3D Graphic Coding?

G

Guest

Guest
Archived from groups: comp.sys.ibm.pc.games.flight-sim,rec.autos.simulators (More info?)

Hi Folks:

A project I'm working on is ready for some animated 3D
graphics. I'm kind of excited about that.

This is an instructional application, and the graphics
will be simple, so I believe this will be a good
introduction the topic.

I need pointers to books and resources on the net,
I believe there is a newsgroup dedicated to this 3d
developers.

People are always bitching about something in their
favorite sim. I wonder if there isn't talent enough in
RAS and the flight sim group to roll our own. With a
nice client/server multiplayer architecture, physics
just the way we want, multiple monitors, even
abstracting the APIs for deployment to non-Windows
platforms. I have no interest in flying against AI,
but those who care can tailor it to suit their needs.

I'm guessing there would be some code common to both
a flight and racing sim. People have built many plane
and race car models and terrains for existing sims.

Just a thought.

Larry
 
G

Guest

Guest
Archived from groups: comp.sys.ibm.pc.games.flight-sim (More info?)

Hi Larry ,

> I need pointers to books and resources on the net,
>I believe there is a newsgroup dedicated to this 3d
>developers.

Do you want to use OpenGL or DirectX?
BTW, do you speak German? I ask because I am mod in a German game
coding forum.

If openGL, the NeHe tutorials are good, if DirectX, the only thing I
can think of the top of my head are the tutorials that come with the
DX SDK. But I am sure there are more for both.

> People are always bitching about something in their
>favorite sim. I wonder if there isn't talent enough in
>RAS and the flight sim group to roll our own. With a
>nice client/server multiplayer architecture, physics
>just the way we want, multiple monitors, even
>abstracting the APIs for deployment to non-Windows
>platforms. I have no interest in flying against AI,
>but those who care can tailor it to suit their needs.

There are about 20-25 open source or freeware sims. From the top of my
head:
- FlightGear
- BoB + MA
- EECH
- glHorizon
- Michael Wessel's sim ( Flight Dynamics Simulator)
- Project Phoenix
- CSG (sp?)
- There is a book on how to write one, including source, but it is
quite old.
- The Finish project that itmo works on (Sorry, I forgot the name)
- The Heli sim of a single guy (again, I forgot the name).
- I have no idea on the Wings with Wire status
- Target for Tonight is dead now. However people want to start a sort
of successor.

Many more I can't think of.

There are some absolutely top notch people in the community, normally
working on very similar fields or even the same field in their day
job. However their number is limited. Just like players, most game
developers find flightsims completely boring and would never work on
one for free when instead they can work on a sexy game. Also, again
like players, most are interested in a certain area and for example
many that work for free on a jet sim will even for money not work on a
prop sim. Now divide the limited number of capable devs that want to
create something through the number of sims and the effect is that 80%
of the open source / freeware sims have no chance to compete with
commercial sims and the rest is struggling very hard. You could easily
put more hours into a flightsim than was actually used to create the
P51 (or probably any other WWII fighter), it's a massive undertaking,
especially if done from scratch.



> Just a thought.
>
> Larry

Bye bye,
Wolfram.
 
G

Guest

Guest
Archived from groups: comp.sys.ibm.pc.games.flight-sim (More info?)

Wolfram Kuss wrote:
> Hi Larry ,
>
>
>> I need pointers to books and resources on the net,
>>I believe there is a newsgroup dedicated to this 3d
>>developers.

Thanks Wolfram:

Isn't there a game developers newsgroup?

> Do you want to use OpenGL or DirectX?

For this application, I'd lean toward DirectX, unless
there was a particular advantage to OpenGL.

Is there a site that compares the two standards from
a developers perspective?

As a Unix programmer toiling in the dark side, Windows,
for the last year and a half I have no love for Microsoft.
But I find less sophisticated users reluctant to venture
outside of Microsoft's world.

Can the graphics API be abstracted to the point of
making it practical to support both standards?

> BTW, do you speak German? I ask because I am mod in a German game
> coding forum.

No, but I appreciate the invitation.

