#emc-devel | Logs for 2009-05-06

[00:39:33] <jepler> hm, no fallout from my nml and var changes yet?!
[00:45:29] <SWPadnos> -ETOOMANYCOMMITS
[00:45:39] <SWPadnos> and also -ETOOMANYCOMMITMENTS
[00:46:02] <jepler> yeah
[00:54:43] <CIA-47> EMC: 03jepler 07TRUNK * 10emc2/src/libnml/inifile/inifile.cc: detect DOS-style line endings and warn; detect CRs without LFs and error
[01:30:03] <mshaver> logger_dev: help
[01:31:00] <SWPadnos> bookmark, perhaps?
[01:31:10] <jepler> logger_dev: bookmark
[01:31:10] <jepler> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2009-05-06.txt
[01:35:51] <mshaver> jepler: looking at this morning's irc traffic, I noticed you talking about the beagleboard
[01:36:57] <jepler> mshaver: yes indeed
[01:37:03] <mshaver> jepler: I've been talking to Dale Grover about porting EMC2 to ARM & he's ordered a beagle board & stuff to make it run
[01:37:18] <jepler> mshaver: he's going to port rtai?
[01:37:20] <mshaver> Betwixt us all, maybe it can be done!
[01:37:50] <SWPadnos> doesn't RTAI already have an ARM port?
[01:37:54] <mshaver> There is some ARM stuff in the rtai sources
[01:38:02] <SWPadnos> (though it may be meant for big chips with MMUs)
[01:38:08] <jepler> SWPadnos: as far as I have seen, it's a different "port" for each different board, or maybe each different CPU
[01:38:15] <jepler> this thing has an mmu, doesn't it?
[01:38:19] <jepler> (I don't really know :-P)
[01:38:20] <mshaver> hopefully it'll "just work"
[01:38:22] <SWPadnos> yah, I guess it does
[01:38:25] <jepler> that's optimistic
[01:38:35] <SWPadnos> yeah, it sure is :)
[01:38:42] <mshaver> ARM Cortex A8 MMU = yes
[01:38:54] <SWPadnos> ask ds3 to help, he's a beagleboard expert
[01:39:07] <mshaver> it also has a hardware floating point "thing" as well
[01:39:10] <jepler> I'm making one problem into two and looking at rt_preempt
[01:39:20] <SWPadnos> jepler, the "hunyue" partition was from his ESC presentation
[01:39:24] <jepler> so it's slow going and I don't expect resultssoon
[01:39:40] <mshaver> jepler: taking the included DSP into account maybe?
[01:40:04] <jepler> mshaver: no, but that's a good idea to make it more complicated still
[01:40:29] <jepler> right now I'm still trying to build my own kernel with working usb, though, so it's early days
[01:40:35] <mshaver> bringing it to KS?
[01:41:49] <mshaver> did you get Angstrom or Ubuntu running? I'm supposing that this is something that can be done by following directions.
[01:42:32] <jepler> mshaver: I am doing debian, following these instructions: http://elinux.org/BeagleBoardDebian#Native_Install_via_debian-installer_.28NON_QEMU_based.29
[01:42:40] <jepler> I used a USB network card in the install process
[01:42:43] <jepler> "card" hah
[01:43:03] <mshaver> "bulge", "lump"?
[01:43:09] <SWPadnos> dongle
[01:43:47] <jepler> and fwiw my goal is to run slow-thread-only (e.g., with an fpga board), not do 50kHz software stepping
[01:44:00] <SWPadnos> mshaver, there was a 2GB flash image with 4 partitions on it at ESC, I stuck it on my website if you want it (and I think it's on the beagleboard site as well)
[01:45:29] <mshaver> jepler: even that would be a _considerable_ accomplishment!
[01:46:08] <mshaver> SWPadnos: sure! Link?
[01:46:12] <SWPadnos> one sec
[01:46:39] <jepler> --- arch/arm/Kconfig.orig.orig 1970-01-01 00:00:00.000000000 +0000
[01:46:39] <jepler> +++ arch/arm/Kconfig.orig 2009-05-04 21:38:43.000000000 +0000
[01:46:48] <jepler> hm I don't think this guy put exactly what he intended in his patch
[01:47:22] <mshaver> damned time travel guys! always messing things up...
