Conference 501clb::flight

Are you offering to help, Jim?

I'm the only one who has done anything since the release of V3.1, and I
didn't really finish what I started.  I was working on a full performance
overhaul, trying to really 'up the bar' for how many people could join into
one world and how fast they could interact.  I'd made some significant
progress, but no one seemed interested even in V3.1, so I pretty much shelved
it.

If there were some interested folks who'd be willing to put in some time, I
could see making an effort to make FLIGHT into a viable contender in today's
Flight Sim world.

Paul

	Well, here's my thoughts. I haven't been in here in a long time
	and noticed a bunch of new replies.

	I think you should consider the following:

	Go public domain, outside of Digital. Get others from all
	over the world involved, similar to the Linux movement.
	Use newsgroups for discussion and moderated code updates.
	(person submits code to moderated newsgroup, it gets checked
	into source control by the moderators on his/her behalf)

	Start with a Win95/Windows NT based product. Use 
	the DirectX API's for things like video and joysticks and
	sound. Code to API's and let Microsoft worry about the 
	underlying stuff. Same with networking.

	Use an existing flight editor, like the one developed for
	Flight Simulator.

	Those are just a couple of thoughts. I'd prefer PC-based
	only because of the market share it holds. VMS/Unix/MAC
	are good systems, but not everyone has them.

							mike

    I have to admit, its intriguing.  But my hidden agenda is expanding my
    exposure to C++/MFC on Win95 platforms.  I wouldn't be interested in
    Pascal or VMS.  Depending on scope, such a project might require a
    Unix-based server that used TCP/IP, but it would have to have a
    PC-based "front-end" which is the area I'd be most interested in.
    
    But, yes, I am intrigued by multi-player flight sims.  I was a member
    of Genie for some time just for access to Air Warrior.  My complaint is
    that everyone wants to make a buck on the sim fanatics, and although I
    wouldn't mind appearing on lifestyles of the rich and famous, I'd
    rather see the concept of Flight moved to the public domain where
    everyone could enjoy it.
    
    I agree that moving outside the fire wall is the final goal, I don't
    know if thats realistic in the short-term.  But if someone has a T1
    line and they are willing to sponsor development, let me know!
    
    jeb
    

Well here's my 2 cents worth.
I'd avoid the DirectDraw and Direct3d interfaces.
The directdraw interface doesn't run or runs poorly
on most modern 3d boards (it depends on a linear frame buffer
and most high end boards hide the frame buffer).
Direct3d is a nightmare to program especially because 
what is available in the interface varies from board to board.
A better 3d interface is OpenGL 1.1 (which conveniently comes
on all NT and windows 95 boxes). I'd stick with MFC or 
even use GLUT for the 2d part (GLUT is an interface library
built on top of OpenGL) The nice thing is that
a GLUT/OpenGL 1.1 port would work on all NT/ windows 95 boxes 
(better on stuff with a 3d card) and on UNIX boxes with
a 3d card. With reasnable 3d cards (Matrox Millenium) at about $400
and  better ones coming  this might be the way to go.

I of course assume you'll want to do surfaces and textures. 
That means new worlds, new vehicles etc etc. You're talking
alot of work. Spare time is something most of us have
darn little of these days :-).

Thanks,
Keith

I'm glad to see interest in flight simulation still around in DIGITAL. Here are
some miscellaneous thoughts.

John Buehler (the original FLIGHT author, now working for Microsoft) and Kevin
Farlee have spent some time in the last year on converting the FLIGHT Pascal
source code to C. I don't know the current state of this effort, but it was
well underway last I knew (using an automated translator of John's).

Depending on one's goals, it might be better to port the FLIGHT code or to
start a new code base:

     o	If it's basically to have FLIGHT on a new platform (e.g., Win32 client
	and Win32 or UNIX server), then porting would be faster.

     o	If it's to get a better simulator (visuals, physics modeling, etc.),
	then starting a new effort would perhaps be better. It would allow
	much more room for creativity, which might appeal to the software
	developers in the crowd. The FLIGHT sources could be consulted for
	ideas and perhaps algorithms.

RE Keith's reply:

>I of course assume you'll want to do surfaces and textures. 
>That means new worlds, new vehicles etc etc. You're talking
>alot of work.

You ain't kidding! It's been many years now, but I think that for me to make a
solid model of the AV-8B Harrier II aircraft (which I originally designed for
solids, including about twice as many model vertices as are used in the wire
frame representation) required on the order of 100 hours. Of course, it's a
more detailed model than would typically be necessary (containing some 1600
surface triangles), but it still gives an idea of the effort involved.
						-- Tom

	I'll yield to Keiths more intimate knowledge of the DirectX api's.
	OpenGL sounds fine. The goal being that some of the drawing
	is offloaded to the video card.

	RE: port of existing FLIGHT

	I know I'd be happy with that! Provided, of course, that support
	is put in for PC joysticks. I don't think I could go from my
	Thrustmaster setup back to a mouse. Also, some support for
	sound would be cool. 

