more useful info regarding non-manifold objetcs

classic Classic list List threaded Threaded
19 messages Options
Reply | Threaded
Open this post in threaded view
|

more useful info regarding non-manifold objetcs

Felipe Sanches
I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Whosawhatsis
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
ds
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

ds
In reply to this post by Felipe Sanches
What I've been doing is judiciously commenting out code, then putting things back until it upchucks.  Once you get systematic it starts to get better, but I fully understand your pain.




On 04/23/13 18:13, Felipe Sanches wrote:
I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

donbright
In reply to this post by Whosawhatsis
"I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries"

No, it is entirely possible to go directly from a non-manifold CGAL Nef Polyhedron to an STL file. CGAL's Nef Polyhedron is just a bunch of flat polygons (with holes sometimes) in 3d space. You can do a triangle-tessellation of these polygons and dump them to STL triangles directly if you want. In fact, this is the process that is done to draw the Nef Polyhedra into OpenGL, because as you may know, OpenGL basically wants triangles as input.

However, when it dumps STL, OpenSCAD uses a different method than for drawing OpenGL. It uses a conversion from CGAL Nef Polyhedron to 'ordinary' CGAL Polyhedron, and that is where the errors get thrown. Ordinary CGAL Polyhedron's can only be 2-manifold, so the conversion fails sometimes.

I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .


-DB


On Tue, Apr 23, 2013 at 9:45 PM, whosawhatsis <[hidden email]> wrote:
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Taylor Alexander
So... would it be possible to have an alternate export option that exports the messy one? NetFabb is really good at cleaning up STL files, but only if you can get a file out of your system!


On Tue, Apr 23, 2013 at 8:11 PM, Don Bright <[hidden email]> wrote:
"I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries"

No, it is entirely possible to go directly from a non-manifold CGAL Nef Polyhedron to an STL file. CGAL's Nef Polyhedron is just a bunch of flat polygons (with holes sometimes) in 3d space. You can do a triangle-tessellation of these polygons and dump them to STL triangles directly if you want. In fact, this is the process that is done to draw the Nef Polyhedra into OpenGL, because as you may know, OpenGL basically wants triangles as input.

However, when it dumps STL, OpenSCAD uses a different method than for drawing OpenGL. It uses a conversion from CGAL Nef Polyhedron to 'ordinary' CGAL Polyhedron, and that is where the errors get thrown. Ordinary CGAL Polyhedron's can only be 2-manifold, so the conversion fails sometimes.

I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .


-DB


On Tue, Apr 23, 2013 at 9:45 PM, whosawhatsis <[hidden email]> wrote:
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Whosawhatsis
In reply to this post by donbright
That was my suspicion. So it would be nice if there was a way to override the error and export an STL from the nef polyhedron and repair it externally. There should definitely be a message warning that the geometry is non-manifold and asking if you want to continue with the export anyway.

On Tuesday, April 23, 2013 at 8:11 PM, Don Bright wrote:

"I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries"

No, it is entirely possible to go directly from a non-manifold CGAL Nef Polyhedron to an STL file. CGAL's Nef Polyhedron is just a bunch of flat polygons (with holes sometimes) in 3d space. You can do a triangle-tessellation of these polygons and dump them to STL triangles directly if you want. In fact, this is the process that is done to draw the Nef Polyhedra into OpenGL, because as you may know, OpenGL basically wants triangles as input.

However, when it dumps STL, OpenSCAD uses a different method than for drawing OpenGL. It uses a conversion from CGAL Nef Polyhedron to 'ordinary' CGAL Polyhedron, and that is where the errors get thrown. Ordinary CGAL Polyhedron's can only be 2-manifold, so the conversion fails sometimes.

I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .


-DB


On Tue, Apr 23, 2013 at 9:45 PM, whosawhatsis <[hidden email]> wrote:
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Taylor Alexander
Yeah, that sounds like a great way to do it. Something like:

"Export STL Error! Would you like to export the raw file as STL? Repair of the file using other tools may be possible, but the file may cause errors in some programs if it is not properly repaired."

It may help in tracking down the problem too.


On Tue, Apr 23, 2013 at 8:15 PM, whosawhatsis <[hidden email]> wrote:
That was my suspicion. So it would be nice if there was a way to override the error and export an STL from the nef polyhedron and repair it externally. There should definitely be a message warning that the geometry is non-manifold and asking if you want to continue with the export anyway.

