Back
[01:27:31] <KGB-linuxcnc> 05arc-tests 95562ab 06linuxcnc 04. branch deleted * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=95562ab
[01:27:59] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.7 40b01c9 06linuxcnc 10tests/interp/good/test tests: if the interp/good tests fail, show what's wrong * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=40b01c9
[01:37:34] <KGB-linuxcnc> 03Sebastian Kuzminsky 05feature/spiral-arc-handling-2.7 7498180 06linuxcnc 10src/emc/rs274ngc/interp_convert.cc 10src/emc/rs274ngc/interp_internal.hh 10src/emc/rs274ngc/rs274ngc_pre.cc make arc radius tolerance an ini setting * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=7498180
[01:37:34] <KGB-linuxcnc> 03Sebastian Kuzminsky 05feature/spiral-arc-handling-2.7 ff3710e 06linuxcnc Merge remote-tracking branch 'origin/2.7' into feature/spiral-arc-handling-2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ff3710e
[01:37:34] <KGB-linuxcnc> 03Sebastian Kuzminsky 05feature/spiral-arc-handling-2.7 9e0af7a 06linuxcnc 10(12 files in 2 dirs) tests: fix interp arc tests for new arc radius tolerance * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=9e0af7a
[02:00:24] <linuxcnc-build> build #2954 of 1202.rip-lucid-amd64 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1202.rip-lucid-amd64/builds/2954 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[02:00:46] <seb_kuzminsky> what, that wasn't supposed to fail...
[02:02:26] <seb_kuzminsky> ok, that's another "dropped mdi" bug that we've been getting occasionally, unrelated to the arc radius tolerance work
[02:04:19] <linuxcnc-build> build #2964 of 0000.checkin is complete: Failure [4failed] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2964 blamelist: Sebastian Kuzminsky <seb@highlab.com>
[02:04:29] <seb_kuzminsky> someone's going to have to fix that some day
[02:04:37] <seb_kuzminsky> but what do we say to the god of bugs?
[08:59:20] <skunkworks> Hi skunksleep..
[09:16:49] <seb_kuzminsky> oops, my commit to feature/spiral-arc-handling-2.7 is wrong :-/
[09:17:25] <micges> skunkworks: feel better?
[09:17:44] <skunkworks> micges, yes - finally back to work
[09:18:03] <seb_kuzminsky> also: that branch rejects many arcs that are accepted by 2.7, and i'm not sure we want that
[09:18:25] <seb_kuzminsky> skunkworks: i'm glad you're better :-)
[09:19:38] <skunkworks> Thanks
[09:24:22] <seb_kuzminsky> in 2.7, large arcs can have radius errors of up to 0.050 inches/1.27 mm, but those are rejected by the feature branch
[09:24:51] <seb_kuzminsky> i wonder if rob intended to change that, he sure didn't tell me about it
[09:24:59] <seb_kuzminsky> (not that i asked him, either)
[09:26:01] <skunkworks> was it some sort of scaled error?
[09:28:01] <KGB-linuxcnc> 03Sebastian Kuzminsky 05feature/spiral-arc-handling-2.7 1e1a6d1 06linuxcnc 10src/emc/rs274ngc/interp_convert.cc 10src/emc/rs274ngc/interp_internal.hh 10src/emc/rs274ngc/rs274ngc_pre.cc fixup "make arc radius tolerance an ini setting" * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=1e1a6d1
[09:28:10] <seb_kuzminsky> skunkworks: sort of, yeah
[09:29:26] <seb_kuzminsky> here's the function in 2.7 that validates center-format arcs:
http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=blob;f=src/emc/rs274ngc/interp_arc.cc;h=501143046370251e15c278b1911468b6331690de;hb=refs/heads/2.7#l246
[09:29:42] <seb_kuzminsky> line 278 does the check
[09:33:41] <seb_kuzminsky> arcs are rejected if the radius error is 100 * the specified tolerance (0.0005 inch), or if the error is larger than the tolerance *and* the relative error (error divided by the longer of the two radii) is greater than 0.001 (ie, if the two radii differ by more than 0.1%)
[09:55:57] <skunkworks> question for rob I guess
[11:51:01] <seb_kuzminsky> the future is getting closer:
http://www.phoronix.com/scan.php?page=news_item&px=96Boards-HiKey-8-Core-ARMv8
[14:05:29] <cradek> an update on
http://bodgesoc.blogspot.com/ !
[15:17:15] <CaptHindsight> seb_kuzminsky: I'll make ARM64 boards with 16-32GB of RAM, The SOC's just need to be made available
[16:15:38] <KGB-linuxcnc> 03Robert W. Ellenberg 05hotfix/final-velocity-estimation 893db5f 06linuxcnc 10src/emc/tp/tp.c tp: fix for incorrect estimation of final velocity when deciding ramped vs. trapezoidal acceleration * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=893db5f
[16:20:32] <seb_kuzminsky> cradek: that dude just never stops, it's awesome
[16:22:22] <seb_kuzminsky> CaptHindsight: at some point in the future we'll need an Arm hypervisor host, to run a gaggle of arm buildslave vms on
[16:22:45] <seb_kuzminsky> 2 cores & 2 gb ram per VM seems about right