rec.autos.simulators

Hullo there,

A friend and I are trying to get GP2 working over a modem link.
Just so there's no confusion, here's the necessary information
about the 2 machines (both IBM compatibles, ofcourse) :-

Cyrix 5x86 (100Mhz?)                    Pentium 133Mhz
Cirrus Logic chipset                    DS64 VRAM
14400 vc/fax/modem (IBM MWAVE)          28800 external fax/modem

Obviously one machine has more processor power than the other,
and we can only run the link at 14400bps. But the game is
intolerably slow for anyone who knows that the framerate is
vitally important in making the game playable. There is an
unnerving 'delay' between control input and response by the
car. With either machine running as lead, it makes no difference.
We have tried it in VGA with everything turned off, whereby it
got marginally better but the game then seriously lacked enough
detail to judge bends/etc.

This is a shame because as most of you chaps/chapesses will know,
GP2's a great sim. Has anyone found this problem and, more
importantly, cured it?

A reply via e-mail would be greatly appreciated. Many thanks,

Dan Howell, UK

P.S. NFS rocks also! Much better as a "game" and still quite
fun in the handling/dynamics stakes.

I found that it worked best if you connected at 9600 baud and had only
your 2 cars racing (i.e. no computer ai cars).  With this setup, I
found it tolerable.

my $0.02 worth...

Hullo there,

A friend and I are trying to get GP2 working over a modem link.
Just so there's no confusion, here's the necessary information
about the 2 machines (both IBM compatibles, ofcourse) :-

Cyrix 5x86 (100Mhz?)                    Pentium 133Mhz
Cirrus Logic chipset                    DS64 VRAM
14400 vc/fax/modem (IBM MWAVE)          28800 external fax/modem

Obviously one machine has more processor power than the other,
and we can only run the link at 14400bps. But the game is
intolerably slow for anyone who knows that the framerate is
vitally important in making the game playable. There is an
unnerving 'delay' between control input and response by the
car. With either machine running as lead, it makes no difference.
We have tried it in VGA with everything turned off, whereby it
got marginally better but the game then seriously lacked enough
detail to judge bends/etc.

This is a shame because as most of you chaps/chapesses will know,
GP2's a great sim. Has anyone found this problem and, more
importantly, cured it?

A reply via e-mail would be greatly appreciated. Many thanks,

Dan Howell, UK

P.S. NFS rocks also! Much better as a "game" and still quite
fun in the handling/dynamics stakes.

Make sure you set the graphics detail level "down" to the lowest common
denominator machine.  then, you might try use the "use16550" command.

gp2 use16550

This forces gp2 to use the 16550 UART serial i/o chip found on most
modern motherboards.  Otherwise, it defaults to the 8250 (I think) mode,
which is slower in terms of serial throughput.

-r

--
Robin Chung                    
Senior Performance Architect, MTI-IC    Aspiring F-1 Vehicle Dynamicist
 AMA#699255 HRC#HM402849                        FAX:(510) 642 4769
My Next Dream Job at Williams F1: Wheel Locknut Specialist for Villeneuve.

You wrote modem play slow,

I have not done modem play, but I have hooked two Pentiums(150 and
166Mhz) up direct connect(SVGA mode) and it is like playing in slow
motion or very slowed and is very hard to judge when to break in the
corners. We gave up in direct connection between the computers. I tried
different baud rates but with no acceptable racing conditions.

I know you probably dont want to give up, but maybe someone else can
give some insight on how to speed up play. If it is slow with direct
connect, I would hate to see what it is like over modem.

