Ditching Linux CNC & why

Guys, I mentioned I has a success in setting up Linuxcnc in under 1 minutes far quicker than Mach3.

It took an hour or two to get the VFD working. That was OK.

The only thing that really wasted my time was gmoccapy. Regardless on how much I dig up info, how many times I had to reread the article, it just wont work. Eventually after 1 day & weekend "3 days" I gave up on gmoccapy.

Axis was fantastic. No complaints, it worked. But unfortunately after getting the buttons configured the VFD would not work. There were no error messages to indicate a problem. If there was I would have understood the problem and solved it.

In terms of buggy, it might be harsh to use buggy but the fact is in my case when you have something that works and then it does not work, in the industry of IT world where standardization is the key to IT success, that is referenced as buggy software, or unstable or pre-alpha.

You guys do not see what goes on behind the screen to change an interface on a CNC router.
1. 1 hour+ downloading, installing LinuxCNC + trouble shooting USB BIOS +booting utility and other issues.
2. Configuring basic LinuxCNC and confirming motors turning
3. Learning to use LinuxCNC, router dimensions, avoiding crashing & breaking bits.
4. Changing the interface was a fail.
5. 5 minutes to configure VFD and 2 hours trouble shooting it up until I got it working.
6. Install buttons few hours of reading, rereading, programming, trouble shooting, analyzing the issues and finally got it to work.
7. VFD will not start after the button configurations.
8. Reconfigure everything from Zero replug in all the numbers and get the router + motors + limit switches + dimentions + estop all working
9. VFD will not start.
Yes buggy software. If it was not buggy then the new configuration will not be affected in any way especially that I was only altering ini files and xml.

Its a lot easier to criticize me than to extend a helping hand. Your actions in fact have swayed me to run away faster from LinuxCNC than to give it another try. You as a community dedicated to LinuxCNC have tainted this software with your comments not caring about people's efforts that went into LinuxCNC and the time spent on it.

KalKat wrote: Its a lot easier to criticize me than to extend a helping hand. Your actions in fact have swayed me to run away faster from LinuxCNC than to give it another try. You as a community dedicated to LinuxCNC have tainted this software with your comments not caring about people's efforts that went into LinuxCNC and the time spent on it.

From an outside observer's perspective I think these guys are still asking you to ask them for help!

A lot can get lost in translation on internet forums like these. Especially with us engineering types. I managed to inadvertently ruffle someone's feathers on my second post here...

As a newbie still in the middle of the learning curve I'll be very curious to learn what you find does work for you in the end. I hope you'll report back. Good luck!

I was not criticizing your efforts.
In general power users and developers forget how much experience they have behind them.
I do understand the learning curve - I have spent much time on configuration wizards to help people get started.

I was hoping that the next person who may read your experience may change a few things to help them avoid some of your troubles.

Without the config files I couldn't tell you what went wrong with the VFD, I could only guess. INI and XML files are part of the core of configuration - more then likely a HAL file wasn't loaded, because it wasn't defined in the INI.

To know for sure would require you to spend some more time on linuxcnc, and it seems you have had enough.

Maybe we could add some popular VFD options in the stepconf wizard - this is probably a common device to use.

Chris M

I think you've made your decision and taking your bat and ball and going home. Everybody here who have commented are willing to help. But as yet you've not given one useful piece of information that would allow us to help you.

Moving to from Axis to Gmoccappy was indeed a one line change config for me from a working axis config. But once you add a few custom features, there is no going back.

Adding interface buttons is an advanced method but I found easier in Gmocappy than Axis which I never attempted to alter. You really need to install Glade and this will let you paint those buttons on the screen but you do need to edit the python screen handler in a text editor. Rather than change the base screens, its probably better to add a tab or follow the example in the gmocappy plasma sim.

Also gmoccappy has a series of macro buttons that can be enabled and if your features can be written in gcode, this becomes very easy to achieve. By the sound of it, you could have achieved most of your basic features with a series of 10 line gcode subroutines without touching the glade XML.

And always copy your config to another folder before you add a new feature so you have a fallback position and just do one thing at a time in each version so you always keep a fallback position.

I think many of us have spent far more than 40 hours playing with Linuxcnc for a non-trivial build. A lot of that is learning not doing and you put that down to inexperience.

meh

I must say that if the OP wants a quick fix this is not the correct place to be. When I started on my plasma machine it was a working commercial unit. I knew or thought I knew exactly how I wanted it to operate. So I bought a Mesa THCad card and was going to use a small micro card to read its output into a PC and use the commercial program. Well I started reading about LinuxCNC and saw the Gmoccapy plasma examples and thought this is it! Well it took me 2 1/2 years to build a working system with all the custom bells and whistles. The free PC hardware I had was not happy with the iso available at the time and wouldn't work for more the a few minutes. I finally tried Linux Mint Debian Edition 2 and it would run on my hardware (still does). Now the problem was no RT. So I searched and search some more to find instructions on patching the kernel for RT. It only took about 6 or 7 maybe more tries before I got it to work. Of course by then I needed the master branch because of my dual drive gantry. More figuring out how the download and build. Now that wonderful Gmoccapy plasma example has no guts behind the GUI, off into first getting the correct Glade downloaded and configured then to building new screens. I am not a programmer or IT guy so along the way I got to learn rudimentary Python reading then parroting and actually building the handlers for the Glade panels. I will not bore you with the adventures into external offsets land (now part of the master branch) but this ate up several months of THC trials time. When I was finally happy with the operation of my system I got rid of the commercial controller and THC. Then six months later this guy in Oz starts posting about his Plasmac branch for running a plasma cutter. Because I had created such a can of code worms my system wasn't really maintainable or easy to change. I needed a new feature and couldn't get it work work. I cloned the Plasmac branch and started to adapt it to my machine. I found I really liked how the THC and other plasma specific stuff worked. Thanks to Phill C. I now have a really nice CNC plasma cutting machine using his Plasmac branch. All of this has taken about 3 1/2 years to date but I have really nice custom CNC plasma machine that cuts as good or better than machines that cost many many thousands of dollars more. I guess the moral of this story is that unless YOU are willing to put in the time and effort to build a custom machine controller this isn't the place to be. Everyone here was more than helpful and patient when helping me along the way. If all else fails RTFM until you understand how the feature works!

John