ChatGPT LinuxCNC Assistent

Has anyone used the wizard yet? The answers and support are good, but the data transfer hasn't been working for days. I've tried all possible options. Has it ever worked? Or is it disabled for whatever reason?


chatgpt.com/g/g-vfeQnX43z-linuxcnc-assistant?ref=img.pt

 

The URL posted to the mailing list was
chat.openai.com/g/g-vfeQnX43z-linuxcnc-assistant
but I can't check that from work as they are worried I will ask questions that leak data.

The link for the same Page.

I tested all of the public file services offered; some of the alleged URLs were not assigned; a private FTP server or cloud was allegedly tested positive, but no data arrived.

I've played around with the wizard a lot and actually put together a good configuration. Unfortunately, LinuxCNC won't start, and after a day of back and forth, I'm pretty frustrated because the AI ​​seems to constantly assume an RTAI kernel and, despite my reminder, can't quite figure out how to use the RT kernel. Attached are the .ini and custom.hal files. How do I modify the files to make it run, or where can I get the specific information?

  Unfortunately I couldn't add anything more, here is the error report from linuxcnc.

Problem is here:

Note: Using POSIX realtime
./custom.hal:5: execv(stepgen): No such file or directory
./custom.hal:5: waitpid failed stepgen stepgen
./custom.hal:5: stepgen exited without becoming ready

I didn't try it, but I have a good experience with Gemini 2.5 pro to configure (text based) a midi controller with custom firmware just by uploading the manual and explain what I wanted. You could try the same with LinuxCNC, upload the right PDF and tell that this is the manual and explain what you want. If it doesn't give a good result try without the manual but add the right prompts.
In the PDF you can see the result with adding the manual, I didn't look at it throughout but it seems already make more sense. (If you indeed will use remora)

 

This is just wrong (typical AI error?)
stepgen is realtime, and should be loaded with "loadrt"
I would rather expect that the same is true of remora-eth-3.0 (if that even exists)

linuxcnc.org/docs/stable/html/man/man9/stepgen.9.html

whole halfile is wrong. its trying to mix other linuxcnc stepgen with remora

from ai

from a remora example

same with the inputs, its mashing up the other linuxcnc inputs with remora inputs

from AI

remora code

I dont think it did a single halpin correctly in the file, maybe the arduino one but im not sure.

How do I modify the files to make it run, or where can I get the specific information?

You should start by not using the entire thing. The thing you generated is essentially a broken version of the default/premade configurations we provide for remora The only thing it actually did was assign signals to the remora IO to add the buttons. The encoder signals arent connected to anything, and many of the parameters in the INI file arent real or connected to anything either, its mixing between joint C and Y, the speeds look insane and most of the ini entries are completely made up

Based on what I assume is the intended halfile, you have barely changed anything from the default remora halfile to begin with.


Compare that file to this one, known good and working
github.com/cakeslob/RRW_LAB/blob/main/Li...olset/remora-xyz.hal

it looks like i didnt upload my lathe config, but ill post that if you promise not to use that ai file