Re: [Emc-developers] question about interpreter
On Mon, 1 Jun 2020 at 15:00, Reinhard wrote: > Motion id 80 took 13.393856 seconds. > Motion id 0 took 0.00 seconds. > Motion id 3771 took 28.790746 seconds. > Motion id 0 took 0.00 seconds. > That does look like it aborted on motion id 80 then for some reason carried on with motion id 3771. ie abort is almost, but not quite, clearing the queue. -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
Hi Andy, > Do the logs show anything interesting? I don't know. Is the fact that gcode-commands don't have feedback from motion ok for you? Is it an old bug? Should I know about it? Don't know, for me its not ok, so I would like to here your opinion. Don't know what is interesting for you. I could send you the log, if you want. A grep for "Motion id" gives this output: Motion id 21 took 1.050300 seconds. Motion id 22 took 3.157045 seconds. Motion id 23 took 59.914013 seconds. Motion id 26 took 1.795147 seconds. Motion id 30 took 0.556928 seconds. Motion id 33 took 0.557085 seconds. Motion id 36 took 0.556588 seconds. Motion id 39 took 0.556267 seconds. Motion id 42 took 0.556746 seconds. Motion id 45 took 0.350573 seconds. Motion id 55 took 37.693853 seconds. Motion id 56 took 0.101479 seconds. Motion id 57 took 0.094618 seconds. Motion id 58 took 0.088539 seconds. Motion id 60 took 0.166279 seconds. Motion id 62 took 0.163056 seconds. Motion id 64 took 0.170251 seconds. Motion id 66 took 0.159595 seconds. Motion id 68 took 0.166556 seconds. Motion id 69 took 0.088413 seconds. Motion id 70 took 0.093762 seconds. Motion id 71 took 0.105196 seconds. Motion id 80 took 13.393856 seconds. Motion id 0 took 0.00 seconds. Motion id 3771 took 28.790746 seconds. Motion id 0 took 0.00 seconds. I did not remove a line from greps output. cheers Reinhard ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
On Mon, 1 Jun 2020 at 07:34, Reinhard wrote: > I run 3D_Chips.ngc from axis with logging enabled. After having enuf logs collected I hit stop. First I thought: ok, this time it stops. ... but then it started to run again (stop button of axis was already shown as pressed) and the tool run across the workpiece. Do the logs show anything interesting? -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
Hi, > Is 2.8 the same? I am wondering if this is related to > https://github.com/LinuxCNC/linuxcnc/issues/865 I had the chance to take a snapshot of that accident =:O When I tested about the omitted lines from backend, I run 3D_Chips.ngc from axis with logging enabled. After having enuf logs collected I hit stop. First I thought: ok, this time it stops. ... but then it started to run again (stop button of axis was already shown as pressed) and the tool run across the workpiece. I uploaded a screenshot: https://www.imagebanana.com/s/1774/QnA1w8ks.html cheers Reinhard ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
On Montag, 1. Juni 2020, 01:00:15 CEST andy pugh wrote: > On Sun, 31 May 2020 at 17:58, Reinhard >> > > > Which LinuxCNC version? > > > > master from repository. > > Is 2.8 the same? I am wondering if this is related to > https://github.com/LinuxCNC/linuxcnc/issues/865 Hm, wasn't aware of that entry. Yes, looks quite the same. But does not happen everytime. I always test with master and hit stop on a running job is quite normal. Sometimes (more often than a hazardous run) I get the Error-Message: "need to be enabled, in coord mode for linear move" ... but most times it works just as expected. Yesterday it was the first time, that I noted that behaviour. When it happened first, it looked like motion will terminate current command/ segment. Second time, it looked like motion performs a command, that's not in the file. So may be random, don't know. regards Reinhard ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
On Sun, 31 May 2020 at 17:58, Reinhard wrote: > Hi Andy, > > > Which LinuxCNC version? > > master from repository. Is 2.8 the same? I am wondering if this is related to https://github.com/LinuxCNC/linuxcnc/issues/865 -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
Hi Andy, > Which LinuxCNC version? master from repository. cheers Reinhard ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
On Sun, 31 May 2020 at 17:12, Reinhard wrote: ... but backend was stil executing motion and sending coordinate updates - > for > about one second or more. Which LinuxCNC version? -- atp "A motorcycle is a bicycle with a pandemonium attachment and is designed for the especial use of mechanical geniuses, daredevils and lunatics." — George Fitch, Atlanta Constitution Newspaper, 1912 ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers
Re: [Emc-developers] question about interpreter
Hi, can someone please verify my understanding? >From recent discussions I got that interpreter is doing all coordinate calculations and that taskmanager does not change them. So if that is true, than I would consider it a bug, that interpreter prints different coordinate values from the same file(s). If my understanding is wrong, please tell me, where my misunderstanding is. Actually I was shocked by testing. I tested longtime running with 3D-chips.ngc and during execution I hit stop, which is send to backend by nml. Axis was running in parallel and axis changed button states to stopped too - ... but backend was stil executing motion and sending coordinate updates - for about one second or more. That can't be ok. Stop means everything should stop immediately - without question or condition. Well, at least in my world. cheers Reinhard ___ Emc-developers mailing list Emc-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/emc-developers