rec.autos.simulators

l have been trying to connect to pcs via a serial cable and
null modem. l am consistantly getting data mismatch errors.

Has anyone been successful? If so can you let me know.
Thank-you advance!!

As I said in mail, lower the frame rate. A lot, if you have less than
a 486DX33.

Just out of interest, what processors & normal (unlinked) occupancy/
frame rate do you have?

Chisa
-- Gizmo
--
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
Launchpad is an experimental internet BBS. The views of its users do not
necessarily represent those of UNC-Chapel Hill, OIT, or the SysOps.
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

If you are using any editors, then BOTH GP.EXE files MUST match EXACTLY.  
If one of you uses the F1ED editor and the other GPED (the windows one),
then they edit slightly differently and the GP.EXE files will be different.

The best thing to do is to copy one of the GP.EXE files to the other
computer.  That should do the trick.

Dave,

This frame rate slow down thing - I have seen absolutely no slow down
when connected directly.  This was done on two 486/25's.  Of course, this
was before GPLAP, CCPIT and GPPERF.  Those programs do effect frame rate
and would probably ***down 2 computers that are connected, whether
it's serially or by modem.

Have these machines got 16550 UARTS or 16540/8540s? What I can tell
you is that (a) a 486DX2/66 and a P90 when directly connected both
experience BIG rises in occupancy, with or without TSRs, and
that a 386DX/33 and 486SX/25 were so bad we had to almost HALVE
the frame rate to run without constant link data mismatches.
(This was in the days before I even heard of GPlap & co.)

I use a RAMdisk when running GPperf linked. I tried logging to
disk but even the P90 doesn't have enough horsepower to cope.

The 386 definitely has 16540/8540s. The P90 has 16550s. I _think_
the 486 (it's the same machine with an upgraded CPU) has 16550s,
but I will check.

One other possibility is that if the machines are not a good match,
the timing skews I talked about earlier may be emphasized. Are
the two 486SX/25s _very_ similar or only CPU-wise?

We've tried in both the eastern and western hemispheres, but only
in the northern, not the southern. :-)

If version 2 includes IPX or better still TCP/IP I will cease to
care. Isn't ethernet wonderful?

Chisa
-- Gizmo
--
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
Launchpad is an experimental internet BBS. The views of its users do not
necessarily represent those of UNC-Chapel Hill, OIT, or the SysOps.
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

We use 2 486DX2/66's with 9600 modem link. And we have to drop to 18
fps with NO texturing to get smooth game play. Both PC's get about
55-60% occupancy when unlinked. The caller is about 95-100% when
linked at 18fps.

And yeh we both have 16550 uarts, no compression, correction etc.

so what gives???

Well  we've tried it in the southern hemisphere :-)

: We use 2 486DX2/66's with 9600 modem link. And we have to drop to 18
: fps with NO texturing to get smooth game play. Both PC's get about
: 55-60% occupancy when unlinked. The caller is about 95-100% when
: linked at 18fps.

Consider yourself lucky.  We are conecting with 2 486/33's at 14400 and
can only go to 12.0fps.

If you were to put both of those 486/66's in the same room and go with a
direct connection, you would be able to go with 25fps and full texture ON
(including track).

World Circuit/F1GP just does not play well over the modem.

I have got the serial link going at the highest rate between a P90 with
flashy UART (16640 or whatever it's called), and a 486/33 with the older
style UART with no problems.  I initially got the data mismatch errors
and found they were caused by one of the addons (probably different
driver/car lists). I got it to work when I re-installed (to a different
directory), so both machines were running a "***" copy of the s/w.

Perhaps you should try and prove the physical link first by trying
intersrv (should come with DOS), or DOOM or something else.

Failing that, I'm due in NZ late this year, if you're willing to pay my
air fair, I'll pop over to oz and sort it out for you :-)

Neal Rayner.

: l have been trying to connect to pcs via a serial cable and
: null modem. l am consistantly getting data mismatch errors.

: Has anyone been successful? If so can you let me know.
: Thank-you advance!!
Something all these articles miss... Is there really a data mismatch?
If the executeable on one end has been 'modified' by gpedit or similar
and is not absolutley identical (car powers, car colours etc) to the one
on the other end, there is an unavoidable data mismatch.. try syncing up
the exes.
Other than that, I have connected 386's to 486's and 486's to 486's
without problem.

M.Sempers

--
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
_/ Mark Sempers                                         _/

_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/

We have even tried a 19200 connect using 28k modems, with little
improvement. Yep, we are lucky to play 18fps over the modem.

Have you tried 13.6 fps without texturing? We have run two 386dx/33's
at 12.0/9600 and it was pretty good. The two 386s used to run 15.7fps
without texturing and full detail when unlinked.

