#emc | Logs for 2004-12-26

[00:08:32] <cradek> you don't have to escape the space
[00:09:11] <alex_joni> I agree
[00:09:23] <alex_joni> but it's escaped in rcslib/etc/generic.def
[00:09:30] <cradek> or the "
[00:09:45] <alex_joni> probably some strange awk on a strange platform (sun, vxworks, win... who knows)
[00:09:54] <alex_joni> that needs it
[00:10:04] <alex_joni> cradek: hi
[00:10:08] <cradek> hello
[00:10:17] <cradek> if they're that strange they won't accept an RE for -F
[00:10:17] <alex_joni> you know more about compiling (make) than me ;)
[00:10:23] <cradek> I think that's a gnuism
[00:10:37] <alex_joni> never mind
[00:10:49] <alex_joni> anyways... I wanted to ask you one thing
[00:10:56] <alex_joni> that's not fully clear to me
[00:11:08] <alex_joni> the foo.def gets included from the Makefile
[00:11:10] <alex_joni> right?
[00:11:32] <cradek> I've avoided the emc makefiles so far
[00:11:35] <alex_joni> first generic.def gets included, then determineplat.def, then plat.def
[00:11:41] <cradek> are you talking about emc1 or 2?
[00:11:44] <alex_joni> 1
[00:12:10] <alex_joni> things inside foo.def are things that make takes into consideration ... I think
[00:12:39] <alex_joni> like CFLAGS=...
[00:12:49] <alex_joni> CC=gcc
[00:12:50] <alex_joni> etc.
[00:12:57] <cradek> sure, the .def are just makefile parts
[00:13:03] <alex_joni> ok...
[00:13:11] <alex_joni> I added a line: RTAI_PLAT=1
[00:13:27] <alex_joni> which doesn't get taken into consideration
[00:13:43] <cradek> don't know what you mean by that
[00:13:50] <alex_joni> if I run 'make PLAT=realtime RTAI_PLAT=1 all' it works
[00:13:51] <cradek> it sets the make varilable RTAI_PLAT to the string 1
[00:14:10] <alex_joni> but running 'make PLAT=realtime all' doesn't
[00:14:20] <cradek> so it's probably set after it's referenced
[00:14:26] <alex_joni> even if RTAI_PLAT=1 is defined inside realtime.def
[00:14:40] <alex_joni> you might have a point there
[00:14:43] <alex_joni> lemme check
[00:15:42] <alex_joni> yup... you are right
[00:15:53] <alex_joni> the include is at the bottom of the Makefile :(
[00:16:13] <alex_joni> that's pretty bad
[00:16:42] <alex_joni> till now the Makefiles figured stuff out by looking at the PLAT name (e.g. rtai_foo.def)
[00:17:07] <cradek> yay jmk fixed multi-axis jogs in emc2
[00:17:13] <alex_joni> if you would have a rtai.def that's actually for rtlinux (inside it), it would get messy
[00:17:38] <alex_joni> Makefile decides based on the PLAT-name, what it should do :(
[00:17:58] <alex_joni> * alex_joni is trying to make a autoconf (./configure) for emc1
[00:18:08] <alex_joni> without touching the Makefiles
[00:18:08] <cradek> why??
[00:18:15] <alex_joni> why what?
[00:18:28] <cradek> why add autoconf to emc1?
[00:18:30] <alex_joni> why the ./configure?
[00:18:34] <cradek> yeah
[00:18:37] <alex_joni> to make it more easy to use
[00:18:50] <alex_joni> the ./configure generates the .def files
[00:18:57] <cradek> ah
[00:19:20] <alex_joni> kinda like the Makefile.inc on emc2
[00:19:34] <cradek> this sounds very hard and with not much payoff...
[00:20:12] <alex_joni> rcslib is kinda compiling
[00:22:28] <alex_joni> can you take a look at a file?
[00:22:37] <cradek> which one?
[00:22:39] <alex_joni> rcslib/src/posemath/Makefile
[00:22:59] <cradek> ok
[00:23:26] <alex_joni> there is a include ../Makefile.rcs
[00:23:38] <cradek> ok
[00:23:38] <alex_joni> which actually includes my def file
[00:23:57] <alex_joni> but if I move it to the top.. make complains about no sources
[00:24:33] <cradek> what's the actual error?
[00:24:50] <alex_joni> No sources.
[00:25:49] <cradek> "No sources" isn't a make error - it must come from the makefile somwhere
[00:25:55] <cradek> somewhere
[00:26:35] <cradek> ifndef SRCS
[00:26:35] <cradek> depend_commands:
[00:26:35] <cradek> $(IGNORE_MARK)$(ECHO_MARK)echo No sources.
[00:27:22] <cradek> so that means the make variable SRCS is not set
[00:27:30] <alex_joni> but SRCS should be $(CSRCS)
[00:27:35] <alex_joni> if RTAI_PLAT is set
[00:27:41] <cradek> it is set at line 277 of your Makefile
[00:27:42] <alex_joni> how can I test RTAI_PLAT?