[01:48:06] <SWPadnos> http://www.cncgear.com/Files/beagle.bz2
[01:48:57] <jepler> rename {arch/arm/plat-omap/include/mach => include/linux}/omapfb.h (94%)
[01:49:02] <jepler> boy, that's too clever by half, git
[01:52:33] <SWPadnos> taht is cool
[01:52:35] <SWPadnos> that
[01:52:46] <SWPadnos> lets you know that the file itself didn't change, just its location
[02:01:00] <jepler> but the file did change a bit (6%)
[02:01:12] <jepler> it was specifically the way the old and new names were given that I think is too clever
[02:02:12] <SWPadnos> yep, I agree
[02:15:55] <jmkasunich> mshaver: are you and steves_logging coming to KS, or is it still up in the air?
[02:21:44] <mshaver> jmkasunich: yes. I'm driving to Ann Arbor on Monday the 11th and to KS the next week. Steve is flying to Wichita and together we will be driving back at the end.
[02:22:05] <jmkasunich> good - I owe steve $5
[02:24:03] <mshaver> lol. really, that actually made me laugh out loud!
[02:24:34] <jmkasunich> I spent too much at NAMES, had to bum $5 for the turnpike toll to get home
[02:27:59] <cradek> ugh, everyone wants "us" (alex) to build a new kernel.
[02:28:40] <cradek> that's only a major pain, but also makes us incompatible with previous packages.
[02:31:04] <jepler> yeah, it sucks
[02:31:16] <jmkasunich> who is asking for the kernel?
[02:31:33] <jmkasunich> I've been kind of oblivious lately
[02:31:34] <cradek> atom users and ibm laptop users
[04:35:41] <CIA-47> EMC: 03cmorley 07TRUNK * 10emc2/src/emc/usr_intf/pncconf/pncconf.py:
[04:35:41] <CIA-47> EMC: Fixed loading of pyvcp panels with other GUIs besides AXIS - loads them stand alone
[04:35:41] <CIA-47> EMC: start of adding ability to select different mesa boards / firmware
[04:35:41] <CIA-47> EMC: Will have to reorganize mesa routines to help speed up display of mesa page.
[05:35:49] <steves_logging> OK, we can back out the IBM laptop users from the new kernel requests, I got a USB dongle to work with it and that is good enough for now.
[05:37:57] <steves_logging> I will hopefully be arriving at Fest late Friday evening. Gave up on sending out my minimill with Matt. Not enough time to install all the limit switches and that was the major reason to want to play with it.
[05:39:50] <SWPadnos> have him bringthe mill, and you bring the switches
[05:40:11] <SWPadnos> they may find their way to the right spots on the machine
[05:40:14] <SWPadnos> good night :)
[05:40:38] <steves_logging> right, and they will magically attach themselves without needing to disassemble the machine to drill holes in the right places
[05:40:55] <SWPadnos> oh, well that could be a problem
[05:40:57] <steves_logging> nope, but I may bring a single axis test bed instead
[05:41:15] <SWPadnos> anyway, see you later. It is officially bedtime
[05:41:42] <steves_logging> given time and ability, I would like to try out some ideas for enhanced homing accuracy
[05:42:28] <steves_logging> taking into account the known state of stepper drivers upon power up
[05:42:50] <steves_logging> then keep track of steps in hal
[05:43:42] <steves_logging> should be possible to be at least 4x as accurate as just hitting the switch alone
[08:25:54] <Lerman____> Lerman____ is now known as Lerman
[09:04:13] <Lerman_____> Lerman_____ is now known as Lerman
[12:32:53] <Roguish> alex_joni: good morning. question/problem.
[12:33:30] <Roguish> alex_joni: please see my post in the forum. (like the forum thing, looks good)
[12:33:54] <Roguish> all about the Intel Atom board and 8.04
[12:42:13] <Roguish> SWPadnos: jepler: cradek: jmkasunich: skunkworks:
[12:42:27] <Roguish> good morning. question/problem.
