#flex-dev Logs

Sep 03 2025

04:36 AM Tom_L: morning
04:50 AM JT-Cave: morning
06:11 AM JT-Cave: 61 89
06:16 AM JT-Cave: fixed that label I think
07:44 AM Tom_L: PR #343
07:44 AM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/JT-SHOP/flexgui/touch-probe3-calcs3.png
08:46 AM Tom_dev: axis & keyboard jog are both enabled during probing. the original jog buttons are not
08:47 AM Tom_dev: probing is misbehaving on 2 pcs. on the bar routine it never finishes probing like in the video i posted
08:47 AM Tom_dev: dunno if something changed there....
08:48 AM Tom_dev: i'm testing a few other probe routines
08:48 AM Tom_dev: corner probing does not finish now either. stops with no warnings or errors
08:51 AM Tom_dev: none of the probe routines except maybe the angle ones will finish running the subroutines now
08:51 AM Tom_dev: no idea what might have changed
08:55 AM Tom_dev: behaves similar on both PCs
09:28 AM Tom_L: it's as if you pushed pause during a program run
09:47 AM JT-Cave: do any of the modes change during the probe move?
09:48 AM JT-Cave: I assume you mean keyboard and axis jog buttons work while probe enable is on?
09:48 AM Tom_L: i didn't check for modes
09:48 AM Tom_L: i can check again after bit
09:48 AM Tom_L: axis style & keyboard jog both are enabled during probing
09:49 AM Tom_L: yes
09:50 AM Tom_L: the original jog buttons behave like they should. no jog during probing with them
09:52 AM JT-Cave: during a move you should not be able to jog because keyboard jog is only during manual mode
09:53 AM JT-Cave: I don't know about axis style jog I'll have to check that
10:05 AM Tom_dev: probe enable button enabled, manual mode
10:06 AM Tom_dev: during probing it's mdi, as soon as it stops probing (not finished) it switches to manual mode
10:07 AM JT-Cave: is this simulated probing or on a real machine?
10:07 AM Tom_dev: i'm out on my mill right now
10:08 AM * JT-Cave heads to town for a bit
10:08 AM rdtsc-w: checked auto on the rpi5 recently, that worked fine... did not try mpg or keyboard jog though, will have to check that
10:08 AM JT-Cave: check the other traj modes
10:08 AM JT-Cave: when do ing a probe
10:08 AM Tom_dev: how do i switch modes?
10:09 AM Tom_dev: during active probing it's mdi but as soon as it stops working it goes to manual
10:10 AM Tom_dev: mid probing
10:11 AM Tom_dev: before the probe is tripped
10:12 AM JT-Shop: also run from a terminal to see if any error messages show up
10:13 AM JT-Shop: https://gnipsel.com/linuxcnc/flexgui/labels.html#id1
10:13 AM JT-Shop: exec_state_lb
10:13 AM JT-Shop: interp_state_lb
10:13 AM JT-Shop: interpreter_errcode_lb
10:13 AM JT-Shop: motion_mode_lb
10:13 AM JT-Shop: motion_type_lb
10:14 AM Tom_dev: i'll try to fit those in somewhere on the screen
10:33 AM Tom_dev: ok i tried to capture it
10:56 AM lcnc-relay: <roguish000@> good morning. matrix.org is down
10:58 AM Tom_dev: JT-Cave, http://tom-itx.no-ip.biz:443/~webpage/cnc/JT-SHOP/flexgui/probe_test/
10:58 AM Tom_dev: crap. gotta change read/write etc
11:01 AM Tom_dev: ok
11:29 AM rdtsc-w: I think matrix is very sensitive to rants, and silently folds at the first whiff of one
11:48 AM Tom_L: JT-Cave, the motion stops after the plot but before it finishes. you can hear the steppers
11:48 AM Tom_L: then when it does, the plot catches up
11:56 AM JT-Shop: do any of the states change when the probe move stops or the screen locks up
12:04 PM Tom_L: not that i could see
12:06 PM JT-Shop: revert back to 1.2.2 and test
12:06 PM Tom_L: the motion type changes from probing to feed
12:07 PM Tom_L: do you save those in the repo?
