Back
[07:33:08] <jepler> invalid command name "set_rapidrate"
[07:33:08] <jepler> while executing
[07:33:08] <jepler> "set_rapidrate 0"
[07:33:08] <jepler> (command executed by scale)
[07:37:00] <jepler> oh nevermind, I had switched branches but not rebuilt
[09:15:31] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args b29b80d 06linuxcnc 10src/hal/utils/comp.g halcompile: 'option extra_link_args' needs a string * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b29b80d
[09:15:31] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 6e8f95c 06linuxcnc 03tests/comp-extralib/.gitignore 03tests/comp-extralib/extralib_test.comp 03tests/comp-extralib/test.sh halcompile: test that option extra_link_args works * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=6e8f95c
[09:15:31] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 30c7ed5 06linuxcnc 10docs/src/hal/comp.txt halcompile: document halcompile extra_link_args * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=30c7ed5
[09:15:34] <jepler> seb_kuzminsky: OK for 2.6? jepler/2.6/halcompile-extra-link-args
[09:15:49] <jepler> seb_kuzminsky: comp was intended to support 'option extra_link_args', but the implementation never worked and it wasn't tested or documented..
[09:18:28] <seb_kuzminsky> with a commit like 6e8f95c6 how can i say no? go ahead
[09:19:33] <cradek> seb_kuzminsky: now we all know your secret
[09:19:33] <jepler> OK, if the buildbot doesn't detect a problem I'll merge it later today.
[09:22:38] <seb_kuzminsky> :-)
[09:22:49] <jepler> seb_kuzminsky: I don't know if you saw it a few weeks ago, but I asked you about the NML serial number fix for 2.7. I think it should go in, even though we know it breaks nml-over-tcp users.
[09:23:03] <seb_kuzminsky> i saw it and then spaced it
[09:23:07] <jepler> OK
[09:23:30] <jepler> I wouldn't really want to be the final decider about that issue
[09:23:57] <jepler> anyway, the alternative is to put it in master, giving more time for some nml-over-tcp user to fix that up before it is broken in a released version
[09:24:48] <cradek> I forget how your thread about it on -devel went
[09:24:51] <jepler> .. since the bug is only "fatal" in a few circumstances, like when you have an infinite "wait complete" timeout
[09:25:59] <cradek> not finding it ... didn't you announce this and ask if anyone was using tcp?
[09:26:34] <seb_kuzminsky> it's basically sascha ittner's for for #328, right?
https://sourceforge.net/p/emc/bugs/328/
[09:26:59] <cradek> oh found it
[09:27:02] <jepler> cradek: Sascha Ittner didn't turn up again, Ebo doesn't have time to work on it, W. Martinjak / matsche didn't either
[09:27:18] <jepler> Subject: ... Patch for issue #328 rebased onto master
[09:28:36] <cradek> ... and probably nobody else cares
[09:28:53] <jepler> probably should merge jepler/test-nml-over-tcp somewhere either way
[09:29:10] <jepler> the 64-bit problems in nml-over-tcp are orthogonal to the sequence number / wait-for-complete problem
[09:29:22] <cradek> yeah
[09:29:44] <cradek> but you fixed the 64-bit problems (aka compatibility between 32 and 64 bit talking to each other over tcp)?
[09:30:59] <seb_kuzminsky> so wait, there's two branches? one that fixes ml over tcp between different-sized hosts, and one that fixes serial numbers from multiple uis while breaking nml over tcp?
[09:31:08] <jepler> seb_kuzminsky: yeah so confuse
[09:31:14] <cradek> that's funny but I think true
[09:32:49] <jepler> after matsche chewed me out for giving him an "impressive assigmnent" I figured I should do it myself
[09:33:04] <jepler> though I don't think I actually tested 32-to-64, just 32-to-32 and 64-to-64
[09:34:31] <linuxcnc-build_> build #798 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/798 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:34:43] <jepler> wooo build failures
[09:34:47] <seb_kuzminsky> ok we probably want the "fix serial numbers" one sooner rather than later, probably in 2.7
[09:35:04] <seb_kuzminsky> jepler: i reread it and i dont think matsche chewed you out
[09:35:15] <jepler> seb_kuzminsky: ok, that was hyperbole
[09:35:43] <seb_kuzminsky> test.sh: line 4: halcompile: command not found
[09:35:43] <jepler> oh failure because I wrote that branch against 2.7 with halcompile, then decided it should probably go in 2.6..
