I’ve got objects in OpenSCAD that I think are ok. They don’t produce errors or warnings and I can do boolean operations with them that work without giving a not manifold error. I can export them as .stls and they slice and print ok in OrcaSlicer. In PrusaSlicer they produce warning messages about numerous errors auto corrected but without description of the nature of the errors and then also print ok. I’m guessing this is just PrusaSlicer being overly fussy, perhaps about vertices rather close to each other. I want to distribute some models and I’d prefer they not generate warnings for anyone if it is straightforward to fix. I’m seeking suggestions for how I might investigate and hopefully address this issue. I’m interested in steps I might take in OpenSCAD and possibly in other tools.

Try 3MF export, STL may break topology. On 3/21/25 23:32, Todd Allen via Discuss wrote: > I’ve got objects in OpenSCAD that I think are ok. They don’t produce > errors or warnings and I can do boolean operations with them that work > without giving a not manifold error.  I can export them as .stls and > they slice and print ok in OrcaSlicer.  In PrusaSlicer they produce > warning messages about numerous errors auto corrected but without > description of the nature of the errors and then also print ok. > > I’m guessing this is just PrusaSlicer being overly fussy, perhaps > about vertices rather close to each other.  I want to distribute some > models and I’d prefer they not generate warnings for anyone if it is > straightforward to fix. > > I’m seeking suggestions for how I might investigate and hopefully > address this issue.  I’m interested in steps I might take in OpenSCAD > and possibly in other tools. > > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org

Ok I'll try it, but can you elaborate on what changes? I imagined the geometry contained in a .3mf export was identical to that in an .stl. Although I am doing binary .stls, could this lead to a warning that an ascii format would avoid? On Fri, Mar 21, 2025 at 10:39 AM Chun Kit LAM via Discuss < discuss@lists.openscad.org> wrote: > Try 3MF export, STL may break topology. > > On 3/21/25 23:32, Todd Allen via Discuss wrote: > > I’ve got objects in OpenSCAD that I think are ok. They don’t produce > > errors or warnings and I can do boolean operations with them that work > > without giving a not manifold error. I can export them as .stls and > > they slice and print ok in OrcaSlicer. In PrusaSlicer they produce > > warning messages about numerous errors auto corrected but without > > description of the nature of the errors and then also print ok. > > > > I’m guessing this is just PrusaSlicer being overly fussy, perhaps > > about vertices rather close to each other. I want to distribute some > > models and I’d prefer they not generate warnings for anyone if it is > > straightforward to fix. > > > > I’m seeking suggestions for how I might investigate and hopefully > > address this issue. I’m interested in steps I might take in OpenSCAD > > and possibly in other tools. > > > > _______________________________________________ > > 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 >

 Try 3MF export, STL may break topology.

 On 3/21/25 23:32, Todd Allen via Discuss wrote:

 produce

 that work

 and

 some

 it is

 OpenSCAD

 _______________________________________________
 OpenSCAD mailing list
 To unsubscribe send an email to discuss-leave@lists.openscad.org

STL stores triangles and vertex coordinates only, which loses data about how vertices are connected or duplicated. Some meshes may be converted to manifold mesh by duplicating vertices and connecting them in a specific way, and this information is stored with 3MF. For STL, the software has to guess when to duplicate vertices. You can check this for the definition of manifoldness: https://github.com/elalish/manifold/wiki/Manifold-Library#manifoldness On 3/21/25 23:50, Todd Allen via Discuss wrote: > Ok I'll try it, but can you elaborate on what changes?  I imagined the > geometry contained in a .3mf export was identical to that in an .stl.  > Although I am doing binary .stls, could this lead to a warning that an > ascii format would avoid? > > On Fri, Mar 21, 2025 at 10:39 AM Chun Kit LAM via Discuss > <discuss@lists.openscad.org> wrote: > > Try 3MF export, STL may break topology. > > On 3/21/25 23:32, Todd Allen via Discuss wrote: > > I’ve got objects in OpenSCAD that I think are ok. They don’t > produce > > errors or warnings and I can do boolean operations with them > that work > > without giving a not manifold error.  I can export them as .stls > and > > they slice and print ok in OrcaSlicer.  In PrusaSlicer they produce > > warning messages about numerous errors auto corrected but without > > description of the nature of the errors and then also print ok. > > > > I’m guessing this is just PrusaSlicer being overly fussy, perhaps > > about vertices rather close to each other.  I want to distribute > some > > models and I’d prefer they not generate warnings for anyone if > it is > > straightforward to fix. > > > > I’m seeking suggestions for how I might investigate and hopefully > > address this issue.  I’m interested in steps I might take in > OpenSCAD > > and possibly in other tools. > > > > _______________________________________________ > > 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 todiscuss-leave@lists.openscad.org

