#emc | Logs for 2007-05-14

[00:03:28] <cradek> going from "the cam makes me set a surprising setting before it works right" to "emc's docs are wrong" is quite a jump, that's all I was saying
[00:03:41] <cradek> oops, wrong channel
[00:25:57] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[00:59:33] <Jymmm> SWPadnos is now known as the artist formaly known as SWPadnos
[00:59:53] <SWPadnos> that's "Mr. Symbol" to you!
[01:00:06] <Jymmm> heh
[01:00:30] <Jymmm> Still no email from that seller =(
[01:00:46] <SWPadnos> bummer
[01:01:38] <Jymmm> Oh, it's only 9am there, eeeesh
[01:05:12] <Jymmm> I hope he has three of them <crosses finger>
[01:06:06] <SWPadnos> heh
[01:07:03] <eric_u> I thought the step/dir version was the 670
[01:07:17] <eric_u> I don't think I would want the indexer version
[01:07:18] <Jymmm> 670 is servos iirc
[01:07:24] <eric_u> ok
[01:07:45] <eric_u> forgot about that
[01:08:40] <Jymmm> skunkworks said the indexer daughterboard can be removed.
[01:09:17] <eric_u> I would guess so, those drives are modular
[01:09:31] <eric_u> for example, the step/dir board comes off the servo version
[01:10:52] <Jymmm> speak of the devil....
[01:11:03] <Guest308> uh oh
[01:11:08] <Guest308> oops
[01:11:17] <Guest308> Guest308 is now known as skunkworks_
[01:13:17] <Jymmm> skunkworks_ nah, just the parker drives
[01:27:41] <skunkworks_> skunkworks_ is now known as skunkworks
[04:22:24] <ohiopctechDOTcom> hey i think my e-mail to the dayton linux user group sparked a little bit of interest in emc2. there was a reply about someone (a forensic engineer) wanting to build a cnc for thier lab!
[04:22:29] <ohiopctechDOTcom> sounds interesting eh?
[04:22:52] <ohiopctechDOTcom> the guy said he downloaded the iso and is gonna scope the project out..
[04:50:57] <eric_u> I was away getting my r/c simulator to work
[04:56:05] <Jymmmm> FreeRTOS.org
[04:59:32] <ds2> nothing but pure Linux
[05:00:50] <Jymmmm> EMC on atmel uC!!!
[05:02:48] <CIA-8> 03jmkasunich 07TRUNK * 10emc2/src/hal/drivers/mesa_5i2x/firmware/businterface.vhd: add decode logic
[05:03:25] <CIA-8> 03jmkasunich 07TRUNK * 10emc2/src/hal/drivers/mesa_5i2x/firmware/stepgen.vhd: add enable, other minor tweaks
[05:04:04] <CIA-8> 03jmkasunich 07TRUNK * 10emc2/src/hal/drivers/mesa_5i2x/firmware/Makefile: FPGA makefile improvements
[05:19:06] <ds2> sure
[05:19:14] <ds2> Amtel uC's run Linux =)
[05:21:53] <ds2> EMC on the AVR32 would be cool... their $70ish eval board includes lots of GPIO
[05:34:50] <Jymmmm> 100% self contained on a uC, less the console (serial/ip)
[05:51:33] <Jymmmm> Actually, I think that's how the gecko thing works too
[06:15:57] <Skullworks_PGAB> Well I just ordered a Pluto-P... no I need some amps...
[06:16:06] <Skullworks_PGAB> now
[06:16:34] <ohiopctechDOTcom> * ohiopctechDOTcom ponders building a power supply for the new machine or just buying one
[06:17:55] <ohiopctechDOTcom> i need at least 12v 3a
[06:19:34] <Skullworks_PGAB> for the hobby cnc board?
[06:20:42] <ohiopctechDOTcom> yes
[06:21:00] <ohiopctechDOTcom> or should i pump it with the full 42v
[06:21:06] <ohiopctechDOTcom> 42v @ 3a?
[06:21:22] <ohiopctechDOTcom> or something more like 36-40v?
[06:22:16] <ohiopctechDOTcom> *i haven't got motors yet.. but thier voltagte shouldn't matter much right?
[06:22:34] <ohiopctechDOTcom> only amps.. from what i gather from the info i have read
[06:23:42] <Jymmmm> Hey, I thought you could dev AVR's in C ??????
[06:25:31] <Skullworks_PGAB> make a 24v un reg with 10 amp
[06:25:54] <Skullworks_PGAB> this will give like 34v @ 10a
[06:26:04] <Skullworks_PGAB> after the cap
[06:26:13] <ohiopctechDOTcom> i ordered a mouser catalog...
[06:26:21] <ohiopctechDOTcom> to start shopping for parts
[06:26:24] <Skullworks_PGAB> I just got mine
[06:26:35] <Skullworks_PGAB> #629
[06:29:04] <ohiopctechDOTcom> so it's pretty huge?