	Hmmm.. I just had an idea. If the simulator could make certain
	events available to other programs, it would be rather easy to
	write some VB code that could produce sounds based on an
	event.

	If plane_type equals "JET" and flight_mode equals "straight_level"
	then play jet1.wav

	If plane_type equals "jet" and model equals "F-16" and g_force
	.gt. "4" play g4.wav and groan.wav

	Hopefully someone gets my drift. :)

	Yea, taking the existing FLIGHT and taking advantage of some
	of the technology available in the past few years could be
	very cool and ALOT of fun.

							mike

RE: <<< Note 1430.6 by 2954::FOLEY "http://axel.zko.dec.com" >>>

>	Hmmm.. I just had an idea. If the simulator could make certain
>	events available to other programs....

This would probably not be too hard. FLIGHT supports the reverse today, that
is, FLIGHT can be signalled from external programs to do certain things. I
recall that X events are the mechanism for communication.

This was developed so that the FLIGHTstick software could control an object,
and so that the Polhemus head-sensor software (that John played with) could
control the view. 
						-- Tom

	This would also allow for alot of other neat things that
	could be done. If it was incorporated into the world
	server, then a simple ActiveX or Java applet could be
	written to show who is in the world from a web page.

	There's a whole host of other ideas, but I'll defer till
	something comes out to play with.

							mike

    Anyone in touch with John?  Is his C port available, or would the
    effort have to be replicated here?
    
    jeb
    

RE: <<< Note 1430.9 by PCBUOA::BAYJ "Jim, Portables" >>>

>    Anyone in touch with John?  Is his C port available, or would the
>    effort have to be replicated here?

I converse via e-mail fairly frequently. His address is [email protected].
Kevin Farlee (DECwest) was working on the C translator with John last year.

I'm sure the current state of the C port (note: machine generated from the
Pascal original) could be obtained.
						-- Tom

A couple of reasons why:

1. No rayshade => fast play (render time + network communication time)
2. Wireframe descriptions may feed well into the VRML world.

BTW, I can't seem to find the latest kit anywhere... If I can get a full
up-to-date one with all the worlds, planes, etc, I'll try a convert to
VRML experiment.

$ dir LHOTSE::DKB0:[FLIGHT.KIT]
%DIRECT-E-OPENIN, error opening LHOTSE::DKB0:[FLIGHT.KIT]*.*;* as input
-RMS-E-FND, ACP file or directory lookup failed
-SYSTEM-F-UNREACHABLE, remote node is not currently reachable
$

Craig.

RE: <<< Note 1430.11 by CRAIGA::SCHOMP "Lord of the Rings" >>>

>BTW, I can't seem to find the latest kit anywhere... If I can get a full
>up-to-date one with all the worlds, planes, etc, I'll try a convert to
>VRML experiment.

LHOTSE went away a few months ago; sorry about that. I've got the V3.1 kit on
another machine, but it's not world-readable. I'll try to make it available,
and post another note.
						-- Tom

    Tom, any news on the availability of the FLIGHT kit?
    
    I managed to put my hands on an old VAX.....
    
    Thanks,
    Micky

RE: <<< Note 1430.13 by TAEC::BALLADELLI "Surfing with the Alien" >>>

>    Tom, any news on the availability of the FLIGHT kit?

Yes; sorry for the delay. The FLIGHT V3.1 kit is on node AIKAHI (61.634) in
directory DKA100:[FLIGHT.KIT]. 
						-- Tom

    Gee, that died off quick...
    
    Any updates from John about the automated port?
    
    jeb
    

RE: <<< Note 1430.15 by PCBUOA::BAYJ "Jim, Portables" >>>
    
>    Any updates from John about the automated port?

He hasn't mentioned it in a while. I'll ask him.
						-- Tom

I asked John about his C-port activities. He hasn't done anything on it in
quite a while, and doesn't expect to any time soon (if ever).

Any and all interested Digits are welcome to work on FLIGHT, be it enhancing
the current code base or in using the current code as a jumping off point. 
						-- Tom

    Well, there you have it!  Question is, is that carte blanche, or death
    knell?
    
    I don't know where all the potential developers are located, but if
    there's any way to get enough of us in one room to try to develop some
    synergy, I think thats the way to start.
    
    I'll start by recommending Acton as being halfway between Marlboro and
    Nashua, specifically, Crossroads restaurant, since they have well hidden
    booths that seat 6-8.
    
    Anyone interested in a tete-a-tete?  I could be wrong, but I'd guess
    that if we can't get everyone in the same room at least once, the odds
    of doing a joint development effort by remote control are slim.  Or
    perhaps I just haven't caught up with the cyberspace generation.
    
    Though it shouldn't be hard to find my DTN, I'll deposit it here for
    convenience, since sometimes landline is more expedient than mail or
    notes:  DTN 244-6789.
    
    Anyone up for it?
    
    jeb