The .3mf export fixed it! At least the first test case I've tried. I did not expect this. Thanks! Also curiously when I did the .3mf export a dialog box popped up which I havee never seen before. I don't know if it is a fairly recent addition since the last time I've done a .3mf export or if it had merely been hidden by a config option. On Fri, Mar 21, 2025 at 10:39 AM Chun Kit LAM via Discuss < discuss@lists.openscad.org> wrote: > Try 3MF export, STL may break topology. > > On 3/21/25 23:32, Todd Allen via Discuss wrote: > > I’ve got objects in OpenSCAD that I think are ok. They don’t produce > > errors or warnings and I can do boolean operations with them that work > > without giving a not manifold error. I can export them as .stls and > > they slice and print ok in OrcaSlicer. In PrusaSlicer they produce > > warning messages about numerous errors auto corrected but without > > description of the nature of the errors and then also print ok. > > > > I’m guessing this is just PrusaSlicer being overly fussy, perhaps > > about vertices rather close to each other. I want to distribute some > > models and I’d prefer they not generate warnings for anyone if it is > > straightforward to fix. > > > > I’m seeking suggestions for how I might investigate and hopefully > > address this issue. I’m interested in steps I might take in OpenSCAD > > and possibly in other tools. > > > > _______________________________________________ > > 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 >

You can also try right-clicking on the object in PrusaSlicer and Fix it, then re-export it from PrusaSlicer. On Fri, Mar 21, 2025 at 10:55 AM Chun Kit LAM via Discuss < discuss@lists.openscad.org> wrote: > STL stores triangles and vertex coordinates only, which loses data about > how vertices are connected or duplicated. Some meshes may be converted to > manifold mesh by duplicating vertices and connecting them in a specific > way, and this information is stored with 3MF. For STL, the software has to > guess when to duplicate vertices. > > You can check this for the definition of manifoldness: > https://github.com/elalish/manifold/wiki/Manifold-Library#manifoldness > On 3/21/25 23:50, Todd Allen via Discuss wrote: > > Ok I'll try it, but can you elaborate on what changes? I imagined the > geometry contained in a .3mf export was identical to that in an .stl. > Although I am doing binary .stls, could this lead to a warning that an > ascii format would avoid? > > On Fri, Mar 21, 2025 at 10:39 AM Chun Kit LAM via Discuss < > discuss@lists.openscad.org> wrote: > >> Try 3MF export, STL may break topology. >> >> On 3/21/25 23:32, Todd Allen via Discuss wrote: >> > I’ve got objects in OpenSCAD that I think are ok. They don’t produce >> > errors or warnings and I can do boolean operations with them that work >> > without giving a not manifold error. I can export them as .stls and >> > they slice and print ok in OrcaSlicer. In PrusaSlicer they produce >> > warning messages about numerous errors auto corrected but without >> > description of the nature of the errors and then also print ok. >> > >> > I’m guessing this is just PrusaSlicer being overly fussy, perhaps >> > about vertices rather close to each other. I want to distribute some >> > models and I’d prefer they not generate warnings for anyone if it is >> > straightforward to fix. >> > >> > I’m seeking suggestions for how I might investigate and hopefully >> > address this issue. I’m interested in steps I might take in OpenSCAD >> > and possibly in other tools. >> > >> > _______________________________________________ >> > 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 > > _______________________________________________ > OpenSCAD mailing list > To unsubscribe send an email to discuss-leave@lists.openscad.org >

