rec.autos.simulators

I'm pretty shocked at how little discussion this fantastic site has
generated here at r.a.s. It really is one of the best additions to the sim
racing community I've ever seen. The ranking system for online racing, and
it's hotlap utility are tremendous.

Schumi delivered on everything he promised, and it's only going to get
better.

David G Fisher

David Cook

David G Fisher

For instance, some people ripped (and never retracted their criticisms) Ubi
Soft for not providing better TCP/IP racing, and denied FIA restrictions
were the cause. They were very, very wrong.

David G Fisher

How come UbiSoft is the only one (that I'm aware of) that was prevented by
license to offer multiplayer in their F1 sim?   Not being a jerk here, I
just don't remember reading about why they were prevented from offering
multiplayer, yet EA Sports offered it in their title ?

I know from speaking with a few of the guys at OLR, that adding GPL would be
awesome.  However, it would probably be a waste of effort as I doubt they
could ever get anyone to 'pay' to play GPL or even if it was free, to give
up VROC as there is really no reason to do so outside having OLRs
connections for those who don't have the capability to host.

Relay is a great piece of software though.   I just think its being 'wasted'
on the titles it's supporting right now.  I happen to be one who is not a
fan of F1 really at all (no - REALLY <G>), but happened to enjoy F1CS2K from
EA.  Although I admit to holding off the purchase of the latest version due
to all the troubles I have read.  I will wait to see if there is a patch and
how things pan out.  I know many of the problems have work arounds.. but
I'll still wait.

--
Scott B. Husted
ICQ# 4395450
http://www.PowerSimsMag.com

Hopefully I can provide some insight on the "Online Racing Genre Pitfalls."
Perhaps even dispel some myths that seem to persist amongst the community.

1) Ubisoft was not permitted to even "promote" F1RC online, or any online
capabilities thereof. That was a contractual limitation in their license. As
such, Ubisoft could not supply us with an iota of support (without taking a
serious business risk), <even though they really wanted to after seeing what
we had to offer ;]>. So we basically had to hack the game to get it played
online... which is why many of the features we have yet to show are not
included or viewable currently.

2) First Person Shooters. The *main* reason that these games have such
extensive multiplayer support and codebase, is basically due to the target
market and target platform of the game. Meaning, for the most part they are
designed for the PC specifically. So... PC specific requirements, and target
audience demands govern how much development time is spent on certain
aspects of the game (like multiplayer). Multiplayer is a very cognate
requirement for the FPS genre, and the development time and effort on the
multiplayer codebase reflects this.

I know a few FPS game programmers will hate me for the next comment
though... but I will try to explain why I feel this way:

Making multiplayer for a FPS game is a LOT easier than it is for a Racing
Game. The fundamental *issues* with online support remain the same
throughout any genre (player synchronization being the #1 problem). However,
the issues and lack of solutions for the issues are greatly amplified and
much more noticeable in Racing Games.

FPS can rely heavily on *good* prediction code. They can really fudge a few
things and get around a number of the issues if the prediction code is
sufficient enough to get away with it, and of course if the netcode itself
is sufficiently good thought out (can't simply rely on prediction code).

What I mean is:

In Driving games (particularly racing sims) you have inches between players
going through chicanes at simulated 200 KM/hr. speeds where even a slight
twitch of the wheel can send you meters off course(exaggerated speed but you
get the point). Accuracy (positional accuracy) of the players in driving
sims is CRUCIAL.... heck its the name of the game... So even a slight
mismatch on prediction code could spell disaster for the game. Driving Sims
are MUCH more reliant on a steady and accurate stream of data.
Unfortunately, the Internet as a whole is a very fragile and often
unreliable beast. Dedicated Servers can certainly help out a lot, since many
of the online issues with many racing games is that they are P2P. Where the
host machine must also display the graphics, sound, etc. with little CPU
time for netcode. We all know that everyone here pushes the puter to the max
to get the most frames/sec. out, which is a DISASTER if that is the mindset
of the host machine. Also a dedicated server can really streamline the sync.
of the players if it is on a sufficient pipe, to reduce lag, etc.

