rec.autos.simulators

by **John Walla** » Tue, 27 Oct 1998 04:00:00

On Thu, 22 Oct 1998 18:41:39 -0400, "David G Fisher"

>Your post is interesting. There must be a bug somewhere and I hope Papyrus
>is looking into it. Obviously some are having success but many others
>aren't. It's not my connection. I've always had great success with online
>play. If they had released a demo with multi play ability they might have
>found these problems before the release of the full game.

They did AFAI remember - 250 people got a copy to try (or was it 150)?
I'm not sure we can even say that it's not your connection since it
may well end up being that. Perhaps it is not so stable, occasional
spike or whatever. It certainly seems that the same people have the
same problems, and since others have the same version of GPL and see
none it must be too do with how GPL interacts with each person's
hardware and internet connection.

Have you tried leaving Pingplotter running for a while and pointing at
Randy's server, perhaps that Mike give some idea of connection
stability or any problems? I know I have experienced problems with GPL
online, sometimes it's just impossible to join a race with even three
people. OTOH I've joined the Papy Cup races and several races on Ken
or Randy's servers with 14-20 people and had no problems at all -
that's from Europe with a 28.8 modem during peak time.

I've certainly encountered the odd problem, but the majority of the
time the racing is incredible. Some investigation into the
inconsistencies would certainly help though. I'm sure there's some
logic to why connects are good and why some are bad, but thus far we
don't know what it is. I can well understand that it would be very
frustrating if the connect tended to be bad.

Cheers!
John

by **John Walla** » Tue, 27 Oct 1998 04:00:00

On Sat, 24 Oct 1998 18:04:07 -0400, "David G Fisher"

>I
>am starting to think that is an answer someone gives simply because that's
>what someone told them and they choose to accept it as fact.

Hmmmm, okay read on....

I talked to

Looks like your first comment could be true - Q.E.D. :-)

Seriously though, I've seen the effects of the physics calculations in
GPL and they _are_ hefty, very hefty. I don't think that's the issue
though. I've had enough top quality online races with my Celeron 300
and 28.8 modem to see that there is no problem system wise, and I
raced online during beta with a P225MMX and same 28.8 modem.

Cheers!
John

by **John Walla** » Tue, 27 Oct 1998 04:00:00

On Sat, 24 Oct 1998 18:43:48 -0400, "David G Fisher"

>Papyrus said themselves they were not too interested in GPL online. That
>says it all. If you want to ignore that and think they still did their best
>with that attitude, fine.

I'm not aware of such a comment by Papyrus, and in any case the work
they put in to transform online play during development shows it not
to be the case.

You're right, it doesn't. What's more though, it isn't, at least not
for the majority of people racing. I'm painfully aware that this could
come across as the standard usenet "I'll defend my position and slag
off any contrary one", and I'm trying to word this to avoid that. I
know that I would be mightily hacked off at being in your position -
booted all the time and with poor performance. This especially true
given your experience with, and obvious enjoyment of, online racing.
That's primarily why I'm writing this, because I have had so much
enjoyment with online racing that I'd hate you to miss out on it :-)
I still have most of my races saved and occupying gigabytes of hard
disk space, since it's fun to review them sometimes - like having
videotapes of old F1 races. It's incredible to sit in someone else's
***pit and watch what they did and saw (like Mike Laskey and I
outbraking each other into Tarzan on Sunday - there must have been,
oh, an inch or so between our cars!).

Given that the vast majority are experiencing excellent online play an
investigation of why maybe could be a good use of your time? Did you
try a different ISP, a different modem, turning off compression,
changing DUN settings, turning off "poll with interupts enabled",
using DirectInput rather than generic, even maybe visiting someone
local who does think he has a good connection and see what the
difference is?

The same game that works fine on my system? It's the combination of
the game, your system and the ISP. The game is the same for everyone,
so if it was solely at fault no-one would be racing at all.

Cheers!
John

by **Andrew MacPhers** » Tue, 27 Oct 1998 04:00:00

In GPL's defence I imagine Quake1/2 is the biggest Net game in recent
history. Anyone on a 33/56k modem can only just about run a server for two
or three players at best. I remember experimenting with local users last
year and we could only get one player at a time onto a 28k hosted
connection.

Ok, I've no idea about the relative data formats etc of the various online
games, but ID sweated buckets to get their Net code optimised, and it's
never improved beyond the point where a central, fast server is required.

