#linuxcnc-devel | Logs for 2014-04-04

[01:03:26] <KGB-linuxcnc> 03Sebastian Kuzminsky 052.6 e1fe90b 06linuxcnc Merge remote-tracking branch 'origin/v2.5_branch' into 2.6 * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=e1fe90b
[01:03:26] <KGB-linuxcnc> 03Michael Haberler 052.6 d17419a 06linuxcnc 10lib/python/pyvcp_widgets.py new pyvcp_widgets * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=d17419a
[01:03:26] <KGB-linuxcnc> 03Sebastian Kuzminsky 05master 388e8ae 06linuxcnc Merge branch '2.6' * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=388e8ae
[02:18:04] <linuxcnc-build> build #132 of 4017.deb-wheezy-armhf is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-armhf/builds/132 blamelist: Dewey Garrett <dgarrett@panix.com>, Sebastian Kuzminsky <seb@highlab.com>, Andy Pugh <andy@bodgesoc.org>, Michael Haberler <git@mah.priv.at>, John Thornton
[02:18:04] <linuxcnc-build> <jthornton@gnipsel.com>
[04:02:57] <linuxcnc-build> build #133 of 4017.deb-wheezy-armhf is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-armhf/builds/133 blamelist: Dewey Garrett <dgarrett@panix.com>, Sebastian Kuzminsky <seb@highlab.com>, Andy Pugh <andy@bodgesoc.org>, Michael Haberler <git@mah.priv.at>, John Thornton
[04:02:57] <linuxcnc-build> <jthornton@gnipsel.com>
[06:19:40] <skunkworks> anyone else seeing the garbage at the top of the page
[06:19:41] <skunkworks> http://www.linuxcnc.org/linuxcnc/index.php/russian/forum/18-computer/27640-asus-m5a-78l-mlx-build
[06:20:59] <skunkworks> huh- some reason I am russian
[06:21:51] <skunkworks> ok - changing to english fixed it
[11:50:12] <memleak> skunkworks, where are you from actually?
[11:51:10] <skunkworks_> Wisconsin..
[11:51:29] <skunkworks_> where did you think I was from?
[11:52:27] <skunkworks_> I missed you at the fest - (didn't actually make it - did you?)
[11:52:43] <skunkworks_> Your dad was there though.
[11:59:10] <memleak> i didn't make it no..
[11:59:27] <memleak> i'
[11:59:39] <memleak> d like to go next time though
[12:00:25] <skunkworks_> it was fun
[12:06:01] <skunkworks_> memleak: http://electronicsam.com/images/KandT/Fest2013/DSC_3884.JPG
[12:47:23] <memleak> am i the only kid?..
[12:51:04] <skunkworks> Well - we are all kids at heart..
[12:54:49] <memleak> that's true :) nobody here is ever grumpy or anything.
[13:22:37] <KGB-linuxcnc> 03Dewey Garrett 052.6 b964ad5 06linuxcnc 10(11 files in 2 dirs) configs/sim/axis/twopass new sim demo * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=b964ad5
[13:22:37] <KGB-linuxcnc> 03Dewey Garrett 052.6 115dee7 06linuxcnc 10(5 files in 4 dirs) docs: updates for configs reorganization * 14http://git.linuxcnc.org/?p=linuxcnc.git;a=commitdiff;h=115dee7
[13:23:34] <seb_kuzminsky> memleak: there's a kid at the hackspace i go to who's 17 and is working on his 2nd or 3rd cnc machine, but he doesn't participate in the linuxcnc community as far as i know
[13:32:02] <MrHindsight> skunkworks: who are all those old people in that pic?
[13:32:24] <skunkworks> well - if you look in the mirror...
[13:47:05] <andypugh> Is the EtherCAT hal driver intended for 2.6?
[13:48:39] <cradek> andypugh: where is it?
[13:49:26] <cradek> isn't ethercat the one with the incompatible license? or did I miss some important change?
[13:50:44] <seb_kuzminsky> yes, jepler pointed out in the "[Emc-developers] EtherCAT" email thread that their license puts restrictions on use, and i thought the consensus back then was that we couldn't use it for that reason
[13:51:18] <seb_kuzminsky> andypugh: do you have other information, or another opinion? i'm all ears
[13:57:05] <seb_kuzminsky> adding stand-alone drivers is pretty low risk, and generally acceptable even in stable branches, imo
[13:58:13] <Connor> I would say make the driver a stand-alone driver.
[14:00:31] <seb_kuzminsky> Connor: you mean, maintained in a separate repo from linuxcnc, and distributed independently? that agrees with jeff's recommendation in that thread
[14:00:34] <seb_kuzminsky> bbl, lunch
[14:01:36] <Connor> Exactly.
[14:02:13] <Connor> just make sure that it's existence is well documented so people know how to get it if they need it.
[14:37:54] <linuxcnc-build> build #134 of 4017.deb-wheezy-armhf is complete: Failure [4failed shell_3] Build details are at http://buildbot.linuxcnc.org/buildbot/builders/4017.deb-wheezy-armhf/builds/134 blamelist: Dewey Garrett <dgarrett@panix.com>
[14:43:37] <andypugh> cradek, seb_kuzminsky Ah, yes, I remember now. There is a thread here: http://linuxcnc.org/lucid/emc2/emc2/emc2/index.php/german/forum/24-hal-components/22346-ethercat-hal-driver?start=60
[14:43:48] <andypugh> It just looks like it is harder than it probably ought to be.
[14:44:09] <andypugh> There is probably an identical problem with the ASAP3 library that I am writing for Arduino.
[14:59:33] <CptMemleakShop> what the way to check the mode the LPT is actually running in?
