puzzling hang
Interesting that others have replicated this while you cannot. I
installed the latest build before posting to the list.
On 9/14/2025 12:28 AM, Jordan Brown wrote:
Not duplicated on 2025.09.12 on Windows.
But there is a related weird behavior.
With that h=1, it reports that the normalized tree has 507 elements.
With h>5, it reports 1008 elements.
And indeed, the "CSG Products Dump Dump" [sic] grows from ~23K to ~38K.
That behavior I see in 2025.09.12, and in 2021.01.
I don't understand that end of the program, so I don't know why that
would be, but it seems strange. And it's not new.
Jon, you don't say which OS you're running on. I'm running on Linux and my
snapshot is 2025.8.17. And as noted, with the default limit of 100k I get
the warning and hang in both that version and in 2021.01. I noticed that
the number of elements changed, which seemed odd to me. But also, if the
number of elements is around 1000, how is it over 200k? I didn't
understand that.
On Sun, Sep 14, 2025 at 7:34 AM jon jonbondy.com via Discuss <
discuss@lists.openscad.org> wrote:
Interesting that others have replicated this while you cannot. I
installed the latest build before posting to the list.
On 9/14/2025 12:28 AM, Jordan Brown wrote:
Not duplicated on 2025.09.12 on Windows.
But there is a related weird behavior.
With that h=1, it reports that the normalized tree has 507 elements.
With h>5, it reports 1008 elements.
And indeed, the "CSG Products Dump Dump" [sic] grows from ~23K to ~38K.
That behavior I see in 2025.09.12, and in 2021.01.
I don't understand that end of the program, so I don't know why that
would be, but it seems strange. And it's not new.
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Windows 11
On 9/14/2025 9:16 AM, Adrian Mariano via Discuss wrote:
Jon, you don't say which OS you're running on. I'm running on Linux
and my snapshot is 2025.8.17. And as noted, with the default limit
of 100k I get the warning and hang in both that version and in
2021.01. I noticed that the number of elements changed, which seemed
odd to me. But also, if the number of elements is around 1000, how is
it over 200k? I didn't understand that.
On Sun, Sep 14, 2025 at 7:34 AM jon jonbondy.com
https://urldefense.proofpoint.com/v2/url?u=http-3A__jonbondy.com&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AsrE-c7ZR7B2Kyr3qgfvvppkCEBVsNmwEMndcrRSuOI&m=J0TH421qmFpS9OwAvKK1G8leRWCL8ewYb6kkSsDIqxNMSFjTvQUJVEEspiXcDm7F&s=NVPsqP7dXbdzjVYSFjJpjZ8sz6PZ4hZXBsGL-1h4MzA&e=
via Discuss discuss@lists.openscad.org wrote:
Interesting that others have replicated this while you cannot. I
installed the latest build before posting to the list.
On 9/14/2025 12:28 AM, Jordan Brown wrote:
Not duplicated on 2025.09.12 on Windows.
But there is a related weird behavior.
With that h=1, it reports that the normalized tree has 507
elements.
With h>5, it reports 1008 elements.
And indeed, the "CSG Products Dump Dump" [sic] grows from ~23K
to ~38K.
That behavior I see in 2025.09.12, and in 2021.01.
I don't understand that end of the program, so I don't know why
that
would be, but it seems strange. And it's not new.
_______________________________________________
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email todiscuss-leave@lists.openscad.org
--
This email has been checked for viruses by AVG antivirus software.
www.avg.com
Jon, do you get the error and hang in 2021.01 under Windows?
I deleted lines from Jon's example to obtain this minimal case:
$fn=8;
difference() {
union() {
cylinder(h = 1, d = 2);
cube(2);
}
for (z = [1:446]) cylinder(h = 11, d = 3);
}
If you change the number of overlapping cylinders to 445 there is no
error. Original code had the cylinders translated so they didn't all
overlap. You can add back a small translation based on z and the error
still occurs.
On Sat, Sep 13, 2025 at 11:14 AM nop head nop.head@gmail.com wrote:
With it set to 10 million I never see the error. It does seem a bug if it
stops OpenSCAD working though.