Bottom line is any decent games server has to be at the centre of a Net
hub sitting on a fat pipe with the relevant packet protocols given highest
priority. But even then you can't overcome individual user issues (like
line quality) or the release of big, popular demos which bring the Net to
its knees.

To add another analogy, I've flown Su27 online for well over a year now.
On a really good day you can get four people together. But usually three's
pushing it and two's often a nightmare given the world-wide spread of the
Flanker flying community. Even under ideal circumstances you're at the
mercy of local Net conditions.

Overall I've learned to be realistic rather than optimistic about Net
***. The Net wasn't/isn't designed for it and more people are getting
online every day, clogging it up.

Of course that doesn't explain why stuff like POD works well (not that I'd
know :-), and GPL could probably be coded better -- I don't know for sure.
But I do know that with just two opponents (on average) I've had some
really enjoyable online racing over the last few weeks.

Andrew McP

by **Andrew MacPhers** » Tue, 27 Oct 1998 04:00:00

Are you sure he's not implying that *historically* physics hasn't been a
big CPU hog in sims? I'd imagine -- without wanting to put words in
anyone's mouth -- that's what he meant. Graphics used to be the biggest
issue, but now acceleration frees up more CPU time, developers can dare to
do the hefty sums necessary to generate a "real" driving model.

Of course, if your friend's saying authentic physics calculations don't
take up much CPU time, then I suggest any astronauts reading RAS take up a
safer job immediately! :-)

Andrew_McP

by **David G Fishe** » Tue, 27 Oct 1998 04:00:00

>Of course that doesn't explain why stuff like POD works well (not that I'd
>know :-),

POD worked well for the time it was released (year and a half ago). Not by
today's standards. If the smiley face is supposed to represent that you are
above a game like POD, that's nice. I used POD because of it's relatively
new online ability and for fun. Art A. is as good as you'll see in GPL and
he raced POD online too.

If you read my original post, what I'm really comparing is the TCP/IP
performance of GPL, on a 33.6 or 56k modem, to CPR and MTM2. For GPL to only
offer 2-3 player ability while CPR and MTM2 offered up to 8 (CPR is a year
old), is not acceptable in my view. Especially with it's instability.

David G Fisher

by **David G Fishe** » Tue, 27 Oct 1998 04:00:00

>>Are you aware that Papyrus (according to beta testers) stated that
internet
>>play with GPL was something they were not too interested in and that the
>>beta testers had to push them for improvements to the online play? This
>>attitude by them was discussed both here at r.a.s. and at another forum
>>awhile back.

>No, that was simply your interpretation - read Mike's reply.

The betas tester said, very clearly, what I wrote in the above paragraph. It
was impossible to misinterpret. The quote was also discussed here and at
another GPL forum. Mike wasn't the one who said it. Interesting that his
view is so different from the other tester.

David G Fisher

by **David G Fishe** » Tue, 27 Oct 1998 04:00:00

>> According to him, physics (and AI) do not eat up that much
>> CPU in PC sims.

>Are you sure he's not implying that *historically* physics hasn't been a
>big CPU hog in sims? I'd imagine -- without wanting to put words in
>anyone's mouth -- that's what he meant.

No. He said that the CPU's available today are more than powerful enough to
handle the physics and AI calculations in a sim like GPL, and provide
stunning graphics, especially with graphics accelerators. The percentage of
CPU requirements for the AI and physics are not as high as some people at
r.a.s. are (for some reason) willing to believe.

The physics model in GPL is still a toy compared to what this man does for
NASA. If he were to start making car sims for the PC, he'd be taking a
gigantic step backward in his career. I'm going to take his word rather
than blindly follow the often ridiculous misinformation which is common in
newsgroups.

David G Fisher

by **John Walla** » Wed, 28 Oct 1998 04:00:00

On Mon, 26 Oct 1998 16:05:26 -0500, "David G Fisher"

>The betas tester said, very clearly, what I wrote in the above paragraph. It
>was impossible to misinterpret. The quote was also discussed here and at
>another GPL forum. Mike wasn't the one who said it. Interesting that his
>view is so different from the other tester.

I think this has been misquoted - as you can imagine this was a big
discussion point in testing and a lack of interest was FAR from the
truth.

