RE: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-05-01 Thread Norman Vine
Andy Ross writes:
> 
> Again, this is an artifact of the primitive collision handling.
> Ideally, every contact point would test against all scenery polygons
> without a notion of "ground plane".  That's simple to imagine, but for
> performance reasons a little hard to implement.

Yes, the FDMs assume a "ground plane" but note a pointer to the 
SSG Entity intersected and the index to the triangle in that entity
are stashed in the hitlist after elevation determination.

Hence one can efficiently determine the local terrain's slope for the
Location queried.

Norman

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


RE: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-05-01 Thread Norman Vine
Wolfram Kuss writes:
> 
> >Also being able to fly through buildings isn't really such
> >a problem, 
> 
> BTW, I remember at LinuxTag, when we taxied the Cessna and by chance
> sometimes came under the wing of the 747 of the scenery, then the
> Cessna would try to "jump up" and the program would crash.

This shouldn't happen any more.

Norman

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-05-01 Thread Gene Buckle
Nah, how about:

+---+
| We regret to inform you that your son was |
| killed because he was stupid. |
|   |
|  +--+ |
|  |  OK  | |
|  +--+ |
+---+


You get bonus points for knowing the quote. *laughs*

g.


On Sat, 1 May 2004, Ampere K. Hardraade wrote:

> Yes, that is nice.  However, I would like to see the plane get torn into
> pieces, explosions, fire and blacksmoke.  So we should have something like
> this:
>   +-+
>   |  Collision detected.  Crash scene is
>   |  being played.
>   |  Press OK when you finish watching.
>   |
>   |  +--+
>   |  |  OK  |
>   |  +--+
>   +-+
>
> Regards,
> Ampere
>
> On April 30, 2004 05:34 pm, Melchior FRANZ wrote:
> > * Erik Hofman -- Friday 30 April 2004 23:16:
> > > I was thinking more small Nasal script, in the line of:
> > >
> > >   +-+
> > >
> > >   |  Collision detected. Simulation halted. |
> > >   |
> > >   |  Press OK to restart.   |
> > >   |
> > >   |  +--+   |
> > >   |
> > >   |  |  OK  |   |
> > >   |
> > >   |  +--+   |
> > >
> > >   +-+
> >
> > Looks good. (But now I want to keep my crash sound. :-)
> >
> > m.
> >
> > ___
> > Flightgear-devel mailing list
> > [EMAIL PROTECTED]
> > http://mail.flightgear.org/mailman/listinfo/flightgear-devel
>
> ___
> Flightgear-devel mailing list
> [EMAIL PROTECTED]
> http://mail.flightgear.org/mailman/listinfo/flightgear-devel
>


___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-05-01 Thread Gene Buckle
> Arnt Karlsen wrote:
>
> >..dude.  This is another "common wisdom"?  I can understand 'not
> >allowing it with paying passengers'.  But I won't ever put my ass in
> >a spam can driven by some clueless burger flippers who has never
> >been _allowed_to_learn_ how to get out of trouble.
> >
> >..the IMHO appropriate way, is produce a "report on violation on
> >regulation with a big nice fat fine to pay."
> >
> >
>
> Crashes generate obscene forces.  Many of these high end simulators are
> connected to motion platforms.  People don't want to break their hardware.
>

This is quite correct.  The MD-83 sim cab has a 5000lb counterweight to
help balance the weight of the flight deck itself.  It's on a hydraulic
hexapod that's driven with a 4" diameter main feed hydraulic line that's
at 2150 PSI.  It can fling that 10,000lb+ simulator cab around like it
weighed nothing.  It is mounted to a 10 foot thick concrete pad that is
isolated from the rest of the building foundation by a 1 inch thick rubber
seal.  This prevents the vibration and motion of the simulator from
collapsing the building.

g.