[12:42:35] <Roguish> please see my post in the forum. (like the forum thing, looks good)
[12:42:48] <Roguish> all about the Intel Atom board and 8.04, not installing
[12:44:31] <CIA-47> EMC: 03jepler 07v2_3_branch * 10emc2/debian/changelog: note some new fixes
[12:46:33] <skunkworks> have you tried http://www.linuxcnc.org/hardy/ - aj13?
[12:47:23] <skunkworks> that is the newest livecd that has 2.3 on it. (built a week or 2 ago)
[12:47:27] <alex_joni> Roguish: 8.04 vs. 8.04.2 has absolutely nothing to do with it
[12:47:36] <alex_joni> it's the kernel version that's important
[12:47:42] <alex_joni> (you can find that out with uname -a)
[12:47:55] <Roguish> ok, thanks for the reply. what does??????????
[12:48:16] <alex_joni> but building a new kernel (newer source) for emc2, means that every emc2 user has to update to the newer kernel
[12:48:32] <alex_joni> and that is something that needs a couple weeks of work and testing
[12:49:22] <alex_joni> skunkworks: I doubt the latest LiveCD will help
[12:49:49] <alex_joni> skunkworks: if he installed emc2 on a vanilla hardy, then he'll get the latest emc2/kernel versions
[12:49:52] <skunkworks> ok. (I should read the links)
[12:50:12] <alex_joni> and since that doesn't work, the LiveCD will fail too
[12:50:32] <Roguish> i have the 'latest' liveCD. no worky.
[12:50:44] <alex_joni> Roguish: there is a workaround in the Intel page you linked
[12:50:52] <alex_joni> Need workaround to do OS installation1:
[12:50:57] <alex_joni> 1 Need to disable LAN support in BIOS to install Operating System (OS)
[12:51:01] <Roguish> let me look again.
[12:52:05] <Roguish> ok. i did not see that. i will try that.
[12:52:48] <Roguish> if it does indeed work, i will let you know.
[12:53:26] <alex_joni> thanks
[12:54:52] <skunkworks> I have installed dapper and hardy on my 330 atom with no issues. a few times ;)
[13:04:54] <cradek> didn't you commit that patch already?
[13:05:01] <cradek> jepler: didn't you commit that patch already?
[13:07:10] <alex_joni> cradek: I think he has, but at some point it got reverted
[13:07:33] <cradek> got reverted??
[13:08:30] <jepler> he's confused about history
[13:08:34] <alex_joni> hmm.. it looks like it wasn't commited
[13:08:35] <jepler> I applied it on TRUNK after the branch
[13:09:44] <alex_joni> the -msse2 one?
[13:10:40] <jepler> yes
[13:10:46] <jepler> bbl
[18:30:57] <RobJansen> Hi, I have a strange problem: when using G73 it says that I use a "q word without G code using it" but when I remove the Q word, EMC tells me that G73 needs one. What am I doing wrong ? (cvs version from yesterday)
[18:32:42] <cradek> uh-oh
[18:36:47] <CIA-47> EMC: 03cradek 07TRUNK * 10emc2/src/emc/rs274ngc/interp_check.cc: fix incorrect "Q word with no g code that uses it" error message with G73, caused by an oversight in the nurb/spline additions.
[18:37:05] <cradek> update your cvs and retest please
[19:08:14] <RobJansen> Thanks, machine is now running but I'll test it tomorrow.
[19:09:15] <cradek> ok
[20:23:05] <skunkworks> THis is interesting. You can read the messageboard as a pdf.. /
[20:23:06] <skunkworks> http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/id,189/catid,16/func,fb_pdf/lang,it/
[20:27:29] <alex_joni> wonder how you ended up with the lang,it in there
[20:35:56] <skunkworks> alex_joni: http://www.google.com/search?hl=en&q=D%26M+Computing+048-83970+&btnG=Google+Search&aq=f&oq=