On Sat, 13 Sept 2025, 15:00 Adrian Mariano via Discuss, <
discuss@lists.openscad.org> wrote:
Increasing the limit does seem to work, but the behavior here seems
strange and broken. I confirmed that in both 2021.01 and a recent snapshot
that it fails with the warning about too many elements. If you increase
the limit to 10 million then it succeeds and displays a message:
Normalized tree has 1020 elements!
which is a lot fewer than the 200k that triggered the problem. And like
Jon said, after you get that warning, OpenSCAD is useless. Clearing the
cache has no effect. Nothing will run, not even other simple files. So
these questions call out for answers::
-
Why do we get a warning about too many elements when there aren't very
many elements? -
Why does OpenSCAD go into a hung state after this warning?
On Sat, Sep 13, 2025 at 9:23 AM nop head via Discuss <
discuss@lists.openscad.org> wrote:
Increase the limit in the preferences. I have it set to 10 million.
On Sat, 13 Sept 2025 at 13:54, Jon Bondy via Discuss <
discuss@lists.openscad.org> wrote:
If you F5 the program, below, it should run without error. Then change
the "h = 1" to "h = 10" at the "<<<<" comment, and you get this
======
Parsing design (AST generation)...
Saved backup file: C:/Users/jon/OneDrive/Documents/OpenSCAD/backups/Pick Holder-backup-agrqbqpZ.scad
Compiling design (CSG Tree generation)...
Compiling design (CSG Products generation)...
Geometries in cache: 6
Geometry cache size in bytes: 48096
CGAL Polyhedrons in cache: 0
CGAL cache size in bytes: 0
Compiling design (CSG Products normalization)...
WARNING: Normalized tree is growing past 200000 elements. Aborting normalization.
Execution aborted
======
I am puzzled by the tree being too big (any recommendations?), but
after this happens, if I change h back to 1, and press F5, nothing
happens. I have to quite out of OpenSCAD and re-start it in order to try
again. That is, OpenSCAD is essentially dead.
I am running OpenSCAD https://www.openscad.org/ version 2025.09.12
========
eps = 0.01;
$fa = 3;
$fs = 0.2;
inches = 25.4;
di = 15.3 - 1;
th = 2;
ht = 100;
difference() {
union() {
cylinder(h = ht, d = di + 2th);
translate([-1.4di, 0, 0])
scale([2, 1, 1])
rotate([0, 0, -45])
cube([0.9di, 0.9di, ht]);
}
rotate([0, 0, -45])
cube(110);
cylinder(h = ht + eps, d = di);
for (z = [0:0.2:100]) {
le = 0.5 * sin(60*z);
translate([-di, le, z])
rotate([0, 90, 0])
cylinder(h = 1, d = 1); // <<<<<
}
}
http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient
Virus-free.www.avg.com
http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient
<#m_-4505123097453822859_m_-3208644596519836369_m_7424113970590188521_m_3532377620774460456_m_-7671437648204620332_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email to discuss-leave@lists.openscad.org
Parsing design (AST generation)...
Saved backup file:
C:/Users/jon/OneDrive/Documents/OpenSCAD/backups/unsaved-backup-firnpdqo.scad
Compiling design (CSG Tree generation)...
Compiling design (CSG Products generation)...
Geometries in cache: 3
Geometry cache size in bytes: 3336
CGAL Polyhedrons in cache: 0
CGAL cache size in bytes: 0
Compiling design (CSG Products normalization)...
WARNING: Normalized tree is growing past 200000 elements. Aborting
normalization.
Execution aborted
On 9/14/2025 10:38 AM, Adrian Mariano via Discuss wrote:
Jon, do you get the error and hang in 2021.01 under Windows?
I deleted lines from Jon's example to obtain this minimal case:
$fn=8;
difference() {
union() {
cylinder(h = 1, d = 2);
cube(2);
}
for (z = [1:446]) cylinder(h = 11, d = 3);
}
If you change the number of overlapping cylinders to 445 there is no
error. Original code had the cylinders translated so they didn't all
overlap. You can add back a small translation based on z and the
error still occurs.
--
This email has been checked for viruses by AVG antivirus software.
www.avg.com
Seems weird that it's happening to you under Windows but not Jordan. It
seems unlikely it depends on Windows version given that it's also happening
under Linux, which leaves differences in the settings as the explanation
for why Jordan doesn't reproduce it.
On Sun, Sep 14, 2025 at 11:01 AM Jon Bondy jon@jonbondy.com wrote:
Parsing design (AST generation)...
Saved backup file:
C:/Users/jon/OneDrive/Documents/OpenSCAD/backups/unsaved-backup-firnpdqo.scad
Compiling design (CSG Tree generation)...
Compiling design (CSG Products generation)...