[09:36:15] <linuxcnc-build_> build #798 of 1400.rip-wheezy-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/798 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:36:29] <seb_kuzminsky> haha its the commit that i said i liked that broke everything :-/
[09:37:25] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 1b1b03f 06linuxcnc 10src/hal/utils/comp.g comp: 'option extra_link_args' needs a string * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=1b1b03f
[09:37:25] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 3937ec4 06linuxcnc 03tests/comp-extralib/.gitignore 03tests/comp-extralib/extralib_test.comp 03tests/comp-extralib/test.sh comp: test that option extra_link_args works * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=3937ec4
[09:37:26] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 460a7b8 06linuxcnc 10docs/src/hal/comp.txt comp: document halcompile extra_link_args * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=460a7b8
[09:38:13] <linuxcnc-build_> build #456 of 1401.rip-wheezy-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1401.rip-wheezy-rtai-i386/builds/456 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:48:52] <linuxcnc-build_> build #2640 of 1300.rip-precise-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1300.rip-precise-i386/builds/2640 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:50:00] <linuxcnc-build_> build #2641 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/2641 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:50:51] <linuxcnc-build_> build #2642 of 1306.rip-precise-amd64 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1306.rip-precise-amd64/builds/2642 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:51:22] <linuxcnc-build_> build #1842 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/1842 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:52:02] <linuxcnc-build_> build #2641 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/2641 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:55:24] <linuxcnc-build_> build #2641 of 1201.rip-lucid-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1201.rip-lucid-rtai-i386/builds/2641 blamelist: Jeff Epler <jepler@unpythonic.net>
[09:55:24] <linuxcnc-build_> build #2652 of 0000.checkin is complete: Failure [4failed] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2652 blamelist: Jeff Epler <jepler@unpythonic.net>
[10:03:02] <seb_kuzminsky> i agree that it's better to fix the serial number collission problem for everyone while breaking nml-over-tcp for hypothetical/silent users, so please put the serial number fix in 2.7
[10:03:13] <cradek> yay
[10:04:03] <seb_kuzminsky> i'm hopeful that we can get rid of nml in the next year or so, and add back in working control-over-the-network at the same time
[10:05:14] <mozmck> so what will replace it?
[10:05:16] <cradek> nml is used for gui->task but also task->task (I think interp list has nml objects on it). have ... any of us thought about how that will work?
[10:05:31] <mozmck> machinetalk? :)
[10:06:10] <seb_kuzminsky> mozmck: i don't know much about machinetalk, but i'm definitely open to being educated about it by someone who does
[10:06:46] <mozmck> I think that's what they called the nml replacement in machinekit
[10:07:01] <seb_kuzminsky> my uneducated impression is that it adds dependencies on still-under-development external libraries that are not packaged for the distributions we want to support, sucah as lucid, precise, and wheezy
[10:07:14] <mozmck> I haven't followed it closely.
[10:07:40] <mozmck> that could be - I don't know
[10:07:41] <seb_kuzminsky> i'm (perhaps naively) imagining that we'll replace it with tcp sockets carrying something like json or protobuf messages
[10:08:05] <mozmck> I think that's what machinetalk does in part anyhow.
[10:08:30] <cradek> machinekit's current master branch has both machinetalk and libnml directories
[10:08:32] <mozmck> but I may be mistaken
[10:08:44] <seb_kuzminsky> mozmck: also... it was really difficult for me to collaborate with michael, and i'm very reluctant to try to cooperate with him again, and that turns me personally away from trying to use their code
[10:09:06] <mozmck> yeah, I don't know that the change is complete.
[10:09:25] <seb_kuzminsky> cradek: that's a good point (about nml being used inside task), i hadn't thought about that.