On Tuesday, April 23, 2013 at 8:11 PM, Don Bright wrote:

"I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries"

No, it is entirely possible to go directly from a non-manifold CGAL Nef Polyhedron to an STL file. CGAL's Nef Polyhedron is just a bunch of flat polygons (with holes sometimes) in 3d space. You can do a triangle-tessellation of these polygons and dump them to STL triangles directly if you want. In fact, this is the process that is done to draw the Nef Polyhedra into OpenGL, because as you may know, OpenGL basically wants triangles as input.

However, when it dumps STL, OpenSCAD uses a different method than for drawing OpenGL. It uses a conversion from CGAL Nef Polyhedron to 'ordinary' CGAL Polyhedron, and that is where the errors get thrown. Ordinary CGAL Polyhedron's can only be 2-manifold, so the conversion fails sometimes.

I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .


-DB


On Tue, Apr 23, 2013 at 9:45 PM, whosawhatsis <[hidden email]> wrote:
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

kintel
Administrator
In reply to this post by donbright
On 2013-04-23, at 23:11 , Don Bright wrote:
>
> I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .
>
I don't know why it was done this way originally, but it is a good error check.
Exporting the raw STL and rather throw a warning if it's not manifold sounds like a good idea.

 -Marius

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Felipe Sanches
In reply to this post by Taylor Alexander
That would be awesome!

On Wed, Apr 24, 2013 at 12:22 AM, Taylor Alexander <[hidden email]> wrote:
Yeah, that sounds like a great way to do it. Something like:

"Export STL Error! Would you like to export the raw file as STL? Repair of the file using other tools may be possible, but the file may cause errors in some programs if it is not properly repaired."

It may help in tracking down the problem too.


On Tue, Apr 23, 2013 at 8:15 PM, whosawhatsis <[hidden email]> wrote:
That was my suspicion. So it would be nice if there was a way to override the error and export an STL from the nef polyhedron and repair it externally. There should definitely be a message warning that the geometry is non-manifold and asking if you want to continue with the export anyway.

On Tuesday, April 23, 2013 at 8:11 PM, Don Bright wrote:

"I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries"

No, it is entirely possible to go directly from a non-manifold CGAL Nef Polyhedron to an STL file. CGAL's Nef Polyhedron is just a bunch of flat polygons (with holes sometimes) in 3d space. You can do a triangle-tessellation of these polygons and dump them to STL triangles directly if you want. In fact, this is the process that is done to draw the Nef Polyhedra into OpenGL, because as you may know, OpenGL basically wants triangles as input.

However, when it dumps STL, OpenSCAD uses a different method than for drawing OpenGL. It uses a conversion from CGAL Nef Polyhedron to 'ordinary' CGAL Polyhedron, and that is where the errors get thrown. Ordinary CGAL Polyhedron's can only be 2-manifold, so the conversion fails sometimes.

I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .


-DB


On Tue, Apr 23, 2013 at 9:45 PM, whosawhatsis <[hidden email]> wrote:
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Peter Falke
Now, technicaly, doesnt Openscad know where the part is non-manifold?
I think it always is, when three faces border on edge.
Could OpenSCAD output the coordinates of this edge in an error message?
Or even draw the edge in a different color (all otherthings appearing transparent)

TakeItAndRun



On 24 April 2013 05:31, Felipe Sanches <[hidden email]> wrote:
That would be awesome!


On Wed, Apr 24, 2013 at 12:22 AM, Taylor Alexander <[hidden email]> wrote:
Yeah, that sounds like a great way to do it. Something like:

"Export STL Error! Would you like to export the raw file as STL? Repair of the file using other tools may be possible, but the file may cause errors in some programs if it is not properly repaired."

It may help in tracking down the problem too.


On Tue, Apr 23, 2013 at 8:15 PM, whosawhatsis <[hidden email]> wrote:
That was my suspicion. So it would be nice if there was a way to override the error and export an STL from the nef polyhedron and repair it externally. There should definitely be a message warning that the geometry is non-manifold and asking if you want to continue with the export anyway.

