Not fully the same but this is how I would create this in Openscad

Regards Hilbrand Klaver

$fn=$preview?20:100;

difference()
{
  minkowski()
  {
    difference()
    {
      scale([1.4,1,1])
        sphere(15);
      hull()
      {
        translate([3,0,8.5])
          rotate([90,0,0])
            cylinder(h=50,r=5,center=true);
        translate([-30,0,8.5])
          rotate([90,0,0])
            cylinder(h=50,r=5,center=true);
      }
      for (i=[-1,1])
        translate([-4,i*10,10])
          scale([1.2,1,1])
            cylinder(h=10,r=9);
    }
    sphere(1);
  }
  translate([0,0,-20])
    cube([50,50,40],true);
}

On 02-Mar-24 6:52 PM, Sanjeev Prabhakar via Discuss wrote:

Thanks Hibrand
This is a good idea, but I rarely use minkowski.
It is extremely slow and therefore almost unusable for me.
Regards
Sanjeev

On Sun, 3 Mar 2024 at 01:46, Hilbrand Klaver Gmail via Discuss <
discuss@lists.openscad.org> wrote:

Hi Sanjeev,

Are you using the latest development snapshot with manifold enabled?
(yes I am asking this again...)

The minkowski code takes about 1 minute to render on my machine (<1s if
$fn=20), not really fast but not unusable...

Best wishes,
John

On 3/3/2024 08:28, Sanjeev Prabhakar via Discuss wrote:

Hi John
I am using 14th Jan snapshot and also ticked the manifold option.
Haven't tried minkowski much with this and if that has made a
significant difference, will definitely try .
Thanks for the info

On Sun, 3 Mar, 2024, 2:34 pm Chun Kit LAM via Discuss, <
discuss@lists.openscad.org> wrote:

On Sun, Mar 03, 2024 at 05:03:57PM +0800, Chun Kit LAM via Discuss wrote:

Are you guys still interested in a model that works in prosaslicer,
but cannot be imported and rendered with openscad?

They are:

cam_upper_lid
cam_upper_part

from https://www.thingiverse.com/thing:3826740

I have the nightly from 19 feb with manifold enabled. Usually works a
charm, but not for these two.

Roger.

--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/ ** +31-15-2049110 **
**    Delftechpark 11 2628 XJ  Delft, The Netherlands.  KVK: 27239233    **
f equals m times a. When your f is steady, and your m is going down
your a is going up.  -- Chris Hadfield about flying up the space shuttle.

In a few of my models with manifold enabled, there is a rendering problem
if you check the option: "stop on the first warning" in the "advanced" tab
in "preferences"
This issue is observed in the latest nightly. Last one I was using, dated
14th Jan 2024, did not have this issue.
An example:
[image: image.png]

On Mon, 18 Mar 2024 at 22:23, Rogier Wolff via Discuss <
discuss@lists.openscad.org> wrote:

Out of curiosity, in the past when I have worked with renders, I always had
some option to color "bad" vertices/triangles/edges to visually call our
attention -- and also give debugging information on what is happening and
possible the location so we can zoom in...  Maybe OpenSCAD has some low
level debug option to do this.  If not, it would be nice to see it added.

BTW, my doing more than commenting is going to be nonexistent until I get a
half dozen dead standing trees felled and taken care of -- which are all
24" to 40" in diameter and leaning towards buildings and other
infrastructure... I'm more used to fighting forest fires (in my distant
past) than standard arborist fare.  Anyway, I plan to be a fly on the wall,
but cannot help for the foreseeable.

On Mon, Mar 18, 2024 at 9:42 PM Sanjeev Prabhakar via Discuss <
discuss@lists.openscad.org> wrote:

On Tue, Mar 19, 2024 at 07:11:45AM +0530, Sanjeev Prabhakar via Discuss wrote:

The "stop on first warning" is meant to help you find "the first problem"
with your openscad file.

Sometimes a single small error near the top will cause an enormous
cascade of further errors down the line.

Somewhere in the mid eighties my first pascal program compiled with
125 errors! I had missed a semicolon near the top. All other 124
errors were caused by that first missing semicolon.

Anyway, in openscad say a "warning module xyz doesn't exist" might
cause the whole thing of not having a purpose. So you can turn this on
and find the first "problem".

This all has to do with the parsing of the openscad source code and
nothing with the rendering backend "manifold" or whatever.

So... you might have "stop on first warning" enabled and not had any
warnings from previous versions and that some check has gotten a bit
more strict, so that where it previously worked it may not now.

You might say: it used to work, so why am I getting a warning now and
not before.

I'm just having this discussion with a friend: Having "strict"
warnings is just useless in 95% of the cases. But it does require you
to do the "quick fix" to get rid of the warning. But the remaining 5%
of the cases, you get handed: "A problem exists on line X of file Y",
which would have otherwise cost you hours, days or weeks to find the
bug if you hadn't spotted it on the first "compile".

My friend has debugged his code already on a previous version of the
software. So he is annoyed he gets to do all those 95% "small fixes"
today as opposed to over the years he's been writing/debugging all
that code.

Roger. 

--
** R.E.Wolff@BitWizard.nl ** https://www.BitWizard.nl/ ** +31-15-2049110 **
**    Delftechpark 11 2628 XJ  Delft, The Netherlands.  KVK: 27239233    **
f equals m times a. When your f is steady, and your m is going down
your a is going up.  -- Chris Hadfield about flying up the space shuttle.

I just tested this on the latest master build and they work fine with Manifold enabled.

-Marius