#linuxcnc-devel | Logs for 2016-09-25

[04:51:47] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #166: For override there is a pin halui.spindle-override.scale adding such a pin should solve that FIXME.... 02https://github.com/LinuxCNC/linuxcnc/issues/166#issuecomment-249411729
[04:57:30] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #130: Still the same behavior! 02https://github.com/LinuxCNC/linuxcnc/issues/130#issuecomment-249411918
[06:42:00] <Roguish> 1
[06:44:02] <jthornton> morning
[07:29:07] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15andypugh commented on issue #166: Norbert: That's a different thing. This request is for a change in the step-size when the + / - buttons are used in GUIs which use the "spindle increase" and "spindle decrease" commands. ... 02https://github.com/LinuxCNC/linuxcnc/issues/166#issuecomment-249417754
[10:03:37] <Roguish_> jepler: Ok, been working at being a developer. so, I added a simple print statement in the emc_interface.py file in the lib/touchy directory. might be a bit crude but it works.
[10:03:38] <Roguish_> if self.emcstat.kinematics_type != emc.KINEMATICS_IDENTITY:
[10:03:40] <Roguish_> print("kinematics_type = %s" % str(self.emcstat.kinematics_type))
[10:03:41] <Roguish_> print("KINEMATICS_IDENTITY = %s" % str(emc.KINEMATICS_IDENTITY))
[10:03:43] <Roguish_> raise SystemExit, "\n*** emc_control: Only KINEMATICS_IDENTITY is supported :lib T\n"
[10:03:44] <Roguish_> so, run linuxcnc and get:
[10:03:46] <Roguish_> kinematics_type = 0
[10:03:47] <Roguish_> KINEMATICS_IDENTITY = 1
[10:03:49] <Roguish_> *** emc_control: Only KINEMATICS_IDENTITY is supported :lib T
[10:05:10] <Roguish_> from the .ini file:
[10:05:12] <Roguish_> [KINS]
[10:05:13] <Roguish_> KINEMATICS = trivkins
[10:05:15] <Roguish_> JOINTS = 6
[10:05:16] <Roguish_> [TRAJ]
[10:05:18] <Roguish_> AXES = 6
[10:05:19] <Roguish_> COORDINATES = X Y Z A B C
[10:05:21] <Roguish_> from my servo.hal file:
[10:05:22] <Roguish_> loadrt trivkins kinstype=KINEMATICS_IDENTITY
[10:15:25] <Roguish_> sure seems to be some kind of out-of-synch somewhere in the code(s)
[10:31:46] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15gmoccapy commented on issue #166: IMHO it is very similar.... 02https://github.com/LinuxCNC/linuxcnc/issues/166#issuecomment-249426664
[11:11:34] <jepler> Roguish: does the problem you're having occur when you use the sample configuration file configs/sim/touchy/touchy.ini ?
[11:14:02] <KGB-linuxcnc> 03Robert W. Ellenberg 05feature/spindle-sync-jitter-fix 6d6f64f 06linuxcnc 03src/emc/tp/math_util.h 10src/emc/tp/tp.c Improve position-mode spindle sync performance with new error calculation. * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6d6f64f
[11:14:02] <KGB-linuxcnc> 03Robert W. Ellenberg 05feature/spindle-sync-jitter-fix 160e55b 06linuxcnc 10src/emc/tp/tp.c tp: new formula for spindle position tracking * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=160e55b
[11:14:02] <KGB-linuxcnc> 03Robert W. Ellenberg 05feature/spindle-sync-jitter-fix 91e85fe 06linuxcnc 03src/hal/components/quant.comp Add a simple component for quantizing a signal (useful for simulating encoders) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=91e85fe
[11:14:05] <KGB-linuxcnc> 03Robert W. Ellenberg 05feature/spindle-sync-jitter-fix e2e0731 06linuxcnc 10lib/hallib/sim_spindle_encoder.hal First cut of spindle encoder simulation with quantized output * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e2e0731
[11:14:09] <KGB-linuxcnc> 03Robert W. Ellenberg 05feature/spindle-sync-jitter-fix 64092ba 06linuxcnc 10src/emc/tp/tp.c Comment cleanups * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=64092ba
[11:15:52] <Roguish_> jepler: no. just ran from a command line and was sure to use the touchy.ini in linuxcnc/configs/sim/touchy
[11:17:05] <jepler> Roguish_: that seems to imply the problem is triggered by something in your configuration files, then
[11:17:08] <jepler> afk
[11:17:37] <Roguish_> could this be the result of some odd or unusual lines in my config?