On Tuesday, April 23, 2013 at 8:11 PM, Don Bright wrote:

"I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries"

No, it is entirely possible to go directly from a non-manifold CGAL Nef Polyhedron to an STL file. CGAL's Nef Polyhedron is just a bunch of flat polygons (with holes sometimes) in 3d space. You can do a triangle-tessellation of these polygons and dump them to STL triangles directly if you want. In fact, this is the process that is done to draw the Nef Polyhedra into OpenGL, because as you may know, OpenGL basically wants triangles as input.

However, when it dumps STL, OpenSCAD uses a different method than for drawing OpenGL. It uses a conversion from CGAL Nef Polyhedron to 'ordinary' CGAL Polyhedron, and that is where the errors get thrown. Ordinary CGAL Polyhedron's can only be 2-manifold, so the conversion fails sometimes.

I always assumed that the purpose of this was as an "error check" to prevent problems with slicers. . . .


-DB


On Tue, Apr 23, 2013 at 9:45 PM, whosawhatsis <[hidden email]> wrote:
I'm wondering if the underlying libraries make it impossible to generate an STL from these non-manifold geometries. If not, it might be a good idea to make that a warning rather than an error, and let the user use netfabb or similar to repair the non-manifold STL.

On Tuesday, April 23, 2013 at 6:13 PM, Felipe Sanches wrote:

I'm trying to export a complex design but I get the famous message saying my object is not a valid 2-manifold.

It would be great if the message could be a bit more verbose giving me a hint regarding which subtrees are the offenders...


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



--
[hidden email]

P.S. Falls meine E-Mail kürzer ausfällt als Dir angenehm ist:
Ich probiere gerade aus kurze Antworten statt gar keine Antworten zu schreiben.
Wenn Du gerne mehr lesen möchtest, dann lass es mich bitte wissen.

P.S. In case my e-mail is shorter than you enjoy:
I am currently trying short replies instead of no replies at all.
Please let me know, if you like to read more.

Enjoy!

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

kintel
Administrator
On 2013-04-24, at 01:00 , Peter Falke wrote:

> Now, technicaly, doesnt Openscad know where the part is non-manifold?

We don't explicitly know it - we get the error message from somewhere within the CGAL library.
It might be possible to parse the error and locate the area of error, but nobody has tried to dig into that - and I fear it's a bit messy.
I could be wrong though.

 -Marius

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Len Trigg

The ability to continue would be fantastic, just yesterday I was thwarted trying to use openscad to import an existing STL and save a mirror image version due to this.

On Apr 24, 2013 5:20 PM, "Marius Kintel" <[hidden email]> wrote:
On 2013-04-24, at 01:00 , Peter Falke wrote:

> Now, technicaly, doesnt Openscad know where the part is non-manifold?

We don't explicitly know it - we get the error message from somewhere within the CGAL library.
It might be possible to parse the error and locate the area of error, but nobody has tried to dig into that - and I fear it's a bit messy.
I could be wrong though.

 -Marius

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

nophead
How come people manage to make non-manifold STLs with OpenScad if CGAL does not allow it?


On 24 April 2013 07:01, Len Trigg <[hidden email]> wrote:

The ability to continue would be fantastic, just yesterday I was thwarted trying to use openscad to import an existing STL and save a mirror image version due to this.

On Apr 24, 2013 5:20 PM, "Marius Kintel" <[hidden email]> wrote:
On 2013-04-24, at 01:00 , Peter Falke wrote:

> Now, technicaly, doesnt Openscad know where the part is non-manifold?

We don't explicitly know it - we get the error message from somewhere within the CGAL library.
It might be possible to parse the error and locate the area of error, but nobody has tried to dig into that - and I fear it's a bit messy.
I could be wrong though.

 -Marius

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Peter Falke
you get an error message only after U F6'ed and export the stl.

Now, if you change and print U'll catch itat once.

TakeItAndRun


On 24 April 2013 09:07, nop head <[hidden email]> wrote:
How come people manage to make non-manifold STLs with OpenScad if CGAL does not allow it?


On 24 April 2013 07:01, Len Trigg <[hidden email]> wrote:

The ability to continue would be fantastic, just yesterday I was thwarted trying to use openscad to import an existing STL and save a mirror image version due to this.

On Apr 24, 2013 5:20 PM, "Marius Kintel" <[hidden email]> wrote:
On 2013-04-24, at 01:00 , Peter Falke wrote:

> Now, technicaly, doesnt Openscad know where the part is non-manifold?

We don't explicitly know it - we get the error message from somewhere within the CGAL library.
It might be possible to parse the error and locate the area of error, but nobody has tried to dig into that - and I fear it's a bit messy.
I could be wrong though.

