Back
[00:08:45] <KGB-linuxcnc> 03Michael Geszkiewicz 05micges/master/loadable-tp e418e7d 06linuxcnc 10(9 files in 9 dirs) Fix runtests after making tp loadable * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e418e7d
[11:22:31] <seb_kuzminsky> micges: neat!
[11:28:17] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 0a960a2 06linuxcnc 10VERSION 10debian/changelog LinuxCNC 2.6.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0a960a2
[11:28:18] <KGB-linuxcnc> 03Sebastian Kuzminsky 05signed tags 9b33147 06linuxcnc 03v2.6.7 LinuxCNC 2.6.7 (tagged commit: 0a960a2) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9b33147
[11:33:47] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 7341c82 06linuxcnc 10debian/changelog 10share/axis/tcl/axis.tcl Merge remote-tracking branch 'origin/2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7341c82
[11:33:47] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 c288583 06linuxcnc 10VERSION 10debian/changelog LinuxCNC 2.7.0~pre4 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c288583
[11:33:47] <KGB-linuxcnc> 03Sebastian Kuzminsky 05signed tags 36e1d71 06linuxcnc 03v2.7.0-pre4 LinuxCNC 2.7.0-pre4 (tagged commit: c288583) * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=36e1d71
[11:50:58] <seb_kuzminsky> i think Google Summer of Code wants two mentors from each project. i'm one does anyone want to be the other? *if* we get accepted, and *if* we get a student, the mentors will have to help the kids with their linuxcnc hacking project
[11:51:35] <seb_kuzminsky> show them around our code, introduce them to the rest of the devs, etc
[11:52:12] <archivist> watch them try to copy each others code
[11:52:39] <archivist> been watching in #brlcad
[11:55:12] <archivist> they had tasks set too, like find n typos
[12:26:07] <seb_kuzminsky> cool, we're piggybacking on the brlcad gsoc application this year
[12:33:17] <seb_kuzminsky> oh man, the gsoc folks really get open source, it's super nice:
http://www.google-melange.com/gsoc/document/show/gsoc_program/google/gsoc2015/help_page
[13:27:59] <kwallace_shop> seb_kuzminsky, A mentor would be helping college students with a major in computer science or engineering?
[14:09:53] <seb_kuzminsky> kwallace_shop: yep
[14:10:36] <seb_kuzminsky> oh, the other thing we need is a list of possible projects for the students to do
[14:11:28] <seb_kuzminsky> i'm going through the feature-requests and bugs on sourceforge, if anyone thinks of anything that'd be useful, please tell me here on irc or in email to the developers list
[14:43:53] <micges> seb_kuzminsky: it's first step before introducing limited jerk to public
[15:00:57] <seb_kuzminsky> micges: i figured - that's awesome
[15:01:13] <seb_kuzminsky> you might be interested in the standalone-motion project that rob ellenberg and i are working on
[15:01:38] <seb_kuzminsky> it'll let you write tests of the trajectory planner more easily than now
[15:02:07] <seb_kuzminsky> it's in the seb/2.7/sam branch (sam == standalone-motion)
[15:22:27] <zeeshan> if add a page like
[15:22:28] <zeeshan> http://wiki.linuxcnc.org/cgi-bin/wiki.pl?CentralizedLubricator
[15:22:36] <zeeshan> will it automatically show up in google searches?
[15:27:29] <Tom_itx> unless there are bot rules preventing it
[15:32:52] <seb_kuzminsky> zeeshan: yes, our wiki is searchable by google
[15:39:44] <Tom_itx> which branch has that pause-resume code in it?
[15:39:54] <Tom_itx> pause - move away - move back - resume
[15:40:09] <Tom_itx> and has that made it's way into mainstream yet?
[15:41:47] <seb_kuzminsky> it's in 2.7
[15:41:56] <seb_kuzminsky> in the hal component called 'moveoff'
[15:42:10] <zeeshan> https://github.com/jepler/linuxcnc-mirror/blob/d937bc9fb1df703fec9977808f436dd87ceb5d69/scripts/moveoff_gui
[15:42:14] <zeeshan> looks like it demonstrates it
[15:42:23] <seb_kuzminsky> here's a video demo:
https://www.youtube.com/watch?v=KY6hx7WBkO8
[15:42:38] <seb_kuzminsky> dgarr's got some other youtube videos that show different ways of using it
[15:43:27] <mozmck> is micges' limited jerk planner a based on rob's? Or something new altogether?
[15:44:31] <Tom_itx> seb_kuzminsky someone in main channel is asking about it
[15:45:17] <micges> mozmck: for now it's based on old tp
[15:45:44] <micges> seb_kuzminsky: can you point me when did you merge rob's planner?
[15:46:35] <mozmck> ok. so does that mean that at some point it could be implemented on the new tp?
[15:47:23] <mozmck> Does limited jerk help speed up execution of many small moves?
[15:47:32] <micges> surely with Rob help
[15:48:43] <seb_kuzminsky> micges: commit 309aa9785c7d996762e76a28099e9780afe84b21 was the first push of rob's circular-arc planner into 2.7
[15:48:55] <seb_kuzminsky> there have been many commits to 2.7 since then to fix bugs that were found
[15:49:27] <micges> thanks, I just want to use newest old as a base for my code
[15:50:08] <seb_kuzminsky> i'd suggest using 2.7.0~pre4 as the new base for your work
[15:50:14] <zeeshan> seb_kuzminsky: how do you guys check for bugs?
