Re: Implement make-bow-stencil, make-tie-stencil for use in markup-commands undertie and overtie (issue 270640043 by address@hidden)

[thomasmorley65]

https://codereview.appspot.com/270640043/diff/20001/scm/stencil.scm
File scm/stencil.scm (right):

https://codereview.appspot.com/270640043/diff/20001/scm/stencil.scm#newcode88
scm/stencil.scm:88: (sqrt (+ (* width width) (* height height))))
On 2015/11/08 21:22:04, dak wrote:
> Is there a reason you are actually taking the absolute values of width
and
> height?  You don't need them for calculating the length here.

Not sure what you mean.
Isn't it simple Pythagoras?

https://codereview.appspot.com/270640043/diff/20001/scm/stencil.scm#newcode100
scm/stencil.scm:100: (/ height length-to-print))))
On 2015/11/08 21:22:04, dak wrote:
> This looks like a total mess.  Do you even need all this for
something?  What
> are you trying to do?

> Perhaps express it in words and we'll try finding a direct
representation that
> does not need all this juggling with angles and stuff.

You ask for quadrant and angle-start/stop-X-axis, right?

I hoped I had put in enough comments to explain what's done and why,
obviously not sufficient.

In general I have to points P(x1 y1) and Q(x2 y2) and want to create a
bow connecting them.
Therefore I first boil down it, to create a bow from Zero(0 0) to a
point on the (positive-part) of X-axis with with the coordinates
(direct-distance-P-Q 0), because it's far easier to calculate.
After this is done the bow is rotated around Zero to have it the same
orientation like the direct connection of P and Q.
Last step is to move the bow from Zero to P.

For the rotating part I need the value of the angle which depends on the
quadrants of Q relative to P. Calculated using Zero and P moved to
'new-stop'

Hope this is clearer now.

In general I regard the section containing quadrant and
angle-start/stop-X-axis the weakest part of the code.
Any suggestion is highly welcome.

https://codereview.appspot.com/270640043/