most forces) is in mid corner with 1g of lateral force being applied it would have some roll over, which would would effectively mean that the outer wall of the tyre would be semi vertical and the inner wall of the tyre would be heavily distorted (ballooned).
Now think how the wieght will be transmitted through the tyre onto the raod... it would be through the air in the tyre and to some extent through the sidewalls of the tyre.
So if some of the wieght is transmitted through the sidewalls of the tyre (forgetting for a moment camber angles excetra and assuming the tyres entire contact patch is flat on the road) would the majority be tranfered through the semi vertical sidewall and not as much through the largely distorted inner sidewall?
And also if the sidewalls do transfer a little of the wieght, wouldnt it stand to reason that there would be slightly less wieght transfer through the centre of the tyre?
We are talking wieght transfer here and not tyre adhesion to the road...

So if those assumptions were correct then the display that you discribe in F9 would be not that far from reality With the slicks they have much stiffer side walls so this effect would not be so noticable.

I drive the XFG alot as well and found it handled wierd in the new patch (felt like i was on cross plys), but after working on setups for a few days have now got it to feel quite natural again without resorting to rediculously high or low tyre pressures (I put the initial wierdness down to not being used to feeling tyre deformation and most of the old setups being unsuitable )... whenever there is a major physics update it is going to take some weeks for things to pan out so I wouldnt draw too many conclusions too quickly

in the attached images you will see basically how tires deform. basically even if sidewalls could theoretically "transimt" a huge amount of the weight of the car, then the outer part of the tire may have an extra load, but the inner part must have a "lift" effect and not the contrary

Quote from tristancliffe :

Sympathic? Not sure if that's an insult or a compliment, so I'll resort to nodding and smiling

compliment! ...meaning that your are a man with sense of humor

Quote from Shotglass :

i tink ill go with tristans approach and just nod and smile

the best approach surely

Quote from Shotglass :

whoah easy there we dont want this to turn ugly ... afaik tristan studies mechanical engineering ... struck me as odd too that he doesnt know the difference but not reason to have a go at him ... just let it slide that he put that lol there

yeah there are few posts for wich the better answer is "LOL"...and i say all

tristan is sympathic anyway, we cannot say that for anyone

Quote from tristancliffe :

LOL

I think there's a good chance you, being one of the few people to think pressure only applied to fluids, is wrong. Perhaps you should read up on it

there is nobody that can pass the first semester of enginering caliming that...i assure you, if you would have studied, you would know what is the conceptual difference, i wonder why every tech article uses pascal instead of N/m2 when talking about gases/fluids and instead uses N/m2 when talkin baout solids in place of pascals...according to you that must be really an amazing coincidence.
please don't get me wrong but a definition is a definition and have a reason behind it

Quote from Shotglass :

kid could you just for once stop flying in other peoples faces just because they critisize lfs ... be my guest and have a go at the guy who started the fubar thread but stop it with people who try to point out flaws and make constructive critisism instead of bashing the sim
and just in case ... this is not my male protection mode your protective behaviour towards lfs annoys me since quite a while

back on topic

so an exam and a bit of thinking about tyres later ive come to believe that my idea of the contact patch folding a little might be correct and that the same thing happens in real life too to some extend not as much as in lfs though (might point to too little stiffness in the contact patch of the tyres)
take this pic for example: http://www.liveforspeed.net/pa ... es/aprilupdate/tires1.jpg looks like the outside sidewall deforms more than the inside which would mean that the contact patch has to either squeeze or fold

looking at pics of tyres deforming in lfs (eg this one: http://www.liveforspeed.net/pa ... es/aprilupdate/tires4.jpg ) one thing that strikes me as odd is that the sidewalls seem to deform in two ways that arent necessarily correct:
1) they appear to tilt to the side instead of deforming and bulging a little which might be caused by the sidewall being modelled as a single straight line (ie two sample points) the effect of which could be that the sidewall is generally too stiff
2) the sidewalls appear to strech out a lot which imho is incorrect or at least too much for rubber with kevlar inlays

once more you are proving to know how to dig into lfs phisics...

as i wanted to "prove" with the link i posted in a previous post (http://www.rsracing.com/tech-tire.htm) sidewall must first of all displace and the center cannot lift more than the inner part, as you clearly noted the source of the problem may be even only the sidewall streching, maybe devs used the real data from slick to simulate that, but if so it is very wrong imho.
imho for road tires the sidewall streching effect should be very negligible and could be even taken out from the lfs equations

Quote from colcob :

I'm sorry, I'm not sure what your physics background is, but pressure is simply a force divided by the area over which it is applied. There is nothing in physics which states it applied only to gases and fluids, that is simply nonsense.

http://en.wikipedia.org/wiki/Pressure have a read here to get yourself up to speed.

