#linuxcnc Logs

May 24 2024

01:06 AM Deejay: moin
02:14 AM DmanT: moin
02:24 AM lcnc-relay: <travis_farmer> Morning
03:13 AM lcnc-relay: <TurBoss> morning
04:10 AM Tom_L: morning
04:16 AM Bushman: morning
04:17 AM Bushman: this is awkward... after update i get this error
04:17 AM Bushman: /usr/share/linuxcnc/hallib/axis_manualtoolchange.hal:1: execv(hal_manualtoolchange): No such file or directory
04:18 AM Bushman: i've seen it broken in many other ways on my system but not this way.
04:18 AM Bushman: any ideas?
04:21 AM Tom_L: my hal_manualtoolchange file has this at the top of the file:
04:21 AM Tom_L: # File saved in /usr/bin with chmod set to 755
04:21 AM Tom_L: # Modded for external tool-change pin while using hal_manualtoolchange dialog
04:21 AM Tom_L: beyond that no idea really
04:21 AM Bushman: hmm...
04:22 AM * Bushman checks the permisions
04:22 AM Tom_L: and i have a copy of that in my config dir
04:22 AM Tom_L: http://tom-itx.no-ip.biz:81/~webpage/cnc/configs/NEW_MILL_SHOP_2.8.4/hal_manualtoolchange
04:23 AM Tom_L: i should rename that to 2.9.2
04:23 AM Bushman: yea, nah. the file exists and has xrxrx- permisions
04:23 AM Bushman: not sure what's the problem
04:24 AM Bushman: but that's how it is when i'm installing it on my everyday OS, as a sim
04:28 AM Tom_L: did the path somehow get messed up?
04:31 AM Bushman: i'm looking at $PATH but can't see anything obviously wrong. the /usr/bin is there
04:32 AM Bushman: everything else works so path must be ok.
04:35 AM Bushman: all my usual suspects are off the table too (tcl lib path in AUR build files was broken in the past but it seems fixed)
04:36 AM Bushman: oh well, i guess i have to go to the shop and check the gcode on the actual machine then :P
04:36 AM Bushman: bummer
06:08 AM -!- #linuxcnc mode set to +v by ChanServ
06:14 AM JT-Mobile: morning
11:28 AM Unterhaus_ is now known as unterhausen
11:28 AM unterhausen: my bender can supposedly bend a 12" radius, but I'm confused about how to do it
11:30 AM unterhausen: it's a flat bar, so collapsing isn't possible
11:30 AM lcnc-relay: <roguish> no problem then, good luch
11:30 AM lcnc-relay: <roguish> luck
11:30 AM unterhausen: Cap'n, that video shows him bending some tiny little thing.
11:31 AM unterhausen: The diameter of my table is only 9"
11:32 AM unterhausen: Maybe I can make an extender, since the forces aren't going to be that big
11:42 AM unterhausen: there is a channel that's called "nothing too strong ever broke"
11:43 AM unterhausen: seems like a challenge
11:47 AM lcnc-relay: <roguish> unterhausen: check this. Tom Lipton started that expression years ago. He worked at a place called CVM, the at Berkeley National Lab. very nice guy. I know him
11:48 AM lcnc-relay: <roguish> https://www.youtube.com/@oxtoolco
11:49 AM unterhausen: I'm subscribed to Oxtool channel
11:49 AM unterhausen: he was smart enough to say that nothing too strong ever broke on his channel
11:59 AM unterhausen: I need his wedge jack
11:59 AM unterhausen: too poor for a nose jack
12:00 PM unterhausen: someone on youtube has done what I want with a bender, but they have it hidden under a bad title
12:07 PM unterhausen: I"m too poor for this bender, just the forming nose costs $300
03:53 PM JT-Cave: home sweet home
03:53 PM XXCoder: woot
03:53 PM XXCoder: nothing like home
03:56 PM Tom_L: how was the drive?
03:59 PM JT-Cave: not bad, better once we passed Little Rock
04:06 PM -!- #linuxcnc mode set to +v by ChanServ
07:33 PM -!- #linuxcnc mode set to +v by ChanServ
07:34 PM Tom_dev: JT-Cave, lib/python/rs274/glcanon.py class GlCanonDraw:
07:35 PM Tom_dev: src/emc/usr_intf/axis/scripts/axis.py: GlCanonDraw.__init__(self, s, None)
07:36 PM Tom_dev: lib/python/rs274/glcanon.py:class GlCanonDraw:
07:36 PM Tom_dev: share/qtvcp/screens/qtplasmac/qtplasmac_handler.py:from rs274.glcanon import GlCanonDraw as DRAW
07:36 PM Tom_dev: maybe not so much that one
07:37 PM Tom_dev: src/emc/usr_intf/axis/scripts/axis.py:from rs274.glcanon import GLCanon, GlCanonDraw
07:39 PM Tom_dev: axis.py ln 401
07:40 PM Tom_dev: 701
07:47 PM Leeloo: I am doing some incremental testing, right now just my Y axis - how can I convince LinuxCNC that all axis are homed without actually homing anything?
07:49 PM Tom_L: NO_FORCE_HOMING = 1 allows the user to make MDI moves and run programs without homing the machine first.
07:49 PM Tom_L: http://linuxcnc.org/docs/stable/html/config/ini-config.html
07:50 PM Tom_L: [TRAJ] section iirc
07:50 PM Leeloo: Oh, great - thank you
10:13 PM Leeloo: Hmm, why would I get joint errors at slow speeds, say anything more than 10mm/s with a Mesa card? I checked latency and I think it's fine below 10 μs seems like. It's a stepper motor and no pid loop.
10:17 PM Leeloo: It's a joint following error specifically, but not sure how can I debug it - since there's no feedback from the stepper, I am not sure what is LinuxCNC not liking exactly.
10:19 PM Tom_L: https://forum.linuxcnc.org/38-general-linuxcnc-questions/36464-joint-following-error-on-stepper-system
10:19 PM Tom_L: i had them on my mill once in a while and slowing accel down fixed mine
10:26 PM Tom_L: also my maxvel is set to 0
10:26 PM Tom_L: http://linuxcnc.org/docs/stable/html/drivers/hostmot2.html
10:26 PM Tom_L: for further reading on the topic
10:27 PM Tom_L: 13.2
10:28 PM Tom_L: make sure the steplen stepspace are set right for your driver too
10:29 PM Tom_L: http://wiki.linuxcnc.org/cgi-bin/wiki.pl?Stepper_Drive_Timing
10:31 PM Tom_L: then work with maxvel maxaccel