will there be free on-line play somewhere? besides with a friend?
--
Joe Conklin
http://members.xoom.com/gplegends/
remove spam to reply
How GPL online play works and what data it sends
Posters explained that GPL would support free TCP/IP play for up to eight players by connecting to a host's IP address, and a journalist said TCP/IP and IPX worked in his preview copy. A long technical argument followed about what data a racing sim must send, such as positions, steering, throttle and gears, and whether bandwidth or latency was the real problem.
online racing/leaguesGrand Prix Legends
Automated summary.
GPL will come with TCP/IP support, you'll be able to race over the net with
up to 8 player for free using an IP address.
Arthur
s t e a l t h
r a c i n g
http://www.***sys.com/stealthracing.html
>will there be free on-line play somewhere? besides with a friend?
>--
>Joe Conklin
>[link lost in import]
>remove spam to reply
>GPL will come with TCP/IP support, you'll be able to race over the net with
>up to 8 player for free using an IP address.
Errr, care to elaborate? How are eight people going to share an ip address?
;-) Or are you trying to say that there will be a dedicated server mode in
gpl, allowing up to eight people to connect?
Thanks to papy for adding TCP/IP support, great move iMHO (as opposed to
that crappy ipx implementation that comes with F1RS).
Cheers,
uwe
--
Spam-proof e-mail: Uwe Schuerkamp <hoover at telemedia . de>
http://www.telemedia.de/~hoover ////// Phone: +49-5241-80-10-66
Best of Scottish Folk: http://home.pages.de/~nsg
>>> Blue Ribbon Campaign: Support Free Speech on the Internet <<<
>>GPL will come with TCP/IP support, you'll be able to race over the net with
>>up to 8 player for free using an IP address.
>Errr, care to elaborate? How are eight people going to share an ip address?
>;-) Or are you trying to say that there will be a dedicated server mode in
>gpl, allowing up to eight people to connect?
FWIK, this is how it works:
People decide who will host the race and who will join. The people thats
joining will then need to know the IP# of the host. The i guess its just a
matter of entering the IP# into the multi-player menu or whatever in GPL and
hopefully connect to the host. :)
I also seem to remembering hearing something about GPL being able to recive
connections over different protocols simultainously (sp?)... in other words:
The host could have 3 drivers connecting thru an inhouse LAN-network and
maybe 3 people via TCP/IP (i.e. internet) at the same time.. making the
total of races to 7 or even more...
Im not quite sure as to how many that can acctually join the race tho.
All the best,
Edwin Solheim
REMOVE SPAM-GUARD IN ADDRESS TO REPLY!
Check out "The Paddock - a legendary site..." for
some cool GPL-stuff!!
[http://home.c2i.net/thepaddock)
something like the internet *** zone(maybe by sierra)
>will there be free on-line play somewhere? besides with a friend?
>--
>Joe Conklin
>[link lost in import]
>remove spam to reply
Sierra has their own *** zone thingy, called won.net
Arthur
s t e a l t h
r a c i n g
http://www.***sys.com/stealthracing.html
Quoted text (8 lines)
>something like the internet *** zone(maybe by sierra) >>will there be free on-line play somewhere? besides with a friend? >>-- >>Joe Conklin >>[link lost in import] >>remove spam to replyIf you want to know how well an IP connection can work, check out MTM2 on
the Zone. Excellent by any standard. TCP/IP is actually nothing new. I
believe F1RS was not allowed to provide TCP/IP racing, but they could of
made it work I'm sure. A dedicated server would be best, but would also not
be a free service.
The only concern I have about GPL's internet racing is that they apparently
didn't make it much of a priority in development and the results were poor.
Then the beta testers pushed for improvements and from what I hear it is
running well. Very good news. I only have so much time for sims, and if GPL
or any other sim is poor on the internet, then I won't have much interest in
it. On-line racing is what I enjoy most with auto sims and this feature
should be as much of a priority as anything else in any new releases.
There's no excuse for poor internet play in a sim anymore. If one developer
can do it right, they all can and must.
Dave
DmndDave
i have faith that the guys at papy will pull through and all aspects of the
sim will be excelent , unlike other sims(read that how ever you want:)
Joe Conklin
http://www.geocities.com/siliconvalley/network/5312/
remove spam to reply
<snip>
The GPL TCP/IP and IPX connections work fine in the preview copy I have and I
don't think that's brand new either. Personally I'll take smooth network play
over jumpy net play any day but that's because I'm spoiled for choice I
suppose.
I can't understand the assertion that a "dedicated server" would cost money.