[00:27:46] <alex_joni> can I echo?
[00:28:55] <alex_joni> SRCS should get set no matter what...
[00:29:56] <cradek> discard := discard := $(shell echo $(RTAI_PLAT) 1>&2)
[00:30:01] <cradek> oops
[00:30:03] <cradek> discard := $(shell echo $(RTAI_PLAT) 1>&2)
[00:30:25] <cradek> this is how you can echo a variable without invoking a rule
[00:32:17] <alex_joni> well RTAI_PLAT is 1
[00:32:28] <alex_joni> so is BUILD_REALTIMEONLY
[00:34:25] <alex_joni> I know why No sources gets printed
[00:34:36] <alex_joni> SRCS is not set when generic.def is called
[00:34:57] <alex_joni> that's why the include must be after the SRCS
[00:35:05] <cradek> right
[00:35:46] <alex_joni> but if it's after I can't figure out what files to build
[00:35:55] <alex_joni> because my PLAT-name is realtime
[00:36:00] <alex_joni> not rtai_realtime
[00:36:06] <alex_joni> nor rtlinux_realtime
[00:36:28] <alex_joni> seems I need to change the Makefile after all
[00:36:40] <cradek> sounds like
[00:38:30] <alex_joni> works now :(
[00:38:34] <alex_joni> don't like it thou
[00:38:56] <alex_joni> I really don't like the way the Makefile is written
[00:39:16] <alex_joni> grepping the platname for the actual RT doesn't seem right
[00:39:16] <cradek> I don't like any of them but I keep quiet because I don't want to redo them
[00:39:34] <alex_joni> don't think it can be redone...
[00:39:39] <alex_joni> it would break too much
[00:39:54] <alex_joni> as far as I see it: emc1 will be fixed
[00:40:01] <alex_joni> and emc2 gets developed
[00:40:14] <alex_joni> emc1: only bugfixes, minor improvement
[00:40:21] <alex_joni> emc2: all the good stuff ;)
[00:40:35] <cradek> do others not share that vision? That's the way I see it too
[00:40:50] <alex_joni> I think all share this point
[00:41:02] <alex_joni> so... I find it usefull to mock about emc1
[00:41:13] <alex_joni> so that it doesn't get repeated in emc2
[00:41:40] <cradek> well emc1 runs my machine and probably will for some time... I keep putting bugfixes in both versions
[00:41:41] <alex_joni> not mock... not the right word
[00:41:51] <alex_joni> but complain
[00:42:14] <cradek> yeah it's good to understand what's bad about it to avoid repeating the bad stuff
[00:42:52] <alex_joni> yup.. my point exactly
[00:49:09] <alex_joni> hmm.. one more strange thing
[00:49:17] <alex_joni> I get an error when building emc
[00:49:36] <alex_joni> no rule to make cannon.hh (needed by emc.cc)
[00:49:48] <alex_joni> in /emc/src/emcnml/
[00:51:15] <cradek> emc.cc: emc.hh $(DEVP_INCLUDE_DIR)/canon.hh emc.gen
[00:51:22] <alex_joni> yup
[00:51:25] <cradek> that is probably coming from this line
[00:51:27] <alex_joni> that's the one
[00:51:36] <cradek> means emc.hh can't be found and there is no make rule to create it
[00:51:47] <cradek> of course we know make isn't supposed to create it
[00:51:58] <cradek> but make doesn't know that
[00:52:01] <alex_joni> it's complaining about canon.hh
[00:52:16] <alex_joni> make should cp it to $(DEVP_INCLUDE_DIR)
[00:52:18] <cradek> oh right
[00:52:24] <cradek> or, DEVP_INCLUDE_DIR isn't set
[00:53:33] <alex_joni> it is set
[00:53:42] <alex_joni> I tried a make PLAT=foo printenv
[00:55:33] <alex_joni> hmmm.. just a minute
[00:55:45] <alex_joni> for realtime .cc code shouldn't get compiled
[00:59:37] <alex_joni> I really am tired...
[00:59:44] <alex_joni> found a typo :(
[01:18:19] <alex_joni> * alex_joni is fighting with CMS
[01:24:35] <alex_joni> * alex_joni wonders when paul_c returns
[01:27:27] <jmkasunich> maybe he's sleeping?
[01:38:54] <alex_joni> I definately like emc2 better
[01:38:57] <alex_joni> way better ;)
[01:39:35] <alex_joni> jmk: never ever do these things with makefiles ...
[01:41:04] <jmkasunich> I won't - I like the build system like it is
[01:41:18] <jmkasunich> BTW, thank you very much for the autoconf work on emc2
[01:41:46] <alex_joni> I'm glad it's wornking ;)
[01:41:54] <alex_joni> I'm struggling on emc1 now
[01:42:03] <alex_joni> but.. all I can say is... yucky
[01:42:20] <alex_joni> the Makefiles figure out what system is based on the plat name
[01:42:21] <alex_joni> :(
[01:42:50] <alex_joni> e.g. if it's a RTAI system but the PLAT is called rtlinux-foo, you're in BIG trouble
[01:43:12] <alex_joni> my plat is called realtime...
