Adrian Mariano answered that quite well. That's what the technique covers.
Not two random points.

On 29/08/2021 20:19, Michael Möller wrote:

Hi Michael,

I did not understand what was being said, and I guess understanding
whether the 'you' that you were using was a generic 'you', instructions
for the world at large, or a specific you, a confirmation for Bob.  The
example I posed is just to represent two random points. (It would be
easy to calculate the distance from the two lines of code to generate
the cones in this specific case, but that was not what I was looking
for). The eyeball/cube/whatever can be applied anywhere, relatively
easily. I was of the understanding that you were describing a technique
that could be applied  between any two points, but by being calculated,
instead of eyeballing. I was hoping that based on the example two
points, the code could be written to generate the distance between them,
in a way that could be applied to any two points, no matter how much is
involved in generating them. I suppose it is different in specifying the
distance between two points, say,  compared to having two points and
trying to find the distance between them.

Best wishes,

Ray

Ray,

I think you need to distinguish two cases.  You are talking about "random"
points.  What's a "random" point?  To me it's the result of a rands()
computation, which means you have the point value in a variable, and hence
it's easy to compute with them.  If you have points that are defined in
some complex way by a sequence of geometry operations then there's no way
to gain access to those points.  So if, say, you don't understand how to
find the point of a cone (in your example) you'd be left with the trial and
error approach.  But anything that OpenSCAD can compute you can compute
yourself, so you can, in principle, write code to compute the location of
points that result from some complex sequence of operations.  We had a
discussion a while back about a bisection method for solving equations.
I've written a linear solver.  Tools like this can find points defined by
geometry.  But it's not something you can extract automatically from your
model.  It's something you have to do separately.

The question I have is: why do you need to know where a point is?  What are
you trying to do that requires this knowledge in a real model?  I wonder if
there might be alternative approaches that don't require that information,
or where the model can be structured so that it's easy to know where the
point is.  I suspect that at least some of the time, such alternative
approaches may exist.

On Sun, Aug 29, 2021 at 8:02 PM Ray West raywest@raywest.com wrote:

Hi Adrian,

My use of random was meant to refer points that i did not know the
position of (without a lot of work), any point on (or off) the object,
for example an imported dxf file, or stl generated elsewhere. In my
previous post I came  to the conclusion that you have mentioned in your
first paragraph, below.

wrt why? - If something needs to be made to fit something else.
Depending on the tolerances required, then it becomes an iterative
process of making prototypes, which could be reduced if distances were
more accurately known.

The problem seems to be caused (at the higher level) by the fact that
values can not be passed out from a module.  I'm thinking a solution
would be to parse the text openscad file in some other program/language,
merely to calculate points. The question then becomes how to define the
points of interest? e.g, in words I can describe it as 'the tip of the
cone that is not at the origin' (and then it needs a better definition,
since both tips are not at the origin) , but can I identify/name that
point in openscad?

I tend to work around the limitations to produce what I want, I'm not
adverse to reaming holes to the correct size, for example, but my
approach to these problems is more pragmatic than pure.

Best wishes,

Ray

On 30/08/2021 01:57, Adrian Mariano wrote:

On Sunday 29 August 2021 03:47:09 Gene Heskett wrote:

I am preparing to print 1 or two of what I have dl'd from the links above
and below, but first I thought I'd see what happens to a priint by
fiddling with temps and flows. Flow in particular seems to be excessive.
so as a test, using the printers builtin first layer calibration.
adjusted it for an initial single line thickness of .19mm, but the solid
pad at the end of that pattern was .25 to .27 thick, so I repeated that
but at 90% flow from the printers tune menu, and got a final pad .21mm
thick that when microscoped, was still nicely welded to the adjacent
line, but the top surface felt a lot smoother, so I have it doing one of
the spline parts, and its looking cleaner about half done than previous
copies have.  Then I'll do that 100mmX,100mmY,50mmZ thingy. Big enough
to separate scaling errors from nozzle radius errors.

Cheers, Gene Heskett

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene

The nozzle radius has no effect on the object dimensions. What matters is
the volume of plastic flowing through it.

On Mon, 30 Aug 2021 at 15:28, Gene Heskett gheskett@shentel.net wrote:

On 8/29/2021 10:48 AM, Bob Ewart wrote:

Zoom is your friend.

More likely is to incorporate it into OpenSCAD proper, so that the
native transformations track the cumulative transformation and make it
available.