Now you also take into consideration that the majority of these games are
sold for the consoles. Which for the most part don't have any multiplayer
abilities (at least none to speak of). Most racing games are designed for
the consoles, and ported to the PC. Whereas the FPS games are done the other
way around (for the most part). So the main requirements for the consoles is
a good graphics engine. You get that, and the rest is pretty much sold.

This is also why I sometimes poke my head in here and whisper to some who
bash arcade racers "Don't bite the hand that feeds you...". Since it is
these arcade racers and their sales numbers that even accommodate the
funds/efforts of the developers to make us our Sims in the first place.

So my main point here is:

Driving Sims is a tough genre to program multiplayer for, and my empathy
goes out to the programmers who are responsible for coding it. Its a really
tough racket.

A lot of driving sim developers opt for the DirectPlay method of connecting.
Which is a nice and easy SDK to program with, but the results on multiplayer
quality are devastating (the DPlay API is shite... and I'm being diplomatic
when I say that... I think much worse of it...).

Papy have the enviable position of being a developer who took the time
earlier on in their history to expend the time/efforts on developing a solid
a cognate multiplayer codebase that they could use/reuse/adapt/modify at
will. The Papy network codebase is custom. They use sockets. They can do so
much more with their custom code than DirectPlay, which will paint you into
the "Grey Hair Corner". They have that codebase in their back pocket, and do
not *really* have to 'redesign the wheel'. They just reuse that existing
code (which works well). Which makes the milestones/etc. easier for Papy to
meet, whilst still maintaining great multiplayer abilities. They have it
good there basically :)

So... at the end of the day, Relay can only be as good as the game's
netcode. Can we make the game's netcode better? Yes and No. Well we can try,
but the rest is up to the developer if they want our help. We certainly know
that we can streamline it, and help out tremendously in other areas too. It
just takes developer support is all.

So the uphill battle that we really face is:

1) Getting in on the development of the game (it is really easy to port
Relay with embedded support... less than a day really).

2) Convince the developers that it is in their best interest to resist all
temptation to use DPlay as their netcode API, and to take the time to create
a custom netcode that they can reuse on future games, and reap the rewards
that a custom netcode brings.

We've already managed to get in on the ground level at an currently
unmentionable game developer, and things are going VERY good. They are very
happy, and we are very happy. They have taken our advice, and are really
boned on applying full support for Relay, and thus allowing us to unleash
the other 40% of our software, which we think really shows what Relay is
about and can do. Unfortunately it is not a Racing Sim as we would hope for,
but I still have to get Relay through #1 before I can get to #2 with other
developers.

We need to get in the front door. After that, the complete package speaks
for itself (so far all developers who have been given a full demo love our
package... and now picking us over Game Spy packages).

We are still a small development firm, and getting recognized/etc. is always
the hardest part in any business. We still plug away and hopefully we can
provide the full package to a Racing Sim Developer who wants to step to the
plate for the "First Game Free Demo Promotion" we have. Which is basically a
no risk, no financial obligation offer to simply allow us to show what we
can do with full support. So far, everything you see has been done with ZERO
support. We hope things will change soon.

Oh... as far as the comparison to OLR. We are a bit diff. than them in
business model (and they might want to watch if they are attempting to
charge for use... NASCAR is VERY specific about not charging to use their
licenses or products licensed by them... especially online multiplayer...
VERY strict on that unless they get a piece of the pie). We don't charge the
end user for the software or the ability to play online. We basically supply
the "Sierra Type Icon" for the game itself. As opposed to fighting against
existing platforms for the same game. This is another reason we chose
F12001, and F1RC.... no one was doing it, and it is obvious those two
companies need all the help they can get in the multiplayer dept. (through
no fault of the developers I might add).

Anyways... I'm rambling. Sorry for the long post.

Cheers,
Schumi

read more »

Hey !!!

For once someone is EXPLAINING things here instead of throwing various
"words" at eachother's faces....

Don't be SORRY for explainig us, be THANKED for it, at least by me...

Eldred
--
Dale Earnhardt, Sr. R.I.P. 1951-2001
Homepage - http://www.umich.edu/~epickett
GPLRank - under construction...

Never argue with an idiot.  He brings you down to his level, then beats you
with experience...
Remove SPAM-OFF to reply.