[01:43:28] <alex_joni> so neither rtai nor rtlinux don't get found in the name
[01:43:30] <alex_joni> :(
[02:45:14] <paul_c> alex_joni: Mail me the stuff to date and I'll look it over in the morning
[02:45:34] <paul_c> Although if that damned alarm keeps going off....
[03:09:00] <alex_joni> * alex_joni is going to bed
[03:09:03] <alex_joni> night...
[03:13:28] <cradek> jmkasunich: you around?
[03:19:50] <cradek> hello
[03:23:01] <jepler> hi chris
[03:23:31] <jepler> cradek: was it much work to move the spindles around?
[03:23:43] <jepler> cradek: do you still have software spindle control this way?
[03:24:07] <cradek> hi jeff
[03:24:22] <cradek> it was a bit of fiddling to get the mounting to work
[03:24:31] <cradek> it's not pretty but it's at least as solid as the original motor
[03:24:38] <cradek> there's no software control right now.
[03:26:00] <jepler> that's too bad
[03:26:31] <cradek> I could put the knob in the rotary axis and have software speed control!
[03:27:02] <jepler> that's, uh, slightly perverse
[03:27:14] <cradek> I do definitely want to figure out a way to get on/off
[03:27:43] <cradek> in fact I think I'll open up the controller box now and see what's in it
[04:28:06] <jmkasunich> cradek: you still here?
[04:59:06] <cradek> yes
[04:59:15] <cradek> (he said, belatedly)
[04:59:19] <jmkasunich> you were looking for me earlier
[04:59:24] <cradek> oh right
[04:59:34] <cradek> I was going to tell you that you fixed multi-axis jogging
[04:59:38] <cradek> it works great
[04:59:40] <cradek> (with AXIS)
[04:59:43] <jmkasunich> cool
[05:00:18] <cradek> that's all!
[05:00:20] <jmkasunich> do you know anything about burning CD-Rs under Linux?
[05:00:26] <cradek> sure
[05:00:37] <jmkasunich> cdrecord -scanbus gives:
[05:00:59] <jmkasunich> No such file or directory, Cannot open SCSI driver
[05:01:22] <cradek> you have an IDE burner I assume
[05:01:26] <jmkasunich> yes
[05:01:51] <jmkasunich> I burned a CD once, months ago, Paul walked me thru it...
[05:01:58] <cradek> hmm
[05:02:07] <cradek> what kernel do you have?
[05:02:09] <jmkasunich> now, when I desperately need to burn one, I don't remember
[05:02:11] <jmkasunich> 2.4
[05:02:16] <jmkasunich> BDI-TNG
[05:02:30] <jmkasunich> 2.4.18-rtai to be precise
[05:02:33] <cradek> ok I have a 2.4.20 machine here with a burner
[05:02:38] <jmkasunich> basically Redhat 7.2
[05:02:39] <cradek> and it gives me the same error!
[05:02:47] <cradek> hmmmm
[05:02:50] <jmkasunich> that's a good thing ;-)
[05:02:57] <cradek> for you maybe!
[05:03:11] <cradek> now I also have a 2.4.20 machine where it works
[05:03:15] <cradek> let me study them and find the difference
[05:03:49] <jmkasunich> * jmkasunich has been reading the CD-R Howto... it was last updated in 2000
[05:04:14] <jmkasunich> I think a module "ide-scsi" is needed
[05:04:28] <cradek> is your CD hdc?
[05:04:41] <cradek> you may need to have hdc=ide-scsi on your kernel command line
[05:05:43] <jmkasunich> how do I check if the CD is hdc? fstab says "/dev/cdrom /mnt/cdrom iso9660 noauto,users,ro 0 0"
[05:05:53] <cradek> cat /proc/ide/ide1/hdc/model
[05:06:15] <jmkasunich> CD-540E
[05:06:24] <cradek> that sounds like it
[05:06:31] <cradek> also ls -l /dev/cdrom
[05:06:56] <jmkasunich> wtf?!? CD-540E is a CD-ROM, not a burner
[05:07:05] <jmkasunich> no wonder it doesn't work
[05:07:17] <cradek> hmm
[05:07:29] <jmkasunich> I'm 99.9% sure I burned a disk on this box once before
[05:08:05] <jmkasunich> maybe it was on my other box (W95/Linux dual boot, that one definitely has a burner in it)
[05:08:53] <cradek> yeah TEAC CD-540E is definitely not a writer
[05:09:06] <jmkasunich> ok, booting the other box
[05:09:26] <jmkasunich> next question... do you know anything about NFS? ;-)
[05:09:36] <jmkasunich> I have two .iso files on this box
[05:09:53] <cradek> sure I know nfs
[05:09:58] <jmkasunich> 100Mb ethernet between the two
[05:10:07] <jmkasunich> burner in the other one
[05:10:11] <jmkasunich> you see my problem ;-)
[05:10:27] <cradek> do you have ssh, rsh, ftp or anything else between them?
