Hmm, CGAL error only after F5/F6 sequence.
This is a strange one. I'm trying to track down a CGAL error in 2016.10.05
(&.04) which doesn't happen in 2015.03-2.
However it appeared to be intermittent.
As you know GCAL error do not repeat as they are cached. So using
Design/Flush_Caches is needed to reproduce.
I happens after a clean start, then F5 then F6.
It only happens with a F6 AFTER an F5!
Noting that there is use of render() which does CGAL calls on F5, comment
out this and there is no error.
----------------------code-------------------
echo(version());
$fa=0.05;
$fs=1;
s=0.1;
s2=s*2;
!recessOrig(i=1);
rOD=40.25;
rID=36.3;
rThick=2.3;
rDeep=38;
rRoundOffset=6;
module recessOrig(i=0,logo=true) { // i inverted
difference() {
union() {
difference() {
cylinder(r=rID/2+rThick,h=rDeep); // main tube
// -
translate([0,0,-s])
cylinder(r=rID/2,h=rDeep+rRoundOffset+s2); // hollow out tube
} // d
} // u
// -
translate([rThick-0.5,0,0 ])
render(convexity=30) // <<== comment out
intersection() {
translate([rID/2-2.5,(i ? -1 : +1)*1.3,(i ? 17.5 : 17.5)])
rotate([90,(i ? 0 : 90),90])
scale(0.9)
{
linear_extrude(height=15, convexity=30)
text("MH",font="Rockwell",spacing=0.7,halign="center");
}
difference(){
cylinder(r=rID/2+rThick,h=rDeep);
translate([0,0,-s])
cylinder(r=rID/2,h=rDeep+rRoundOffset+s2); // hollow out tube
}
}
} // d
} // recess
Admin - PM me if you need anything, or if I've done something stupid...
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
Also doesn't happen in 2016.04.06
--
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954p18957.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
Doesn’t happen to me with the latest OS X build.
OpenSCAD version 2016.10.16 (git 5cb40a2)
Jean-Paul
N1JPL
On Nov 6, 2016, at 12:07 AM, MichaelAtOz oz.at.michael@gmail.com wrote:
This is a strange one. I'm trying to track down a CGAL error in 2016.10.05
(&.04) which doesn't happen in 2015.03-2.
However it appeared to be intermittent.
As you know GCAL error do not repeat as they are cached. So using
Design/Flush_Caches is needed to reproduce.
I happens after a clean start, then F5 then F6.
It only happens with a F6 AFTER an F5!
Noting that there is use of render() which does CGAL calls on F5, comment
out this and there is no error.
----------------------code-------------------
echo(version());
$fa=0.05;
$fs=1;
s=0.1;
s2=s*2;
!recessOrig(i=1);
rOD=40.25;
rID=36.3;
rThick=2.3;
rDeep=38;
rRoundOffset=6;
module recessOrig(i=0,logo=true) { // i inverted
difference() {
union() {
difference() {
cylinder(r=rID/2+rThick,h=rDeep); // main tube
// -
translate([0,0,-s])
cylinder(r=rID/2,h=rDeep+rRoundOffset+s2); // hollow out tube
} // d
} // u
// -
translate([rThick-0.5,0,0 ])
render(convexity=30) // <<== comment out
intersection() {
translate([rID/2-2.5,(i ? -1 : +1)*1.3,(i ? 17.5 : 17.5)])
rotate([90,(i ? 0 : 90),90])
scale(0.9)
{
linear_extrude(height=15, convexity=30)
text("MH",font="Rockwell",spacing=0.7,halign="center");
}
difference(){
cylinder(r=rID/2+rThick,h=rDeep);
translate([0,0,-s])
cylinder(r=rID/2,h=rDeep+rRoundOffset+s2); // hollow out tube
}
}
} // d
} // recess
Admin - PM me if you need anything, or if I've done something stupid...
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
OpenSCAD mailing list
Discuss@lists.openscad.org
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org
On Nov 6, 2016, at 00:07, MichaelAtOz oz.at.michael@gmail.com wrote:
This is a strange one. I'm trying to track down a CGAL error in 2016.10.05
(&.04) which doesn't happen in 2015.03-2.
Could you describe the error, and how to reproduce it, in more detail?
I’ve tested this on various version (on Mac only so far) without managing to reproduce it..
-Marius
At first I had trouble tracking down what I changed, in my .scad, to trigger
it.
I got to something I changed around the text("MH"... bit, I had moved it
around.
But reverting and adjusting wasn't consistent in generating the error.
Which is when I worked out it only happens for a F6 AFTER a F5.
I can reproduce, just by loading the code above, doing F5, then F6. F6 alone
does not error.
(load a fresh instance of the .exe, I had numerous occasions where, after
doing a few, design/flush_caches no-longer caused the message to occur, you
could tell by the F6 being instant, v's taking a bit of time to think before
error-ing)
Admin - PM me if you need anything, or if I've done something stupid...
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954p18973.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
Sorry, misunderstood, error is:
Compiling design (CSG Tree generation)...
ECHO: [2016, 10, 4]
Rendering Polygon Mesh using CGAL...
ERROR: CGAL error in CGAL_Nef_polyhedron3(): CGAL ERROR: assertion
violation! Expr: pe_prev->is_border() ||
!internal::Plane_constructor::get_plane(pe_prev->facet(),pe_prev->facet()->plane()).is_degenerate()
File:
/opt/mxe/usr/x86_64-w64-mingw32.static/include/CGAL/Nef_3/polyhedron_3_to_nef_3.h
Line: 251
Geometries in cache: 10
Geometry cache size in bytes: 151720
CGAL Polyhedrons in cache: 5
CGAL cache size in bytes: 2820112
Total rendering time: 0 hours, 0 minutes, 0 seconds
Top level object is a 3D object:
Simple: yes
Vertices: 488
Halfedges: 1464
Edges: 732
Halffacets: 492
Facets: 246
Volumes: 2
Rendering finished.