People run plenty of servers for Quake, Unreal et al without charging for it
so why should this be any different.
The problem with internet play is that the more complex a sim, the more data
needs transferring between machines and the more complicated the code needs to
be to get it to work on anything below 56k.
M
Quoted text (39 lines)
>If you want to know how well an IP connection can work, check out MTM2 on >the Zone. Excellent by any standard. TCP/IP is actually nothing new. I >believe F1RS was not allowed to provide TCP/IP racing, but they could of >made it work I'm sure. A dedicated server would be best, but would also not >be a free service. >The only concern I have about GPL's internet racing is that they apparently >didn't make it much of a priority in development and the results were poor. >Then the beta testers pushed for improvements and from what I hear it is >running well. Very good news. I only have so much time for sims, and if GPL >or any other sim is poor on the internet, then I won't have much interest in >it. On-line racing is what I enjoy most with auto sims and this feature >should be as much of a priority as anything else in any new releases. >There's no excuse for poor internet play in a sim anymore. If one developer >can do it right, they all can and must. >Dave >DmndDave >>>Errr, care to elaborate? How are eight people going to share an ip >address? >>;-) Or are you trying to say that there will be a dedicated server mode in >>gpl, allowing up to eight people to connect? >>Thanks to papy for adding TCP/IP support, great move iMHO (as opposed to >>that crappy ipx implementation that comes with F1RS). >>Cheers, >>uwe >>-- >> Spam-proof e-mail: Uwe Schuerkamp <hoover at telemedia . de> >>http://www.telemedia.de/~hoover ////// Phone: +49-5241-80-10-66 >> Best of Scottish Folk: http://home.pages.de/~nsg >> >>> Blue Ribbon Campaign: Support Free Speech on the Internet <<<Quoted text (7 lines)
> I can't understand the assertion that a "dedicated server" would cost money. > People run plenty of servers for Quake, Unreal et al without charging for it > so why should this be any different. > The problem with internet play is that the more complex a sim, the more data > needs transferring between machines and the more complicated the code needs to > be to get it to work on anything below 56k.Sounds about right to me Rafael! Position and orientation of course. A 3D model makes
it a little more complex by adding a 3rd dimension to both orientation and position, but
big deal! Besides, bandwidth is rarely the problem, ping is!
Quoted text (13 lines)
> > I can't understand the assertion that a "dedicated server" would cost money. > > People run plenty of servers for Quake, Unreal et al without charging for it > > so why should this be any different. > > The problem with internet play is that the more complex a sim, the more data > > needs transferring between machines and the more complicated the code needs to > > be to get it to work on anything below 56k. > In the case of racing sims,which else data needs to be sent more than the cars > position in the track.I think the image of the other's car is generated on the > receiving computer own .That's a guess,sorry if this was stupid > RafaelOkay try this on for size.
Paramaters required by a racing sim at connection = setup details, driver
details, car choice
Paramaters during race = steering, throttle, brake positions, gear,
orientation on the track.
This is then acted on by the client machines to draw the car correctly. In GPL
you can hear the engine note, see the steering and the physical effects of
acceleration, braking, changing gear etc on the other cars so this data is
needed. This is without the occasional packets of information to tell the
machine to raise the drivers arm (in GPL) or set up a pitstop (GP2). This is,
as you can imagine a LOT of data. The server has to collect it and send it out
to the clients as well as process it's own info. This is why dedicated servers
are popular, it removes (at least) one layer of processing from the server.
"ping" as you so call it is the time required for a packet of information to
be sent to a machine and return. If your PC is using all it's bandwidth for
the job it is engaged in the "ping" increases. Admittedly there is also a
factor on a) network usage b) "server" load but these are all because of the
amount of traffic vs. the amount of bandwidth/processing time available.
BTW for the techies out there I *know* that's not the "kosher" explanation but
to be honest I don't think people want to read an essay on the cause and
effect of inter-networking overload.
Quoted text (31 lines)
>Sounds about right to me Rafael! Position and orientation of course. A 3D model > makes >it a little more complex by adding a 3rd dimension to both orientation and > position, but >big deal! Besides, bandwidth is rarely the problem, ping is! >> > I can't understand the assertion that a "dedicated server" would cost > money. >> > People run plenty of servers for Quake, Unreal et al without charging for > it >> > so why should this be any different. >> > The problem with internet play is that the more complex a sim, the more > data >> > needs transferring between machines and the more complicated the code needs > to >> > be to get it to work on anything below 56k. >> In the case of racing sims,which else data needs to be sent more than the > cars >> position in the track.I think the image of the other's car is generated on > the >> receiving computer own .That's a guess,sorry if this was stupid >> Rafael> something like the internet *** zone(maybe by sierra)
> >will there be free on-line play somewhere? besides with a friend?