[05:10:37] <cradek> you can copy the file with any of those
[05:10:49] <cradek> (easier than nfs)
[05:11:07] <jmkasunich> probably not, but I'll check
[05:11:20] <jmkasunich> other one still booting
[05:11:35] <cradek> ok I'm going to reboot my server machine (with the burner)
[05:11:43] <cradek> don't think it will affect us but if so, I'll be right back.
[05:13:28] <cradek> hello, jmk's other personality
[05:13:50] <cradek> scsibus0:
[05:13:51] <cradek> 0,0,0 0) 'HP ' 'CD-Writer+ 9100 ' '1.0c' Removable CD-ROM
[05:13:54] <cradek> yay, I fixed mine
[05:14:41] <jmkasunich2> cdrecord -scanbus on this machine works
[05:14:55] <cradek> shows up as 0,0,0?
[05:15:05] <jmkasunich2> 0,0,0 0) '52x24x52 CD-RW'
[05:15:11] <cradek> good
[05:15:16] <cradek> now all you need are your files
[05:15:25] <cradek> do you have disk space for them available?
[05:15:48] <cradek> if not, we will have to use nfs
[05:16:10] <jmkasunich2> /home is full (248M)
[05:16:18] <jmkasunich2> but / has 2.6G
[05:16:25] <cradek> good
[05:16:30] <cradek> is /tmp under /?
[05:16:36] <cradek> cd /tmp; df .
[05:17:00] <jmkasunich2> yep, 2.6G
[05:17:07] <cradek> ok
[05:17:17] <cradek> see if you can ftp otherbox, or ssh otherbox, or rsh otherbox
[05:17:22] <jmkasunich2> I can't ping from here to the machine with the files
[05:17:27] <cradek> hmm
[05:17:39] <jmkasunich2> it's on a subnet - I can disconnect/reconnect
[05:17:54] <cradek> isn't there a route between them?
[05:18:12] <jmkasunich2> the router for the subnet is blocking me...
[05:18:45] <jmkasunich> but I can ping from here (box with files) out to the one with the burner
[05:18:57] <cradek> ok, you can copy that way too
[05:19:06] <cradek> try ssh, ftp, rsh
[05:19:09] <jmkasunich> (I have the router configured for security, it was used at emcFest for the EMC net)
[05:19:21] <cradek> I understand
[05:19:35] <jmkasunich> ssh refused
[05:19:57] <jmkasunich2> ssh isn't running on this box
[05:20:04] <jmkasunich2> should I just try starting sshd?
[05:20:08] <cradek> sure
[05:20:40] <cradek> you won't be able to use NFS since with this setup you need to "push" the files and NFS is a "pull"
[05:21:22] <jmkasunich2> there's a "sshd" script in /etc/rc.d/init.d... I should just be able to do "sshd start" as root, right?
[05:21:29] <cradek> service sshd start
[05:21:52] <jmkasunich2> service: command not found
[05:22:02] <jmkasunich2> I have used other init scripts directly
[05:22:18] <jmkasunich2> starting sshd: OK
[05:22:44] <jmkasunich> coo - I'm in
[05:22:48] <cradek> ok
[05:22:52] <cradek> get back out then
[05:22:58] <jmkasunich> logout
[05:23:00] <cradek> scp thefile.iso othermachine:/tmp
[05:23:03] <jmkasunich> ops
[05:23:21] <jmkasunich> should I do this as user or root?
[05:23:28] <cradek> doesn't matter for the copy
[05:23:32] <cradek> however ssh worked
[05:24:27] <jmkasunich> it's going
[05:24:33] <jmkasunich> 11 minutes eta
[05:24:39] <cradek> ugh
[05:25:08] <jmkasunich> probably would have been faster if I took the burner box down and reconnected it directly to the subnet... the router might be the bottleneck
[05:25:31] <cradek> that must be 10 Mbit
[05:25:49] <jmkasunich> crap... I just remembered it is
[05:26:05] <jmkasunich> the router (a compile farm slot) has a built-in 10/100 port, that one is on the subnet
[05:26:12] <cradek> oh well, it's only ten minutes
[05:26:13] <jmkasunich> the upstream connection is 10MB
[05:27:08] <jmkasunich> actually, I need to move this machine outside the subnet anyway, cause I'm gonna take the farm down tonight... don't want to leave it powered while we're gone for a week
[05:27:29] <jmkasunich> I'll do that before I transfer the second file
[05:27:41] <cradek> you can do that while the first one is burning
[05:27:45] <jmkasunich> right
[05:28:19] <cradek> I think once you get it there, you only need to do (as root) cdrecord -v dev=0,0,0 /tmp/thefile.iso
[05:28:34] <jmkasunich> yep.
[05:29:06] <jmkasunich> I got the iso's from till's website, he had a sample command there...
[05:29:13] <jmkasunich> he uses -pad and -data
[05:29:32] <jmkasunich> the cdrecord man page says -pad is only for audio, and -data is the default I think
[05:29:56] <cradek> I've never used either of those
[05:30:03] <cradek> (never burned a music cd with it)
[05:30:26] <cradek> did you see I fixed helix feed?
