rodw wrote:
bevins wrote: Until you give it a command where it wasn't designed to accept one and it decides to ignore you cause its busy dealing with task. My opinion remap is a band-aid and will always have issues. Every time I do remaps I have to deal with this. I have one machine I cant fix because of it. Have to just cut three or 4 pieces and reset, otherwise it just goes brain-dead. I'm even thinking on putting a commercial PC controller on it.
Unless I am doing something wrong, but that's hard to tell because there are no detailed docs to deal with it nor anyone available to help.
Sorry, ranting...... This belongs in another arena.
Well you could write your own version of task and iocontrol. Both of these are designed to be swapped in and out by ini file settings. There is even an example version of iocontrol v2 in the source that adds additional pins to iocontrol. You can blame Andy but earlier in the week I copied and renamed the task module and tweaked the source to make it compile. It only took maybe half an hour to get my machine running my new (identical but renamed) task module but I really had no idea what to do next. Things that came to mind would be modified spindle handling (M3/M5) for plasma and maybe adding a new gcode or two that updated pins in iocontrol. I wasn't sure if the gcode additions could be done.
iocontrol has nothing to do with it. Nor is it a problem of task.
When I read "ps: to decode the mystery 'wtf is the interpreter doing'," which is a direct quote from the author of remap, I realize I am in over my head. I am not much of a programmer. I can write a remap in python but beyond that, understanding the internals of linuxcnc is way above my head. He clearly states it works in ngc, you do it in python you are on your own. So I guess it was my mistake.
The docs should have some type of warning...... I am left to try and convert a 1,000 line python remap to ngc oword remap which may or may not have the same issues.