I couldn't agree more :-)

: We have even tried a 19200 connect using 28k modems, with little
: improvement. Yep, we are lucky to play 18fps over the modem.

Same here.  In fact, when we connected with 28.8k modems, it actually
played WORSE than with 14.4k's.  Go figure.

: Have you tried 13.6 fps without texturing?

Yes, the rates I was quoted at are with texture off, but full detail
otherwise.  Selecting the lowest detail level only gives us about 5%
difference anyway, so why bother.

|>|> l have been trying to connect to pcs via a serial cable and
|>|> null modem. l am consistantly getting data mismatch errors.
|>|>
|>|> Has anyone been successful? If so can you let me know.
|>|> Thank-you advance!!

|>Something all these articles miss... Is there really a data mismatch?
|>If the executeable on one end has been 'modified' by gpedit or similar
|>and is not absolutley identical (car powers, car colours etc) to the one
|>on the other end, there is an unavoidable data mismatch.. try syncing up
|>the exes.

Yes that's exactly the problem. I did it yesterday and it took us ages to work
out what the problem was! Once I copied his .exe onto my machine all went fine,
except that I drove ***:-)

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

 David Hackett

 De Montfort University, UK.    WWW  : http://www.racesimcentral.net/~c2dh/

 "...I'd settle for a cup of coffee,
     but you know what I really need..."  

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

: >: We use 2 486DX2/66's with 9600 modem link. And we have to drop to 18
: >: fps with NO texturing to get smooth game play. Both PC's get about
: >: 55-60% occupancy when unlinked. The caller is about 95-100% when
: >: linked at 18fps.

: >Consider yourself lucky.  We are conecting with 2 486/33's at 14400 and
: >can only go to 12.0fps.

: We have even tried a 19200 connect using 28k modems, with little
: improvement. Yep, we are lucky to play 18fps over the modem.

: Have you tried 13.6 fps without texturing? We have run two 386dx/33's
: at 12.0/9600 and it was pretty good. The two 386s used to run 15.7fps
: without texturing and full detail when unlinked.

: >World Circuit/F1GP just does not play well over the modem.

I've played successfully at 24fps using a 486DX25 and 386DX16, without
texture but perfectly good.  This was using a standard null-modem
cable (the one supplied with LapLink as it happens).

Neither machine had `decent' I/O ports, ie both used 8250 or whatever
its called.  Why the problems?

Thomas

-  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -  -

Goodricke College
University of York                   "God preserve us from
York, YO1 5DD, UK                       religious nutters"

: I've played successfully at 24fps.....This was using a standard null-modem
: cable (the one supplied with LapLink as it happens).

: Why the problems?

You must have missed the key.  Over a null-modem, THERE IS NO LOSE IN
FRAME RATE.

When you play over a REAL modem, you really lose the ability to have a
decent frame rate.

Doug, you're so sure of this is frightening. The truth seems to be that
FOR SOME SETUPS EVEN NULL MODEM CONNECTIONS result in HIGHER OCCUPANCY.

You talk about "loss of frame rate", but this is just a symptom, not
the problem. You have to drop FPS because occupancy rises. Until you
grasp this simple fact you will not fully understnad the problem.

I have heard of and seen many examples of occupancy rising during null
modem link. In fact, Doug, you must be very lucky; you are the ONLY
person I have heard of who didn't experience SOME RISE IN OCCUPANCY
during null-modem linked play.

Let me reiterate my own experiences:

386DX/33 to 486SX/25, the latter quite capable of 25fps no texture on
its own, the former easily able to do 15 fps with minimum detail on its
own. Connect them, and the game becomes unplayable, with pauses,
mismatches, and frame rate needing to be reduced down to about 8fps
just to play for more than a few minutes. (This was pre-editors.)

Now, P90 and 486DX2/66, both eminently capable of 25fps texture full
detail on standalone without even breaking into a sweat. Connect
them, and the 2/66 needs texture off. If the /66 goes into the pits,
eee-yuck, major slowdown, occupancy at ~120% on it.

Both with identical game .exes, no TSRs.

Or, Doug, am I imagining it?

Wrong again. When you play over modem, in general the rise in occupancy
is even worse than that seen over null modem, typically resulting in
much greating reductions in frame rate being required.

Sorry about all the shouting but I really thought we done this topic
to death over the past few weeks and you seem determined to keep
plugging away at it despite reality.

-- Gizmo
--
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
Launchpad is an experimental internet BBS. The views of its users do not
necessarily represent those of UNC-Chapel Hill, OIT, or the SysOps.
-- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --

: Doug, you're so sure of this is frightening. The truth seems to be that
: FOR SOME SETUPS EVEN NULL MODEM CONNECTIONS result in HIGHER OCCUPANCY.

Fine.  I yield to your superior knowledge of the subject.