[05:30:32] <jmkasunich> saw the email
[05:30:41] <cradek> I'm even pretty sure it's right!
[05:31:03] <jmkasunich> that's deep into code I don't really understand yet, so I can't tell you if it's right or not
[05:31:30] <cradek> I figured out enough to be pretty confident of the change
[05:31:38] <cradek> and, the results are right
[05:31:45] <jmkasunich> great!
[05:32:11] <cradek> I guess nobody else ever noticed, but it was sure wrong. Maybe nobody does helixes.
[05:32:20] <jmkasunich> I'm gonna be away for a week... I'm hoping I can do some EMC hacking while I'm gone
[05:32:42] <cradek> vacations shouldn't involve computers
[05:32:45] <jmkasunich> (that's why I'm burning the CD - need to install BDI-TNG on a laptop
[05:33:02] <jmkasunich> we're going to visit my wife's parents
[05:33:12] <cradek> haha
[05:33:18] <cradek> ideally, vacations don't have in-laws either
[05:33:39] <jmkasunich> past history - about 20% of the day is socializing and other reasonably fun stuff, the rest of the time I'm bored out of my head
[05:33:56] <cradek> I understand completely
[05:34:07] <jmkasunich> I find myself watching the History Channel (we don't have cable here)
[05:34:32] <cradek> we don't have it either - TV is a time suck
[05:34:39] <jmkasunich> yep
[05:34:52] <jmkasunich> I watch about 2 hours a weel
[05:34:55] <jmkasunich> week
[05:35:04] <jmkasunich> Lost and Enterprise ;-)
[05:35:25] <cradek> I watch maybe an hour a month - when visiting someone
[05:35:42] <jmkasunich> when I was single, I went about 5 years without a TV
[05:35:54] <cradek> your file should be ready
[05:36:00] <jmkasunich2> yep
[05:36:54] <jmkasunich2> do I load the disk first, or does cdrecord prompt?
[05:37:01] <cradek> I think you load it first
[05:37:08] <jmkasunich2> darn
[05:37:22] <jmkasunich2> * jmkasunich2 tries to remember which of the unlabeled CD drives is the burner
[05:37:36] <jmkasunich2> probably the one without the earphone jack
[05:38:07] <cradek> eject /dev/hdc
[05:38:22] <jmkasunich2> right one
[05:38:52] <jmkasunich2> but cdrecord isn't happy
[05:38:58] <jmkasunich2> cdrecord: No disk / Wrong disk!
[05:39:08] <cradek> did the lights blink?
[05:39:26] <cradek> maybe just try again
[05:39:34] <cradek> maybe you didn't wait long enough after loading
[05:39:37] <jmkasunich2> blinked for a while when I loaded the disk, then went out...
[05:39:55] <cradek> I mean did it do anything when you ran cdrecord?
[05:39:59] <jmkasunich2> tried cdrecord after it stopped blinking... no lights, error msg immediately
[05:40:12] <cradek> cdrecord -scanbus again?
[05:40:13] <jmkasunich2> less than 1 sec
[05:40:26] <cradek> you are doing it as root, right?
[05:40:31] <jmkasunich2> yep
[05:40:42] <jmkasunich2> * jmkasunich2 is reading the cdrecord output
[05:41:23] <cradek> Sense Code: 0x3A Qual 0x00 (medium not present) Fru 0x0
[05:41:26] <cradek> do you get this?
[05:41:33] <jmkasunich2> working now
[05:41:40] <cradek> what did you change?
[05:41:40] <jmkasunich2> it was the wrong drive
[05:41:46] <cradek> oh!
[05:41:52] <jmkasunich2> apparently 0,0,0 is not mapped to hdc
[05:41:58] <jmkasunich2> I put the disk in the other drive
[05:42:15] <cradek> ok then, I'm going to bed
[05:42:24] <cradek> you're set now, right?
[05:42:28] <jmkasunich2> thanks for the help
[05:42:33] <cradek> no problem
[05:42:34] <jmkasunich2> I would have been here all night
[05:42:44] <cradek> have a good trip
[05:42:49] <jmkasunich2> thanks
[05:43:27] <jmkasunich2> I'll be reading email (probably) but not on IRC until after newyears
[05:44:07] <jmkasunich2> hmmm. got some error messages
[05:46:48] <jmkasunich> disconnecting to reconfigure network
[10:27:42] <jmkasunich> I really hate laptops
[15:44:31] <jepler> not a very busy developer meeting
[15:44:55] <les> pretty quiet
[15:47:46] <robin_sz> its boxing day ffs.