[10:09:26] <mozmck> heh, michael was a little pushy it seemed.
[10:10:15] <seb_kuzminsky> it may be that nml can stay inside task while the communications protocol to the uis gets fixed, and then we can tackle the inside of task separately
[10:10:29] <cradek> that sure seems true
[10:10:41] <cradek> it is a totally separate problem
[10:10:59] <archivist> is the anti nml feeling the right thing anyway
[10:11:03] <cradek> changing the types on the interplist to match the types of interpqueue might actually be pretty easy
[10:11:04] <seb_kuzminsky> or maybe the task update to handle the new protocol will naturally point to refactoring the inside of task at the same time
[10:11:09] <jepler> the structures used internally to task don't need to be NML structures .. they just are
[10:11:39] <seb_kuzminsky> archivist: i dont know anyone who's looked closely at nml and liked what they saw
[10:11:42] <cradek> jepler: exactly
[10:12:49] <jepler> a few weeks ago I looked at their "machinetalk". The structure seemed to have a Python program which talked NML out one end to task, and a new protocol out the other end. This is certainly a sort of transitional step we have contemplated for liblinuxcnc-ui as well.
[10:13:40] <jepler> they chose and implemented a protocol first, while we chose and are implementing an API first.
[10:14:03] <jepler> in parallel we're converting existing UIs; I didn't look to see whether they are converting existing UIs
[10:14:17] <cradek> it's absolutely right to have the api be protocol-agnostic
[10:14:18] <jepler> once our UIs are all converted to the new API, we can change the protocol freely
[10:14:19] <linuxcnc-build_> build #799 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/799 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:14:42] <jepler> geez jeff
[10:14:45] <linuxcnc-build_> build #799 of 1400.rip-wheezy-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/799 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:14:47] <seb_kuzminsky> jepler: or at least freely-er
[10:14:57] <jepler> .. I'll have to fix it later.
[10:15:02] <jepler> hooray for tests
[10:15:20] <seb_kuzminsky> there are some things that the api must incompass to work with nml, that may or may not make sense given an improved future protocol
[10:17:29] <linuxcnc-build_> build #457 of 1401.rip-wheezy-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1401.rip-wheezy-rtai-i386/builds/457 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:23:06] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args a49cf73 06linuxcnc 03tests/comp-extralib/.gitignore 03tests/comp-extralib/extralib_test.comp 03tests/comp-extralib/test.sh comp: test that option extra_link_args works * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=a49cf73
[10:23:06] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args c61975b 06linuxcnc 10docs/src/hal/comp.txt comp: document halcompile extra_link_args * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c61975b
[10:25:44] <linuxcnc-build_> build #2641 of 1300.rip-precise-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1300.rip-precise-i386/builds/2641 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:26:51] <linuxcnc-build_> build #2643 of 1306.rip-precise-amd64 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1306.rip-precise-amd64/builds/2643 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:27:50] <linuxcnc-build_> build #2642 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/2642 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:28:48] <linuxcnc-build_> build #1843 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/1843 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:29:27] <linuxcnc-build_> build #2642 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/2642 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:32:38] <linuxcnc-build_> build #2642 of 1201.rip-lucid-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1201.rip-lucid-rtai-i386/builds/2642 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:32:38] <linuxcnc-build_> build #2653 of 0000.checkin is complete: Failure [4failed] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2653 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:45:23] <seb_kuzminsky> i'm closing in on that dropped-mdi thing in the liblinuxcnc-ui branch
[10:46:15] <seb_kuzminsky> did you know that task accepts mdi when in Manual mode? but that it appears to do it wrong?
[10:46:23] <seb_kuzminsky> i did not know this
[10:46:26] <seb_kuzminsky> task is full of surprises
[10:51:31] <linuxcnc-build_> build #800 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/800 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:52:06] <linuxcnc-build_> build #800 of 1400.rip-wheezy-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/800 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:54:50] <linuxcnc-build_> build #458 of 1401.rip-wheezy-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1401.rip-wheezy-rtai-i386/builds/458 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[10:58:16] <jepler> go home jeff -- you may be sober, but you still haven't made that test work in all these tries
[11:00:25] <seb_kuzminsky> it's ok, it helps keep my house nice & warm in this cold snap
[11:01:00] <seb_kuzminsky> funny, we were talking on the list just recently about mdi in manual mode
[11:01:13] <seb_kuzminsky> turns out, it's there! it's just unused by all our uis and broken in task! :-)
[11:01:21] <seb_kuzminsky> (or something)
[11:01:44] <skunkworks> good read!