Sierra has their own *** zone, and it sucked.
Best bet would be if TEN would support it OR if they
allowed players to set up servers <a la Quake>.
However Sierra <unlike ID> is into money, and doesnt
give a hoot about the quality/networkability of games.
Just my .02
Of course, id aren't even remotely interested in money.....they operate
the business simply to be charitable.
> However Sierra <unlike ID> is into money,
--
Regards,
Bruce.
----------
> Okay try this on for size.
> Paramaters required by a racing sim at connection = setup details, driver
> details, car choice
> Paramaters during race = steering, throttle, brake positions, gear,
> orientation on the track.
Steering????? Why???????
Throttle - optional
Brake positions?????????? Why???????????????
Gear - optional
Orientation - of course
If these are you guesses, I hope your wrong. If not, the programmers are clueless by
the looks of things to me! Way too much unnecessary info here!
Have you ever played a racing simulation? Without these things the car would
just move around a track on rails. The wheels would not turn (I'm not talking
about rotation here) when the car turned a corner, you wouldn't hear the
engine notes change, see the effects of braking etc
I seriously doubt you have given enough thought to what makes a racing
simulation work as it does. The physics involved in modern racing sims is
astounding and is fast catching up racing teams in terms of accurate
simulations of the subject. Not only this but the players demand more from
their sims and patchy multiplayer play as you seem to think would work would
not be acceptable to 90% of people.
Personally I'd be bothered if when racing against the PC in GPL I could see
the suspension move on the other cars but not when I was racing over my LAN or
the internet. Maybe I'm just a wacky perfectionist but I doubt it....
M
Quoted text (20 lines)
>> Okay try this on for size. >> Paramaters required by a racing sim at connection = setup details, driver >> details, car choice >> Paramaters during race = steering, throttle, brake positions, gear, >> orientation on the track. > Steering????? Why??????? > Throttle - optional > Brake positions?????????? Why??????????????? > Gear - optional > Orientation - of course > If these are you guesses, I hope your wrong. If not, the programmers are > clueless by >the looks of things to me! Way too much unnecessary info here!You and some others sound unnecessarily down on internet play in sims today.
Even if you have zero interest in MTM2 or the new Motocross Madness, try the
demos online. I use a 33.6 modem and the results are excellent. MS improved
the performance just in the time between CPR and MTM2. From what I
understand, GPL is better than CPR online, and if it's as good as MTM2,
there will be great net racing with that as well. I think I speak from some
authority as I've had hundreds of online races with these games with people
from all over the world. Even races with people with a poor ping rating are
often acceptable. If I want to guarantee a smooth race, I just look at the
ratings and only include guys with good pings. Then, it's as smooth as a
race with the computer cars. Every movement of the opponents vehicles is
modeled. Nothing has been left out at all.
This is a fairly simple explanation, but apparently some developers have
made online racing a priority and have learned how to have the game
successfully fill in the gaps between the packets of info that is
transferred over the net during a race. This eliminates a lot of the warping
effect. I raced POD last summer, and it was good for that time but compared
to today, it's a joke. You would see POD cars flying off the road, only to
appear a quarter mile down the track ahead of you a few seconds later.
Byron and myself know that sim manufacturers CAN provide smooth internet
play if they choose to. We also have noticed a kind of coolness by some
developers to do so and that's what's bothersome. There's no excuse for that
attitude. Even aviation sims are providing online play now. How can racing
sims not provide it as well? Auto racing (sports) is all about human
competition.
Dave
DmndDave
>Have you ever played a racing simulation? Without these things the car would
>just move around a track on rails. The wheels would not turn (I'm not talking
would
>not be acceptable to 90% of people.
>Personally I'd be bothered if when racing against the PC in GPL I could see
>the suspension move on the other cars but not when I was racing over my LAN or
>the internet. Maybe I'm just a wacky perfectionist but I doubt it....
>M
Quoted text (8 lines)
>about rotation here) when the car turned a corner, you wouldn't hear the >engine notes change, see the effects of braking etc >I seriously doubt you have given enough thought to what makes a racing >simulation work as it does. The physics involved in modern racing sims is >astounding and is fast catching up racing teams in terms of accurate >simulations of the subject. Not only this but the players demand more from >their sims and patchy multiplayer play as you seem to think would workalot of those effects can be handle client side i would think. like the
suspension movement you talk about, the server could send the speen and
track placement ect. to you and your comp. could figure out the suspension
movent etc.
