On 6/5/25 8:02 PM, andy pugh wrote:
Lines 2288 and 2289 might be responsible for this:
https://github.com/LinuxCNC/linuxcnc/blob/fd5be668a61e42e2d8c8ceec75d19e77f3612e2e/src/emc/task/emctaskmain.cc#L2287
It does look like it. It takes a fair bit of scrolling to work out the
structure, but that all seems to be inside the line 2219 case, which
looks to be MDI mode.
Would we just delete all that, or make it INI-controlled defaulting to
the old behaviour?

It needs to be analyzed to see what "side-effects" the abort code has on the machine state and whether it (re)sets some internals that would be required to continue.

From the looks, there are multiple functions that need to be traced.


I tend to think the old behaviour is annoying and wrong, but maybe
leave an option to revert to the old behaviour (we have an outstanding
request to disallow MDI queuing altogether, as someone feels it is in
contravention of the RS274NGC standard. I would guess that he would
favour an abort-on-error)

Disallowing MDI queuing should still not require aborts and are not expected there either. Disabling queuing should probably be done at another level.

But, allowing to revert to (quoting your words) "the old behaviour is annoying and wrong" situation seems a bit like masochism to me.

My guess is that users will be glad the software behaves more like you'd expect...

But, if it is merely a comment-in/comment-out code question (after above analysis), then, by all means, add a masochism configuration in the INI file.

--
Greetings Bertho

(disclaimers are disclaimed)



_______________________________________________
Emc-developers mailing list
Emc-developers@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-developers

Reply via email to