[14:59:46] <CptMemleakShop> what/what's
[15:00:14] <cradek> CptMemleakShop: hal_parport prints that information in dmesg -- unfortunately a lot of parports lie
[15:12:15] <CptMemleakShop> PC-style at 0xe010 (0xe000), irq 16, using FIFO [PCSPP,TRISTATE,COMPAT,EPP,ECP, sure but which one? :)
[15:32:20] <cradek> I think that means it's EPP capable
[15:32:29] <cradek> so if the driver asks for EPP it will use it
[15:32:31] <cradek> I think
[15:32:44] <cradek> but that's not the hal_parport message you pasted
[15:33:00] <cradek> what is the actual problem you are having?
[15:36:53] <CptMemleakShop> we are just wondering what mode the LPT is actually running in
[15:37:07] <cradek> with which hal driver?
[15:39:59] <CptMemleakShop> http://pastebin.com/tpjuNAuu was just the default using stepconfig
[15:40:36] <cradek> so hal_parport
[15:40:41] <cradek> look at the messages it prints when it loads
[15:47:38] <CptMemleakShop> halcmd doesn't run here yet on Gentoo
[15:55:18] <andypugh> The German “Dokument” page still says lots about EMC and Doxygen.
[15:57:16] <seb_kuzminsky> andypugh: which page is this?
[15:57:27] <seb_kuzminsky> our translations are such a mess (except for french)
[15:57:49] <andypugh> http://www.linuxcnc.org/index.php/german/documentation
[15:58:36] <seb_kuzminsky> wow, that's terrible :-(
[15:59:33] <seb_kuzminsky> i hardly ever look at www.linuxcnc.org, but it *is* our front-line public face and i wish it were less out of date
[16:02:03] <andypugh> I have poked about in the Admin sections and can’t see where the content comes from
[16:02:53] <cradek> if I'm logged in, I get a teeny tiny edit icon on those pages
[16:03:19] <seb_kuzminsky> i've never looked under the covers of our website, thanks to you guys for dealing with it
[16:03:21] <cradek> but I am not bilingual enough to know if that's how you do the translations
[16:04:08] <cradek> I only mess with it when it's sploited
[16:05:09] <andypugh> We have numerous Germans available.
[16:06:09] <seb_kuzminsky> the way we do translations today - copying the english to the new language, then editing the new language - is fraught with danger
[16:06:32] <seb_kuzminsky> it requires humans to notice when the translations drift out of sync with each other
[16:06:50] <cradek> which is a terrible system
[16:07:18] <seb_kuzminsky> francis tisserant does a great job of it (i think, my french is way too rusty to tell, really), but i think he may be unique within our project in that regard
[16:07:30] <cradek> yes he is
[16:08:11] <seb_kuzminsky> if someone wants to step up to improve the german docs (or any other language) i would welcome them
[16:08:34] <seb_kuzminsky> but if someone wants to replace our translation system with something more rational and manageable, i think that would be even better
[16:10:18] <andypugh> It turns out that I, too, can edit the web pages. I will try to source a better translation for the German ffront page.
[16:10:30] <seb_kuzminsky> awesome, thanks andypugh
[16:11:33] <andypugh> Even the English “Documentation” page has a section at the bottom “About EMC2”
[16:12:32] <seb_kuzminsky> whoops
[16:13:16] <andypugh> Which is possibly fine, but perhaps should go to a Wiki page that uses the past tense
[16:14:24] <cradek> I bet at some point, no translation becomes better than the out-of-date translation
[16:14:48] <seb_kuzminsky> andypugh: maybe past-tense info should go into the History page, just above
[16:14:58] <seb_kuzminsky> cradek: agreed
[16:15:47] <seb_kuzminsky> gauging where that cut-off point is is difficult now, as the only way to tell how out of date a translation is, is to read it next to the english and compare...
[16:16:17] <cradek> yep
[16:16:25] <seb_kuzminsky> if we used po (the gettext system we use for some of our apps) we could potentially say, any translation that has > 75% (or whatever) up-to-date-ness gets to be released
[16:17:10] <seb_kuzminsky> and it would keep track of what phrases change in the original/master language, and invalidate the translations that no longer match the original
[16:17:12] <cradek> I can imagine using po4a for the docs, but I have no idea how you'd use something like it with joomla
[16:17:20] <seb_kuzminsky> yeah :-/
[16:17:33] <seb_kuzminsky> you're right, i was thinking of the docs, not the website
[16:17:49] <seb_kuzminsky> maybe the website could be built from our git too? i think the asciidoc website does that
[16:18:00] * seb_kuzminsky goes back to $DAYJOB
[16:18:04] <cradek> (modern) websites are much less hackable than we are used to
[16:18:35] <cradek> that's a big question and I'd love to hear from someone who knows about things like that
[16:19:02] <cradek> I have no particular love for joomla, but we need continuity in the forum part
[16:19:07] <cradek> I don't know how closely they're tied together
[20:25:23] <CptMemleakShop> unexpected real time delay with the program stopped? (not in E-Stop) any ideas what would cause that?
[20:34:50] <skunkworks> well - anytime periods are missed - you will get that (even when the program is stopped)
[20:49:00] <CptMemleakShop> what real time events are running when a program is stopped?
[20:49:49] <skunkworks> a bundle of sticks..yes
[20:49:55] <skunkworks> heh -
[20:49:57] <CptMemleakShop> something is making the preempt_rt kernel miss every hour or so
[20:49:57] <skunkworks> yes
[20:50:02] <andypugh> All of them. They just don’t change anything. Pretty much all the code still runs
[21:49:31] <skunkworks> what kind of period are you guys trying? 1ms?
[22:02:01] <CptMemleakShop> skunkworks, base period=120000
[22:02:31] <CptMemleakShop> this is with a LPT PCIe card and steppers
[22:16:49] <CptMemleakShop> the latency test doesn't climb over 50K