> If openGL, the NeHe tutorials are good, if DirectX, the only thing I
> can think of the top of my head are the tutorials that come with the
> DX SDK. But I am sure there are more for both.
>
>
>> People are always bitching about something in their
>>favorite sim. I wonder if there isn't talent enough in
>>RAS and the flight sim group to roll our own. With a
>>nice client/server multiplayer architecture, physics
>>just the way we want, multiple monitors, even
>>abstracting the APIs for deployment to non-Windows
>>platforms. I have no interest in flying against AI,
>>but those who care can tailor it to suit their needs.
>
>
> There are about 20-25 open source or freeware sims. From the top of my
> head:
> - FlightGear
> - BoB + MA
> - EECH
> - glHorizon
> - Michael Wessel's sim ( Flight Dynamics Simulator)
> - Project Phoenix
> - CSG (sp?)
> - There is a book on how to write one, including source, but it is
> quite old.
> - The Finish project that itmo works on (Sorry, I forgot the name)
> - The Heli sim of a single guy (again, I forgot the name).
> - I have no idea on the Wings with Wire status
> - Target for Tonight is dead now. However people want to start a sort
> of successor.
>
> Many more I can't think of.

Is there any effort to unify these projects into one
killer engine that can serve as the base of different
types of sims?

> There are some absolutely top notch people in the community, normally
> working on very similar fields or even the same field in their day
> job. However their number is limited. Just like players, most game
> developers find flightsims completely boring and would never work on
> one for free when instead they can work on a sexy game. Also, again
> like players, most are interested in a certain area and for example
> many that work for free on a jet sim will even for money not work on a
> prop sim. Now divide the limited number of capable devs that want to
> create something through the number of sims and the effect is that 80%
> of the open source / freeware sims have no chance to compete with
> commercial sims and the rest is struggling very hard. You could easily
> put more hours into a flightsim than was actually used to create the
> P51 (or probably any other WWII fighter), it's a massive undertaking,
> especially if done from scratch.

What is the difference between a jet sim and a prop sim?
I understand the different performance envelopes would
dictate different level of terrain detail, but isn't that
a value to be dialed in? Instrument panels have also
undergone several revolutions over the years, but I'd
think these can be isolated.

One of the posts in the "Pacific Fighters - Any Torpedoes?"
thread quotes a press release, which my browser can't find,
"The engine of IL-2 was designed in the same manner and it
still have at least 2 years in it, it`s very flexible".

What happens after two years? Newtonian physics hasn't
changed much recently, thrust, lift and drag are numbers.

I believe the IL2 team turned to the broader flight sim
community for aircraft, and perhaps other vehicle, modeling.
Would these models be the property of their creators? Are
they available for other sims?

My profound ignorance of this topic should be obvious,
I appreciate your advice.

Thanks
Larry
 
G

Guest

Guest
Archived from groups: comp.sys.ibm.pc.games.flight-sim,rec.autos.simulators (More info?)

"Larry Lindstrom" <nobody@aracnet.com> wrote in message
news:ItCdneTho-v5SanfRVn-rQ@comcast.com...
> Hi Folks:
<snip>
> People are always bitching about something in their
> favorite sim. I wonder if there isn't talent enough in
> RAS and the flight sim group to roll our own.

http://www.motorsport-sim.org/
 
G

Guest

Guest
Archived from groups: comp.sys.ibm.pc.games.flight-sim,rec.autos.simulators (More info?)

"Steve Blankenship" <steveNOSPAMblankenship@comcast.net> wrote in message
news:KJ-dnSSMfNRzhavfRVn-rQ@comcast.com...
> "Larry Lindstrom" <nobody@aracnet.com> wrote in message
> news:ItCdneTho-v5SanfRVn-rQ@comcast.com...
>> Hi Folks:
> <snip>
>> People are always bitching about something in their
>> favorite sim. I wonder if there isn't talent enough in
>> RAS and the flight sim group to roll our own.
>
> http://www.motorsport-sim.org/
>
>
I like what has been done to the old Falcon 4.0 simulator by various
teams over the years. Once Micropose announced no more releases and released
the source code, real F-16 pilots got together with developers and have kept
it the most realistic combat flight simulator to date. Very little of the
original is still there. Granted, the learning curve of the sim is very
steep and will turn off people who expect a point and shoot gaming
experience. You need to read the 400+ page F-16 training manual just to get
off the ground. This kind of thing could be done with N2003, but the current
holders of the code are somehow against it at this point in time. (Wasn't
the source code of N2003 released at some point? I really don't remember.
Maybe it was, but not legally). N2003 and GPL are, IMO, the most robust
simulators for modifications out there right now. Between the two, the
physics can be modified for almost any type of racing. I think it would
actually help with sales of a new release of either if the restrictions on
modification of codes were relaxed. I know that if some company wanted to
re-release a new version of Falcon it would go like hotcakes (as long as the
company kept the realism of the original). I always liked the way Papy
encouraged mods and even linked to them on their site. That flexibility of
having add-ons always kept us on the edge waiting for the next release for
the entire course of the sim. It certainly did not hurt the sales of
subsequent releases all the way to N2003 and GPL. We definitely have the
talent in the modding community to do almost anything with them, but the
current atmosphere seems to point to a gradual end of future improvements.
The GTP mod is a good example of what to expect when developers work hard to
give us a better racing experience. I know that the major problem with
racing vs. flight simulators is apples and oranges because of the extreme
cost of product licenses and labeling in racing today. With licenses for
tracks, automakers, sponsors and even drivers, it makes it very hard to
"roll your own" racing simulator.

