#linuxcnc-devel | Logs for 2014-12-09

[00:03:59] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 66c4b2b 06linuxcnc 10VERSION 10debian/changelog LinuxCNC 2.6.5 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=66c4b2b
[00:03:59] <KGB-linuxcnc> 03Sebastian Kuzminsky 05signed tags c6ff74f 06linuxcnc 03v2.6.5 LinuxCNC v2.6.5 (tagged commit: 66c4b2b) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c6ff74f
[00:04:13] <cradek> woo
[00:07:57] <seb_kuzminsky> wee
[00:43:21] <seb_kuzminsky> holy crackamole, git-imerge is rocking my world tonight
[00:50:39] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 37f73ab 06linuxcnc 10(9 files in 7 dirs) Merge origin/2.6 into 2.7 (using imerge) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=37f73ab
[00:51:17] <seb_kuzminsky> i <3 pairwise conflict resolutions
[00:54:36] <KimK> Hi, Seb, I watched the imerge video you posted earlier, imerge looks very useful, glad it's working out for you (and us!)
[09:14:28] ChanServ changed topic of #linuxcnc-devel to: http://linuxcnc.org | Latest release: 2.6.5 | (this channel is logged by the zlog robot)
[09:15:01] <skunkworks> Yay!
[09:15:17] <seb_kuzminsky> :-)
[09:15:28] <skunkworks> Great work!
[09:15:40] <seb_kuzminsky> thanks :-)
[09:15:46] <cradek> yay!
[09:17:01] <seb_kuzminsky> skunkworks: thanks for talking with Karl Jacobs on the list even though he's using mk, and especially thanks for verifying that G95 works right in master
[09:17:48] <seb_kuzminsky> we have a g76 test, but not g95 or g33
[09:17:54] <skunkworks> No problem. That is something that I didn't test with the new TP... Is G95 tested in the run tests?
[09:17:56] <skunkworks> ah
[09:17:57] <skunkworks> ok
[09:18:21] <skunkworks> there was a bug in G33 that was fixed quite a while ago.
[09:18:31] <seb_kuzminsky> in rob ellenberg's tp?
[09:18:35] <skunkworks> yes
[09:18:57] <skunkworks> (it would cap the treading velocity to the last F word..
[09:19:12] <seb_kuzminsky> 6 test only tests the interp part, not the actual motion
[09:19:26] <seb_kuzminsky> err... our g76 test does that
[09:19:58] <seb_kuzminsky> one of these days we really need to write a bunch of motion tests
[09:32:06] <skunkworks> How would you do that?
[09:32:21] <skunkworks> wouldn't you need real realtime?
[09:33:22] <cradek> it's hard, but no you probably don't need realtime
[09:35:00] <cradek> at fest we were talking about the ability to single-step hal threads. it seems like you could use that to test motion, but it would be tricky
[09:41:31] <seb_kuzminsky> a while back i started tests/motion/g0, but abandoned it before i got it working
[09:41:53] <seb_kuzminsky> it uses halsampler to record the output of motion in realtime, then validates it in non-realtime
[09:42:05] <cradek> ah
[09:44:17] <seb_kuzminsky> in 2011 i made commit that says "I'm disabling this until I figure out why it's failing on the buildbot..."
[09:44:20] <seb_kuzminsky> :-/
[09:45:37] <seb_kuzminsky> heh, it doesn't run any more because: test.sh: line 6: emc: command not found
[09:46:39] <seb_kuzminsky> other than that it runs and passes
[09:46:56] <seb_kuzminsky> it'd probably be pretty easy to get it to work reliably
[09:47:13] <seb_kuzminsky> it needs to be changed to not use linuxcncrsh any more
[09:47:27] <seb_kuzminsky> then it might serve as a foundation to build more complicated/interesting motion tests on
[09:47:30] <seb_kuzminsky> another day...
[09:49:55] <jepler> seb_kuzminsky: thank you for the 2.6 release
[09:50:31] <seb_kuzminsky> yep! thanks for the comp & s-o-b fixes :-)
[09:52:38] <seb_kuzminsky> huh, look at that, i did start to reanimate the motion tests, about a year ago...
