Gene,

Try moving the include statements above (outside) of the module definition.

Len

On Thu, Jun 26, 2025 at 8:32 AM gene heskett via Discuss <
discuss@lists.openscad.org> wrote:

On 6/26/25 09:32, gene heskett via Discuss wrote:

And the formatting has been totally removed.  WTF?????

Cheers, Gene Heskett, CET.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis

On 6/26/25 11:04, Leonard Martin Struttmann via Discuss wrote:

That worked, thank you. Is this a required rule?

But I had to modify both the z offset, and the length of the
threaded_rod by quite a bit to get it to project from the cylinder at
both ends, so that line now looks like this:

    #translate([0,0,25])threaded_rod(d=12, l=51, pitch=2,
internal=false, bevel=0.5);

And I don't believe I see any bevel.

Threading is limited to the top by the 2nd cylinder and a bit of Z
tweaking made it clean to the bottom of the hole.

I see now that the line above this one clears the bore below the threads
so the added length of threaded_rod is not needed. The correct Z
translation would work as the OP seems to have a good understanding of
but using exactly the same for both subtractions was part of he error.

So my version with the stretched rod looks like this:

include </home/gene/.local/share/OpenSCAD/libraries/BOSL2/std.scad>
include </home/gene/.local/share/OpenSCAD/libraries/BOSL2/threading.scad>
module test();
{
    difference()
    {
        cylinder(h=50, d=16, $fn=150);
        translate([0,0,0])cylinder(h=32.5, d=12);
        translate([0,0,25])threaded_rod(d=12, l=55, pitch=2,
internal=false, bevel=0.5);
    }
}
test();

I hope this helps, it sure has me.

[...]

Cheers, Gene Heskett, CET.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis

That worked, thank you. Is this a required rule?

In this case, yes.  The OpenSCAD docs say "use <filename> is allowed only
at the top level of a file."

...and BOLS2 color.scad contains a use statement:

color.scad:use <builtins.scad>

On Thu, Jun 26, 2025 at 11:38 AM gene heskett via Discuss <
discuss@lists.openscad.org> wrote:

On 6/26/2025 9:37 AM, gene heskett via Discuss wrote:

Mostly, yes.

The include directive copies the specified file into your program,
almost exactly as if you replaced the include with the text of the
file.  (There's a little bit of magic to allow overriding variables in
the same scope.)

Usually, you want a library to be at the top level.  Some constructs may
require it - in this case, a "use" that BOSL2 uses internally.

That's not the only problem.

Here's your original text.  I've reformatted it for readability, but
changed only the whitespace, not anything that actually changes the
semantics.

module test();
{
    include </home/gene/.local/share/OpenSCAD/libraries/BOSL2/std.scad>
    include </home/gene/.local/share/OpenSCAD/libraries/BOSL2/threading.scad>
    difference()
        cylinder(h=50, d=16, $fn=150);
    translate([0,0,1.5])
        cylinder(h=32.5, d=12);
    translate([0,0,34])
        threaded_rod(d=12, l=16, pitch=2, internal=false, bevel=0.5);
}
test();

Reformatting like this makes a couple of things more obvious.

First is the semicolon after "module test()".  That makes the module be
empty.  The stuff in braces is some non-module stuff that's in what the
manual calls an "anonymous scope" which mostly isn't a scope at all.  An
anonymous scope mostly isn't a scope at all, but it's enough to make
"use" fail.

Second is that the difference() has only one child, so is kind of
boring.  I think it wants an open brace immediately after "difference()"
and a close brace after the call to threaded_rod().

So, going back to your original question, yes, libraries usually want
"include" or "use" at the top level.

There may be special cases where "include" might be useful at other points.

For instance, if you have a file that has a bunch of numbers separated
by commas, you could put it in an OpenSCAD array like so:

array = [
include <datafile.txt>
];

(Note:  this will not process general CSV.  The file must be valid
OpenSCAD syntax, which means commas between the elements, quotes around
strings, and so on.)

But these special cases are very much the exception.

For super extra fun, try this program.  Hint:  don't use "include" as a
variable name.

include = 2;
something = 5;
if (include < 3 && something > 4) {
    echo("surprise!");
}

The include provides definitions.  By putting the include inside the
functions that were included, the call to the functions doesn't know
where to find what to call.  Putting the include outside provides the
definitions of where to find the stuff you call.  All programming
languages (IIRC) need this kind of order.  Define it first, then call it.

Compilers hate surprises.

Harvey

On 6/26/2025 12:37 PM, gene heskett via Discuss wrote:

On Thu, 2025-06-26 at 12:37 -0400, gene heskett via Discuss wrote:

It's there, but it's a small bevel. Try changing it to 1.5 and put a
'#' at the beginning of that line. Of course, since the bevel is not
inside the cylinder, it will now be seen anywhere in the final render.

Right, but the cylinder which clears the bore needs to be smaller than
the bottom end of the threaded rod in order for the bevel to be in the
final render.

Because interior=false, the bevel is the wrong direction.  If you take a
rod and bevel the ends and then subtract that you're going to get a hole
with a little lip at the top.  You need to set inside=true, which reverses
the bevels.  Or you need to set a negative bevel so that it flares outward
so that it bevels the hole.

You can always try making the bevel really big, like how about 4.  You
wouldn't want such a big bevel, but you'll be able to see it!  A bevel on
a threaded rod can be subtle.

Really the threads will come out better if you use the interior argument
correctly.  In fact, I think with interior=false they won't work unless you
run the threads all the way off the end.

