Axis Bug or something else?

I have a weird behaviour, veryfied in 2.8.3 and 2.8.4.
My G-code (generated by CamBam) looks perfectly sound, all moves are within the parameters I would expect.
But when I load it into Axis, I see a much larger bounding box for the piece and moves way outside the capabilities of the machine.
Also, this is not just a display bug: when I try to run the G-code, the machine actually moves to the positions shown by Axis.
I attached the Axis screenshot and the G-Code loaded, no secrets there.

I have been using Linuxcnc for a while now and not encountered this behaviour before - but this is the first time I use a g-Code with many small, finicky moves.

In case it matters: the machine controller is a RaspberryPi4 with Mesa FPGA (Ethernet) and a self-made HAT for the spindle interface..
The same G-code loaded into LinuxCNC 2.7.15 displays as expected (no moves outside the expected bounding box). But this other machine I still run on the old version is limited and not suitable for this job, so sadly not an option.

How would I downgrade my Rasbpi to 2.7? In the repositories I only find 2.8.0 and up. Would a downgrade even help?

Merci
hase

 

If i load your file, it looks like this. My machine has Z0 at Z full retracted. You try to run a different file with the same name, with additional vertical item at 722.6
If you don't select a coordinate system in your gcode, it will run with whatever is active.

 

Thanks for the quick reply.
I tried renaming - following your hint - to get the same result: an uncommanded move at x=722.6. I also see the lowest Z-move way below any move in the file (file minimum is Z=-0.7, Axis shows -113.0 in the bounding box).

I move my origin by touching off all 3 axis every time, so the coodinate system should not matter.

And I can assure you: I am loading exactly the G-code I attached to the post, exact same file.

And the weird thing: when I click on the vertical move at 722.6, Axis shows that to be line 35 of my file which has the move G1 X98.95778 Y97.89562 in it, preceeeded by G1 X98.97112 Y97.96341 in line 34.
And deleting the line in question gives the exact same result (uncommanded move at 722.6) with Axis indicating line 35 (the new line 35) again. 

Loads fine on my simulation in version 2.9.
Just a stab in the dark but have you tried removing the tool change command in line 9:
T11 M6

Thanks a lot for the comment.
I found my problem, it is *not* with LinuxCNC.
It's a layer 8 error.

I recently started an upgrade for my machine (tool height sensor). And that sensor is at - you guessed it - x=722.6...
Too bad that "recently" for my Hobby shop amounts to "longer than my flaking memory" or

Anyway: your tip was on the mopney: there is more G-Code executed than there is in the file.

Good catch: it was indeed an override of the M6 code (intended for tool length probing whenever I get to that upgrade) that injected the vertical Z move.
But as usual: such thing happen if - and only if - the job must be finished today. Then the unexpected stuff happens.

And also as usual: once the reason was found, I was embarrased
hase