Geometries in cache: 3
Geometry cache size in bytes: 3336
CGAL Polyhedrons in cache: 0
CGAL cache size in bytes: 0
Compiling design (CSG Products normalization)...
WARNING: Normalized tree is growing past 200000 elements. Aborting
normalization.
Execution aborted
On 9/14/2025 10:38 AM, Adrian Mariano via Discuss wrote:
Jon, do you get the error and hang in 2021.01 under Windows?
I deleted lines from Jon's example to obtain this minimal case:
$fn=8;
difference() {
union() {
cylinder(h = 1, d = 2);
cube(2);
}
for (z = [1:446]) cylinder(h = 11, d = 3);
}
If you change the number of overlapping cylinders to 445 there is no
error. Original code had the cylinders translated so they didn't all
overlap. You can add back a small translation based on z and the error
still occurs.
http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient
Virus-free.www.avg.com
http://www.avg.com/email-signature?utm_medium=email&utm_source=link&utm_campaign=sig-email&utm_content=emailclient
<#m_-3051569756275543422_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
Is there any easy way to dump all of my settings into a text file?
On 9/14/2025 11:31 AM, Adrian Mariano via Discuss wrote:
Seems weird that it's happening to you under Windows but not Jordan.
It seems unlikely it depends on Windows version given that it's also
happening under Linux, which leaves differences in the settings as the
explanation for why Jordan doesn't reproduce it.
On Sun, Sep 14, 2025 at 11:01 AM Jon Bondy jon@jonbondy.com wrote:
Parsing design (AST generation)...
Saved backup file:
C:/Users/jon/OneDrive/Documents/OpenSCAD/backups/unsaved-backup-firnpdqo.scad
Compiling design (CSG Tree generation)...
Compiling design (CSG Products generation)...
Geometries in cache: 3
Geometry cache size in bytes: 3336
CGAL Polyhedrons in cache: 0
CGAL cache size in bytes: 0
Compiling design (CSG Products normalization)...
WARNING: Normalized tree is growing past 200000 elements. Aborting
normalization.
Execution aborted
On 9/14/2025 10:38 AM, Adrian Mariano via Discuss wrote:
Jon, do you get the error and hang in 2021.01 under Windows?
I deleted lines from Jon's example to obtain this minimal case:
$fn=8;
difference() {
union() {
cylinder(h = 1, d = 2);
cube(2);
}
for (z = [1:446]) cylinder(h = 11, d = 3);
}
If you change the number of overlapping cylinders to 445 there is
no error. Original code had the cylinders translated so they
didn't all overlap. You can add back a small translation based
on z and the error still occurs.
<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.avg.com_email-2Dsignature-3Futm-5Fmedium-3Demail-26utm-5Fsource-3Dlink-26utm-5Fcampaign-3Dsig-2Demail-26utm-5Fcontent-3Demailclient&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AsrE-c7ZR7B2Kyr3qgfvvppkCEBVsNmwEMndcrRSuOI&m=cI5OQvjofAcGkyb8nkEn2lygu_l97ybIqlbf4H_7UiRRls5XV_iRbpIzdAxCSamd&s=C_6ofJ9k9o6KWlQHPu26hdtMqXPYACoe755TA1Tjqik&e=>
Virus-free.www.avg.com
<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.avg.com_email-2Dsignature-3Futm-5Fmedium-3Demail-26utm-5Fsource-3Dlink-26utm-5Fcampaign-3Dsig-2Demail-26utm-5Fcontent-3Demailclient&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AsrE-c7ZR7B2Kyr3qgfvvppkCEBVsNmwEMndcrRSuOI&m=cI5OQvjofAcGkyb8nkEn2lygu_l97ybIqlbf4H_7UiRRls5XV_iRbpIzdAxCSamd&s=C_6ofJ9k9o6KWlQHPu26hdtMqXPYACoe755TA1Tjqik&e=>
<#m_-3051569756275543422_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
OpenSCAD mailing list
To unsubscribe send an email todiscuss-leave@lists.openscad.org
--
This email has been checked for viruses by AVG antivirus software.
www.avg.com
It doesn't hang or crash for me, but with the h set to 10, iirc, the
pointed end is separated from the other part. Also, a bit of z fighting
for both h's.
On 14/09/2025 16:54, Jon Bondy via Discuss wrote:
Is there any easy way to dump all of my settings into a text file?
On 9/14/2025 11:31 AM, Adrian Mariano via Discuss wrote:
Seems weird that it's happening to you under Windows but not Jordan.