Joe Conklin
For Central PA Online Gamers:
http://www.geocities.com/siliconvalley/network/5312/
For Grand Prix Legends:
http://members.xoom.com/gplegends/
remove spam to reply
>Have you ever played a racing simulation? Without these things the car would
>just move around a track on rails. The wheels would not turn (I'm not talking
would
>not be acceptable to 90% of people.
>Personally I'd be bothered if when racing against the PC in GPL I could see
>the suspension move on the other cars but not when I was racing over my LAN or
>the internet. Maybe I'm just a wacky perfectionist but I doubt it....
>M
<snip>
Quoted text (8 lines)
>about rotation here) when the car turned a corner, you wouldn't hear the >engine notes change, see the effects of braking etc >I seriously doubt you have given enough thought to what makes a racing >simulation work as it does. The physics involved in modern racing sims is >astounding and is fast catching up racing teams in terms of accurate >simulations of the subject. Not only this but the players demand more from >their sims and patchy multiplayer play as you seem to think would work>There's no excuse for that
attitude.
Ummm.....there is, you know, David.
"Economic viability" is the excuse: if only a minority of consumers are
buying the product for that feature then it is uneconomical to spend
time and resources on the in-house development of that feature. far
wiser to spend those resources elsewhere in the "manufacture" of the
product.
--
Regards,
Bruce.
----------
The GP Legends Historic Motor Racing Club is located at:-
http://www.netspeed.com.au/brucek/legends/
Well not really... the suspension dip on your road car is a function less of
speed more of change in velocity be it braking, accelerating or cornering. The
same applies to race cars.
The client *does* calculate it but based on the parameters I stated rather
than something as simple as a position triplet and velocity.
m
Quoted text (36 lines)
>alot of those effects can be handle client side i would think. like the >suspension movement you talk about, the server could send the speen and >track placement ect. to you and your comp. could figure out the suspension >movent etc. >Joe Conklin >For Central PA Online Gamers: >http://www.geocities.com/siliconvalley/network/5312/ >For Grand Prix Legends: >http://members.xoom.com/gplegends/ >remove spam to reply >>Have you ever played a racing simulation? Without these things the car >would >>just move around a track on rails. The wheels would not turn (I'm not >talking >>about rotation here) when the car turned a corner, you wouldn't hear the >>engine notes change, see the effects of braking etc >>I seriously doubt you have given enough thought to what makes a racing >>simulation work as it does. The physics involved in modern racing sims is >>astounding and is fast catching up racing teams in terms of accurate >>simulations of the subject. Not only this but the players demand more from >>their sims and patchy multiplayer play as you seem to think would work >would >>not be acceptable to 90% of people. >>Personally I'd be bothered if when racing against the PC in GPL I could see >>the suspension move on the other cars but not when I was racing over my LAN >or >>the internet. Maybe I'm just a wacky perfectionist but I doubt it.... >>M ><snip>I'm lucky enough to have a LAN connection to the net at work which moves me
into the land of LPB's when I play Quake or Quake 2 and, to be honest, I've
never had much problem with any racing game over the net because of that. At
home where I have a 33.6k connection I never bother with net *** because
it's not worth the time and money (1 to 1.7 pence per minute) to play laggy
games. That said, I LOVE multiplayer *** and think it's the way forward.
I think you'll find that it isn't easy to provide GOOD internet ***
otherwise everyone would do it. MS games are good at it (MXM is stunning), but
Unreal is the worst I have ever seen. Why? It's not easy. Do we want to wait
another 4-6 months for flawless netplay when 90% of the time we'll be playing
against the PC?
Anyway, the point of my post was to make the issue that there is a lot more
data than car position and orientation required to make a sim work right on
network play.