Dave Thayer

Are you using a standard init-string? Well, it's very simple, don't.
They're probably not fully compatible with one or both modems.
I don't know the command-sets for your modems, but if they are Rockwell-bas=
ed and
Hayes-compatible, then one of the following init-strings might work :

  9600 : ATS37=3D9N0S46=3D0&Q0&K0%C0 ..(Zoom 14k4 fax/modem) ............mo=
dem fixed at 9600
  9600 : AT%C0\N0B8 ...............(markless 28k8 fax/modem) ........modem =
fixed at 9600
  14k4 : ATS37=3D11N0S46=3D0&Q0&K0%C0 .(Zoom 14k4 fax/modem) ............mo=
dem fixed at 14k4
  28k8 : AT%C0\N0B11 ..............(markless 28k8 fax/modem) ........modem =
fixed at 14k4
  28k8 : AT%C0\N0B15 ..............(markless 28k8 fax/modem) ........modem =
fixed at 28k8
  28k8 : AT%C0\N0&K0 ..............(Dynalink 1428VQH 28k8 fax/modem) modems=
peed determined
                                                                       by n=
egotiation with
                                                                       the =
other modem

  If you need the strings with B8 or B15, the value sets the speed,
  and 10 or 11 will set it to 14k4 (I don't have this modem, so I
  don't know the exact value for sure), you should be able to find
  this in your modem-manual.

  (In all strings the 0's are zero's, not capital o's)

All these init-strings turn off datacompression (%C0) and DTE/DCE-flowcontr=
ol (&K0)
and select normal, speed buffering, mode (\N0). () for Dynalink.

Some information about the above init-strings to help you find your way in
your modem-manuals :

Zoom-modem commands :
  S37=3D9 : sets 9600 baud
  S37=3D11: sets 14k4 baud
  N0    : modem will only handshake with other modem if connectionspeed
          matches S37 register.
  S46=3D0 : disables datacompression.
  &Q0   : asynchronous operation in direct mode, DTE (pc) speed must match
          telephoneline-speed.
  &K0   : disable flowcontrol (RTS/CTS).
  %C0   : disable datacompression.

marklees modem :
  %C0 : disable datacompression.
  \N0 : selects normal, speedbuffering, mode.
  B8  : selects 9600 baud
  B11 : selects 14k4 baud (might be 10)
  B15 : selects 28k8

Before I had the Zoom-modem (14k4), and I used the 9600 init-string.
When selecting a baudrate of 19200 in GP2 it slowed down the gameplay,
what probably means that GP2 tried to get rid of it's data at 19200,
while the modem couldn't serve it, so to be on the save side, select 9600 i=
n GP2.
I don't think GP2 'sees' that the modem is fixed at 9600 by the init-string=
,
so it won't see it beeing set at 14k4 either, if you would select 19200.
In other words : if you use an init-string that sets your modem fixed at a
certain baudrate, select a baudrate below or equal in GP2.

When you're using two different Pentiums, then the slower must be the
leader. The 'follow'-er gets a hard load to pull and probably will let
off steam every now and then.

While in menu's pressing Ctrl-M brings up a chatwindow.

The graphics-settings I'm using for other circuits than Monaco are :
(P120, 16MBEDO, DS64_2MBDRAM/S3SPDUP)
  While not linking : SVGA, Fences, Track, Hills, Trackside objects
                      No plot and no texture in mirrors
                      All trackside objects
                      Estimated framerate 17.0 , I set it at 13.4
  While direct-linking : SVGA, Fences, Hills, Trackside objects
                         No plot and no texture in mirrors
                         All trackside objects
                         Framerate set to 12.8
  While modem-linking : Same as with direct, but fps set to 12.1 or 11.8

Hope this helps,
Ren=E9.
-- =

*****************************     *****************************************=
**************
*  Dreammaster rules again  *     * Windows.....brings you yesterdays speed=
 today.      *

             *
* When Dream And Day Unite  *     *        .....the choice of a brainwashed=
 generation. *
*****************************     *****************************************=
**************

Thanks to everyone who mailed a reply, or answered here.
I've not had the chance to try it out as my friend has
disappeared over to the States for a couple of weeks, but
it sounds like the problem will be due to the lack of using
the 16550 option on loading.

As most have pointed out, it seems daft to default to 8250
UART settings when it surely can't be that hard to auto-detect
a 16550 UART (MSD does it fine, although I admit this isn't
anything to go by!!). Init strings used were, as most of you
pointed out, fine (compression off, etc). We will certainly
give it a go when he returns.

Many thanks to all,
Dan Howell