Rounding library

I recently been giving some love to a library I wrote to help with radii in
OpenSCAD.
Personally, the library is pretty bound to how I think about cading parts
now, I like to think it both helps with adding fillets as well as leveraging
the battle test paradigm of extruding from sketches of other cad packages
since it heavily uses polygons.
There are a number of function/modules, but the main api is
<http://forum.openscad.org/file/t1892/easy-api.png>
Though with some of the more recend editions I've been diving into
polyhedrons in order to round the end of polygons.
<http://forum.openscad.org/file/t1892/polyRound-extrude.png>

The feature list can be found here:
https://kurthutten.com/blog/round-anything-a-pragmatic-approach-to-openscad-design/
Github repo: https://github.com/Irev-Dev/Round-Anything

And originally I posted a thread here about it and got some really useful
feedback: http://forum.openscad.org/Rounded-Polygon-td21897.html

I would love to hear your thoughts/feedback.



--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

I am very interested in rounding and your original library inspired me to
write a bunch of rounding functions, some of which are similar to yours.
Take a look at:

https://github.com/revarbat/BOSL2/wiki/rounding.scad

The BOSL2 library also has rounding parameters to cuboid() and cyl() and
reguar_ngon() to produce these basic shapes with rounding already applied.
I find this general capability indispensable and use the various rounding
features in the BOSL2 rounding library all the time.  

I took a look at your listed examples and am curious what "point translation
helpers" does.  Also, what is the difference between polyRoundExtrude and
radius extrude?  I implemented something called offset_sweep that is similar
to this, though it takes arbitrary curves for the ends instead of just
fillet or roundover.   You might want to show code examples on your page so
people can really see the API.






--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

Oh dam,
Nice, looks like some good functionality there. Glad to see my work live on
in another library. Sorry you had to dig through my code. I think/hope I've
improved since then.

Some of your example smoothing in 3d is really nice.

I do have some examples of the api here:
https://kurthutten.com/blog/round-anything-api/

And up offset_sweep looks like it very similar to extrudeWithRadius, the
difference between that and polyRoundExtrude is the latter uses polygon
point (or radiiPoints) instead of a 2d child. It make it less flexible 2d
children certainly cover more cases. Though by using polygon points directly
mean it's able to make the shape using the polyhedron and it makes it much
more performant, and I don't have to use the step method of extruding small
slices with offset. Here's a gif of the fn value changing. It makes for a
much smooth transition at much lower values of fn as well since there are no
steps.
https://twitter.com/IrevDev/status/1292223481728208897



--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

I think I understand:  you're using a stair-step type construction to create
the shape with extrudeWithRadius and you're constructing a polyhedron for
polyRoundExtrude.  

My offset_sweep() will take any polygon, extrude it, and add rounding (or
some other end treatment) to either end.  It sounds like it's the same
method as your polyRoundExtrude.   I compute offsets of the input polygon
and use them to construct a polyhedron.  The real challenge in doing this
was in writing an offset() function, and book keeping when points disappear
during offset.   (It's very hard to write a robust offset function.)  I also
have convex_offset_extrude, which is like your stair-step approach but it
uses hull to make slices with sloped boundaries.  This means it won't work
on concave children.  (Well, it will hull them.)   I never thought the
stair-step modeling approach was acceptable.  

What do you see as the missing functionality for rounding?  What do you wish
you could do but you can't?






--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

Yup, stair-step. I did try the hull as well, but preferred the stair-step
over loosing the ability to use concave children. 👍

I don't have anything in mind that's missing.

Some times I wish the polygon could be used more like a "sketch" in other
cad packages, where you could add constrains instead of defining every
point, but that's got nothing to do with rounding.



--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

On 16.08.20 22:05, irevdev via Discuss wrote:
> Some times I wish the polygon could be used more like
> a "sketch" in other cad packages, where you could add
> constrains instead of defining every point, but that's
> got nothing to do with rounding.

How would you extend the language to express that?

ciao,
  Torsten.


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

I'm not familiar with this "sketch" feature.  What is it?  How does it work?


OpenSCAD mailing list-2 wrote
> I don't have anything in mind that's missing.
>
> Some times I wish the polygon could be used more like a "sketch" in other
> cad packages, where you could add constrains instead of defining every
> point, but that's got nothing to do with rounding.





--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

I'm not sure who you're addressing here, or which method, exactly.  Creating
a rounded cube by hulling spheres is a good method, and perhaps the fastest
way to do that job.   I haven't done any time testing on this.  You don't
need sophisticated rounding if you want to create a rounded cube.  The
intention of my rounding library (and the other one mentioned) is to apply
rounding in more sophisticated ways, to more complex objects.   What if you
want only parts of the cube rounded?  (Probably your approach still
works...in some form...but it begins to get complicated.)  What if you want
a concave object with rounding?   You cannot round a concave polygon with
hull.  What if you want a cube that is rounded at one end and flared out at
the other end for a smooth transition onto a plane?  

I actually implemented a large polyhedron library that will compute all of
the platonic, archimedean and catalan solids.  And I could not figure out a
correct way to round these solids---the method I implemented fails on the
catalan solids.  Doing this as a hull of spheres is difficult because it's
hard to figure out where to put the spheres.      


drxenocide wrote
> I do rounding for cubes by hull()ing around spheres placed at the corners
> (pulled in by the rounding radius, so my rounded cube has coplanar faces
> with a regular cube).
>
> Is your method faster than this?
>
> On Sun, Aug 16, 2020 at 6:42 PM adrianv &lt;

> avm4@

> Discuss@.openscad

>> http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org
>>
>
> _______________________________________________
> OpenSCAD mailing list

> Discuss@.openscad

> http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org





--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

drxenocide, Your hulled sphere's sounds fine. Though I think it becomes
difficult for more complex shape is all. I'm not sure about speed.