Basically it was as Mike described it, but as you've said, people will
believe what the want to regardless of what they are told to the
contrary.

Cheers!
John

by **John Walla** » Wed, 28 Oct 1998 04:00:00

On Mon, 26 Oct 1998 16:32:11 -0500, "David G Fisher"

>The physics model in GPL is still a toy compared to what this man does for
>NASA. If he were to start making car sims for the PC, he'd be taking a
>gigantic step backward in his career. I'm going to take his word rather
>than blindly follow the often ridiculous misinformation which is common in
>newsgroups.

So you choose to believe one opinion which has no basis in fact for
you over another opinion which has equally scant factual support. In
what way are you not "blindly following"? :-)

Cheers!
John

by **David G Fishe** » Wed, 28 Oct 1998 04:00:00

John,

The tester's exact words were that "the beta testers had to push for
improvements" due to the attitude of Papyrus towards online GPL racing. The
post was here at r.a.s. If you'd like, I'll give you the name of the tester
and you can ask them yourself. Deja news probably has the post also.

David G Fisher

>I think this has been misquoted - as you can imagine this was a big
>discussion point in testing and a lack of interest was FAR from the
>truth.

>Basically it was as Mike described it, but as you've said, people will
>believe what the want to regardless of what they are told to the
>contrary.

>Cheers!
>John

by **John Walla** » Thu, 29 Oct 1998 04:00:00

On Tue, 27 Oct 1998 15:11:06 -0500, "David G Fisher"

>The tester's exact words were that "the beta testers had to push for
>improvements" due to the attitude of Papyrus towards online GPL racing. The
>post was here at r.a.s. If you'd like, I'll give you the name of the tester
>and you can ask them yourself. Deja news probably has the post also.

I probably saw the post too, but I also know that the way it has been
taken is not the case.

You now have comments from ONE beta-tester who you interpret as saying
multiplayer was not important for Papyrus and TWO telling you
categorically that the way you present this is not how it was. You
choose to believe the one which agrees with your pet theory.

Whichever way you slice it it's not very consistent.

Cheers!
John

by **Paul Jone** » Fri, 30 Oct 1998 04:00:00

I was racing MGP at Suzaka over a direct IPX phone line the other day. I got off to a
faster start than he did and plunged into the first corner confident in my lead when
bang - he appeared to knock me off the road - and that was me done for. After the race
I did the usual Schume rant "what the hell were you doing" etc. over the phone and he
said he never touched me but wandered how I managed to lose the car the way I did.
Next race, also at Suzuka (practising it so I know it for the forthcoming GP), he did
me when I over cooked the hairpin. Seeing red I flew down the track after him and in a
fantastic overtaking maneouve out-braked him at the Spoon. I felt great - revenge is
so sweet. But after the race he asked me what the **** I thought I was doing knocking
him off the circuit like that. And MGP looks relatively smooth - the AI must drive in
the gaps. TOCA, that's horrid except for modem to modem. Last night a guy pulled up on
my right, fine, no worries, a split second later he's on my left - what? Even GPL
dances a bit. Until we've got 1 megabit lines there is no solution, but it's still
more fun to race humans, however good the AI gets. It's like chess programs, you
haven't got an earthly. But the real fun comes from playing a human.
Paul

> On Fri, 23 Oct 1998 02:34:41 +0100, Paul Jones

> >> It would be a relatively simple matter to make an online
> >> sim with no warping at all. Whenever the sim lacks constant positional
> >> information it smoothly substitutes AI lines or a fixed line to avoid
> >> any jumps or warps - the player car is smoothly moved from one to the
> >> other.

> >But you would still have the possibility of one instance of the program believing
> >its car had crashed into the other one's because it had got the position of the
> >other wrong.

> Exactly, that's what I was saying. It can be done quite easily, but
> since the condition of the internet at the time of the race cannot be
> predicted the program itself must compensate by creating the illusion
> of a solid connect - that's all it is though. an illusion. Or, it may
> not be an illusion - until reviewing the two replays we don't know.
> It's basically a slider between smooth racing and accurate
> positioning, and everyone would place it somewhere different - I
> strongly favour only being able to race people with good pings but
> always knowing exactly where they are, but others will surely differ.
> Depending which path a program follows it will appeal to one or other
> camp.

> Cheers!
> John

rec.autos.simulators

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor

GPL Multiplayer is Poor