[15:47:57] <robin_sz> coding shoud be illegal on boxing day :)
[15:48:13] <les> hey robin
[15:50:18] <les> I'm just giving a shop tour today
[15:50:25] <les> otherwise goofing off
[15:51:03] <jepler> looks like someone is trying to write a g-code-like language with looping (but not getting very far): http://mail.python.org/pipermail/python-list/2004-December/257253.html
[15:52:14] <les> the inability to do loops made for some awfully nasty gcode in this production thing I have
[15:53:11] <les> 45 up...six tool changes...all with G5x and G10
[15:54:05] <les> but I understand the problem with loops in emc...it's the queue I think
[15:54:57] <robin_sz> hey les
[15:55:12] <jepler> It's tempting to add a gcode preprocessor to AXIS, but I doubt I can design it right
[15:55:18] <robin_sz> going to go fly random model planes with the kids in a minute :)
[15:55:28] <robin_sz> lets check the gear:
[15:55:31] <robin_sz> plane yep
[15:55:36] <robin_sz> tx yep
[15:55:47] <robin_sz> plastic back to carry the remains back .. yep
[15:56:02] <robin_sz> s/back/bag/
[15:56:14] <les> I have problems with rc planes...they turn into a dot then the dot goes down
[15:56:53] <les> I could do those new ones with video though
[15:58:26] <les> Jepler: I am still having the most problems with emc's trajectory planner
[15:58:35] <les> it's just not good
[15:59:30] <les> I can easily do the maths for a good planner but I just can't make heads or tails of the code with all those globals
[16:00:25] <cradek> morning
[16:00:51] <les> hi chris
[16:01:13] <robin_sz> les: heres a good test ...
[16:01:23] <robin_sz> if it has lots of globals ... its broken
[16:02:01] <les> no argument there heheh
[16:03:32] <les> it drives me nuts because emc is so filled with good features...but has an anomalous garbage traj planner
[16:05:04] <les> and a good planner is what...100 lines of code?
[16:08:29] <paul_c> Heey Robin - Can that laser of yours nail a target given GPS coords ?
[16:09:53] <cradek> haha
[16:11:09] <les> I think we still have SA dithering on with gps
[16:11:40] <les> We were supposed to turn it off but I think that quietly died
[16:11:49] <cradek> so it takes several shots
[16:12:14] <les> yeah...with 100 meters uncertainty I guess
[16:12:26] <paul_c> Don't care how many shots it takes...
[16:12:57] <paul_c> As long as someone can take out that damned burgler alarm that's been going off for the last two days.
[16:13:14] <les> car?
[16:13:19] <paul_c> House
[16:13:37] <cradek> urgh
[16:13:45] <paul_c> Car would have been easy - A can of gas & a match..
[16:13:52] <les> heh
[16:14:30] <cradek> well they're advertising that there's lots of time to break in to disable it, and nobody is watching
[16:14:58] <les> I don't even lock the doors when I go out
[16:14:59] <paul_c> Be my guest
[16:15:13] <les> but always a car in the driveway and leave tv on
[16:35:43] <robin_sz> * robin_sz suggest builders PU foam in a can
[16:36:02] <robin_sz> or a .308 cal
[16:36:25] <les> haw
[16:37:00] <les> well better go straighten up tha shop for the little tour....
[16:37:25] <les> I need to build some more space and get the electronics lab area elsewhere
[16:48:56] <jepler> this is an interesting stepper control idea that I had not seen before: nkm3517 uses two supply voltages. when a winding turns on, it is given Vss >> Vss for a short time, then Vss for the remaining on-time. (Vmm allows current to rise fast)
[16:49:51] <jepler> er, Vss >> Vmm for a short time, then Vmm for the remaining on-time
[16:50:00] <jepler> (Vss allows current to rise fast)
[16:50:42] <jepler> and the chip is njm3517, not nkm...
[16:50:51] <jepler> is there anything I can say right the first time?
[17:12:24] <paul_c> dave-e: Going to have some new stuff for you to breack^H^H^Htest
[17:12:52] <dave-e> I'll probably be able to break it. ;-)
[17:13:18] <paul_c> Thought you might..
[17:13:34] <dave-e> what is the date of the cvs/emc included?
[17:14:02] <paul_c> don't follow...
[17:14:20] <dave-e> how current is the emc src
[17:14:41] <paul_c> in SF ?
[17:14:59] <dave-e> no in the new bdi or what ever you want me to break
[17:15:24] <paul_c> You are NOT breaking the latest BDI !
[17:15:48] <dave-e> Oh! you mean I can't break it or don't get a chance.
[17:16:01] <paul_c> Got some autoconfig stuff going in to rcslib shortly
[17:16:16] <dave-e> Oooooh!
[17:16:27] <paul_c> and a simple make script for the top level emc dir
[17:16:37] <dave-e> it is sounding better
[17:17:23] <dave-e> simple is good
[17:17:42] <dave-e> after my main objective is to make chips not test software
[17:17:49] <dave-e> ...after all...
[17:18:40] <dave-e> I've moved my test bench....motenc + single axis servo to a friends electronic test bench...
[17:18:55] <dave-e> we'll at least change enviroments that way
[17:19:19] <dave-e> and do a very structured reconnect of the system
[17:20:58] <paul_c> You want BDI-4.05 to test it on ?
[17:22:02] <dave-e> well...is should include a runnable vital... (right?) and therefore be a test from the box without a compile???