12:07 PM JT-Shop: when it stops moving
12:07 PM JT-Shop: yes all are in the repo
12:07 PM Tom_L: i will after lunch
12:07 PM JT-Shop: https://github.com/jethornton/flexgui/releases/tag/1.2.2
12:07 PM JT-Shop: nap time here
12:08 PM Tom_L: i'll probably have to remove 1.3 to revert
12:09 PM JT-Shop: yup
12:15 PM Tom_L: ok installed, will test in a bit
12:22 PM JT-Shop: I just ran my probe hole routine and had no problems with 1.3.0
12:23 PM JT-Shop: not sure if it's up to the minute or not
12:38 PM Tom_dev: 1.2.0 works, 1.2.1 fwd are broke
12:39 PM Tom_dev: i'll do a comparison mp4
12:47 PM Tom_dev: http://tom-itx.no-ip.biz:443/~webpage/cnc/JT-SHOP/flexgui/probe_test/
12:47 PM Tom_dev: working test marked 1.2.0
01:05 PM Tom_L: something changed between 1.2.0 and 1.2.1
01:48 PM JT-Shop: 1.3.0 ran fine on my bp with a circle probe that does 10 probe routines to find center
02:14 PM lcnc-relay: <roguish> finished my ride........ 21 miles. still cool, in the high 70's
02:23 PM JT-Shop: Tom_L, the bp is running flex 1.3.0 that is less than a week old and probing works fine
04:26 PM lcnc-relay: <roguish> JT-Shop: article in Popular Mechanics says Corvettes are getting recalled....
04:26 PM lcnc-relay: <roguish> Z06 and ZR1
04:26 PM JT-Shop: those are pretty new so teething pains I guess
04:27 PM lcnc-relay: <roguish> says some hidden flaw
04:28 PM JT-Shop: the basic design is 6 years old so pretty hard to imagine a hidden flaw
04:29 PM lcnc-relay: <roguish> unfortunately it's paywalled, so can't read much of the article
04:30 PM JT-Shop: that kind of stuff is usually speculative and I don't pay much attention to them
04:31 PM lcnc-relay: <roguish> https://www.chevrolet.com/performance/corvette-recall
04:39 PM JT-Shop: looks like a ralf nader crap
04:41 PM Tom_L: just posted what mine did
04:41 PM Tom_L: on 2 different PCs
04:44 PM JT-Shop: can't tell anything from videos and can't reproduce it here
04:46 PM Tom_L: well 1.2.0 runs the probe routines without hanging but 1.2.1 does not
04:46 PM Tom_L: i've no idea what changed
04:46 PM JT-Shop: try setting up some watch labels to see if the variables change somehow
04:46 PM Tom_L: this is on a 7i90 with a parport interface to it
04:47 PM JT-Shop: I ran my hole probe program and it does 10 probe moves without an issue
04:47 PM JT-Shop: on 1.3.0
04:47 PM Tom_L: yeah i saw thta
04:47 PM Tom_L: that
04:48 PM Tom_L: i think rdtsc was going to test his
04:48 PM lcnc-relay: <roguish> you guys have been doing so much work, I'll download and try the latest.....
04:48 PM Tom_L: he's on a rpi iirc
04:48 PM JT-Shop: so unless I can duplicate the issue I can't diagnose what it might be
04:49 PM Tom_L: have you tried running any of the touch-probe3 probe routines?
04:49 PM JT-Shop: no
04:50 PM JT-Shop: you use different variables than I do
04:50 PM Tom_L: that shouldn't matter, they are all available as user
04:52 PM Tom_L: when i get a chance i'll swap out the 7i90 for a 7i80 and try the rpi5
04:53 PM Tom_L: or the dev pc for that matter
04:53 PM JT-Shop: the bp uses a 5i25 and 7i77 but that should not matter
04:53 PM Tom_L: i have the configs set so it's easy to change the interface
04:53 PM JT-Shop: I did get the n100dc setup with a 7i96s today
04:54 PM Tom_L: saw that
04:54 PM JT-Shop: ended up using a usb nic
04:54 PM JT-Shop: I'm old school so my first thought was pice nic but that turned out to be not the best option
04:54 PM Tom_L: hope that doesn't cause problems
04:55 PM JT-Shop: the usb nic is just to connect to the lan
04:55 PM * rdtsc-w has a USB wifi nic (and 10ft USB cable) - no apparent issues
04:56 PM JT-Shop: I avoid wifi when possible
04:56 PM rdtsc-w: ditto... only for getting a file to it, then have a disconnect-wifi button :)
04:56 PM Tom_L: i could take the subroutine wrapper off one and run it as a program to see if that makes a difference
04:57 PM JT-Shop: ok
04:57 PM * rdtsc-w is still at work... mandatory overtime
04:57 PM JT-Shop: fun
04:57 PM * JT-Shop is not at work for a few weeks now
04:58 PM Tom_L: could be something about the parport interface and RTAI
04:58 PM Tom_L: no idea
04:59 PM lcnc-relay: <roguish> JT-Shop: you retiring again ???
04:59 PM JT-Shop: LOL
04:59 PM Tom_L: it would be easier to test with the rpi5 then i wouldn't have to change kernels on the test PC
05:03 PM * JT-Shop calls it a day
05:40 PM lcnc-relay: <roguish> me too