[11:18:26] <Roguish_> yeah. i think i am going to take the sample and slowly modify it. seem reasonable?
[11:19:40] <Roguish_> took me a while to figure the correct way to add the print in the emc_interface.py. (not a coder.... yet)
[11:22:43] <Roguish_> hey, if i recompile and add the install menus to the make, what does it actually do? does it over write the menus in the current installation??
[11:25:41] <cradek> try removing that kinstype=KINEMATIC_IDENTITY
[11:26:07] <cradek> I think that should say =I which is the default
[11:29:10] <Roguish_> cradek: pretty sure i've done that, but I will do it again.
[11:31:41] <Roguish_> cradek: ok, took out the kinstype=KINEMATIC_IDENTITY, and not change. same error and output of print statements.
[12:04:29] <dgarr> kinstype=KINEMATIC_IDENTITY is not a valid parameter specifier
[12:04:38] <dgarr> kinstype=1 (the numeral one not I is valid (see $man trivkins))
[12:05:47] <dgarr> kinstype=I (the letter I) means KINEMATICS_INVERSE_ONLY -- not a commonly used ortested option
[12:06:11] <dgarr> Roguish_ it would be more efficient/productive if you would post a complete tar or zip file of your complete configuration directory
[12:25:06] <Roguish_> dgarr: thanks. i am rebuilding the config from scratch for 3 axes (though i need 6). most of the way now and it's still functioning.
[12:47:08] <cradek> oops, thanks dgarr, I misread it
[12:48:26] <cradek> I think he still should have gotten IDENTITY because K (first letter) is not any of bBfFiI1
[12:49:09] <cradek> I wonder if you have a bogus or incomplete build. might try building master from clean.
[12:55:47] <dgarr> cradek: i agree but unless all files are posted i don't know what is wrong
[12:55:57] <cradek> heh I also agree with that
[13:04:48] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15c-morley commented on issue #166: The problem is that command.c doesn't seem to have status for spindle increase increment.... 02https://github.com/LinuxCNC/linuxcnc/issues/166#issuecomment-249434794
[13:30:48] <cradek> I don't agree with the fixme
[13:30:59] <cradek> if we need increase/decrease the amount should come in the nml message from the gui
[13:31:23] <cradek> but of course we don't need it, because the gui can already look at the current speed and add or subtract from it
[13:32:14] <cradek> these increase/decrease messages are from the pre-hal days, probably invented for mshaver's bridgeport with an airmotor on the varispeed, and no speed feedback other than your eyeballs looking at the dial on the head
[13:32:38] <cradek> those bridgeports had go-faster and go-slower buttons, and you'd hold them down while watching the dial
[13:32:53] <cradek> today we'd use hal to do that, and there would not be an nml message for it at all
[13:48:54] -linuxcnc-github:#linuxcnc-devel- [13linuxcnc] 15cradek commented on issue #166: I agree with a lot of what cmorley says. I don't agree with the FIXME in the code. It makes no sense to have "increase now" come from the GUI and "how much" come from HAL.... 02https://github.com/LinuxCNC/linuxcnc/issues/166#issuecomment-249437040
[17:01:38] <andypugh> terkaa has a .diff now. I have suggesed he make a pull rquest on Github, but actually have no idea how to do that myself
[17:02:17] <andypugh> The alternative is that someone else just merges it. It’s an undocumented G-code, what could it break?
[17:02:21] <cradek> once you have it in a branch on github somewhere, there's a button
[17:02:49] <cradek> ... so it's either 1 step or 100 steps depending on what you already know how to do
[17:02:52] <cradek> have a link?
[17:03:01] <andypugh> https://forum.linuxcnc.org/forum/49-basic-configuration/31588-g84-without-spindle-feedback?start=10#80893
[17:09:11] <cradek> does it ... work?