___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-05-01 Thread Gene Buckle
> > At a minimum, the simulator should freeze with a message denoting a
> > destructive contact or out of bounds attitude.  For instance, the
> > MD-83 sim at Alaska Airlines is configured to freeze if the bank angle
> > exceeds 45 degrees - they don't want their pilots doing that unless
> > it's absolutely necessary.
>
> ..dude.  This is another "common wisdom"?  I can understand 'not
> allowing it with paying passengers'.  But I won't ever put my ass in
> a spam can driven by some clueless burger flippers who has never
> been _allowed_to_learn_ how to get out of trouble.
>
It's strictly for paying-passengers kind of procedures.  The bank angle
limiter is not active during FAA checkouts.

g.


___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-05-01 Thread Wolfram Kuss
>Also being able to fly through buildings isn't really such
>a problem, 

BTW, I remember at LinuxTag, when we taxied the Cessna and by chance
sometimes came under the wing of the 747 of the scenery, then the
Cessna would try to "jump up" and the program would crash.

>m.

Bye bye,
Wolfram.


___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Ampere K. Hardraade
Yes, that is nice.  However, I would like to see the plane get torn into 
pieces, explosions, fire and blacksmoke.  So we should have something like 
this:
  +-+
  |  Collision detected.  Crash scene is
  |  being played.
  |  Press OK when you finish watching.
  |
  |  +--+
  |  |  OK  |
  |  +--+
  +-+

Regards,
Ampere

On April 30, 2004 05:34 pm, Melchior FRANZ wrote:
> * Erik Hofman -- Friday 30 April 2004 23:16:
> > I was thinking more small Nasal script, in the line of:
> >
> >   +-+
> >
> >   |  Collision detected. Simulation halted. |
> >   |
> >   |  Press OK to restart.   |
> >   |
> >   |  +--+   |
> >   |
> >   |  |  OK  |   |
> >   |
> >   |  +--+   |
> >
> >   +-+
>
> Looks good. (But now I want to keep my crash sound. :-)
>
> m.
>
> ___
> Flightgear-devel mailing list
> [EMAIL PROTECTED]
> http://mail.flightgear.org/mailman/listinfo/flightgear-devel

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Al West
On Friday 30 April 2004 21:52, Curtis L. Olson wrote:
> Norman Vine wrote:
> >Melchior FRANZ writes:
> >>* Al West -- Friday 30 April 2004 20:05:
> >>>Is collision detection part of the model/dynamics set per aircraft? Just
> >>> I've been flying through the buildings in San Fran all afternoon in the
> >>> bo105.

I'd just like to add that it wasn't my intention to fly into buildings all 
afternoon - just I saw trying to land *on* buildings as a good next step from 
being able to pull into a hover.

> >>
> >>Contact points are per model, but the behavior is AFAIK the same for all
> >>models: you can fly through walls, but not through roofs (neither up nor
> >>down). I've no idea if this was intended. :-)
> >
> >FlightGear does not do collision detection.
> >FlightGear does Height Above Terrain determination.
>
> Right, and also notice that FG samples current terrain elevation once
> each frame.  So when you are one frame short of flying into an object,
> you still get the normal ground elevation.  Now the next frame you are
> inside the object.  FG looks for a highest terrain intersection point
> that is below your current elevation.
>

Forgive my ignorance...  but I think I might be volunteering here.  If the 
path of any contact point over those two frames positions intersects a 
polgyon then we have a collison yes?

Or per frame, if a contact point is on the opposite side of a polygon to any 
other contact point.   

Though at high velocities there is the case where you can pass through an 
object without being detected.

Personally I'm looking for things to do with the other processor on my box.  
And this seems a likely candidate.  Perhaps I should spend some time actually 
looking at some code.

> Because it samples only once per frame, it will almost never see the
> wall.  And as Norman and other's have pointed out, this logic allows us
> to taxi into hangers, fly under bridges, even taxi along the lower level
> of the bay bridge.
>

Although it does - in addition it allows you to taxi through hanger walls, 
bridge struts etc.