Later
M
Quoted text (54 lines)
>You and some others sound unnecessarily down on internet play in sims today. >Even if you have zero interest in MTM2 or the new Motocross Madness, try the >demos online. I use a 33.6 modem and the results are excellent. MS improved >the performance just in the time between CPR and MTM2. From what I >understand, GPL is better than CPR online, and if it's as good as MTM2, >there will be great net racing with that as well. I think I speak from some >authority as I've had hundreds of online races with these games with people >from all over the world. Even races with people with a poor ping rating are >often acceptable. If I want to guarantee a smooth race, I just look at the >ratings and only include guys with good pings. Then, it's as smooth as a >race with the computer cars. Every movement of the opponents vehicles is >modeled. Nothing has been left out at all. >This is a fairly simple explanation, but apparently some developers have >made online racing a priority and have learned how to have the game >successfully fill in the gaps between the packets of info that is >transferred over the net during a race. This eliminates a lot of the warping >effect. I raced POD last summer, and it was good for that time but compared >to today, it's a joke. You would see POD cars flying off the road, only to >appear a quarter mile down the track ahead of you a few seconds later. >Byron and myself know that sim manufacturers CAN provide smooth internet >play if they choose to. We also have noticed a kind of coolness by some >developers to do so and that's what's bothersome. There's no excuse for that >attitude. Even aviation sims are providing online play now. How can racing >sims not provide it as well? Auto racing (sports) is all about human >competition. >Dave >DmndDave >>Have you ever played a racing simulation? Without these things the car >would >>just move around a track on rails. The wheels would not turn (I'm not >talking >>about rotation here) when the car turned a corner, you wouldn't hear the >>engine notes change, see the effects of braking etc >>I seriously doubt you have given enough thought to what makes a racing >>simulation work as it does. The physics involved in modern racing sims is >>astounding and is fast catching up racing teams in terms of accurate >>simulations of the subject. Not only this but the players demand more from >>their sims and patchy multiplayer play as you seem to think would work >would >>not be acceptable to 90% of people. >>Personally I'd be bothered if when racing against the PC in GPL I could see >>the suspension move on the other cars but not when I was racing over my LAN >or >>the internet. Maybe I'm just a wacky perfectionist but I doubt it.... >>MSo the orientation of your opponent's wheels can be seen while
steering.
Not if you want to be able to hear his engine. This is important from
positional aspect, and also for atmosphere.
To show lockups, generate appropriate level of smoke, skidding etc.
Means no gears in the replay.
Not at all, in fact if you are looking at a _serious_ sim then you
need more. There's a difference between what is required for a real
simulation and what is required for a game - that's not intended to be
disparaging, just an observation of fact. How can you know how much
draft you're going to get unless you know my setup and thus can
generate how big a hole I'm making in the air? In a game you
approximate this and it's not a problem, but in a sim it is, at least
of you want to do it correctly.
Cheers!
John
On Thu, 27 Aug 1998 14:25:42 -0400, "David G Fisher"
>This is a fairly simple explanation, but apparently some developers have
>made online racing a priority and have learned how to have the game
>successfully fill in the gaps between the packets of info that is
>transferred over the net during a race. This eliminates a lot of the warping
>effect.
It does, but at a cost. Let's say we're both exiting Parabolica and
you want to draft me all the way up Retifilio and make the pass into
Curva Grande - I will jink side to side to keep you out of the draft
as much as I can, but with warping and a computer filling in the gaps
you don't know where I am - at 200mph that is not good. You pull out
to pass, my position is updated, and we both crash.
It is a very fine art to this, and whereas a Quake pushlatency will be
lost in the melee, a race between two good drivers will likely be
settled by one moment, one moment where you really, really need to
know where everything is. The balance between absorbing latency
problems and providing accurate update is very delicate. IMO it's
often better to know that there may be a problem (and back off
accordingly) than be lulled into a false sense of security and ruin
everyone's race.
MTM2 seems to have it pretty good, the latency control was one of it's
finest aspects IMO.
Correct me if I'm wrong, but a lot of aviation sims don't model plane
to plane collisions for the very reasons described above - if heading
adjustments are acted upon immediately in a head to head the planes
will surely crash, and that can't be guaranteed. That would be a
ludicrous solution in auto-sims, being able to drive through someone
lowers the challenge enormously.
Cheers!
John
You raise some good points, but OTOH I still question their *necessity*. If the
difference between *all* of this information and just the *necessary* stuff is the
difference between unplayable and playable multiplay then I still think my point stands.
It seems your point is that you want it all or you want nothing! Btw, I doubt you would be
in a 90% majority here.
Quoted text (42 lines)
> Have you ever played a racing simulation? Without these things the car would > just move around a track on rails. The wheels would not turn (I'm not talking > about rotation here) when the car turned a corner, you wouldn't hear the > engine notes change, see the effects of braking etc > I seriously doubt you have given enough thought to what makes a racing > simulation work as it does. The physics involved in modern racing sims is > astounding and is fast catching up racing teams in terms of accurate > simulations of the subject. Not only this but the players demand more from > their sims and patchy multiplayer play as you seem to think would work would > not be acceptable to 90% of people. > Personally I'd be bothered if when racing against the PC in GPL I could see > the suspension move on the other cars but not when I was racing over my LAN or > the internet. Maybe I'm just a wacky perfectionist but I doubt it.... > M > >> Okay try this on for size. > >> Paramaters required by a racing sim at connection = setup details, driver > >> details, car choice > >> Paramaters during race = steering, throttle, brake positions, gear, > >> orientation on the track. > > Steering????? Why??????? > > Throttle - optional > > Brake positions?????????? Why??????????????? > > Gear - optional > > Orientation - of course > > If these are you guesses, I hope your wrong. If not, the programmers are > > clueless by > >the looks of things to me! Way too much unnecessary info here!Quoted text (11 lines)
> >There's no excuse for that > attitude. > Ummm.....there is, you know, David. > "Economic viability" is the excuse: if only a minority of consumers are > buying the product for that feature then it is uneconomical to spend > time and resources on the in-house development of that feature. far > wiser to spend those resources elsewhere in the "manufacture" of the > product. > --> > Steering????? Why???????
