Button vs. Message box scripts AND things rotational

[Jim Hurley]

>  Jim Hurley <jhurley at infostations.com>
>  > Subject: Re: Button vs. Message box scripts
>
>>  Suppose I wanted to alter
>>  revRotatePoly  (or use it as a base for a new routine) so that the
>>  center of rotation was a point of my choosing rather then the
>>  graphic's loc.
>----------
>Well, I don't think that'd be a native thing, but here is a couple
>alternatives I can think of:
>
>1) Stretch the box that the graphic is in so that the image area is offset
>from the actual loc by some dimension that will work for you.
>
>2) Use the Animation Manager. Select it and open the Animation Manager,
>click 'New', and set it to frame 1. Then rotate it manually (by message or
>whatever), and move it over so that it appears to have moved around a
>different axis, and apply that to whatever frame you like and test the
>animation to see if it's the effect you want. Adjust it until it's right,
>then put a [  revPlayAnimation "MyCrank" (or whatever)  ] handler in your
>button.
>
>HTH,
>Ken N.
>


Ken,

Thanks for the suggestions. Actually it is possible, with minor 
variations to use a routine published in the digest a few issues ago 
to write a rotation routine that will rotate any polygon (or line) 
about any pivot point. This is  a variation on a routine written by 
Derek Huby:

On mouseUP
   --Rotate a polygon "myPoly" about a button "Pivot" by 45 degrees
   rotatePoly "myPoly",the loc of button "Pivot", 45
end mouseUP

On rotatePoly theGraphic, thePivot, theAngle
   put the points of graphic theGraphic into tPoints

   put item 1 of thePivot into xPivot
   put item 2 of thePivot into yPivot

   put empty into newPointList

   repeat for each line tLine in tPoints --Set points relative to pivot
     put (item 1 of tLine)- xPivot & "," & (item 2 of tLine)- yPivot & 
return after tRelPoints
   end repeat

   put sin(theAngle* pi/180) into S
   put cos(theAngle * pi/180) into C

   put empty into rotptlist

   repeat for each line tLine in tRelPoints
     put round(C*(item 1 of tLine)+ S*(item 2 of tLine) + xPivot) 
after rotPtlist
     put "," after rotptlist
     put round(-S*(item 1 of tLine) + C*(item 2 of tLine)+ 
yPivot)after rotPtlist
     put return after rotPtList
   end repeat

   set the points of graphic theGraphic to rotPtlist

end rotatePoly

And a good tip from Cubist at aol.com. To avoid distortions through 
repeated rotations one might store the original points set in a 
custom property of the graphic. You would also store the net angle of 
rotation as a property. So  you start from scratch with each new 
rotation and avoid any accumulated distortion through round off 
errors.

Alejandro: I wasn't  able to access the utilities you suggest form 
web site you mentioned. Problems with my version of Stuffit. I'll 
work on it. But thanks for the tip.

BTY I have found a matrix inversion routine in the archives. Works 
well and fast. 60 milliseconds to invert a 4x4. Of course the speed 
decreases rapidly with size.

Undocumented archania: Also, in examining the RR version of 
revRotatePoly I finally figured out what " get  revPoints (graphic)" 
means.  It takes a list of items and converts them to a single item, 
presumably for speed. That is a  graphic pionts set obtained from 
"get the points of graphic...."

1,2
3,4
5,6

Looks like this when derived from "get  revPoints(myGraphic)"

1,2,3,4,5,6

And the points are subsequently summed over k where k takes on the 
values 2*k-1 and 2*k, that is the odd and even entries in the item 
set. The  odd number are item 1's and the even are item 2's. This 
would  be valuable for all the speed demons out there.

Jim