 -Marius

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



--
[hidden email]

P.S. Falls meine E-Mail kürzer ausfällt als Dir angenehm ist:
Ich probiere gerade aus kurze Antworten statt gar keine Antworten zu schreiben.
Wenn Du gerne mehr lesen möchtest, dann lass es mich bitte wissen.

P.S. In case my e-mail is shorter than you enjoy:
I am currently trying short replies instead of no replies at all.
Please let me know, if you like to read more.

Enjoy!

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

kintel
Administrator
In reply to this post by nophead
On 2013-04-24, at 03:07 , nop head wrote:

> How come people manage to make non-manifold STLs with OpenScad if CGAL does not allow it?
>
People manage that? Perhaps we should start collecting examples of that..

The only thing I can think of is that floating point numbers get collapsed, causing self-intersections on export, as the internal representation has higher resolution but the STL output will be forced to double precision.

 -Marius
_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Triffid Hunter
On 24 April 2013 23:01, Marius Kintel <[hidden email]> wrote:
On 2013-04-24, at 03:07 , nop head wrote:
> How come people manage to make non-manifold STLs with OpenScad if CGAL does not allow it?
>
People manage that? Perhaps we should start collecting examples of that..

yes absolutely, the commonly available involute gear script is notorious for producing non-manifold STLs. I've had countless occasions where openscad is perfectly happy to render and export, but then it turns out that the STL is a complete mess.

fwiw, I'd say that it's quite common for openscad to export bad STLs unless the designer is very careful about how they construct things

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

nophead
I have had a case where Netfabb says there is a degenerate face I wrote a script to check the three points of each triangle were unique and they all were. Skeinforge slices it fine, but Slic3r fill in the hollow interior unless Netfabb "fixes" it.




On 24 April 2013 14:13, Triffid Hunter <[hidden email]> wrote:
On 24 April 2013 23:01, Marius Kintel <[hidden email]> wrote:
On 2013-04-24, at 03:07 , nop head wrote:
> How come people manage to make non-manifold STLs with OpenScad if CGAL does not allow it?
>
People manage that? Perhaps we should start collecting examples of that..

yes absolutely, the commonly available involute gear script is notorious for producing non-manifold STLs. I've had countless occasions where openscad is perfectly happy to render and export, but then it turns out that the STL is a complete mess.

fwiw, I'd say that it's quite common for openscad to export bad STLs unless the designer is very careful about how they construct things

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

Kliment Yanev
On 04/24/2013 03:34 PM, nop head wrote:
I have had a case where Netfabb says there is a degenerate face I wrote a script to check the three points of each triangle were unique and they all were. Skeinforge slices it fine, but Slic3r fill in the hollow interior unless Netfabb "fixes" it.

They might be unique but collinear.



On 24 April 2013 14:13, Triffid Hunter <[hidden email]> wrote:
On 24 April 2013 23:01, Marius Kintel <[hidden email]> wrote:
On 2013-04-24, at 03:07 , nop head wrote:
> How come people manage to make non-manifold STLs with OpenScad if CGAL does not allow it?
>
People manage that? Perhaps we should start collecting examples of that..

yes absolutely, the commonly available involute gear script is notorious for producing non-manifold STLs. I've had countless occasions where openscad is perfectly happy to render and export, but then it turns out that the STL is a complete mess.

fwiw, I'd say that it's quite common for openscad to export bad STLs unless the designer is very careful about how they construct things

_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566



_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566
Reply | Threaded
Open this post in threaded view
|

Re: more useful info regarding non-manifold objetcs

kintel
Administrator
In reply to this post by Triffid Hunter
On 2013-04-24, at 09:13 , Triffid Hunter wrote:

> yes absolutely, the commonly available involute gear script is notorious for producing non-manifold STLs. I've had countless occasions where openscad is perfectly happy to render and export, but then it turns out that the STL is a complete mess.
>
I guess there have been examples of this in the past, but I cannot find any atm.
If anyone has got concrete example, I'll start collecting them.

It could be that the manifold-check doesn't check for self-intersections, only for edge/point intersections.

 -Marius


_______________________________________________
OpenSCAD mailing list
[hidden email]
http://rocklinux.net/mailman/listinfo/openscad
http://openscad.org - https://flattr.com/thing/121566