> So the orientation of your opponent's wheels can be seen while
> steering.
Host computer? (Like a shopping trolley - ie the wheels point where they're pushed)
Host computer? (Acceleration measurement)
Well, I doubt using brake positions would be efficient for doing this. More like a
parameter for each corner, only sent when applicable ie not rolling smoothly.
Transfer all opponents gear ratios before race? Or maybe entire setup?
Why is it that a positional and orientational triplet is insufficient to deal with susp
travels?
Btw, how did we get into this anyway. Ping is the issue, not bandwidth. I dare say,
completely ignorant as I am to packet sizes, that all of the info anyone could want could
be contained in a single packet anyway! Yes or no?
Quoted text (48 lines)
> Well not really... the suspension dip on your road car is a function less of > speed more of change in velocity be it braking, accelerating or cornering. The > same applies to race cars. > The client *does* calculate it but based on the parameters I stated rather > than something as simple as a position triplet and velocity. > m > >alot of those effects can be handle client side i would think. like the > >suspension movement you talk about, the server could send the speen and > >track placement ect. to you and your comp. could figure out the suspension > >movent etc. > >Joe Conklin > >For Central PA Online Gamers: > >http://www.geocities.com/siliconvalley/network/5312/ > >For Grand Prix Legends: > >http://members.xoom.com/gplegends/ > >remove spam to reply > >>Have you ever played a racing simulation? Without these things the car > >would > >>just move around a track on rails. The wheels would not turn (I'm not > >talking > >>about rotation here) when the car turned a corner, you wouldn't hear the > >>engine notes change, see the effects of braking etc > >>I seriously doubt you have given enough thought to what makes a racing > >>simulation work as it does. The physics involved in modern racing sims is > >>astounding and is fast catching up racing teams in terms of accurate > >>simulations of the subject. Not only this but the players demand more from > >>their sims and patchy multiplayer play as you seem to think would work > >would > >>not be acceptable to 90% of people. > >>Personally I'd be bothered if when racing against the PC in GPL I could see > >>the suspension move on the other cars but not when I was racing over my LAN > >or > >>the internet. Maybe I'm just a wacky perfectionist but I doubt it.... > >>M > ><snip>Believe me, in GPL you are rarely travelling where your wheels are
pointing...
This varies depending upon gearing, uphill/downhill, fuel weight,
clutch engaged or not and a host of other things.
So you need to send the rotational speed of each wheel, same result.
Doesn't help. You squeeze every last rev out of your engine whereas I
shift up 2k early to avoid damage and lower fuel consumption.
Different shifting points despite same gearing.
Like I said, you can either approximate it or you can do it right. One
involves compromises, the other bandwidth issues. There is no
shortcut, you just weigh it up and set the balance as you see ift.
Cheers!
John
> Why is it that a positional and orientational triplet is insufficient to deal with susp
> travels?
> Btw, how did we get into this anyway. Ping is the issue, not bandwidth. I dare say,
> completely ignorant as I am to packet sizes, that all of the info anyone could want could
> be contained in a single packet anyway! Yes or no?
Think so, no large amount of data in there
Beers and cheers
(uncle) Goy
UncleGoy on TEN
No sim will be even close to reality even if in the future we have good feedback seats and
real 3d vision(things like lasers sending a 3d images directly to your retine).The point
is ,it's a game and and i think that most of psycological aspects of motorsports comes
from the fact that's it's very very life(or at least ,injure) risky.
So ,what are the risks when you're playing on your pc for hours?No chance for a single
cut(unless your wife throws a can on you ,asking you for goto bed or giving her the
computer to play freecell(mine does play,but doesn't throw :)).