>You can also try right-clicking on the object in PrusaSlicer and Fix it, then re-export it from PrusaSlicer. Thanks!!! My first thought was, "Not interested..." because I may be dealing with many files and I prefer to automate as much of my workflow as possible, which is why I'm using OpenSCAD. But a quick look finds PrusaSlicer has a command line interface and if it supports fixing and re-export then this could work for me. Possibly a bigger win is I might switch to doing some slicing via command line scripting. Hopefully other slicers support this too. I don't know why I never considered this before... On Fri, Mar 21, 2025 at 11:02 AM Leonard Martin Struttmann via Discuss < discuss@lists.openscad.org> wrote: > You can also try right-clicking on the object in PrusaSlicer and Fix it, > then re-export it from PrusaSlicer. > > On Fri, Mar 21, 2025 at 10:55 AM Chun Kit LAM via Discuss < > discuss@lists.openscad.org> wrote: > >> STL stores triangles and vertex coordinates only, which loses data about >> how vertices are connected or duplicated. Some meshes may be converted to >> manifold mesh by duplicating vertices and connecting them in a specific >> way, and this information is stored with 3MF. For STL, the software has to >> guess when to duplicate vertices. >> >> You can check this for the definition of manifoldness: >> https://github.com/elalish/manifold/wiki/Manifold-Library#manifoldness >> On 3/21/25 23:50, Todd Allen via Discuss wrote: >> >> Ok I'll try it, but can you elaborate on what changes? I imagined the >> geometry contained in a .3mf export was identical to that in an .stl. >> Although I am doing binary .stls, could this lead to a warning that an >> ascii format would avoid? >> >> On Fri, Mar 21, 2025 at 10:39 AM Chun Kit LAM via Discuss < >> discuss@lists.openscad.org> wrote: >> >>> Try 3MF export, STL may break topology. >>> >>> On 3/21/25 23:32, Todd Allen via Discuss wrote: >>> > I’ve got objects in OpenSCAD that I think are ok. They don’t produce >>> > errors or warnings and I can do boolean operations with them that work >>> > without giving a not manifold error. I can export them as .stls and >>> > they slice and print ok in OrcaSlicer. In PrusaSlicer they produce >>> > warning messages about numerous errors auto corrected but without >>> > description of the nature of the errors and then also print ok. >>> > >>> > I’m guessing this is just PrusaSlicer being overly fussy, perhaps >>> > about vertices rather close to each other. I want to distribute some >>> > models and I’d prefer they not generate warnings for anyone if it is >>> > straightforward to fix. >>> > >>> > I’m seeking suggestions for how I might investigate and hopefully >>> > address this issue. I’m interested in steps I might take in OpenSCAD >>> > and possibly in other tools. >>> > >>> > _______________________________________________ >>> > 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 >> >> _______________________________________________ >> 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 >

> > Also curiously when I did the .3mf export a dialog box popped up which I havee never seen before. I don't know if it is a fairly recent addition since the last time I've done a .3mf export or if it had merely been hidden by a config option. That was added in the last couple of weeks. You can hide it in the preferences to save you a click on export. -Marius

On 3/21/2025 8:50 AM, Todd Allen via Discuss wrote: > Although I am doing binary .stls, could this lead to a warning that an > ascii format would avoid? I believe that 3MF is ASCII XML, under a frosting of ZIP.

>That was added in the last couple of weeks. You can hide it in the preferences to save you a click on export. Thanks! And it looks like the command line options allow setting all fields of the new export dialog. Which unfortunately for me doesn't allow setting object names. Is there any plan for allowing object names to be set in .3mf files? If not, perhaps it could be added to the export options? On Fri, Mar 21, 2025 at 12:01 PM Marius Kintel <marius@kintel.net> wrote: > > > > Also curiously when I did the .3mf export a dialog box popped up which I > havee never seen before. I don't know if it is a fairly recent addition > since the last time I've done a .3mf export or if it had merely been hidden > by a config option. > > That was added in the last couple of weeks. You can hide it in the > preferences to save you a click on export. > > -Marius > >