[11:02:37] <linuxcnc-build_> build #2642 of 1300.rip-precise-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1300.rip-precise-i386/builds/2642 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:03:30] <linuxcnc-build_> build #2644 of 1306.rip-precise-amd64 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1306.rip-precise-amd64/builds/2644 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:04:57] <linuxcnc-build_> build #2643 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/2643 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:05:48] <linuxcnc-build_> build #2643 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/2643 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:06:13] <linuxcnc-build_> build #1844 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/1844 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:09:45] <linuxcnc-build_> build #2643 of 1201.rip-lucid-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1201.rip-lucid-rtai-i386/builds/2643 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:09:45] <linuxcnc-build_> build #2654 of 0000.checkin is complete: Failure [4failed] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2654 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:22:40] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 120c4bc 06linuxcnc 03tests/comp-extralib/.gitignore 03tests/comp-extralib/checkresult 03tests/comp-extralib/extralib_test.comp 03tests/comp-extralib/test.sh comp: test that option extra_link_args works * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=120c4bc
[11:22:40] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args ffe4f5e 06linuxcnc 10docs/src/hal/comp.txt comp: document halcompile extra_link_args * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=ffe4f5e
[11:41:40] <linuxcnc-build_> build #801 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/801 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:42:22] <linuxcnc-build_> build #801 of 1400.rip-wheezy-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1400.rip-wheezy-i386/builds/801 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:44:53] <linuxcnc-build_> build #459 of 1401.rip-wheezy-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1401.rip-wheezy-rtai-i386/builds/459 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:53:15] <linuxcnc-build_> build #2643 of 1300.rip-precise-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1300.rip-precise-i386/builds/2643 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:53:51] <linuxcnc-build_> build #2645 of 1306.rip-precise-amd64 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1306.rip-precise-amd64/builds/2645 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:54:36] <tinkerer> "(16:07:13) jepler: W. Martinjak / matsche didn't either": maybe I've time to make some tests around christmas.
[11:54:53] <linuxcnc-build_> build #2644 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/2644 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:55:59] <linuxcnc-build_> build #2644 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/2644 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[11:56:11] <linuxcnc-build_> build #1845 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/1845 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[12:00:28] <linuxcnc-build_> build #2644 of 1201.rip-lucid-rtai-i386 is complete: Failure [4failed compile runtests] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/1201.rip-lucid-rtai-i386/builds/2644 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[12:00:28] <linuxcnc-build_> build #2655 of 0000.checkin is complete: Failure [4failed] Build details are at
http://buildbot.linuxcnc.org/buildbot/builders/0000.checkin/builds/2655 blamelist: dummy, Jeff Epler <jepler@unpythonic.net>
[12:11:04] <jepler> this time I don't immediately see why it's failing
[13:04:19] <KGB-linuxcnc> 03Michael Geszkiewicz 052.7 5d99bc0 06linuxcnc 10src/hal/drivers/mesa-hostmot2/sserial.c hostmot2: Fix reporting sserial remote faults * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=5d99bc0
[13:04:19] <KGB-linuxcnc> 03Michael Geszkiewicz 052.7 4136cbd 06linuxcnc 10src/hal/drivers/mesa-hostmot2/sserial.c hostmot2: Fix driver not reporting all sserial remote faults * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4136cbd
[13:18:36] <jepler> tinkerer: hi, that would be great.
[13:20:17] <tinkerer> yep! I hope I can find some timeslots. :)
[13:20:19] <jepler> cradek: can you build my branch and see if the newly-added test fails for you as well? It's not failing here.