[15:50:27] <zeeshan> do you have a mailinglist for testing?
[15:50:30] <zeeshan> *is there
[15:50:43] <pcw_home> Typically, limited jerk allows higher accelerations within the same error budget
[15:52:07] <mozmck> interesting.
[15:52:08] <seb_kuzminsky> zeeshan: we have the test suite in the git repo, develoeprs run that when they change stuff, and the buildbot runs it on every push to the git.linuxcnc.org git repo
[15:52:16] <micges> mozmck: can't tell yet about any differencies, just simple test for now
[15:52:21] <zeeshan> ah
[15:52:59] <mozmck> micges: I might do some testing between the two sometime. I should be getting a plasma table set up soon.
[15:53:45] <seb_kuzminsky> zeeshan: bug reports can go to the emc-developers mailing list, or to the bug tracker on source forge (
http://sf.net/p/emc/bugs)
[15:54:09] <micges> mozmck: surely, here tests are also on fast plasma table
[15:54:09] <pcw_home> One way of looking at limited jerk is that in the frequency domain, the high frequencies that excite vibrations are reduced
[15:54:49] <micges> yes they are, 0.5g were barely noticeable from machine sound
[15:56:08] <pcw_home> In the time domain you dont have steps in force
[15:56:49] <mozmck> micges: that is interesting - how does your limited jerk compare to the new tp?
[15:57:14] <micges> mozmck: velocity scope from limited jerk:
http://ibin.co/1sKDvNiWAGYO
[15:57:27] <micges> mozmck: in what way?
[15:58:48] <mozmck> oh, speed, smoothness, etc
[16:00:46] <micges> mozmck: I'll get to it, code just runs after few months, more than few rewrites, many nights and I'm just making branch with working code for starting tests (mine and others)
[16:02:03] <mozmck> I see, well, I should be able to help with testing soon.
[16:12:36] <micges> sure
[16:12:49] <seb_kuzminsky> that's cool micges, i look forward to testing it
[16:14:14] ChanServ changed topic of
#linuxcnc-devel to:
http://linuxcnc.org | Latest release: 2.6.7 | (this channel is logged by the zlog robot)
[16:15:45] <micges> seb_kuzminsky: are you sure this time it's fixed?
[16:21:36] <seb_kuzminsky> pretty sure :-)
[16:21:53] <seb_kuzminsky> the guy who reported the problem tested the fix and said it works for him now
[16:22:13] <micges> cool
[16:23:01] <seb_kuzminsky> also i ran 'git grep' to look for instances of the bug, from the top level of the git repo this time instead of just from src/, and there are none left
[16:38:40] <KGB-linuxcnc> 03Sebastian Kuzminsky 05master 516e754 06linuxcnc 10debian/changelog Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=516e754
[18:06:46] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/histo 5dfd15d 06linuxcnc 03lib/hallib/hal_procs_lib.tcl 10lib/hallib/hookup_moveoff.tcl 10lib/hallib/xhc-hb04.tcl 10scripts/moveoff_gui hal_procs_lib.tcl: consolidate common procs * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5dfd15d
[18:06:46] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/histo 5c427cd 06linuxcnc 10src/hal/utils/halcompile.g halcompile.g: provide rtapi_math64.h * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5c427cd
[18:06:46] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/histo 9ed446d 06linuxcnc 10docs/src/config/images/latency-histogram.png 10scripts/latency-histogram 10src/hal/components/latencybins.comp latency-histogram: include min,max,stddev * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9ed446d
[18:06:49] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/histo 8e62cdc 06linuxcnc 10(13 files in 7 dirs) hal-histogram: histogram util for s32,float pins * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=8e62cdc
[18:46:03] <robinsz> hi guys, where can I browse the 2.7 source? I need to have a look at hal moveoff examples
[18:51:19] <andypugh> git.linuxcnc.org
[19:11:07] <seb_kuzminsky> robinsz: if you have 2.7.0~pre3 or newer installed, it's also available in the sample configs
[19:11:30] <seb_kuzminsky> sim/axis/moveoff
[19:11:32] <robinsz> found it I think
[19:11:35] <seb_kuzminsky> cool
[19:11:43] <seb_kuzminsky> let us know how they work for you!
[19:11:57] <robinsz> trying to figure out how to set it up for a router
[19:12:04] <atom1> so it is implemented as a comp?
[19:12:16] <robinsz> it looks like the no_gui thing will doit
[19:12:28] <robinsz> anyway, I'll let you know if I get it to go :)
[19:12:39] <seb_kuzminsky> atom1: yes, it's a hal component that sits between Motion and the motors, and inserts offsets to the motors
[19:13:16] <atom1> one think i plan to try once i get this darn machine back together
[19:13:21] <atom1> getting closer...
[19:14:12] <seb_kuzminsky> cool
[19:14:29] <atom1> got spindle control set up last night finally
[19:14:47] <atom1> now trying to link the spindle index to motion
[19:15:02] <atom1> that looks like a one liner
[19:25:40] <atom1> what's tclx?
[22:13:47] <KGB-linuxcnc> 03Dewey Garrett 052.7 b8ab242 06linuxcnc 10(5 files in 3 dirs) moveoff: fixed-offset, single-axis retract demo * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b8ab242