[06:29:06] <ohiopctechDOTcom> lol
[06:29:48] <Skullworks_PGAB> 1838 is last page of index
[06:30:18] <Skullworks_PGAB> big price diff between brands of caps
[06:30:57] <ohiopctechDOTcom> i have a boatload of electrical parts catalogs here they are either all computer or all arcade/vending machine stuff though
[06:31:40] <Skullworks_PGAB> mame keyboard emulation is an easy way out for a control panel...
[06:31:48] <ds2> you can dev AVRs in C
[06:31:51] <ohiopctechDOTcom> i was hoping to be able to salvage all the parts i needed out of atx power supplies but i don't wanna muck around with tearing them apart and checking for parts
[06:31:57] <ds2> gcc at least
[06:32:40] <ohiopctechDOTcom> anyone tried emc2 on a Gumstix PC?
[06:32:46] <ohiopctechDOTcom> that would be one tiny controller!
[06:33:21] <ohiopctechDOTcom> http://gumstix.com/
[06:33:33] <Skullworks_PGAB> I want to use the Pluto for PWM servo control
[06:33:53] <Skullworks_PGAB> and reading the Quad feedback
[06:34:22] <Skullworks_PGAB> BTW if your going to build your own PWS...
[06:34:46] <ohiopctechDOTcom> ok, i don't see a gumstix paralell i/o board
[06:35:01] <Skullworks_PGAB> the Vishay/Sprague caps are by far the best price.
[06:41:32] <Skullworks_PGAB> Mariss from Gecko offered up a link to a local PCB house that has great prices
[06:42:13] <Skullworks_PGAB> now I just need a optimal BOM and I can order PCBs made
[06:44:57] <Skullworks_PGAB> I may just make up my ganged L298 board and try that for a while.
[06:46:54] <ohiopctechDOTcom> http://www.linuxdevices.com/articles/AT8498487406.html
[06:49:02] <ohiopctechDOTcom> http://www.newmicros.com/
[06:49:35] <Skullworks_PGAB> most of MESA systems stuff is PC104
[06:51:43] <Skullworks_PGAB> I'd use the 5i@0 + the 7i33 if I could buy that Hurco KM3 that is dead.
[06:51:55] <Skullworks_PGAB> 5i20
[06:52:03] <Skullworks_PGAB> PCI card
[06:52:11] <ohiopctechDOTcom> i wish my brain didn't shutdown everytime i tried to learn any kind of programming besides basic or html
[06:52:59] <Skullworks_PGAB> I hear that
[06:53:56] <Skullworks_PGAB> never could "write C" but could debug others code... go figure.
[06:54:33] <ohiopctechDOTcom> i can compile it.. thats about it...
[06:54:38] <ohiopctechDOTcom> i got that far
[06:54:51] <ohiopctechDOTcom> and i can read and understand tiny bits of what stuff is going on
[06:55:42] <Skullworks_PGAB> yeah
[06:57:08] <Skullworks_PGAB> well its late - and I will hate life enough as is come "morning"... time to try and grab some ZZZzzz's
[06:57:57] <Skullworks_PGAB> I'm out
[06:58:01] <alex_joni> g'night
[06:58:57] <alex_joni> azi m-am trezit pa la 7:10
[06:58:59] <alex_joni> argh
[06:59:10] <alex_joni> * alex_joni sometimes hates synergy
[07:10:56] <Jymmmm> ????? http://www.google.com/search?q=define%3Asynergy
[07:12:38] <alex_joni> Jymmmm: http://synergy2.sourceforge.net/
[07:12:55] <Jymmmm> I love it, what did you do?
[08:28:57] <alex_joni> Jymmmm: I love it too, most of the time
[08:29:12] <alex_joni> but sometimes it doesn't take teh focus from a window when I leave that PC
[08:29:35] <alex_joni> so I look at the window, see it's got focus, cursor is blinking.. so I start typing (which ends up in a different window ;)
[08:30:05] <Jymmmm> how many machines?
[08:30:10] <alex_joni> only 2
[08:30:15] <Jymmmm> OS?
[08:30:22] <alex_joni> win+linux
[08:31:16] <Jymmmm> Hmmm, both mine are won, but I had to play with the settings a bit. But I have NEVER seen that happen before. My biggest issue has been reconnecting after I close th elid on my laptop, but I've got the resolved now.
[08:31:23] <Jymmmm> s/won/win/
[08:31:31] <alex_joni> s/win/loss/
[08:31:32] <alex_joni> :P
[08:31:54] <Jymmmm> You might have an issue with X cursor
[08:31:59] <alex_joni> gotta run for a while.. g'night (if I don't catch you when I'm back)
[08:32:04] <Jymmmm> G'Night
[08:32:11] <alex_joni> (no, it happens on the win box)
[08:32:27] <Jymmmm> But the X is the server?
[08:32:38] <alex_joni> right now it's another win as server
[08:32:41] <alex_joni> win+win
[08:32:44] <alex_joni> same issue
[08:32:51] <Jymmmm> what ver?
[08:32:55] <Jymmmm> same on both?