[17:22:03] <les> dave: motenc fully functional?
[17:22:24] <dave-e> les...not yet ... I still have hopes
[17:22:31] <les> haha ok
[17:22:37] <paul_c> * paul_c emails dave-e a copy of the iso...
[17:22:53] <dave-e> I'll make lots of noise when I get it to run correctly
[17:23:22] <dave-e> paul...sure glad I have cable
[17:24:05] <paul_c> Most mail servers have a limit of 10Megs
[17:24:39] <dave-e> can't squeeze any decent iso thru that
[17:28:07] <dave-e> paul...you going to hang the 4.05 out on a server?
[17:30:44] <paul_c> 4.06 will probably be the first public release
[17:39:44] <paul_c> * paul_c has to go and throw some food at a wrinkly
[17:40:09] <dave-e> enjoy...I'm just making the first pot of coffee for the day
[17:40:20] <dave-e> slow start this morning
[17:44:54] <dave-e> les ... what input (voltage) will just move your axis...on the gantry?
[17:54:09] <dave-e> I'm off to the shop....catch people later.
[19:36:21] <les> was out in the shop...but if you see this...
[19:37:11] <les> about a volt or two will break friction
[20:39:06] <CIA-6> 03paul_c * 10rcslib/ (6 files in 3 dirs):
[20:39:06] <CIA-6> Autoconfigure support with minimal changes to the source code - This should mean
[20:39:06] <CIA-6> that existing platforms will work whilst allowing newer configurations to be
[20:39:06] <CIA-6> added with minimal effort.. Should also cut down on the confusion when building
[20:39:06] <CIA-6> EMC.. (Thanks to Alex Joni for the *def.in work.)
[21:15:31] <alex_joni> hello
[21:17:04] <alex_joni> anyone around?
[21:58:48] <Imperator_> Hi Laex
[21:59:06] <Imperator_> alex_joni: Hi Alex
[22:02:19] <alex_joni> hi martin
[22:03:40] <alex_joni> what's up?
[22:04:58] <Imperator_> I have survived christmas :-)
[22:05:50] <Imperator_> do you need a nice CNC-Controller ???
[22:05:52] <Imperator_> http://cgi.ebay.de/ws/eBayISAPI.dll?ViewItem&rd=1&item=3863000699
[22:10:36] <alex_joni> no thanks, I'll stay with emc ;)
[22:10:40] <alex_joni> something fun to watch:
[22:10:41] <alex_joni> http://www.allmad.shyper.com/movies/ihatec.html
[22:13:17] <alex_joni> hmmm... that CNC-Controller is in Aalen
[22:13:26] <alex_joni> thought you were there.. right?
[22:36:19] <alex_joni> hello ray
[22:36:31] <rayh> Hi Alex.
[22:36:56] <alex_joni> seems after all the elections were pretty ok
[22:37:03] <alex_joni> down here
[22:37:20] <rayh> That's good news.
[22:37:28] <alex_joni> yeah...
[22:37:47] <alex_joni> I sent paul some stuff, and he commited it to rcslib
[22:37:56] <alex_joni> it's some autoconf stuff for emc1
[22:41:07] <rayh> I heard that he was going to work with auto and 1.
[22:41:36] <rayh> Been a bit cold here -34C and some wind.
[22:41:53] <alex_joni> huh... it's around 10+ here
[22:42:15] <rayh> Went walking on the lake with the grandkids and dogs.
[22:43:45] <Imperator_> very warm in romenia, here are about 0� with some snow
[22:44:21] <rayh> Nice. The kids went sledding down the hill and on the lake.
[22:45:18] <alex_joni> * alex_joni would like that too
[22:45:27] <alex_joni> hope I'll have some snow in the mountains
[22:45:28] <Imperator_> * Imperator_ hopes he gets his spicke tires before everything is covered with ice here
[22:46:27] <Imperator_> for the Mountainbike !!!
[22:46:50] <rayh> Spikes on a mountainbike. That sounds like fun.
[22:46:59] <alex_joni> yeah...
[22:48:15] <Imperator_> drivin down frozen rivers :-)
[22:49:08] <rayh> Got a few of those around here.
[22:49:16] <Imperator_> but you don't have to break on a normal street, otherwhise you here something jingle
[22:50:00] <ray_lan> Well. I've got nat working.
[22:50:19] <ray_lan> under bdi 4.02
[22:50:48] <alex_joni> coo
[22:51:08] <alex_joni> * alex_joni can't wait to try BDI 4.0x
[22:51:15] <ray_lan> Guess I don't need two of me.
[22:51:37] <alex_joni> never know
[22:51:41] <alex_joni> ;)
[22:51:47] <alex_joni> meep?
[22:52:03] <ray_lan> Two would be much to much for many folk.
[22:54:10] <ray_lan> Hi Robin
[23:02:06] <Imperator_> has anybody here experience how to open a laptop batterie ?