It seems unlikely it depends on Windows version given that it's also
happening under Linux, which leaves differences in the settings as
the explanation for why Jordan doesn't reproduce it.
On Sun, Sep 14, 2025 at 11:01 AM Jon Bondy jon@jonbondy.com wrote:
Parsing design (AST generation)...
Saved backup file:
C:/Users/jon/OneDrive/Documents/OpenSCAD/backups/unsaved-backup-firnpdqo.scad
Compiling design (CSG Tree generation)...
Compiling design (CSG Products generation)...
Geometries in cache: 3
Geometry cache size in bytes: 3336
CGAL Polyhedrons in cache: 0
CGAL cache size in bytes: 0
Compiling design (CSG Products normalization)...
WARNING: Normalized tree is growing past 200000 elements.
Aborting normalization.
Execution aborted
On 9/14/2025 10:38 AM, Adrian Mariano via Discuss wrote:
Jon, do you get the error and hang in 2021.01 under Windows?
I deleted lines from Jon's example to obtain this minimal case:
$fn=8;
difference() {
union() {
cylinder(h = 1, d = 2);
cube(2);
}
for (z = [1:446]) cylinder(h = 11, d = 3);
}
If you change the number of overlapping cylinders to 445 there
is no error. Original code had the cylinders translated so they
didn't all overlap. You can add back a small translation based
on z and the error still occurs.
<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.avg.com_email-2Dsignature-3Futm-5Fmedium-3Demail-26utm-5Fsource-3Dlink-26utm-5Fcampaign-3Dsig-2Demail-26utm-5Fcontent-3Demailclient&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AsrE-c7ZR7B2Kyr3qgfvvppkCEBVsNmwEMndcrRSuOI&m=cI5OQvjofAcGkyb8nkEn2lygu_l97ybIqlbf4H_7UiRRls5XV_iRbpIzdAxCSamd&s=C_6ofJ9k9o6KWlQHPu26hdtMqXPYACoe755TA1Tjqik&e=>
Virus-free.www.avg.com
<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.avg.com_email-2Dsignature-3Futm-5Fmedium-3Demail-26utm-5Fsource-3Dlink-26utm-5Fcampaign-3Dsig-2Demail-26utm-5Fcontent-3Demailclient&d=DwMFaQ&c=euGZstcaTDllvimEN8b7jXrwqOf-v5A_CdpgnVfiiMM&r=AsrE-c7ZR7B2Kyr3qgfvvppkCEBVsNmwEMndcrRSuOI&m=cI5OQvjofAcGkyb8nkEn2lygu_l97ybIqlbf4H_7UiRRls5XV_iRbpIzdAxCSamd&s=C_6ofJ9k9o6KWlQHPu26hdtMqXPYACoe755TA1Tjqik&e=>
<#m_-3051569756275543422_DAB4FAD8-2DD7-40BB-A1B8-4E2AA1F9FDF2>
OpenSCAD mailing list
To unsubscribe send an email todiscuss-leave@lists.openscad.org
OpenSCAD mailing list
To unsubscribe send an email todiscuss-leave@lists.openscad.org
On 9/14/2025 4:33 AM, jon jonbondy.com wrote:
Interesting that others have replicated this while you cannot. I
installed the latest build before posting to the list.
Ah... it looks like I set the limit to 200K at some point in the dim
past, while the default is 100K. The true limit is (for some unknown
reason) twice the parameter setting, so I saw the 200K in your error
message and it matched the 200K in my settings, so I thought we had the
same settings.
Indeed, for the h=1 case it fails if the parameter is set less than
about 65K, and for the h=10 case it fails if the parameter is less than
about 130K.
Sorry for the bad evidence.
On 9/14/2025 2:44 PM, Jordan Brown via Discuss wrote:
On 9/14/2025 4:33 AM, jon jonbondy.com wrote:
Interesting that others have replicated this while you cannot. I
installed the latest build before posting to the list.
Ah... it looks like I set the limit to 200K at some point in the dim
past, while the default is 100K. The true limit is (for some unknown
reason) twice the parameter setting, so I saw the 200K in your error
message and it matched the 200K in my settings, so I thought we had
the same settings.
Indeed, for the h=1 case it fails if the parameter is set less than
about 65K, and for the h=10 case it fails if the parameter is less
than about 130K.
Sorry for the bad evidence.
But I still don't see a hang, just the unexplained variation in the tree
size (507 vs 1008) and the unexplained failure (growing past 200K when
it only ever reaches ~1K).