That technique has advantages, but it also has serious weaknesses:

• It can't measure the distance between points on two sibling objects,
  because of the "black hole" effect.  (You can echo coordinates out
  to the log and then externally do math on them, but bleah.)
• It has no direct way to measure from anything but the origin. 
  Measuring against simple objects like cubes is easy math.  Measuring
  the height of a cylinder is easy.  Measuring from a particular point
  on the perimeter of a circle is not too bad.  Measuring between the
  two closest points of two circles starts to get ugly.  Measuring
  anything non-trivial involving boolean operations gets hard.

On 8/29/2021 6:44 AM, Ray West wrote:

Slight simplification:  the 3D extension of the Pythagorean Theorem is
straightforwardly x^2 + y^2 + z^2 = d^2, so you can get the distance
with sqrt(x^2 + y^2 + z^2); no need for the intermediate square root. 
(Note that you can derive this from your expression; substitute db into
the second expression and then cancel the square against the square root.)

Better simplification:  this is the definition of the norm() function,
so all you need is norm([xd,yd,zd]).  Note that this also means that if
you have two points p1 and p2, you can get the distance between them
with norm(p1-p2).

On 8/29/2021 5:57 PM, Adrian Mariano wrote:

I think the problem with straight "calculation" schemes is that they can
get very hard, very fast.

Quick, what's the distance between the closest points in these squares?

rotate(a) square(10);
translate([20,20]) rotate(b) square(10);

You're a much stronger math guy than I am, so maybe you can do that off
the top of your head, but it would take me some pretty serious
thinking.  And that's a simple case.

In a real project... as I've mentioned, a lot of my work is in scale
models.  I was designing an armchair:
https://cdn.thingiverse.com/assets/ed/00/42/d1/65/featured_preview_WIN_20200407_22_06_02_Pro.jpg

Here's what generates the sides and arms:

        // Sides
        for (s=[-1, 1]) {
            translate([s*w/2, d, leg_h+seat_h]) {
                rotate([90, 0, -90])
                    clip(v=[0,0,s]) {
                        pipe_polygon(r=wing_t, fill=true, open=true, [
                            [3*inch, 0],
                            [18*inch, 0],
                            [18*inch, 8.5*inch],
                            [4*inch, 8.5*inch],
                            [3.5*inch, 17*inch],
                            [4*inch, 26*inch],
                            [-3.8*inch, 28*inch]
                        ]);
                    }
                // Arms
                translate([s*0*inch, -20.75*inch, 8.75*inch])
                    rotate([-86.5,0,4*s])
                    cylinder(d2=wing_t*1.3, d1=wing_t*2, h=16*inch);
            }
        }

All of those dimensions are measured by hand off the real object, and
the angles are eyeballed.

Now, what's the width, measured between the outsides of the two arms?  I
want to tweak the angles and sizes so that's it's pretty close to correct.

On Monday 30 August 2021 11:21:51 nop head wrote:

Do you not get a nozzle dia oversize when you are driving the nozzles
center x distance? It is not and never can be a point src when it
is .4mm in diameter. So if I drive it exactly a mm, I will have -.2 at
the start of that mm. and a +.2mm at the end, for a 1.4mm total length.

The slicer is supposed to know that based on the nozzle dia and
compensate but the success of that is highly dependent on the flow,
which if way off, can exceed 2mm in width for the laid down line even if
its only .19mm thick.  With this printer, the initial prime on the edge
of the build plate as it starts is nearly 3mm wide using the default
preamble in cura. I have cut that by 50%, but still get that initial
line nearly 2mm wide. And that I am positive, contributes a lot to the
plastic buildup on the nozzle that will pull a pat loose and waste it.
Several times I have lost a print, and had to clean a 50 cent coin
diameter

Using a .19 layer, its now around 6mm up on that 100x100y50z test file.
It did one of the splines at 90% flow this morning, showing 5.99mm up as
it parked, but the part measures 6.72mm tall.

I'll know more when this is done, a bit over an hour from now.

Once thats fixed, then I'll do a temp tower to see if I can stop the
bridging. By then it will be close to time for another spool, and start
all over. Need 10 kg spools and driers to fit them... :-)

Thank you nop head.

wrote:

Cheers, Gene Heskett

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis
  Genes Web page http://geneslinuxbox.net:6309/gene