[13:20:57] <jepler> cradek: [origin/]jepler/2.6/halcompile-extra-link-args
[13:31:16] <cradek> *** /home/cradek/emc/tests/comp-extralib: FAIL: checkresult exited with 1
[13:35:31] <jepler> yay
[13:35:35] <jepler> why doesn't it fail for you?
[13:35:47] <jepler> are you on a wheezy-amd64 ?
[13:36:07] <cradek> it does fail for me
[13:36:18] <jepler> why doesn't it fail for me?
[13:36:23] <jepler> sorry, I forgot which one of us was which
[13:36:30] <cradek> I'm the one here
[13:36:44] <cradek> sure wish I could abort runtests with C-c
[13:37:06] <jepler> jepler@parrot:~/local/src/linuxcnc$ runtests tests/comp-extralib/
[13:37:06] <jepler> Running test: tests/comp-extralib
[13:37:06] <jepler> *** tests/comp-extralib: FAIL: checkresult exited with 1
[13:37:09] <jepler> jepler@parrot:~/local/src/linuxcnc/tests/comp-extralib$ runtests .
[13:37:09] <jepler> Running test: .
[13:37:09] <jepler> Runtest: 1 tests run, 1 successful, 0 failed + 0 expected
[13:37:15] <jepler> oh it's a current directory vs test directory problem
[13:37:38] <cradek> yeah I figured that, but I hadn't been able to test it yet because of C-c
[13:38:56] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 3354d1a 06linuxcnc 03tests/comp-extralib/.gitignore 03tests/comp-extralib/checkresult 03tests/comp-extralib/extralib_test.comp 03tests/comp-extralib/test.sh comp: test that option extra_link_args works * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=3354d1a
[13:38:56] <KGB-linuxcnc> 03Jeff Epler 05jepler/2.6/halcompile-extra-link-args 3991f9c 06linuxcnc 10docs/src/hal/comp.txt comp: document halcompile extra_link_args * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=3991f9c
[13:39:55] <jepler> haha "home/cradek/emc"
[13:41:12] <cradek> welll
[13:43:08] <cradek> *sigh*
http://pastie.org/pastes/9717404/text
[13:44:15] <jepler> it looks like you've got some sort of problem that's not going to work out
[13:44:19] <jepler> halrun -U ?
[13:45:05] <cradek> after I killed a lot of things it looks like it might finish
[13:45:13] <cradek> I guess I just need to leave them alone
[13:45:42] <cradek> http://i1.kym-cdn.com/photos/images/newsfeed/000/234/765/b7e.jpg
[13:45:54] <jepler> > The result of that vote will decide whether the definition of the word "day" will change to depend solely on cesium atoms or whether the word "day" will continue to be related to the rotation of the earth as specified in existing international agreements.
[13:51:25] <pcw_home> I must have missed it somehow
[13:52:12] <pcw_home> oops
[13:52:35] <cradek> they didn't ask me either
[13:53:39] <pcw_home> they dont have better things to quibble about?
[13:55:36] <jepler> for those obsessed with timekeeping it's an important question
[14:07:48] <cradek> for those obsessed with [...] it's an important question
[14:07:59] <cradek> (this is universal)
[14:09:32] <jepler> The IAU express to the ITU-R
[14:09:39] <jepler> that the word “universal”, and by extension the abbreviation “UT”, is appropriate
[14:09:42] <jepler> only for a time scale that is linked to the rotation of the Earth,
[14:10:11] <jepler> hmmm of anyone who would know what the definition of "universal" is, surely the IAU (International Astronomical Union) would
[14:21:19] <seb_kuzminsky> i was wrong, it's only emcTaskPlan() that accepts MDI in non-MDI modes, they're rejected further down the call chain :-/
[14:23:39] <cradek> yay?
[14:30:06] <pcw_home> Yay! micges fixed sserial remote fault reporting
[14:33:34] <pcw_home> http://imagebin.org/323785
[14:33:36] <pcw_home> (shorted 7I76e field output)
[14:34:40] <jepler> > This web page at imagebin.org has been reported as an attack page and has been blocked based on your security preferences.