Also, just to be really anal, a pascal is not atmospheric pressure at sea level, that is a Bar*. A pascal is an SI unit and is therefore 1N/m2. Which is actually really quite a small unit, hence a typical tyre pressure is measured in hundreds of kPa.

While I welcome anyone getting involved in physics debates, I would recommend doing a bit of basic checking with google and wikipedia before attempting to state facts, just to stop yourself looking a bit silly.

*To be even more anal, technically an Atmosphere is atmospheric pressure, but a Bar is a convenient rounding to 100,000 pa, and is only a bit off an Atmosphere.

wikipedia is wrong, pressure is a measurement only for gasses/fuilds (in phisics), it may be popular to call a force/area pressure but this by any meaning wrong, as i said pascals have intrinsecally the "metric" (dunno the correct english word for this) of a force per area, but by definition it will never really counted like that (except for special purposes) but rather by pascals which is by definition a fraction of bar wich is by definition the air pressure at sea level (as i said in a previous post).
as you asked: i'm an engineer and i worked bla, bla...that realy doesn't matter since there are enginners, phisicist, etc. that sometimes are damn wrong etc. you don't have to believe me that wikipedia is wrong, but if you have a chance to open a real phisic book you know (after studying a bit ) what pressure is

Quote from KiDCoDEa :

after 4 pages of your rants, im still waiting, in proper english and in one sentence if you dont mind, for you to tell us all what the problem is.

due to my own male limitations please consider in your reply that left is left and that right is right.

anxiously waiting.

i'm not ranting, i'm always calm

left and right were referred to the sketches in the post were "left" and "right" were written, otherwise i've referred to outer or inner part of the tire while cornering, that seemed clear ofr everyone except you, i'm sorry if i cannot halde the english language to explain you better, if you have some hint...you're welcome!

the issue is:

1-a car with road tires (xfg more than others) is always cornering on sidewalls no matter how much pressure you inflate and/or how much negative camber you set

2-a car with road tires (xfg more than others) when cornering tend to lift the central and the inner part of the tire, having the central part lift much more even than the inner part (see my ugly "sketches" on a previous post) and that is "against any phisic law"

Quote from MagicFr :

are u kidding the GTT is faster now, but the RB4 is still way behind

as someone said, it's still soon to make conclusions: i've managed to be way faster with rb4 on a fe-green server and i easily beated every fxo and gtt, both on qualy and race...still to consider that previously nobody drove rb4, so rb4 experience for setups is still "young"

Quote from Frags :

Is it because of the whole postpatch experience, or is the sound a bit better?
I find it more easy to hear things in the game.
Echo from the cars, Spesific cars sound, tires etc. etc.
Maby I just never payed attention.

i'm quite sure thare are improvements in sound also, surely the BF1 emphatizes the wind sound because of very high speeds, but that was already present, surely i think that car sound are improved -> i clearly noticed with FZR: the sound is more rich and "colourful" even if a bit less raw...in conclusion it feels more natural...another for the devs

Quote from DEVIL 007 :

Hi all,

Flame me if you want after my post but Scawen is just a human after this all GREAT work he made and the only one who work on the game code.
I sometimes dont understand people who say like "it was requested many times but its still not included".
There are other simulation made in past or around the date of LFS and guess how big the development team was and how far the simulation went comparing to LFS...see it?!!
And to be honest there could be programmed thousands of tiny details to the LFS but you would need Super Computer to get decent frame rate.
During the time of LFS development the PC computing power is growing so you can get decent frame rate with every upgrade to the phisic code.

Just see what happend after latest update to the phisic code.The actual frame rate had dropped a bit as some people complained.I guess its for 99percent due to update to the phisic code.

And I understand your pointsabout the issues and missing things but please try to consider more the situation in which is LFS developed.

hi, oviously we all thank scawen, eric and victor for their daily hard work that had and will give us much fun.
scawen told months ago that this patch would have fixed tires and aerodynamic...period! we are all happy for the big surprise of a new real car (BF1) wich is more that he promised...i would love much things inside lfs, but i didn't blame the devs for not including anything, since the announcement was clear -> tire and aero.

what we are trying to understand and put attention on, is that even if (probably because of michelin data) slick tires seems much more realistic (even if it's hard for me to judge since i'm not a slick expert), something with road tires has gone too much unrealistic and smaller cars like xfg make it really evident.

we are not complaining, asking or pretending anything...we are only try to find out the most effective and clear feedback as possible to see if:
1 - there is effectively a big issue in the new tire phisic (i think it's quite evident)
2 - what is actually wrong (to me is only wrong some paramenters of some phisic functions)
3 - summarize and clarify things as much as possible so that in the future if and when the devs think that this have to be fixed they have enough information to know where to start diggin into...

if you read the thread more carefully you will see that we are not blaming devs for anything (except for scawen's bad sarchams ), only trying to do the previously mentioned things.

cheers