> Besides, flying into buildings isn't exactly top form these days.
>

Al is just part of my name, not part of the group I belong to ! ;-)

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Jim Wilson
"Curtis L. Olson" said:

> 
> Crashes generate obscene forces.

According to m-w.com:

One entry found for obscene.
Main Entry: ob·scene
Pronunciation: äb-'sEn, &b-
Function: adjective
Etymology: Middle French, from Latin obscenus, obscaenus
1 : disgusting to the senses : REPULSIVE
2 a : abhorrent to morality or virtue; specifically : designed to incite to
lust or depravity b : containing or being language regarded as taboo in polite
usage  c : repulsive by reason of crass disregard of moral or
ethical principles  d : so excessive as to be
offensive  


I think we just got culled by Google's SafeSearch (tm) ;-)

Best,

Jim


___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Curtis L. Olson
Arnt Karlsen wrote:

..dude.  This is another "common wisdom"?  I can understand 'not
allowing it with paying passengers'.  But I won't ever put my ass in 
a spam can driven by some clueless burger flippers who has never 
been _allowed_to_learn_ how to get out of trouble.

..the IMHO appropriate way, is produce a "report on violation on
regulation with a big nice fat fine to pay."
 

Crashes generate obscene forces.  Many of these high end simulators are 
connected to motion platforms.  People don't want to break their hardware.

Curt.

--
Curtis Olsonhttp://www.flightgear.org/~curt 
HumanFIRST Program  http://www.humanfirst.umn.edu/
FlightGear Project  http://www.flightgear.org
Unique text:2f585eeea02e2c79d7b1d8c4963bae2d



___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Arnt Karlsen
On Fri, 30 Apr 2004 14:27:20 -0700 (PDT), Gene wrote in message 
<[EMAIL PROTECTED]>:

> At a minimum, the simulator should freeze with a message denoting a
> destructive contact or out of bounds attitude.  For instance, the
> MD-83 sim at Alaska Airlines is configured to freeze if the bank angle
> exceeds 45 degrees - they don't want their pilots doing that unless
> it's absolutely necessary.

..dude.  This is another "common wisdom"?  I can understand 'not
allowing it with paying passengers'.  But I won't ever put my ass in 
a spam can driven by some clueless burger flippers who has never 
been _allowed_to_learn_ how to get out of trouble.

..the IMHO appropriate way, is produce a "report on violation on
regulation with a big nice fat fine to pay."

-- 
..med vennlig hilsen = with Kind Regards from Arnt... ;-)
...with a number of polar bear hunters in his ancestry...
  Scenarios always come in sets of three: 
  best case, worst case, and just in case.


___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Erik Hofman
Melchior FRANZ wrote:
* Curtis L. Olson -- Friday 30 April 2004 22:52:

Because it samples only once per frame, it will almost never see the 
wall. [...] Besides, flying into buildings isn't exactly top form
these days. 
That's OK. Also being able to fly through buildings isn't really such
a problem, although I generally prefer max realism (minus blood and
pilot disintegration). Still, I configured the bo105 to play a crash sound
if appropriate, and I would like to see the other aircrafts do the
same. This was desirable, because when flying from chase view with the
viewer position under the aircraft and looking in the sky, I often didn't
even notice a crash for a few seconds.  Not even the engine sound
stopped. Fooled me quite some times. A crash is only a "problem" in
real life, it can be educative in a simulator.  :-)
I was thinking more small Nasal script, in the line of:

 +-+
 |  Collision detected. Simulation halted. |
 | |
 |  Press OK to restart.   |
 | |
 |  +--+   |
 |  |  OK  |   |
 |  +--+   |
 +-+
Erik

