I've never worked with a wire foam cutter, but I have made some Linuxcnc configurations that can use some custom M-code commands that will temporarily slave two axis together such that one will move with the commands of the other. The machines I was using this on were routers with two spindles each mounted on their own axis and configured XYZW. I had ran these for years just using XYZW g-code, but when the multi-line look ahead planer was first introduced in Linuxcnc, it only supported XYZ moves, any moves with a W command reverted to the old single line tool planner, and significantly slowed the wood carving we did with these. So I built my config with an M-code that slaved the W axis to the Z at their current positions when the command was issued. The W would then follow the Z until I issued the M-code command that released the W. (This had to be done when Z and W were commanded to be at the same positions relative to the position they were slaved at, or the W would give a following error.) Maybe this technique could be useful for you? Here is a copy of one of my configurations.
 

Ok try this one:

The first problem is Python.
Python is powerfull, but permanently developed, so python scripts are version sensitive.
LinuxCNC depends on Python...
Debian is conservative, Ubuntu is progressive, so there must be version mismatches.
Sometimes old programs not works with new OS, but worst is when works, but not properly, become unstable or some functionality gone.... 
Believe me, experiments like LinuxCNC+Ubuntu are not worth wasted time.

Mass - thanks so, so much!

This is sweet. I will dive deeper into it in the days to come and the project you linked to. But I am almost sure I will buy a (used) SK4/6 panel. I wouldn't be surprised if they all come with the PS/2 + matrix combo. (Some will possibly be USB+matrix but that would just be a win). I know, no guarentees - also in terms of pinout and such, but the main foundation is certainly there as you have shown. And most important, nothing locked behind a proprietary serial protocol either, it seems.

But any chance you could share the code, too?
Having looked at the github for the Arduino-LCNC project a bit, I now realize that the matrix never gets turned into USB-HID or gamepad protocol, at all. Interesting. Maybe a better more direct approach.

Not that I really know what I will be looking at but with the help of Gemini, I managed to make a nice custom version of the NanoELS FW (an open source electronic leadscrew project) for my little lathe. Had to get DRO scales implemented and add some other features - and I actually managed to do so on code level. So, hopefully the robots can help me on this project, too. 

Various Panels
As for SK4/6 I actually think they share the same layout 100%. I think I have even seen some SK6s with SK4 printed on the PCBs - but I could be wrong as I have looked at sooo many pics lately.
But there's also an SK5 which does have a slightly different layout. (Attaching pics of SK4/5/6).
Most often SK4 has membrane buttons whereas the SK6 often has tactile ones. I am leaning towards the latter.
But looking at SK PCBs get confusing very fast as there are so many different itirations of the SK4/5 plus a bunch of clones. 

Encoder Overrides?
BTW, it looks like you swapped out the clicky rotary switches for pots or encoders? So, no hardstops at each end but possibly better "sync" between screen and knob, is that it?
I think I will try to stay with the rotaries and see how they feel. 

sorry, I posted on that one, but when nobody replied, I figured the post was too old and wasn't active anymore

Ha, that's spot on—I really did feel like a detective sometimesimage missing /media/kunena/emoticons/smile.png
The top part is a standard PS/2 keyboard; I connected it to the computer using a PS/2-USB adapter. I remapped the keys, mainly to access lowercase characters (Remap the Shift + button, just like in the original system).
The lower part is connected via an Arduino and an ArduinoConnector ( forum.linuxcnc.org/show-your-stuff/48630-arduino-io-expansion ).
On this keyboard, the pinout is rotated 90 degrees relative to the visual orientation, and not all pins are used. I added a pinout; I hope this helps.
P.S. The SK6 keyboard has a visually different circuit board; I haven't checked it, so all of this applies only to the SK4.

Hi Chris.

Here you have the file.

Now it has let me set it to 5 + Jog, and it displays correctly.

Lcvette wrote:

Sweet are you in the probe basic Facebook group?

Yes, I'm in the Facebook group.

Heh.. the very next search i did after replying lead me to the info that the displaylink driver doesn't handle touch, Linux kernals hud-multitouch does.

always something new to learn eh.. though its pretty obveous in hindsight

To be fair to Tommy, I've used this ISO successfully for a while now, this is the first issue I've come across and its not anything to do with linuxcnc, and sounds like I can resolve it.

In fact the synaptics displaylink drivers are only officially offered for ubuntu, while there is a community Debian option available, and the second computer I have that I set up a Debian build on is struggling to get the screen to show anything, where as this Mint ISO already allows the screen to display and operate with single point touch only, which would actually be sufficient for a cnc mill, but less refined than the screen could be.

This goes to show that there are pros and cons to all options and its not at all accurate to claim we must stick with Debian.

With that said, i am wanting to try Probe Basic, and being that this was developed for Debian may be the way i go, hence the aforementioned debian unit. 

To all that replied with useful information, greatly appreciated. Ill report back with results.

well strange.
If you are content to just ignore the warning that's fine, otherwise I would need you to find the qtdragon_hd_handler.py file from your system so i could see why it's different.

Sweet are you in the probe basic Facebook group?