Adrianv, sketching is very common in cad packages. See this video from about
10s to 40s he makes a sketch https://www.youtube.com/watch?v=CbPIs4MzvzU.
Basically a bunch of points and he adds constraints by adding dimensions,
though not all are dimensions. At one point he selects two lines and
constains them to be the same length. Other things that are common is
defining a line to be vertical or horizontal, or constrain two lines to be
parallel.

tpv, not sure. I did play with an experimental API like this
<http://forum.openscad.org/file/t1892/experimental.png>
Where "L" means take the value from the previous point, and the last line of
the polygon is defining that's it's supposed to be 45 from the Y axis and
each letter of "ayra" means something, maybe like absolute y . .  I can't
remember. Obviously it could be cleaned up but I figure it a) doesn't shine
a light on normal sketches and b) is cumbersome. Ultimately sketching is a
GUI tool and not sure it can be translated to code in an intuitive way.



--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

OpenSCAD mailing list-2 wrote
> Adrianv, sketching is very common in cad packages. See this video from
> about
> 10s to 40s he makes a sketch https://www.youtube.com/watch?v=CbPIs4MzvzU.
> Basically a bunch of points and he adds constraints by adding dimensions,
> though not all are dimensions. At one point he selects two lines and
> constains them to be the same length. Other things that are common is
> defining a line to be vertical or horizontal, or constrain two lines to be
> parallel.

This doesn't really fit the OpenSCAD framework at all.  If you want to lines
the same length...you define them to be the same length.  I liked it when he
just selected some edges and rounded them over.  


There is the concept of turtle graphics, like move forward 10, turn left 100
deg, move forward until y=3, etc.  I implemented something like this and
find it pretty useful.  





--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

[Sketch]

One could imagine having a way to describe a figure and its constraints in the form of a data structure, and having a solver that would analyze the data structure to find a figure that satisfies the constraints.

Something like:

poly = solve([
    [ ABS, 0, [0,0,0] ],         // Point 0 is at the origin
    [ REL, 1, 0, [10,0,0]],      // Point 1 is +10 X from point 0
    [ LENGTH, [1,2], [0,1], 2],  // The distance from point 1 to
                                 // point 2 is twice the distance
                                 // from point 0 to point 1.
    [ ANGLE, [0,1,2], 90 ],      // The angle between points 0, 1, and 2
                                 // is 90 degrees.
    [ ABS, 2, [undef,undef,0] ]  // Point 2 is on the Z=0 plane.
]);

I think there are two solutions to those constraints:  [0,0,0]-[10,0,0]-[10,20,0] and [0,0,0]-[10,0,0]-[10,-20,0].  (Maybe only the -Y one, if the ANGLE rule follows the right-hand-rule.)

I don't know how hard it is to create such a solver.  (It would depend of course on what rules are allowed.)  Probably too hard for me.

I don't know whether anybody would really be interested in such a thing in OpenSCAD.

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

That seems like a reasonable api JordanBrown, But yeah not sure if it suits
OpenScad.



--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

It seems like if you allow a specific limited class of rules that operate
sequentially then it is not hard to do this.  The point basically is that
for every rule you need to ask if you can solve the relevant equation it
implies.  

This actually seems a lot like the turtle graphics concept but without the
turtle state or order:  at every step you create a new point based on a
previous point using some kind of instruction.   I am wondering whether this
scheme is significantly more powerful than my turtle implementation.   I
think the main possible source of extra power in this system is the ability
to refer back to previously generated points, but I can't tell if that
really lets you do anything different.  

If you don't require a sequential process then it is going to get a lot
harder, because the equations will form a large coupled nonlinear system.  A
general solver might be able to find a solution, but finding all solutions
will be hard to do, and the process may not be robust.   And of course there
may be no solutions, or an infinite set of solutions.   If someone gave me
this capability as a black box, how would I use it?  It seems likely that
this is going to be hard to use as a practical tool for constructing shapes.  



JordanBrown wrote
> Discuss@.openscad

> http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org





--
Sent from: http://forum.openscad.org/

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org

It seems like if you allow a specific limited class of rules that operate
sequentially then it is not hard to do this.  The point basically is that
for every rule you need to ask if you can solve the relevant equation it
implies.  

This actually seems a lot like the turtle graphics concept but without the
turtle state or order:  at every step you create a new point based on a
previous point using some kind of instruction.   I am wondering whether this
scheme is significantly more powerful than my turtle implementation.   I
think the main possible source of extra power in this system is the ability
to refer back to previously generated points, but I can't tell if that
really lets you do anything different.  

If you don't require a sequential process then it is going to get a lot
harder, because the equations will form a large coupled nonlinear system.  A
general solver might be able to find a solution, but finding all solutions
will be hard to do, and the process may not be robust.   And of course there
may be no solutions, or an infinite set of solutions.   If someone gave me
this capability as a black box, how would I use it?  It seems likely that
this is going to be hard to use as a practical tool for constructing shapes.  

What I was intending to describe - and what I *think* these tools do - is the latter.  I wrote out the rules more or less sequentially, because that's how I was mentally constructing the figure, but the intent was that they were an unordered set of constraints on the various points.

It's a very different way of thinking of design from the much more concrete "transform a shape" model that OpenSCAD uses.

As I said, it's not clear to me that there would be an audience for it in the OpenSCAD context.  I was just trying to describe how a tool like this could be carried over to an OpenSCAD environment.


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org