Re: Marking extrapolated ship locations

---------------------- Forwarded by Jared E Noble/AAI/ARCO on 03/08/99
09:24 AM
---------------------------

"Jared E Noble" <JNOBLE2@MAIL.AAI.ARCO.COM> on 08/14/98 10:45:39 AM

Please respond to FTGZG-L@bolton.ac.uk

To:   FTGZG-L@bolton.ac.uk
cc:    (bcc: Jared E Noble/AAI/ARCO)
Subject:  Re: Marking extrapolated ship locations

The problem with these maneuvers 'outside the envelope' is an incorrect
assumption of what valid orders are.  While the rules allow silly
constructions like this, common sense does not.  I will try to point out
how common sense in orders will alleviate this difficulty.

If we assume you have a main drive of 8, with thrusters of 4, this would
seem to imply that my applying the full power on your main drive over
the
course of the turn will change your velocity by 8 in the direction of
your
facing, right?	And that burning your thrusters over the full course of
the
turn could push you perpendicular to your facing a maximum of 4. (again,
keeping with the magical handwaving to erase the fact that it's really
only
half that, but at least staying with the same inconsistency everyone
else
is.)

*Concurrent Maneuvers*
Orders can be written in any sequence.	That's fine and dandy.	However,
it
is not reasonable to assume that a MD8, TP3,PS3 will change my vector by
11
in the direction of my initial facing. Why?  Because the MD8 burn takes
the
entire length of the turn! Even if you assume that rotations occur
quickly
and will 'give' them to me for free at the end (or even beginning) of my
movement, the PS3 requires another 3/4 turn to execute.  Now of course
everyone is screaming at me that things happen concurrently, not
sequenced
all at once, but they would HAVE to be sequenced to achieve the above
result.

*Vectors over time*
 I contend that since the MD burn is applied over the course of
movement,
you will need to break the vectors down further. and apply them in
smaller
increments. Move MD4, PS1.5 execute the turn, move the remaining MD4,
PS1.5.	now look at where you are at.  They are NOT the same location. 
In
the 'massively broken' hack we began with the velocity is +11 in the
direction of original facing.  In the second the velocity +6 (6.04
actually) in a direction about 30 degrees to port from the original
facing.

To make it more accurate (for a continuous burn, smooth rotation over
the
course of the turn) - Divide both the MD and Pushes by the points of
facing
change+1, and perform the maneuver in those discreet steps, turning one
facing between each movement.  So for the above, we have 4 segments of
MD
1, and 4 segments of PS 3/4.  Applying these as described gives a
velocity
change of right near 7 in a direction of 30 deg to port from original
facing.  Note that what we have done is a simple matter of 'Successive
approximations'.  Sure calculus would be more 'correct', but why bother
-
you are moving minis on the table anyway - turn and move.  The point is
that it is the assumptions of what are valid orders that are causing
some
of these abuses.

*Sequenced Maneuvers*
OK, with that said, you say that you WANT to do sequenced maneuvers? 
Fine
with me.  But remember that the sequence cannot be any longer in 'time'
that one turn.	I propose adding a single character to the normal order
writing conventions - "/".  This separates sequenced movement. 
Continuing
with the above example, you want to burn MD4/TP1/PS3+MD4, you can't.
Sorry.
But you could do MD4/TP1/PS2+MD4.  Do you see why?  The first step in
the
sequence is MD4, that takes half the turn. The next is TP1, which I am
counting here as instantaneous (partly to ease explanation, otherwise is
should be a quarter turn), finally the last step is PS3+MD4 (first
example,
and the PS3 takes 3/4 turn to execute - BZZZZ, sorry, but your time has
expired) or PS2+MD4 (second example, and it fits within the remaining
1/2
turn).

Lasty, about correctly marking what your course is during the turn -
however you do vector movement, pick up an artists or mathematicians
flexible curve tool.  It is a couple pieces of plastic that are joined
with
a sliding joint, so that when you hold 2 points in appropriate
directions,
the piece in between forms a smooth, continuous curve.	(I guess it
approximates a real-world Bezier).  The 2 points should be your initial
location and heading and the your final location and heading.  For
non-complicated (and non-sequenced) movement this should give you a
fairly
good approximation of where you were along your flight path.  I don't
know
what this tool is, but someone should.	Anyone taking Topology courses?

Jared

Jeff Lyon <jefflyon@mail.utexas.edu> on 08/14/98 04:56:29 AM

Please respond to FTGZG-L@bolton.ac.uk

To:   FTGZG-L@bolton.ac.uk
cc:    (bcc: Jared E Noble/AAI/ARCO)
Subject:  Re: Marking extrapolated ship locations

On 12 Aug 98 at 18:39, laserlight wrote:
> BTW it looks like you can have a max thrust of 11 under vector
> rules. "How?" you ask.  Thrust 8, burn 1 of your extra 4 maneuver
> points to rotate 90 degrees, then push port (or starboard) with the
> remaining 3 maneuver points.	Or execute the same steps in reverse
> order.  It's not logical (the 8 Thrust should occupy the entire
> length of the turn, so you don't have extra time to push in the same
> direction) but it is, insofar as I can tell and until the amendment
> I feel sure Jon will add about 30 seconds after reading this, legal.

I went back and re-read the rule and you are right.  As written, it is
perfectly legal.  The relevant passage states that "unlike in cinematic
movement" thruster movement points do not reduce the number of main
thrust
points available.

Was this how it was done in the EFSB?  Or did they still have it where
the
total of main thrust plus thrusters had to be less than or equal to the
ship's overall thrust rating?

I'm more comfortable with doing it that way and it seems like it would
be
the quickest and easiest fix to what seems like an abuse of the system.

In my opinion, if you are going to try and model vector movement, a
ship's
thrust rating should describe its maneuver envelope.  (Actually, using a
continuous acceleration model, it would be half that during the first
turn
of thrust, but let's not get into that...)  No matter how it manuevers,
it
should never be more than that distance from its projected endpoint.  If
a
thrust, turn, push maneuver lets you do something different, then your
model is broken.

Jeff