[image: image.png]

include<BOSL2/std.scad>
include<BOSL2/threading.scad>

threaded_rod(d=12, pitch=2, internal=true, l=25, bevel=2);
right(20)threaded_rod(d=12, pitch=2, internal=false, l=25, bevel=2);

On Thu, Jun 26, 2025 at 4:02 PM larry via Discuss <
discuss@lists.openscad.org> wrote:

On Thu, 2025-06-26 at 12:37 -0400, gene heskett via Discuss wrote:

Some comments:

include <BOSL2/std.scad>
include <BOSL2/threading.scad>
$fn= $preview ? 60 : 180; // automatic preview/render for $fn

// no semicolon after test()
module test() {
difference() {
cylinder(h=50, d=16, $fn=150);
// d=10 to allow for bevel
cylinder(h=32.5, d=10);
up(25+32.5) threaded_rod(d=12, l=50, pitch=2, internal=false,
bevel=1);

// uncomment the following line to see what's happening internally.
// cube(60);
}
}

test();

On 6/26/25 12:58, Leonard Martin Struttmann via Discuss wrote:

So is "include" an alias for "use"?  It seems to act like it is. except
"use" is a show stopper error

Which assumes the default library path is correct. So my use of the full
path should not be required?

It apparently is good, removing the full path still works.  Thank you.

Cheers, Gene Heskett, CET.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis

The "use" and "include" statements are different.  The intention of "use"
is to import a library while somewhat encapsulating it from the rest of the
code.  You're supposed to be able to run the file stand-alone to generate
examples, or run use it as a library.  Unfortunately, due to a bug in
OpenSCAD, if you use "use" then anytime you call a function in the used
file it runs every command that appears at the top level in the file.  This
was making some things in BOSL2 1000x slower, so we abandoned "use". The
"include" statement just directly includes the contents of the file, as if
they appeared right there in the code.

On Thu, Jun 26, 2025 at 5:23 PM gene heskett via Discuss <
discuss@lists.openscad.org> wrote:

On 6/26/25 13:11, Jordan Brown via Discuss wrote:

Actually two, first expands to clear the threads at the bottom of the
central hole,

And the second is the threaded_rod threading the top of the hole.

done.

This likely would not print well because there is no visible
compensation for the expansion caused by the size of the nozzle,
typically .4mm.

In my personal use where I am fitting a 2 start plastic printed nut pair
of halves, so I reduce the image of the maple screw that started as a
27" 2x2 that is just short of 48mm in diameter once carved on my 6040 4
axis mill, I use the same bolt code made nominally .25mm small to
preview the bolt, and make the same bolt code when used as the
difference to make the nuts, is made .25mm bigger when making the nuts.
So the assembly has about .1mm backlash yet turns freely when the screw
is painted with pure carnaba wax. With about 13 teeth engaged, that
maple screw might as well be 4340 steel. A buttress thread is only 4mm
tall, and the pitch is 12mm per turn of the vice handle. So if I were to
make this, and assuming a commercially available bolt of nominally 12mm
diameter, I would measure its actual OD and expand the clearance
cylinder in the bottom by .5mm, and the diameter of the bolt by .45mm to
assure adequate clearance.  Assuming of course that the printer is well
calibrated. This sort of thinking will get a usable part for the OP at
least 1 iteration quicker.

"use" does not work here.

Thank you.

Cheers, Gene Heskett, CET.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis

On 6/26/2025 3:13 PM, gene heskett via Discuss wrote:

Ah, I see that you fixed that in an iteration after the one I quoted. 
Sorry.

Some libraries want to be "included", while others want to be "used". 
BOSL2 wants to be "included".

On 6/27/25 13:25, Jordan Brown via Discuss wrote:

NP. While I was pointing out that the 2 subtraction children existed in
the OP's first post. That snippet did not constrain the scope of the
difference so I added a pair of {} to do so. Makes, in this case, no
difference in the final result as there is not anything that s/b
excluded from the difference beyond the scope of the difference.

A difference I don't see defined. It may be, but my hardcopy printout is
now 4 years old. I suspect that today it would be a dozen+ pages longer. ;)

Thanks Jordan.

Cheers, Gene Heskett, CET.

"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.

• Louis D. Brandeis

On 6/27/2025 11:58 AM, gene heskett via Discuss wrote:

The difference between the two is documented at

https://en.wikibooks.org/wiki/OpenSCAD_User_Manual/Include_Statement

I don't immediately see anything in the BOSL2 documentation that
discusses the distinction, but the examples all use "include".

I don't know why there should be a difference with BOSL2, but just for fun,
I opened an existing file that works fine, changed the "include" to "use",
and got errors.

On Fri, Jun 27, 2025 at 3:14 PM Jordan Brown via Discuss <
discuss@lists.openscad.org> wrote:

BOSL2 defines global variables at the top level that are then referenced by
other files than the one where they are defined.  This works with include
but does not work with use.

On Fri, Jun 27, 2025 at 4:32 PM Father Horton via Discuss <
discuss@lists.openscad.org> wrote:

It seems like the text about "use" is fairly explicit, when it says that if
A uses B then neither A nor B can see global variables defined in the other
file.

It doesn't mean global variable assignments "don't happen" because if B
makes global variable assignments they do get run---in fact they get
(re)run every time you invoke a function defined in B.  But those global
definitions are only global to B, not global to the whole program.

On Fri, Jun 27, 2025 at 7:15 PM Father Horton via Discuss <
discuss@lists.openscad.org> wrote: