Moderator tommylight indicates that no 8255 is installed, instead, there's just a parallel port?  I'm using the "native" parallel port on the Compaq Presario,  Running dmesg on the rtai kernel says:
parport_pc 00:07 reported by plug& play ACPI
parport0 PC-style @ 0x378 irq 7 [PCSPP,EPP]
lp0: using parport0 (interrupt driven)

This could mean there is a version of linuxcnc kernel that already supports a "native" parallel port.  Certainly if there's an ISO image that already supports a parallel port my hardware setup isn't going to have the desired affect of testing against an 8255. 



 

hows it coming?

Hi
Meanwhile, there is Xenomai3 and Xenomai4 support in master branch.

Details:
github.com/LinuxCNC/linuxcnc/issues/3878

Prebuilt Kernels and Linuxcnc 2.9.10 back-port for Xenomai4:
github.com/hdiethelm/xenomai3-linuxcnc
github.com/hdiethelm/xenomai4-linuxcnc

For me, it works well and I fully switched to Xenomai4.

Hi All
Meanwhile, I fixed Xenomai3 support.
Additionally, I added Xenomai4 including support Ethernet mesa cards.

Details:
github.com/LinuxCNC/linuxcnc/issues/3878

Prebuilt Kernels and Linuxcnc 2.9.10 back-port for Xenomai4:
github.com/hdiethelm/xenomai3-linuxcnc
github.com/hdiethelm/xenomai4-linuxcnc

Is someone using Xenomai3 actively? Xenomai3 support for Ethernet mesa cards is available in a private branch but not (yet) merged.

I don't recommend this kind of fun to anyone, but LinuxCNC with Trixie can be installed and run from an 8 GB partition.
For a standard xfce4 installation, 16 GB is enough.
I'm talking about just LinuxCNC without any CAD/CAM or other utilities.
Of course, today it is difficult to buy such a small disk and there is no technical or economic justification for it, but it gives you an idea of ​​how much is needed.

40 GB would be plenty, if you could run gparted and move your partitions around making space for a third system, and install debian trixie, maybe with xfce, we could try and develop support for pci 8255 on the preempt-rt kernel, and you could test it out.

preempt-rt kernels are easy to install,
sudo apt install linux-image-rt-amd64 linux-headers-rt-amd64
Is all you need for that, you will have to run in place, so you want to git pull the repo, and build locally uspace and run in place.
Once you got that working, let me know, and I will prepare something for you to test.

Sure thing, I'll try and put one together in the next couple of days.

I would suggest qtdragon-hd with a 1080p touch screen. Several probing methodologies are included.
Linuxcnc includes a  (large?) number of Sims.
There is an extensive thread about Ethercat jitter which you will see died out once issues were resolved.
forum.linuxcnc.org/ethercat/58103-ethercat-random-jitter-fix
Part of the solution  is tuning the system for RT performance and it was comforting for me to see it acknowledged that my tuning  methods were needed. I might add, the same applies for mesa network cards.

The fundamental difference between mesa and Ethercat is that mesa requires a dedicated point to point railway track. Packets hop on the train and get off at the end. Then packets get on the train and go back to the PC.

Ethercat is more like a streamlined bullet train which goes past a number of stations (slaves). Packets get on and off at each station (slave) without the train stopping. Hence why I say its more efficient.

The other difference is that mesa has pncconf to build your config and ethercat need you to build your config from a template just like I had to do on my first machine that used a Mesa 7i76e that was not supported by LinuxCNC at the time.  Having a similar Ethercat dsonfig tool is unlikely. It hard enough to do for the family if mesa cards, imagine how hard it would be with thousands of devlces, some of which are rarely found on a CNC machine. Then ask a machine to home itself without LinuxCNC's knowledge or control....

It seems to me the lawn is really an XY plane with no motors on the axes and the tyres follow a tangential arc and provide the drive.
I think the new motion.interp.heading pin in master branch could solve this. You give the mmachine fake XY g1/g2/g3 moves including the arc at the end and tell the tyres (eg A axis) to follow the heading. If the front and rear wheels need to follow differing arcs on a turn, you might need to adjust/scale the heading using a real time component. Perhaps the spindle speed (S command) might set the rpm on the wheels.

I think with this new feature, solving your problem is a lot simpler than you think.

Some more questions:
What's the best touch GUI in your opinion? Later I must implement custom multi spindle tool probing etc, so that should be possible.
Can someone give me some tips how to set up a machine simulation? It would be great if that also hast a collision detection. Is that possible?, how does that integrate into a GUI?

Please excuse the crudity of these drawings. I didn't have time to build it to scale. (Doc Brown BTTF reference)

That sounds great, where do I find more information about that?