[14:34:58] <pcw_home> Yeah
[14:51:24] <pcw_home> http://ibin.co/1h8SXBzTacu2
[14:51:42] <jepler> ah cool
[14:56:48] <pcw_home> before the fix, the driver didn't report these errors except at startup
[14:57:20] <cradek> those look useful
[14:57:55] <pcw_home> I think before it would just report remote fault
[17:49:58] <Roguish> good day all. simple question. how do I get a listing of all the pins and signals, etc. into a file, while LCNC is running. and I will write it down this time.
[17:50:29] <PCW> halcmd show all > thebigtamale
[17:54:08] <Roguish> thanks PCW.. I swear I tried that one before. oh well. worked this time. and written down.
[18:20:56] <KGB-linuxcnc> 03Michael Geszkiewicz 052.7 c72e6f6 06linuxcnc 10src/hal/drivers/mesa-hostmot2/sserial.c hostmot2: sserial: Fix .scalemax parameter was ignored on analog inputs * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c72e6f6
[18:38:50] <micges-dev> jepler: does HAL_NAME_LEN == 41 from specific reason? or it was just set for this value on beggining and it was enough? we are out of chars in some hostmot2 configurations
[18:41:20] <seb_kuzminsky> it's because jepler's floppy disk was full and his mom said he couldnt have any more until he mowed the yard
[18:41:48] <PCW> can we buy a vowel?
[18:42:20] <micges-dev> haha I see :)
[18:42:29] <micges-dev> seb_kuzminsky: we got hm2_7i76e.0.7i76.0.0.fieldvoltage-scalema
[18:59:07] <seb_kuzminsky> does anything break if you bump it up to... 57? (that's the next logical option after 41)
[19:00:39] <PCW> Heinz 57?
[19:02:50] <PCW> or are these parts of lumps allocated on 16 byte boundaries?
[19:03:01] <seb_kuzminsky> PCW: can't you just make shorter part numbers?
[19:03:26] <PCW> the 7I76e broke a lot of things
[19:04:30] <seb_kuzminsky> wow that's a neat board!
[19:05:30] <PCW> and with linuxcnc 2.7 coming out maybe even usable
[19:07:36] <PCW> (Ive had it running almost a year now so I'm pretty sure linuxcnc network stack/Preemt-RT is trustworthy)
[19:17:10] <jepler> HAL_NAME_LEN + 1 is a multiple of 4, so it doesn't introduce gaps in structures when followed by a pointer
[19:17:33] <PCW> OK
[19:17:57] <jepler> but ideally it HAL_NAME_LEN + 1 would be a multiple of 8, since some types have 8 bit alignment requirements
[19:18:40] <jepler> I think names take up a substantial portion of the HAL shared memory area, so I'd recommend to expand it by the same fraction as names are expanded
[19:20:14] <cpresser> when i want to get this fix: "hostmot2: sserial: Fix .scalemax..". do i need to compile myself, or can i just get a binary from the buildbot?
[19:20:16] <jepler> HAL_NAME_LEN was apparently 31 from 2003 (originally) until 2005, that would have given the desired 8-alignment. it was increased to 41 by alex_joni
[19:20:21] <jepler> and I lied when I said 42 was divisible by 4
[19:22:20] <PCW> 43?
[19:22:38] <jepler> I bet that's not divisible by 4 either
[19:22:49] <PCW> 43 +1 is
[19:23:25] <jepler> is 2 more characters enough?
[19:23:35] <PCW> for now
[19:24:57] <jepler> while we're dabbling, I am pretty sure HAL_SIZE should be a multiple of 4096
[19:25:03] <PCW> I will try to be less extravagant with names
[19:26:33] <jepler> cpresser: you can see from the "kgb" message that the commit went on the 2.7 branch, so within a couple of hours it'll be in the buildbot's 2.7~pre packages
[19:27:38] <cpresser> jepler: thats what i assumed.. however i cant seem fo find a 2.7 repository:
http://buildbot.linuxcnc.org/
[19:28:12] <jepler> cpresser: seb must not have had enough tuits to get that page updated.