[10:04:04] <seb_kuzminsky> uh oh, i think i broke kgb
[10:04:07] * seb_kuzminsky runs away
[10:04:31] <cradek> really? it was working last night
[10:20:25] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests b97534b 06linuxcnc 04tests/motion/g0/skip un-skip the g0 motion test * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b97534b
[10:20:25] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests 8c8c795 06linuxcnc 10tests/motion/g0/test.sh tests: sweep some of the dust & cobwebs out of the g0 test * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=8c8c795
[10:20:25] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests b3e2fd2 06linuxcnc 10tests/motion/g0/core_sim.hal try to work around period inexactness on realtime systems (thank jepler!) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b3e2fd2
[10:20:27] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests 0e009e8 06linuxcnc 10(5 files) remove the motion/g0 test * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0e009e8
[10:20:31] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests cbc9497 06linuxcnc 03lib/python/linuxcnccontrol.py make LinuxcncControl a reusable python library * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=cbc9497
[10:20:35] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests a275b85 06linuxcnc 10tests/halui/jogging/test-ui.py 10tests/toolchanger/m61/test-ui.py tests: use the new LinuxcncControl python module * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=a275b85
[10:20:39] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests 316a838 06linuxcnc 10(9 files in 2 dirs) tests: add a motion test of the X-axis g0 vel/acc profile * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=316a838
[10:20:43] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests 0d5e7df 06linuxcnc 10(5 files) tests: add a motion test of jogging the X axis * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0d5e7df
[10:20:47] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests 9dabfb3 06linuxcnc 10tests/motion/README README * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9dabfb3
[10:20:50] <KGB-linuxcnc> 03Sebastian Kuzminsky 05motion-tests b38c0ec 06linuxcnc 10tests/motion/jog/test-ui.py 10tests/motion/mdi/g0/test-ui.py handle realtime errors better * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b38c0ec
[10:29:56] <seb_kuzminsky> cradek: oh, it just took a really long time to accept my push
[15:01:48] <bjmorel_work> pcw: For the ethernet cards like the 7176E, can you put in another nic card and have internet on the same machine, and does that affect the performance?
[15:02:02] <cradek> yes, no
[15:05:27] <skunkworks> bjmorel_work, I have tried quite a few ways.. (dual nic motherboards, onboard nic + pci nic, onboard nic + usb wireless...) all work well - although you want to test to see which nic works better of the two. (I have had better luck with realtime ethernet using the onboard nic vs pci)
[15:07:53] <bjmorel_work> Great, thanks.
[15:14:14] <jepler> wow this is terrible: pay $9000 for a laser engraver, then pay by the hour after you've used it for 10 hours "similar to pre-paid mobile phones" http://www.rayjetlaser.com/en-US/Product_Locator/Documents/The_clever_choice.pdf
[15:19:05] <pcw_home> bjmorel_work I have a USB --> Ethernet adapter for internet on my test machine, works fine
[15:19:12] <mozmck> it's plug and (p)ray
[15:19:33] <pcw_home> Is that in the fine print?
[15:20:12] <mozmck> that does not sound like a really good price to start with!
[15:38:31] <pcw_home> One think I have learned with Preemt-RT Ethernet: CPU speed and cache size are important
[15:39:42] <pcw_home> Best performing machine I have now is my old core duo desktop upgraded to a E8500 CPU (dual core 6M cache)
[15:41:01] <pcw_home> other trick is setting Ethernet IRQ priority to 99
[15:49:21] <pcw_home> I had no idea these were so cheap, wonder if it is actually usable:
[15:49:22] <pcw_home> http://www.ebay.com/itm/High-Precise-40W-CO2-Laser-Engraving-Cutting-Machine-Engraver-Cutter-USB-Port-/321430750448?pt=LH_DefaultDomain_0&hash=item4ad6c404f0
[16:45:01] <Tom_itx> those are 40 chinese watts
[16:46:05] <Tom_itx> if i can't toaster oven those smps parts this board ain't gonna fly no more
[16:46:08] <Tom_itx> i'm not sure i cna
[16:46:10] <Tom_itx> can*
[16:46:35] <Tom_itx> pcw_home, is there a way to test the 7i47 without a host board?
[17:11:21] <skunkworks> does joint_axis6 have the new TP?
[17:12:46] <micges> yes
[17:12:51] <skunkworks> awesome
[17:12:58] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/moveoff 4f598a9 06linuxcnc 10(16 files in 3 dirs) moveoff: clarify pin names for other use cases * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4f598a9
[17:46:28] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/moveoff_squashed 19a0e38 06linuxcnc 10(30 files in 7 dirs) moveoff: Hal-only offsetting * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=19a0e38
[17:50:31] <seb_kuzminsky> haw yea
[17:53:04] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/moveoff_squashed 0ade81d 06linuxcnc 10(30 files in 7 dirs) moveoff: Hal-only offsetting * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0ade81d
[22:49:36] <KGB-linuxcnc> 03Dewey Garrett 052.7 73f195a 06linuxcnc 10docs/src/config/ini_config.txt 10src/libnml/inifile/inifile.cc 10src/libnml/inifile/inifile.hh inifile.cc: support extending lines with backslash * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=73f195a
[22:49:37] <KGB-linuxcnc> 03Dewey Garrett 052.7 d6db5ca 06linuxcnc 10scripts/sim_pin sim_pin: better default for bit types * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d6db5ca
[22:52:07] <KGB-linuxcnc> 03Sebastian Kuzminsky 05master 7118806 06linuxcnc 10debian/changelog Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7118806
[23:07:12] <KGB-linuxcnc> 03Dewey Garrett 052.7 6b5437d 06linuxcnc 10(30 files in 7 dirs) moveoff: Hal-only offsetting * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6b5437d