[17:09:30] <cradek> it sure does mess with spindle-at-speed, for which we probably don't have good test coverage
[17:10:50] <cradek> needs doc work too, but that can be separate
[17:13:53] <cradek> if you have the OVERRIDES disabled before the cycle, it will erroneously reenable them
[17:15:41] <cradek> block->p_number =
[17:15:47] <cradek> block->p_number == -1.0 ? settings->cycle_p : block->p_number;
[17:15:59] <cradek> I don't understand how the P word is supposed to work (without docs)
[17:18:26] <cradek> this doesn't just call the rigid tap cycle because it's supposed to be used without spindle feedback?
[17:18:48] <andypugh> Indeed. It’s a tapping-head cycle
[17:19:16] <andypugh> And the docs don’t even mention it.
[17:19:35] <andypugh> It one of the “accepted but undocumented” ones
[17:19:50] <cradek> oh ok, for a nonreversing tapping head
[17:20:13] <cradek> darn the old ngc doc at nist.gov has bitrotted
[17:21:07] <cradek> https://www.nist.gov/customcf/get_pdf.cfm?pub_id=823374
[17:21:12] <cradek> ngc docs had g84
[17:23:09] <cradek> it didn't use P
[17:26:53] <cradek> if the author doesn't want to do the doc writing, maybe s/he could link us to a spec (I assume it's implemented how some other control does it?)
[17:27:44] <cradek> this looks at least 80% done and that's awesome. people have requested these cycles for a long time.
[17:32:41] <andypugh> he’s talking about it on the dev list too, perhaps make your observations there?
[17:34:22] <cradek> aha will do
[17:40:58] <KGB-linuxcnc> 03Dewey Garrett 052.7 9ea8327 06linuxcnc 10tcl/twopass.tcl twopass,bugfix: support all ini var substitutions * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9ea8327
[17:55:33] <cradek> andypugh: done, hope it helps him
[18:00:40] <linuxcnc-build_> build #2707 of 1405.rip-wheezy-armhf is complete: Failure [4failed garbage-collect git repo] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1405.rip-wheezy-armhf/builds/2707 blamelist: Dewey Garrett <dgarrett@panix.com>
[18:00:40] <linuxcnc-build_> build #4528 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4528 blamelist: Dewey Garrett <dgarrett@panix.com>
[18:00:41] <linuxcnc-build_> build forced [ETA 1h34m23s]
[18:00:41] <linuxcnc-build_> I'll give a shout when the build finishes
[18:50:41] <andypugh> Actually, G84 isn’t for a tapping head is it? It’s for a floating tap. A tapping head reverses when you swich Z direction.
[19:06:04] <linuxcnc-build_> Hey! build 0000.checkin #4529 is complete: Success [3build successful]
[19:06:04] <linuxcnc-build_> Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4529
[19:20:33] <linuxcnc-build_> build #851 of 4021.deb-jessie-i386 is complete: Failure [4failed shell_1] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4021.deb-jessie-i386/builds/851
[20:06:15] <cradek> andypugh: tension/compression head, not reversing
[20:15:11] <linuxcnc-build_> build #2675 of 1403.rip-wheezy-amd64 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1403.rip-wheezy-amd64/builds/2675 blamelist: John Morris <john@zultron.com>, Chris Radek <chris@timeguy.com>
[20:31:18] <linuxcnc-build_> build #4530 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/4530 blamelist: John Morris <john@zultron.com>, Chris Radek <chris@timeguy.com>
[21:06:58] <cradek> oof
[21:12:21] <jepler> a blank line vanished from an output file?
[21:13:36] <jepler> cradek: what does happen if you try to single-step in a g33 series?
[21:18:53] <Tom_L> the sync motion should still follow the spindle
[21:19:29] <Tom_L> i can turn my spindle manually and the axis will follow it
[21:20:31] <Tom_L> i would think the G33.1 would finish it's cycle before stopping though
[21:43:12] <cradek> fwiw, I didn't get that test failure here
[23:10:19] <linuxcnc-build_> build #3727 of 1301.rip-precise-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1301.rip-precise-rtai-i386/builds/3727 blamelist: John Morris <john@zultron.com>, Chris Radek <chris@timeguy.com>