#flex-dev Logs

Nov 02 2025

04:03 AM JT-Cave: stupid time change
05:26 AM Tom_L: morning
07:41 AM Tom_L: i'd dislike it more if i'd lost an hour
09:36 AM lcnc-relay: <ebe_lyn@> good morning, I picked up a probe and am trying to set up touch probe 3 to use it and the the metric probing subroutines. do i just go in designer and change the o<whatever> call to o<whatever_m> call
09:40 AM Tom_L: the metric suroutines may need a litle work
09:40 AM Tom_L: i've not completely tested those
09:40 AM Tom_L: i have however tested all of the inch ones
09:41 AM Tom_L: check them to see if they are different than the inch routines
09:41 AM Tom_L: i'd say 80% of them are ok
09:42 AM Tom_L: during developement it got too hard to keep a working set of both
09:58 AM Tom_L: yes you can change the name in designer but i would test them as you go
09:59 AM Tom_L: the metric ones will need to be put in the subroutines folder
10:00 AM Tom_L: the parameter file .var will also need adjustd for metric
10:51 AM JT-Cave: finally figured out what was messing me up making a hal led button
10:51 AM Tom_L: what was that?
10:52 AM Tom_L: got a touch screen comin...
10:52 AM Tom_L: i may go thru those metric routines one of these days now that all the inch stuff works
10:53 AM JT-Cave: after subclassing the led button it changes from PyQt6.QtWidgets.QPushButton object to libflexgui.led.LEDButton object
10:54 AM JT-Cave: and I was looking for QPushButton to add the hal to
10:54 AM Tom_L: is this that new stuff we looked at?
10:54 AM JT-Cave: no
10:54 AM Tom_L: ok
10:57 AM JT-Cave: making it so any button can be led and have a hal pin so led labels can be sync
10:58 AM Tom_L: nice
10:58 AM Tom_L: connect the hal pin in the postgui?
11:01 AM JT-Shop: yup
11:01 AM * JT-Shop heads to the Midway Bar and Grill
11:02 AM Tom_L: so you'd create a hal pin for the label too?
11:02 AM Tom_L: input
11:21 AM lcnc-relay: <roguish> good morning
12:29 PM Tom_L: ebe_lyn, i am currently working on those
12:29 PM Tom_L: fixed the UI file and plan to test the routines soon
01:43 PM lcnc-relay: <allaboutjake@> Got motors turning with Flex GUI. Looking forward to experimenting with it. @Tom_L I do see the keyboard jog issue, installed from deb-- "dpkg -l" says:
01:43 PM lcnc-relay: <allaboutjake@> "ii flexgui 1.3.0 amd64 LinuxCNC GUI"
01:43 PM Tom_L: i'm on 1.3.1
01:45 PM Tom_L: https://github.com/jethornton/flexgui/releases
01:46 PM Tom_L: https://forum.linuxcnc.org/41-guis/57372-flex-gui-keyboard-jog-error
02:02 PM lcnc-relay: <allaboutjake@> will install from the releases directly
02:03 PM lcnc-relay: <allaboutjake@> I have probe in my existing hal as "net probe-in <= hm2_7i96s.0.inm.00.input-03-not" Do I need to pipe this signal into "flexhal.probe_trigger_pin"?
02:09 PM lcnc-relay: <allaboutjake@> I can still get the jog to stick with the keyboard when hitting mutliple arrows (Up and left, for example) in rapid succession or at the same time.
02:10 PM Tom_L: yes on the probe pin
02:10 PM Tom_L: one sec
02:25 PM Tom_L: map it to motion.probe.input if you're running live
02:25 PM Tom_L: the sim has a probe button for 'sim'
02:25 PM Tom_L: allows testing at least
02:25 PM Tom_dev: my probe is: net touch-probe hm2_[HOSTMOT2](BOARD).0.gpio.028.in_not => motion.probe-input
02:25 PM lcnc-relay: <allaboutjake@> ahh ok don't need it t all then in the postgui hal becuse its mapped already in the main hal. I think the simultor button was overwriting the connection
02:25 PM Tom_L: maybe
02:25 PM Tom_L: you can dump that button if you're running live but it's never interfered here
02:31 PM lcnc-relay: <allaboutjake@> yeah that seems to fix the probe opertions tab thanks. Just need to figure out the label error next. This is going so much more stress free than Probe Basic 🙂
02:31 PM Tom_L: need help just holler
02:31 PM Tom_L: what label error?