If not ,soon we'll be asking the developers (in the name of realism)to not give us saving
features for setups ,hotlaps,etc ,accerated time in qualifying(will GPL have it ,or will
we have to sit and watch the whole thing?),wait the real time it lasts to change your
gearing setup ,for example, and option for only 100% races.
It sound ridiculous, and in fact it is ,and these above will never be
implemented,Because it's only a game.
Rafael
Statistics provide one irrefutable fact....that 100% of surveys give
100% statistics!
But you are overlooking the fact, in your wholesome enthusiasm for
on-line ***, that to the vast majority of software buyers, it will be
something that they haven't the slightest interest in.
And that *IS* as fact!! :o)
> I still say Bruce that you can talk as many statistics as you like, and I will link the
> results to the consumers knowledge that online sim racing is currently a joke outside of
> NROS. The buzz created by a good online sim racing game could change the results of a
> survey in a flash ie people coming to the party when they hear the news. It's a matter of
> when does the catalyst appear!
--
Regards,
Bruce.
----------
> No sim will be even close to reality... <anip>...
> i think that most of psycological aspects of motorsports comes
> from the fact that's it's very very life(or at least ,injure) risky.
> So ,what are the risks when you're playing on your pc for hours?
I get eye strain... does that count? ;-)
Later...
Andre
LOL, how did that feel, Byron, being lectured like that :o))).
/Christer, really believes car position and orientation should be enough, but
what do I know :o)))
--
http://home4.swipnet.se/~w-41236/ (Read all about the "Global online
racing"-proposal under "For developers". Read it a couple of times, cause noone
has understood it the first time they've read it yet :o))
> Okay try this on for size.
> Paramaters required by a racing sim at connection = setup details, driver
> details, car choice
> Paramaters during race = steering, throttle, brake positions, gear,
> orientation on the track.
> This is then acted on by the client machines to draw the car correctly. In GPL
> you can hear the engine note, see the steering and the physical effects of
> acceleration, braking, changing gear etc on the other cars so this data is
> needed. This is without the occasional packets of information to tell the
> machine to raise the drivers arm (in GPL) or set up a pitstop (GP2). This is,
> as you can imagine a LOT of data. The server has to collect it and send it out
> to the clients as well as process it's own info. This is why dedicated servers
> are popular, it removes (at least) one layer of processing from the server.
> "ping" as you so call it is the time required for a packet of information to
> be sent to a machine and return. If your PC is using all it's bandwidth for
> the job it is engaged in the "ping" increases. Admittedly there is also a
> factor on a) network usage b) "server" load but these are all because of the
> amount of traffic vs. the amount of bandwidth/processing time available.
> BTW for the techies out there I *know* that's not the "kosher" explanation but
> to be honest I don't think people want to read an essay on the cause and
> effect of inter-networking overload.
> >Sounds about right to me Rafael! Position and orientation of course. A 3D model
> > makes
> >it a little more complex by adding a 3rd dimension to both orientation and
> > position, but
> >big deal! Besides, bandwidth is rarely the problem, ping is!
> >> > I can't understand the assertion that a "dedicated server" would cost
> > money.
> >> > People run plenty of servers for Quake, Unreal et al without charging for
> > it
> >> > so why should this be any different.
> >> > The problem with internet play is that the more complex a sim, the more
> > data
> >> > needs transferring between machines and the more complicated the code needs
> > to
> >> > be to get it to work on anything below 56k.
> >> In the case of racing sims,which else data needs to be sent more than the
> > cars
> >> position in the track.I think the image of the other's car is generated on
> > the
> >> receiving computer own .That's a guess,sorry if this was stupid
> >> Rafael
I dont think you believe that the high pulse the F1 drivers has on the grid just
before the start is because of the high risks in F1. I believe you could get the
same high pulse before an indoor kart race. I believe it's in the amount of time
spent preparing. The more time you have put into preparing a race the more you
want that race to go well and there you have true pressure and a high pulse. I
have felt that high pulse before sim races in offline league races. The reason
for this, as I understand it, is because I have practiced a lot for the race and
now I only have one shot at the race and that's spelled pressure :o). I bet a
lot of NROS drivers can tell you some stories about high pulse, pressure and
sweating :o).
/Christer
--
http://home4.swipnet.se/~w-41236/ (Read all about the "Global online
racing"-proposal under "For developers". Read it a couple of times, cause noone
has understood it the first time they've read it yet :o) )
Quoted text (12 lines)
> The point > is ,it's a game and and i think that most of psycological aspects of motorsports comes > from the fact that's it's very very life(or at least ,injure) risky. > Rafael > > Like I said, you can either approximate it or you can do it right. One > > involves compromises, the other bandwidth issues. There is no > > shortcut, you just weigh it up and set the balance as you see ift. > > Cheers! > > JohnQuoted text (8 lines)
> Statistics provide one irrefutable fact....that 100% of surveys give > 100% statistics! > But you are overlooking the fact, in your wholesome enthusiasm for > on-line ***, that to the vast majority of software buyers, it will be > something that they haven't the slightest interest in. > And that *IS* as fact!! :o)> LOL, how did that feel, Byron, being lectured like that :o))).
