"Internal error: vector too long" when right click; select "fill bed with instances" of the provided object with poorly chosen arrange parameters

[u89djt]

Description of the bug

Open as project "maximum pin print.3mf". EDIT: in the arrange settings dialog, set spacing to 20 and spacing from bed to 96 (screenshot in first comment) Right click on the merged group of objects on the plate and select "Fill bed with instances". Popup error dialog appears as follows "PrusaSlicer Error Internal error: vector too long [OK]": Click on OK and the program closes. Objects taken from https://www.thingiverse.com/thing:243278/files by Emmett Lalish Posted here in compliance with https://creativecommons.org/licenses/by-sa/3.0/ No changes were made to the object meshes; they are spatially distributed differently.

Project file & How to reproduce

u89djt240403.zip

Checklist of files included above

• [X] Project file
• [X] Screenshot

Version of PrusaSlicer

PrusaSlicer-2.7.3+win64-202403280959

Operating system

Edition Windows 10 Home Version 21H2 Installed on ‎16/‎06/‎2021 OS build 19044.3086 Experience Windows Feature Experience Pack 1000.19041.1000.0

Printer model

Original Prusa MK3S with no modifications or upgrades

[u89djt]

I had a crazy offset from bed value under the arrange dialog. I stupidly just changed it without noting it. I'll explore what happens with different values. Right now it's taking forever to perform the fill bed with instances function with object spacing of 0.6 and zero from bed, but I've got it set to try all the options, so I'm asking for trouble. Yep, works with 0.6 and 0. And I remembered correctly, it was spacing of 20 and 96 from bed. I must have randomly clicked on the from bed slider or something. Presumably the internal error is interesting, though. Here are the settings in situ: And that leads to the originally reported error in the circumstances described.

[corywagner]

I just had the same error. It happened during slicing. Happens everytime on the same model. Attached.
The castle J4.zip

[u89djt]

Hi @corywagner I've loaded the stl you provided into prusaslicer and sliced with Prusa XL defaults and it doesn't throw any errors. If you can post the 3mf file, I'll slice that and report back. We should note that you've had the same error /message/, not necessarily the same error. It's a tedious fact in software engineering that most error messages are very, very specific about what's happening in the code at one instant and depressingly vague in relation to anything a user might care about :) It could be the same bit of code giving up, though. Giving them the 3mf will enable them to run the code in a debugging environment to find it.

[u89djt]

@corywagner I just ran it with the default MK3S profile and got the error. So it happens with MK3S but not with XL 5 tool input shaper 0.4 nozzles. Sorry, I should have thought more. But we shouldn't leave the devs wondering. (Aside: the error message is different, it just has some of the same words in it - every detail counts.) Here's the castle in the MK3S profile: MK3S The castle J4.zip Here's the error dialog: Hit OK on the error dialog and we see: and if I hit the undo arrow, I see this:

[u89djt]

@corywagner If you hit the arrange button before slicing, it succeeds in slicing it. You get a print stability warning because of all the overhangs, but it gives a printable result. Here's a 3mf file of it after hitting arrange: arranged MK3S The castle J4.zip It sits at 30% preparing infill as it did when emitting the error, but it continues and succeeds. That's either going from scratch placing the stl then hitting arrange, or loading this 3mf file as a fresh execution of prusaslicer. I guess that says it's something in the way the data is organized, possibly also arguing against there being a transient state in the executable.

[corywagner]

sorry, didn't work . Same error. I've included the 3MF file as requested The castle B4.zip

[corywagner]

BTW, u89djt, thank you very much for your help and insight.

[u89djt]

Aha, details to chase! You've posted B4 there, and the one you had before was called J4. Are those from exactly the same stl file? I've just had another few goes at slicing the original stl file as follows with interesting results: 1) Open prusaslicer, change to MK3S profile, close the program. Reopen it and it's in the MK3S profile. Drag and drop the stl. Hit slice now. It gives the error. Close the program. 2) Open prusaslicer, it's in MK3S, drag and drop stl, change to XL 5T IS, Hit slice now, it gives the error. Close slicer, which will now default to XL. 3) Open slicer, it's in XL, drag and drop stl, hit slice now and it succeeds without error, just warning about stability. Close the program. 4) Open slicer, it's in XL, drag and drop stl, change to MK3S, hit slice now and it succeeds. <- workaround for now In the context of these attempts, it seems the error occurs when the data is loaded into the slicer while the MK3S environment is active, whatever that might mean. I haven't engaged with the code, so I don't know any more. Re hitting arrange before slicing, that still works for me from scratch in the MK3S environment, but I've changed the values for spacing in various ways. Try setting them both to zero before arranging and see what happens?

[corywagner]

same model, different file name.

On Thu, May 23, 2024 at 1:38 AM Dave T. @.***> wrote:

Aha, details to chase! You've posted B4 there, and the one you had before was called J4. Are those from exactly the same stl file?

— Reply to this email directly, view it on GitHub https://github.com/prusa3d/PrusaSlicer/issues/12544#issuecomment-2126547784, or unsubscribe https://github.com/notifications/unsubscribe-auth/ARUJRWXOMGLXAPCS6K33X43ZDWTHLAVCNFSM6AAAAABFVIXFXSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMRWGU2DONZYGQ . You are receiving this because you were mentioned.Message ID: @.***>