02:33 PM Tom_L: you will want to change the probe settings page for your own setup AND add all the #2000 numbered vars to your .var file just above the #5000 ones
02:33 PM Tom_L: if they are not there, they won't be saved on exit
02:33 PM Tom_L: odd quirk about linuxcnc
02:34 PM lcnc-relay: <allaboutjake@> https://jauriarts.org/_heisenbridge/media/jauriarts.org/ABBshyzgettOPovCypokQJNI/aUyKIJauVSw/image.png
02:35 PM Tom_L: https://www.gnipsel.com/linuxcnc/flexgui/labels.html#velocity-labels
02:35 PM Tom_L: see if that helps
02:35 PM Tom_L: Name the label two_vel_lb and add two int type Dynamic Properties called joint_0 and joint_1 and set the values to the perpendicular joint numbers you want to calculate.
02:36 PM lcnc-relay: <allaboutjake@> you know no that you post this I may not have copied the .ui file from the examples after going from 1.3->1.3.1
02:36 PM Tom_L: copy my personal one
02:36 PM Tom_L: it's most up to date
02:37 PM Tom_L: http://tom-itx.no-ip.biz:443/~webpage/cnc/JT-SHOP/flexgui/uploads/
02:37 PM Tom_L: i added some led labels may be the only difference
02:37 PM Tom_L: testing flood / mist coolant
02:40 PM lcnc-relay: <allaboutjake@> no error
02:40 PM Tom_dev: http://tom-itx.no-ip.biz:443/~webpage/cnc/JT-SHOP/flexgui/2velocitylabel.png
02:41 PM Tom_dev: the one below it would have 3
02:41 PM lcnc-relay: <allaboutjake@> odd behavior with probe_enable though I can enabel the probe just fine but after an estop activation and clear the probe_enable button doesn't work anymore...
02:41 PM lcnc-relay: <allaboutjake@> but I'
02:43 PM Tom_dev: estop kills the power button too
02:43 PM lcnc-relay: <allaboutjake@> home seems to bring it bcak
02:43 PM Tom_L: the probe enable is a safety thing
02:44 PM Tom_L: you'll find some buttons disabled with it on
02:44 PM Tom_L: jog etc
02:47 PM Tom_L: gonna go try and test some metric routines...
02:47 PM Tom_L: afk
02:47 PM lcnc-relay: <allaboutjake@> Going to have to take break for today. The doggo wants dinner and a walk.
02:48 PM lcnc-relay: <allaboutjake@> thanks for the help
02:48 PM Tom_L: np
02:48 PM Tom_L: you can customize this any way you like
02:52 PM lcnc-relay: <allaboutjake@> yeah UI. I don't have a controllable spindle, for example, may tke those settings out... though I'm getting task 0 delay errors too, now I see... so thats concerning. More to do! 🙂
03:31 PM rdtsc-wtf is now known as rdtsc
06:02 PM Tom_L: ebe_lyn, i have finished testing the metric routines and will add them to the example soon
06:21 PM Tom_L: i have not finished testing the ops yet so they will not be included in the subdirectory
06:37 PM lcnc-relay: <allaboutjake@> Trying to understand some of the behavior I was getting earlier today. When probing left (p_left.ngc), after finding the edge it goes to "Z #2036". Which is safe-Z. But the default LinuxCNC.var has #2036 set to -2.25. This cause the tip to plunge down towards the table, not what I think I was expecting/want
06:51 PM Tom_L: JT-Shop, PR #381 metric probe routines
06:52 PM Tom_L: allaboutjake, that is the distance from machine z zero down, make that number smaller
06:52 PM Tom_L: it avoids going to machine Z zero after the probe
06:54 PM Tom_L: i included that for a machine that may have alot of Z travel so they can limit the retract
06:54 PM Tom_L: obviously mine is different than yours
06:57 PM Tom_L: if you'd prefer edit your edge files to G53 G0 Z0 and ignore the variable
06:58 PM Tom_L: line 12 of each of those 4 files
07:02 PM Tom_L: actually that appears in 12 of the files
07:02 PM Tom_L: the others are the corner probe routines
07:03 PM Tom_L: line 15
07:04 PM Tom_L: i lowered it during testing on my cnc to make sure it was accounting for the probe radius
07:08 PM Tom_L: allaboutjake@, if you feel it safer to move to G53 Z0 i'm open to suggestions and not opposed to chaning those files
07:09 PM Tom_L: and eliminate that var
07:39 PM lcnc-relay: <allaboutjake@> Well I figured it would be a Z positive move of some sort after the edge is found. Not a Z negative one. 2.5 magnitude might be more than I want but was was puzzling is that it went negative rather than up to some clearance plane as I expected
07:41 PM Tom_L: i changed those var files and set that to zero
07:41 PM Tom_L: i'm not sure how my local var made it into the examples
07:42 PM Tom_L: you were probing above that then
07:43 PM Tom_L: i'm sure it included a pucker factor