Ed
 
G

Guest

Guest
Archived from groups: comp.sys.ibm.pc.games.flight-sim (More info?)

BTW, have a look here:

http://www.flipcode.com/
http://www.gamasutra.com/

Sorry, I do not know the English language newbie groups / forums /
websites.

> Can the graphics API be abstracted to the point of
>making it practical to support both standards?

Yes, but at a cost of course. If you would do it at very low level,
you would get a speed penalty, if you do it at high level (render that
object) then coding the two low level libs will be quite a lot of
work.

> Is there any effort to unify these projects into one
>killer engine that can serve as the base of different
>types of sims?

Unfortunately none with a chance of success, IOW of reducing the
number of projects and increasing the number of people working on a
single project. There is a huge ego problem. Basically all people
starting with this underestimate it. Some obviously think "the
commercial guys needed 30 man years for a small sim covering a tiny
area? If I get a few helpers for the simple stuff, I can do it in 2!".
So if they see a freeware sim with say 5 man years inside, they think
they can replicate that stuff in half a year and offer them to
completely throw away this 5 man years and instead join the new
effort, under the leadership of the new guy of course. These are those
that want to work together with other people. There are also those
that want to make it all by themselves so, when their sim ships and
kicks ass, they know it is all 100% their work. Actually I once spoke
to an even more extreme case. He was a professional developer and had
all the issues solved on paper and "knew" he could do a sim kicking
ass. He said himself he would like to know the end result would be his
alone (coding wise, he had a modeller and a large group of people in
his forum). He went so far as to on purpose not read articles in paper
or on the internet, since he wanted to know he has also found all the
algorithms etc by himself !! He said so to me himself as I wanted to
have a technical discussion with him and he did not know many of the
technical words. With egos like those, there are new projects popping
up (and then dying again after 3 years) all the time.



>> Also, again
>> like players, most are interested in a certain area and for example
>> many that work for free on a jet sim will even for money not work on a
>> prop sim. [...]
>
> What is the difference between a jet sim and a prop sim?
>I understand the different performance envelopes would
>dictate different level of terrain detail, but isn't that
>a value to be dialed in? Instrument panels have also
>undergone several revolutions over the years, but I'd
>think these can be isolated.

I think many flightsim devs at night dream about the sim that can do
everything from Lilienthal to the F22. Coding for say 10 generations
of planes is indeed definitely not 10 times the work of a sim for just
one, but say it is 2 times the work and there is no freeware or
commercial sim team that enough resources to make a one generation sim
as good as the hardware and known algorithms would allow.

It's not just the plane FMs and 3D models you need.
Like you said, the terrain should be optimised for the typical speed
and altitude. In extreme cases (say a hangglider or the Wright plane
etc) even another algorithm could be used. For example, I know one
commercial sim using a non flight sim engine, namely Wings of War.
Each map for them is 10x10 km and IMHO this is not by chance ;).

You need the jet engine model itself. You need the code for guided
missiles. For radar. For the new avionics. OTOH a prop sim needs stuff
like p effect, torque, etc. Complex sims like the Microsoft ones or
the ex Rowan sims do not have data files for the 3D models, but they
contain executable code in a language only these specific sims use and
no one else. For example, the MA/BoB engine has general commands like
"draw a poly", "draw a billboard" etc but also specialised ones like
"draw a gun sight", "choose the correct propellor", "rotate
propellor", "rotate guns", "rotate generic", "generic animation" etc.
A sim of modern jets would probably not support movable guns but would
for example support receptable for in air refuelling.

I certainly would like to exteend the time frame of BoB, but this will
need a community to provide the "data" and will still need quite a bit
of coding, at least going into the modenr area that the engine was
never made for. OTOH, the code stems from Flying Core Gold and there
are still WWI features inside.

> What happens after two years? Newtonian physics hasn't
>changed much recently, thrust, lift and drag are numbers.

It sounds good to the masses to say "our next sim is written from
scratch". Also, I do know coders that prefer to rewrite than to tweak.
I am the other way round, I hate throwing away all that work.

> Thanks
> Larry

Bye bye,
Wolfram.