[23:05:21] <alex_joni> use a chain-saw
[23:05:27] <Imperator_> :-
[23:05:31] <Imperator_> :-)
[23:05:43] <Imperator_> it's realy good closed
[23:05:44] <jepler> I was gonna suggest hammer
[23:06:17] <Imperator_> thats a good tool for everything
[23:07:24] <alex_joni> hammer is ok, but chain-saw is more impressive
[23:07:28] <alex_joni> ;)
[23:09:33] <Imperator_> no the Hammer was a good tip
[23:10:09] <Imperator_> the housing was glued together, some hitts with a hammer has broken the glue
[23:16:46] <jepler> did somebody here have experience with this board? http://www.xylotex.com/#4axis
[23:18:44] <jepler> .. and how do I calculate "back-EMF" to be sure I'm within the 35V?
[23:19:37] <alex_joni> jepler: best for back-EMF is to put some electronics to suppress it
[23:19:55] <alex_joni> I remember seeing some schematics for a gecko powersupply
[23:35:27] <jepler> alex_joni: the xylotex apparently doesn't. Are the "electronics" in question the 8 diodes in the l297/l298 stepper driver?
[23:37:28] <alex_joni> nope
[23:38:03] <alex_joni> as I recall some combination of a transister, a voltage comparing device and a large resistor to disipate the heat
[23:38:10] <alex_joni> e.g. the extra voltage
[23:38:37] <jepler> I must not understand at all what back EMF is, then
[23:39:00] <alex_joni> maybe I misunderstood it
[23:40:27] <alex_joni> it's when you want to stop the motor, and the inertia makes it run like a generator, thus bringing voltage into the power supply?
[23:44:50] <paul_c> Any inductor will store energy - When the magnetic field collapses, this energy is dumped back in to the circuit..
[23:45:10] <paul_c> This gives an initial spike on the power rails...
[23:45:54] <paul_c> and then there is the voltage generated by the "dynamo" effect of back driving the motor.
[23:46:43] <alex_joni> hi paul
[23:46:56] <paul_c> Yo Alex
[23:47:03] <alex_joni> seen your commit
[23:47:27] <paul_c> Yup - Got some work to do on the makefiles.
[23:47:43] <alex_joni> rcslib did compile for me
[23:47:49] <paul_c> Did you get the emails ?
[23:47:49] <alex_joni> I'm struggling with emc now
[23:47:54] <alex_joni> yup
[23:48:22] <alex_joni> I agree on the GTK, message from configure.in (removed those)
[23:48:29] <paul_c> I can't get the realtime stuff to compile on this box..
[23:48:37] <alex_joni> rcslib?
[23:48:50] <alex_joni> or emc?
[23:48:53] <paul_c> Any of the RT code in rcslib/emc
[23:49:01] <alex_joni> strange...
[23:49:03] <alex_joni> 2.6?
[23:49:11] <paul_c> 2.6.9-adeos
[23:49:20] <alex_joni> maybe that's one reason
[23:49:22] <alex_joni> :D
[23:49:30] <paul_c> Nope...
[23:49:32] <alex_joni> how's autoconf on emc2 working?
[23:49:39] <alex_joni> for the 2.6?
[23:49:56] <paul_c> It IS the reason the RT code doesn't compile.
[23:51:21] <paul_c> autoconf in the emc2 tree produces all the rquired flags...
[23:52:01] <alex_joni> jepler: do you have an yahoo account?
[23:54:21] <alex_joni> paul_c: found a flag not suited for emc
[23:54:26] <alex_joni> -D__KERNEL__
[23:54:46] <alex_joni> it was generating errors for emcmot/ sources (e.g. dro.c)
[23:54:46] <paul_c> in nonrealtime or realtime ?
[23:54:50] <alex_joni> realtime
[23:55:02] <paul_c> what errors ?
[23:55:46] <alex_joni> include problems
[23:56:36] <alex_joni> there's a large error from /usr/include/linux/version.h
[23:56:58] <paul_c> No, no, no...
[23:57:11] <alex_joni> "You should not include /usr/include/{linux,asm}/ header files directly for the compilation of kernel modules...
[23:57:21] <paul_c> Never, ever use usr headers in kernel modules
[23:57:42] <alex_joni> I don't
[23:57:45] <paul_c> and never include kernel headers in usr apps
[23:58:02] <paul_c> 'cept for te ones in /usr/include/linux
[23:58:09] <alex_joni> in dro.c there is sys/io.h included (or was it asm/io.h?)
[23:58:21] <alex_joni> I agree
[23:58:45] <alex_joni> but something's fishy with the includes here... can't seem to figure out what
[23:59:15] <alex_joni> jepler: on the yahoo group: geckodrive there is the pdf I was talking about
[23:59:25] <alex_joni> jepler: http://f6.grp.yahoofs.com/v1/cELPQYWBqjLA3Y-ZKKRTdkDdNubxMetZ1S3U84IO90M9g0xWPLt6LkAmW8Yd_4cLm_GTaig1n0AEOzHEpdtmz9fe-wu37RA/G320%20and%20G340/POWER%20SUPPLY.pdf