Well honestly I was looking for something that had about as minimal 
programming approach as possible. The freecad approach doesn't seem to 
implement a hal style approach like setting pins, and signals from the 
client, it appears to be more of an MDI interface. I'd need full hal 
visibillity as I need command and feedback of pins directly. 

As was probably the original intention of LinuxCNC with someone like me in 
mind, I'm good with hal but not traditional programming languages. I have a 
friend that helps me with Python and C but I try not to throw huge projects 
at him. Like I said, I was unable to run halrmt in MK but I believe it 
seems similar to Linuxcncrsh with more hal abillities. I don't suppose that 
anyone has looked into improving these components? 

I might just be talking to the air here but if a userspace or RT component  
(which does not exist yet) could  write telnet commands as pins that could 
be written to in hal, it could exist on the a client machine running a sim 
config that likely would not need an RT kernel. Hal-python could also 
read/write these pins/signals on the client side. On the RT hal side halrmt 
or similar would exist almost exactly as it does, just grabbing the telnet 
commands and commanding RT-hal. This approach would pretty much enable a 
linuxCNC machine to talk to an MK-CNC machine just using hal as a universal 
language. Completely GUI and language independent, if the interface can 
somehow set a hal pin, or give an MDI command it should work with this

I'm certainly not married to telnet as a protocol, I just mention it 
because it exists and this communication path is not meant for realtime 
critical things. If there is some existing way to do similar to what I 
mentioned I'd love to hear about it.

On Wednesday, December 4, 2019 at 11:19:18 PM UTC-5, mlampert wrote:
>
> Not sure if you're looking for python or C++. In case it's the former 
> you could use the base classes from the FreeCAD workbench: 
>
> https://github.com/mlampert/machinekit-workbench 
>
> All files which start with MK... are independent of FreeCAD and/or QT. 
> The final class which ties everything together is in machinekit.py, 
> which can also function without FreeCAD but it does use a QT base 
> object for signal propagation into the UI. You can probably replace 
> that quite easy with whatever GTK2 uses for signals/notifications. 
>
> Other than that you are expected to call _update() in machinekit.py 
> periodically for service discovery and message processing. 
>
> HTH and let me know if you have any questions. 
>
>
> On Wed, 4 Dec 2019 18:17:42 -0800 (PST) 
> justin White <[email protected] <javascript:>> wrote: 
>
> > Every once in a while I sit down and try to get something going in QT 
> > creator for remaking  My project machine LinuxCNC GUI to QT for use 
> > with a remote instance of Machinekit. Right now this machine runs 
> > LinuxCNC with a mesa 7i96, and it works just fine, however my ideal 
> > setup is to use MKSOCFPGA on the DE10-Nano with the interface board I 
> > created to replace the mesa hardware and get the hal side off of the 
> > x86 PC. Getting a QT ui going would just be too much work even if I 
> > could get it started 
> > 
> > Problem is AFAIK all of the remote stuff is baked into qtquickvcp, 
> > and not MK's gladevcp has nothing new. I'm fairly certain I can get 
> > the existing GTK2 UI to run under MK as I've made a GTK2 test UI for 
> > my interface board that runs in both MK and LinuxCNC, but I'd need 
> > remote capabilities. The other day I was playing with linuxcncrsh and 
> > was able to control basic things on my MK test GUI with it but 
> > linuxcncrsh has no ability to set pins. I tried halrmt which 
> > supposedly does set pins but it does not seem to run on MK. From what 
> > I've read, linuxcncrsh and halrmt are both poorly coded so I don't 
> > think they are ideal anyway. I guess the basic question is is there 
> > any way to get machinetalk or any small useable portion of it to work 
> > with a GTK2 GUI? 
> > 
>
>

-- 
website: http://www.machinekit.io blog: http://blog.machinekit.io github: 
https://github.com/machinekit
--- 
You received this message because you are subscribed to the Google Groups 
"Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/machinekit/eb1e124d-049c-44bc-b3a6-99f80579a8c6%40googlegroups.com.

Reply via email to