Re: [CAD_CAM_EDM_DRO] Re: RoboFac / Stewart-Gough Platform / Hexapod Milling Machine

Mike wrote:

>As someone who makes their living writing code for embedded Linux
>devices, I can't completely disagree with him, except when he says
>that EMC doesn't have the capabilities he needed for his machine. It
>almost certainly does.
>

Let me be specific. EMC uses constants for the endpoints of the heapod
actuators. In order to run a hexapod that has different dimensions than
what is defined in advance, you have to re-compile EMC.

The RoboFac software lets you define your own hexapod geometry using the
GUI. It also has a calibration feature so that you don't have to build
your hexapod to exact dimensions.

I tried re-comiling EMC on a BDI box, but the compile didn't work; it
errored out.

I wanted to develop my control software on my regular development
workstation, not on the EMC-BDI machine down in my workshop.

I found EMC's GUI to be confusing and ugly. Only one of the several GUIs
for EMC supported hexapods at all. The configuration files are also
confusing when it comes to setting up hexapods.

These were some of the reasons I chose to develop from scratch, rather
than adding to EMC.

> [...]
>Granted, it's a complex thing to control a machine tool and the code
>will reflect that complexity. But EMC is an example, IMHO, of what I
>call "professor code" or "PhD programming".
>[...]
>
>

If I had an ordinary 3 axis milling machine, I'd probably use EMC to run
it. I was impressed by its features and I occasionally refer to the
source code to see how they deal with certain problems I have with RoboFac.

And I have yet to have a single person ask how to use RoboFac to drive
an actual working hexapod or ask me how to use the calibration system.