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Gene Buckle
> * Curtis L. Olson -- Friday 30 April 2004 22:52:
> > Because it samples only once per frame, it will almost never see the
> > wall. [...] Besides, flying into buildings isn't exactly top form
> > these days.
>
> That's OK. Also being able to fly through buildings isn't really such
> a problem, although I generally prefer max realism (minus blood and
> pilot disintegration). Still, I configured the bo105 to play a crash sound
> if appropriate, and I would like to see the other aircrafts do the
> same. This was desirable, because when flying from chase view with the
> viewer position under the aircraft and looking in the sky, I often didn't
> even notice a crash for a few seconds.  Not even the engine sound
> stopped. Fooled me quite some times. A crash is only a "problem" in
> real life, it can be educative in a simulator.  :-)
>

At a minimum, the simulator should freeze with a message denoting a
destructive contact or out of bounds attitude.  For instance, the MD-83
sim at Alaska Airlines is configured to freeze if the bank angle exceeds
45 degrees - they don't want their pilots doing that unless it's
absolutely necessary.

g.



___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Curtis L. Olson
Norman Vine wrote:

Melchior FRANZ writes:
 

* Al West -- Friday 30 April 2004 20:05:
   

Is collision detection part of the model/dynamics set per aircraft? Just I've 
been flying through the buildings in San Fran all afternoon in the bo105.
 

Contact points are per model, but the behavior is AFAIK the same for all
models: you can fly through walls, but not through roofs (neither up nor
down). I've no idea if this was intended. :-)
   

FlightGear does not do collision detection.
FlightGear does Height Above Terrain determination.
 

Right, and also notice that FG samples current terrain elevation once 
each frame.  So when you are one frame short of flying into an object, 
you still get the normal ground elevation.  Now the next frame you are 
inside the object.  FG looks for a highest terrain intersection point 
that is below your current elevation.

Because it samples only once per frame, it will almost never see the 
wall.  And as Norman and other's have pointed out, this logic allows us 
to taxi into hangers, fly under bridges, even taxi along the lower level 
of the bay bridge.

Besides, flying into buildings isn't exactly top form these days.

Curt.

--
Curtis Olsonhttp://www.flightgear.org/~curt 
HumanFIRST Program  http://www.humanfirst.umn.edu/
FlightGear Project  http://www.flightgear.org
Unique text:2f585eeea02e2c79d7b1d8c4963bae2d



___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


RE: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Norman Vine
Melchior FRANZ writes:
> 
> * Al West -- Friday 30 April 2004 20:05:
> > Is collision detection part of the model/dynamics set per aircraft? Just I've 
> > been flying through the buildings in San Fran all afternoon in the bo105.
> 
> Contact points are per model, but the behavior is AFAIK the same for all
> models: you can fly through walls, but not through roofs (neither up nor
> down). I've no idea if this was intended. :-)

FlightGear does not do collision detection.
FlightGear does Height Above Terrain determination.

If you want to crash when you fly into a building 
in SRC / Scenery / hitlist.cxx  
change the two occurences of 

   if ( alt > hit_elev && alt < max_alt_m ) {
to
if ( alt > hit_elev ) {

Note with this change you won't be able to taxi inside of hangars 
or fly under bridges though.

Cheers

Norman

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel


Re: [Flightgear-devel] Re: COLLISION DETECTION: possible or not?

2004-04-30 Thread Andy Ross
Melchior FRANZ wrote:
> Contact points are per model, but the behavior is AFAIK the same for
> all models: you can fly through walls, but not through roofs
> (neither up nor down). I've no idea if this was intended.

It was.  This is the same rule that allows you to fly under bridges;
before that change happened, you could collide with buildings.  The
ground plane that an aircraft sees is the highest intersection of a
vertical line drawn downward from the aircraft coordinate origin with
the scenery.  That could be a building roof, bridge level, or ground.
But layers "above" the aircraft aren't seen.

Again, this is an artifact of the primitive collision handling.
Ideally, every contact point would test against all scenery polygons
without a notion of "ground plane".  That's simple to imagine, but for
performance reasons a little hard to implement.

Andy

___
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel