#linuxcnc-devel | Logs for 2014-11-03

[08:34:37] <KGB-linuxcnc> 03Sebastian Kuzminsky 05liblinuxcnc-ui 01359d0 06linuxcnc 77 commits pushed, 1043 files changed, 031298(+), 041963(-)
[08:37:44] <seb_kuzminsky> jepler: that's lui rebased onto master, with some reorg of commits & hunks to make it simpler, & to make each commit build & pass tests
[09:04:55] <jepler> seb_kuzminsky: thank you!
[09:05:32] <jepler> I got about 1/3 of the way through converting python to use lui getters in the implementation of linuxcnc.stat .. but it's not done yet
[09:05:38] <jepler> and I don't think there's much point of a partial conversion
[09:06:57] <seb_kuzminsky> cool
[09:07:14] <seb_kuzminsky> i'm picking lui back up again after some release distractions
[09:07:18] <jepler> making every intermediate commit pass tests, that takes dedication and a little bit of ocd
[09:07:50] <seb_kuzminsky> i'll try to get a couple more setters (nml commands) converted this week
[09:08:16] <jepler> > The chip has three different inputs for controlling its power states: /RESET, /SLEEP, and /ENBL.
[09:08:19] <jepler> boy that should be about enough
[09:09:15] <seb_kuzminsky> well, off to work for me, laters
[09:09:28] <jepler> see ya
[13:31:30] <kwallace4> Has anyone tried using a pair of PWM input brushed servo amps to drive a stepper motor? One of my stepper amps is still broken and I'm thinking I could use a pair of my Pico amps with a HAL two-phase brushless component to take over.
[13:39:36] <jepler> I don't think you'll get a very good step rate since the pico amps only get a new PWM command once every servo cycle. 1 step per servo period is way too little to be useful
[13:41:12] <cradek> you'd be time ahead to just buy geckos or parkers wouldn't you?
[13:42:24] <cradek> er, forget it - I bet you know that already
[14:07:18] <kwallace4> jepler, cradek, I just needed someone to slap me in the head and back to reality. It seemed it might be a way to get software micro stepping, which might be interesting. I've blown out a bunch of transistors and don't want to carve out the time and effort to figure out why. It seems to happen when I hit E-Stop to power up the machine. I think I'll try connecting the driver enable to the Power button so the power up and enable happen at diffe
[14:45:11] <cradek> worst bug report ever: http://article.gmane.org/gmane.linux.distributions.emc.user/54201
[14:45:25] <cradek> (looks fine to me)
[14:46:30] <CaptHindsight> linuxcnc.org works fine for me but somebody just posted it being hacked and in Russian on ML, they may have just clicked the Russian link on the bottom of the page
[14:47:05] <jepler> sometimes the website switches me to the wrong language too
[14:47:25] <skunkworks> I have had that happen... But it is an easy fix :)
[14:48:17] <CaptHindsight> often the search results for me are in French
[14:49:54] <jepler> you know, that's probably exactly what it is
[14:50:10] <jepler> a search can take you to a language-specific URL; I bet that sticks
[14:50:36] <jepler> yup.
[14:51:21] <cradek> obnoxious
[14:51:48] <jepler> http defines a way for the user agent to state a preferred language. hardly any websites use it.
[14:55:34] <jepler> hm somebody with an ounce of sense could write a robots file that causes google not to index any URLs with /language/ tags in them
[15:13:05] <jepler> I took the linuxcnc version number poll off the front page
[15:17:39] <cradek> thanks
[15:36:24] <jepler> and with any luck, google will also drop from its indexes URLs like /russian/foo due to a change in robots.txt.
[16:20:55] <jepler> nice, there's now a US reseller of hardkernel products. next time I need one it'll be nice to avoid the expensive international shipping. http://ameridroid.com/t/single-board-computers
[16:24:07] <Tom_itx> any particular one suited for lcnc?
[16:26:41] <jepler> Tom_itx: I have been using odroid u3 as a development platform for linuxcnc. it's the only one I have much experience with, so I stop short of making any recommendations.
[16:27:22] <Tom_itx> the $70 one?
[16:27:42] <jepler> it's far short of turn-key for linuxcnc, it only works with one hardware interface -- and that one requires a board you can't buy anywhere (I designed it and had a few made at osh park), and I have no idea if there'll be a supply of them in a year
[16:28:06] <Tom_itx> oh is that the one you've been working with lately?
[16:28:08] <jepler> yeah
[16:28:40] <jepler> and yes, the $70 one (plus power supply and e-mmc or micro sd for storage, unlike many arm products it has no onboard flash for an OS)
[16:36:47] <KGB-linuxcnc> 03Jeff Epler 052.6 9b4294a 06linuxcnc 10docs/src/code/Contributing-to-LinuxCNC.txt docs: document SOB procedure * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9b4294a
[16:39:16] <KGB-linuxcnc> 03Jeff Epler 052.7 b999008 06linuxcnc Merge remote-tracking branch 'origin/2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b999008
[16:41:01] <KGB-linuxcnc> 03Jeff Epler 05master 246fa4e 06linuxcnc Merge branch '2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=246fa4e
[16:54:15] <memfrob> I thought preempt_rt was dead after about 8 or 10 weeks of no development, then 3 days ago about a hundred commits hit preempt_rt
[17:18:37] <jepler> memfrob: that's one style of development!
[17:19:35] <jepler> I've been meaning to try the debian "armmp" kernel to see if it works on my arm board (mp = multiplatform, uses devicetree I guess)
[17:19:44] <seb_kuzminsky> memfrob: http://lwn.net/Articles/604695/
[17:21:55] <jepler> and more recently http://lwn.net/Articles/617140/
[17:22:34] <seb_kuzminsky> ah, that was the one i was looking for, thanks
[18:33:34] <linuxcnc-build_> build #1810 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/1810 blamelist: Jeff Epler <jepler@unpythonic.net>
[18:40:08] <seb_kuzminsky> huh, the tlo test failed
[18:40:52] <seb_kuzminsky> heh
[18:41:20] <seb_kuzminsky> it ran an invalid gcode and looked for the expected error, but instead got "Unexpected realtime delay: check dmesg for details" and got scared
[18:42:29] <seb_kuzminsky> i think this is why halcmd has setexact_for_test_suite_only...
[18:45:11] <seb_kuzminsky> wfm
[18:49:53] <linuxcnc-build_> build #2620 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2620 blamelist: Jeff Epler <jepler@unpythonic.net>
[22:12:10] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 41276aa 06linuxcnc 10tests/tlo/core_sim.hal tests: fix a spurious false failure in the tlo test * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=41276aa
[22:12:11] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 687d493 06linuxcnc 10(17 files in 3 dirs) tests: reorganize the halui jogging test dir layout * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=687d493
[22:12:11] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 5efacc6 06linuxcnc 10(6 files) halui test: add an mdi test * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5efacc6
[22:13:06] <seb_kuzminsky> that should fix the test failure, and add a test i want for lui development
[22:13:09] <seb_kuzminsky> bbl
[22:46:34] <linuxcnc-build_> build #2611 of 1200.rip-lucid-i386 is complete: Failure [4failed compile runtests] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/1200.rip-lucid-i386/builds/2611 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[22:51:23] <linuxcnc-build_> build #2622 of 0000.checkin is complete: Failure [4failed] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2622 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[23:04:05] <aluminum> we just installed machinekit, is there a quick start guide we can look at to get started?