Reproduced as mentioned above.
Admin - PM me if you need anything, or if I've done something stupid...
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954p18974.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
I suppose what I'm reporting here is the render() seems to be the cause of
the F5 first F6 error bit.
As just F6 does not error, and the 3d object is correct, I'm assuming the
exact CGAL error is spurious. ??
Admin - PM me if you need anything, or if I've done something stupid...
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954p18976.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
On Nov 6, 2016, at 18:56, MichaelAtOz oz.at.michael@gmail.com wrote:
I suppose what I'm reporting here is the render() seems to be the cause of
the F5 first F6 error bit.
This sort of stuff can happen:
o When doing F5 and we encounter geometry without any CSG ops (i.e. just geometry + transformations), we calculate and cache the transformed object by applying the transformation ourselves (i.e. without CGAL).
o When we subsequently do an F6, this cached object is used.
o However, if doing F6 from the start, CGAL is used all the way.
This is not optimal and we should probably manage that sort of caching in a better way.
I belive we have an issue for this somewhere but I cannot find it atm…
I’ll try to reproduce the problem under Windows..
-Marius
When I try the code, I get after F5:
Saved backup file: /Users/jean-paullouis/Documents/OpenSCAD/backups/unsaved-backup-xJL31134.scad
Compiling design (CSG Tree generation)...
ECHO: [2016, 10, 16]
Compiling design (CSG Products generation)...
Geometries in cache: 9
Geometry cache size in bytes: 117552
CGAL Polyhedrons in cache: 2
CGAL cache size in bytes: 848032
Compiling design (CSG Products normalization)...
Normalized CSG tree has 3 elements
Compile and preview finished.
Total rendering time: 0 hours, 0 minutes, 1 seconds
After F6, I get:
Saved backup file: /Users/jean-paullouis/Documents/OpenSCAD/backups/unsaved-backup-xJL31134.scad
Compiling design (CSG Tree generation)...
ECHO: [2016, 10, 16]
Rendering Polygon Mesh using CGAL...
Geometries in cache: 10
Geometry cache size in bytes: 137864
CGAL Polyhedrons in cache: 5
CGAL cache size in bytes: 3361744
Total rendering time: 0 hours, 0 minutes, 0 seconds
Top level object is a 3D object:
Simple: yes
Vertices: 660
Halfedges: 2080
Edges: 1040
Halffacets: 764
Facets: 382
Volumes: 2
Rendering finished.
No problem at all. Must be platform dependent.
Jean-Paul
N1JPL
On Nov 6, 2016, at 12:07 AM, MichaelAtOz oz.at.michael@gmail.com wrote:
This is a strange one. I'm trying to track down a CGAL error in 2016.10.05
(&.04) which doesn't happen in 2015.03-2.
However it appeared to be intermittent.
As you know GCAL error do not repeat as they are cached. So using
Design/Flush_Caches is needed to reproduce.
I happens after a clean start, then F5 then F6.
It only happens with a F6 AFTER an F5!
Noting that there is use of render() which does CGAL calls on F5, comment
out this and there is no error.
----------------------code-------------------
echo(version());
$fa=0.05;
$fs=1;
s=0.1;
s2=s*2;
!recessOrig(i=1);
rOD=40.25;
rID=36.3;
rThick=2.3;
rDeep=38;
rRoundOffset=6;
module recessOrig(i=0,logo=true) { // i inverted
difference() {
union() {
difference() {
cylinder(r=rID/2+rThick,h=rDeep); // main tube
// -
translate([0,0,-s])
cylinder(r=rID/2,h=rDeep+rRoundOffset+s2); // hollow out tube
} // d
} // u
// -
translate([rThick-0.5,0,0 ])
render(convexity=30) // <<== comment out
intersection() {
translate([rID/2-2.5,(i ? -1 : +1)*1.3,(i ? 17.5 : 17.5)])
rotate([90,(i ? 0 : 90),90])
scale(0.9)
{
linear_extrude(height=15, convexity=30)
text("MH",font="Rockwell",spacing=0.7,halign="center");
}
difference(){
cylinder(r=rID/2+rThick,h=rDeep);
translate([0,0,-s])
cylinder(r=rID/2,h=rDeep+rRoundOffset+s2); // hollow out tube
}
}
} // d
} // recess
Admin - PM me if you need anything, or if I've done something stupid...
Unless specifically shown otherwise above, my contribution is in the Public Domain; to the extent possible under law, I have waived all copyright and related or neighbouring rights to this work. Obviously inclusion of works of previous authors is not included in the above.
The TPP is no simple “trade agreement.” Fight it! http://www.ourfairdeal.org/ time is running out!
View this message in context: http://forum.openscad.org/Hmm-CGAL-error-only-after-F5-F6-sequence-tp18954.html
Sent from the OpenSCAD mailing list archive at Nabble.com.
OpenSCAD mailing list
Discuss@lists.openscad.org
http://lists.openscad.org/mailman/listinfo/discuss_lists.openscad.org
I’ve tried to reproduce this without success on 32-bit Windows (I don’t have a 64-bit Windows lying around..)
I need some help - could someone:
- Verify that this issue only happens using the 64-bit Windows binaries?
- Post a Help->Library Info dump from a failing version of OpenSCAD?
-Marius