[19:28:25] <cpresser> ah nevermind. its there:
http://buildbot.linuxcnc.org/~buildmaster/dists/wheezy/2.7-rt/binary-i386/
[19:28:49] <jepler> OK
[19:29:03] <cpresser> but the page needs an update :)
[19:29:09] <jepler> I think if you take a 2.6 -rt line and change it to 2.7, that's the right one
[19:31:48] <cpresser> the patch which worksaround the kernel bug (enabled vs enable in sysfs) doesnt seem to be in the 2.7 brancht?
[19:33:24] <jepler> I don't remember what branch(es) I put it on
[19:34:36] <jepler> http://emergent.unpythonic.net/files/sandbox/0001-hal-increase-some-limits.patch
[19:34:50] <jepler> this seems to work, though I haven't run the full testsuite yet
[19:34:56] <cpresser> this one:
http://git.linuxcnc.org/gitweb?p=linuxcnc.git;a=shortlog;h=refs/heads/jepler/uspace-pci-sysfs-enabled
[19:35:00] <jepler> name length to 47, shared memory area to (75*4096)
[19:49:12] <jepler> seb_kuzminsky: WDYT about increasing the hal limits like in
http://emergent.unpythonic.net/files/sandbox/0001-hal-increase-some-limits.patch ?
[19:49:18] <jepler> any later and we wouldn't want to do it for 2.7..
[19:56:53] <KGB-linuxcnc> 03Jeff Epler 052.7 d6c114a 06linuxcnc 10src/rtapi/uspace_rtapi_app.cc uspace: don't try to use rt hardening except on a realtime kernel * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d6c114a
[19:56:53] <KGB-linuxcnc> 03Jeff Epler 052.7 1b1ab10 06linuxcnc 10src/rtapi/rtapi_pci.cc uspace: compensate for kernel API breakage * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=1b1ab10
[19:56:53] <KGB-linuxcnc> 03Jeff Epler 052.7 174a0a3 06linuxcnc 10(10 files in 2 dirs) tests: test nml-over-tcp * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=174a0a3
[19:56:55] <KGB-linuxcnc> 03Jeff Epler 052.7 4da1da8 06linuxcnc 10(13 files in 7 dirs) nml: convert arch-dependent types to fixed-width types * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=4da1da8
[19:56:59] <KGB-linuxcnc> 03Jeff Epler 052.7 c4c7fb9 06linuxcnc 10(30 files in 8 dirs) nml: implement command queue with reliable reception * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=c4c7fb9
[20:06:24] <PCW> hmm my J1800 definitely does not like isolcpus=1 with preemt-RT
[21:06:14] <KGB-linuxcnc> 03Jeff Epler 052.6 3991f9c 06linuxcnc fast forward * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=3991f9c
[21:07:39] <KGB-linuxcnc> 03Jeff Epler 052.7 0378170 06linuxcnc 10docs/src/hal/comp.txt 03src/hal/utils/halcompile.g Merge remote-tracking branch 'origin/2.6' into 2.7 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=0378170
[21:08:41] <KGB-linuxcnc> 03Jeff Epler 05master bf22e30 06linuxcnc Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=bf22e30
[22:13:04] <KGB-linuxcnc> 03Jeff Epler 052.7 66dd596 06linuxcnc 10tests/comp-extralib/test.sh tests: comp was renamed halcompile * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=66dd596
[22:13:13] <KGB-linuxcnc> 03Jeff Epler 05master 49a9ed2 06linuxcnc Merge remote-tracking branch 'origin/2.7' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=49a9ed2
[22:26:52] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/jog_ignore_fo 11f61af 06linuxcnc 10(10 files in 6 dirs) inifile: support jog_ignore_feed_override * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=11f61af
[23:00:27] <KGB-linuxcnc> 03Dewey Garrett 05dgarr/jog_ignore_fo 2b5c9ce 06linuxcnc 10(10 files in 6 dirs) inifile: support jog_ignore_feed_override * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=2b5c9ce
[23:21:05] <seb_kuzminsky> jepler: 0001-hal-increase-some-limits.patch looks good for 2.7