> /Christer, really believes car position and orientation should be enough, but
> what do I know :o)))
> --
Great, since it was educational :) As I just said elsewhere though, I dunno how we got
into this because it is a bandwidth thing rather than the critical ping aspect.
Nothing....just the new circuit.
--
Regards,
Bruce.
----------
The GP Legends Historic Motor Racing Club is located at:-
http://www.netspeed.com.au/brucek/legends/
Quoted text (11 lines)
> > No sim will be even close to reality... <anip>... > > i think that most of psycological aspects of motorsports comes > > from the fact that's it's very very life(or at least ,injure) risky. > > So ,what are the risks when you're playing on your pc for hours? > I get eye strain... does that count? ;-) > Later... > AndreJesus, Byron.....*EVERYTHING* is just "for now", mate!! :o)
> For now, maybe! :)
--
Regards,
Bruce.
----------
The GP Legends Historic Motor Racing Club is located at:-
http://www.netspeed.com.au/brucek/legends/
>features for setups ,hotlaps,etc ,accerated time in qualifying(will GPL have it ,or will
>we have to sit and watch the whole thing?),wait the real time it lasts to change your
>gearing setup ,for example, and option for only 100% races.
> It sound ridiculous, and in fact it is ,and these above will never be
>implemented,Because it's only a game.
>Rafael
Actually that's given me an idea . How about when qualifying if you decide
to change the car setup, during the time in the garage you lose say 5 to 10
minutes for simple changes, or 10 -20 minutes for more drastic changes. Also
if you wreck the car in an accident , depending on whereabouts on the
circuit you crash you lose more time due to the walk / run back to the pits
, and hopefully the spare car is setup for you and not your team mate. If
you were then to go out and wreck the spare maybe you wouldn't qualify for
the race (or further back on the grid) . The single spare car idea may cause
some arguments for online racing, if both driver's in a team wreck their own
race cars:-)
Ian P
Wont it always be an approximation?
/Christer
--
http://home4.swipnet.se/~w-41236/ (Read all about the "Global online
racing"-proposal under "For developers". Read it a couple of times, cause noone
has understood it the first time they've read it yet :o))
> Like I said, you can either approximate it or you can do it right. One
> involves compromises, the other bandwidth issues. There is no
> shortcut, you just weigh it up and set the balance as you see ift.
> Cheers!
> John
On Sun, 30 Aug 1998 17:21:23 GMT, Christer Andersson
>Wont it always be an approximation?
To all intents and purposes uyes, but not necessarily. When playing
Quake online I used to use an office connection to the UK's
SuperJANET, which gave ping times of 12ms - that is as close to real
time as makes no difference and therefore not an approximation.
As ping time increases you can either plot only when a packet
succesfully arrives, or you can extrapolate in one way or another
until you get the information and then update the position instantly
or via some smoothing transition. Whichever way you choose will be an
approximation, but how much of an approximation depends upon
connection quality and how you program the code to handle such
instances.
Which way programmers should be headed depends upon their thoughts on
what is "real" racing or not, and whether the internet can in future
provide suitable latency for racing (or whether software companies
need to be coding to take account of long term deficiencies). I think
the latency will improve and I also prefer knowing where my opponent
actually is, so I err toward less approximation.
Cheers!
John
My point was not "all or nothing" but what is acceptable and what is
prefereable. I know there has to be a balance between the two but to be
honest, why can't a programmer manage to send the required info?
Okay, you have to set a minimum connection speed but if that is 33.6k or
56k then fine. Most people are happy to splash out on a 3d card or whatever
but are a little less keen to spend a little extra on a faster modem.
The point I'm making is that whatever complexity the sim takes, any
multiplayer form of the sim HAS to support the same otherwise you have 2
games. This is great for Quake and QuakeWorld as people can pick the option
best suited to them (either NetQuake or QW) when playing online but I can't
see any racing sim company moving to provide 2 choices in online play
technology if only because of the complexity and expense. id develop QW for
free and have just released the last version after working on it for God knows
how long (18-24 months?).
M


