[08:33:02] <alex_joni> 1.3.1
[08:33:12] <alex_joni> yup.. same on both
[08:33:14] <Jymmmm> I'll check when I get home.
[08:33:21] <alex_joni> not that important
[08:33:28] <alex_joni> (it only happens on the client)
[08:33:49] <Jymmmm> Eh, still... might be enviroment. running anything funky? moue drivers, etc?
[08:35:36] <martin_lundstrom> Dallur: Hello any news on the THC?
[10:00:19] <Dallur> martin_lundstrom: hey Martin, no work done over the weekend i'm afraid :(
[10:03:02] <anonimasu> hey
[10:05:05] <Dallur> Anyone know how to backtrace a major/minor number to a hardware port ?, or do I have to patch the driver to make a map of hardware vs major/minor ?
[12:03:46] <danielbr> Hello, 100% cpu usage RIP Trunk?????
[12:04:45] <danielbr> 5% RIP 2.1-Branch
[12:06:40] <danielbr> I need use the new oneshot.comp
[12:08:40] <Dallur> danielbr: I think it would be helpful if you can provide a better explanation pls, are you saying that using oneshot.comp in trunk causes 100% CPU utilisation ?
[12:09:28] <danielbr> No running any config in trunk use 100% cpu
[12:09:58] <alex_joni> danielbr: there was a change in 2.1 which has been reverted for TRUNK
[12:10:04] <danielbr> it seems is broken
[12:10:06] <alex_joni> there is a 'nice' before the GUI
[12:10:15] <alex_joni> but that caused other problems for some people
[12:11:49] <alex_joni> http://cvs.linuxcnc.org/cvs/emc2/scripts/emc.in.diff?r1=;r2=
[12:13:19] <alex_joni> the main thing there is the nice before $EMC2_BIN_DIR/$EMCDISPLAY
[12:13:37] <jepler> without more details I wouldn't jump to the conclusion that this is the cause
[12:14:33] <jepler> what is the name of the process that is using 100% CPU? Do you know whether this problem has only recently begun to affect TRUNK?
[12:14:44] <danielbr> I only need use some timer, i tried copy/paste oneshot.comp from trunk to 2.1-branch but i get errors
[12:15:05] <danielbr> jepler let me see
[12:16:44] <danielbr> milltask
[12:17:23] <jepler> OK, then this is *not* due to the change alex_joni mentioned, which affects how much of the CPU the display program (e.g., axis) uses.
[12:18:19] <jepler> have you looked on the terminal or in the output of "dmesg" to see whether there are any messages that indicate a problem?
[12:20:19] <danielbr> i am not a expert i can pastebin dmesg to you if you want
[12:20:38] <jepler> OK
[12:20:44] <jepler> I will look at it
[12:21:26] <jepler> does this happen no matter what configuration file you use? I am currently using TRUNK with the sim/axis.ini configuration and I do not see 100% CPU usage from milltask (or any other part of emc)
[12:22:19] <danielbr> http://pastebin.ca/487348
[12:22:58] <danielbr> any config i tested is the same problem
[12:23:14] <jepler> OK, what about messages on the terminal?
[12:23:25] <danielbr> also axis have other errors messages
[12:24:06] <danielbr> bad number format
[12:24:16] <danielbr> in axis
[12:24:38] <danielbr> when running axis emc
[12:27:14] <alex_joni> can you pastebin the ini you are running?
[12:27:27] <alex_joni> and the exact errors?
[12:28:56] <danielbr> running sample axis mm
[12:29:31] <danielbr> xyza do not open
[12:31:23] <CIA-8> 03jepler 07TRUNK * 10emc2/share/axis/tcl/axis.tcl: underlines must exactly match the names used when the menu is created, so that translated messages match
[12:33:30] <jepler> please do a "cvs up"; the change I just made fixes errors I got on the terminal when running the AXIS GUI in the portugese language pt_BR.
[12:34:09] <jepler> this might have caused an error message similar to: bad menu entry index "Parar"
[12:35:06] <danielbr> bad menu entry index "Passo"
[12:35:06] <danielbr> while executing
[12:35:08] <danielbr> "$m entrycget $idx -state"
[12:35:10] <danielbr> (procedure "state" line 9)
[12:35:12] <danielbr> invoked from within
[12:35:14] <danielbr> "state {$task_state == $STATE_ON && $taskfile != ""} .toolbar.program_step {.menu.machine "_Step"}"
[12:35:18] <danielbr> (procedure "update_state" line 18)
[12:35:22] <danielbr> translation problem?
[12:35:34] <jepler> please do a "cvs up"; I just fixed that problem.
[12:36:51] <danielbr> How exactly i can do cvs up
[12:37:21] <jepler> Later, to incorporate any changes made to the emc2 source by other developers, run
[12:37:20] <jepler> cvs -z5 update -Pd
[12:37:20] <jepler> cd src
[12:37:20] <jepler> make
[12:37:21] <jepler> in the top emc2 directory.
[12:37:28] <jepler> (from http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?CVS)
[12:42:43] <danielbr> the problem continue after cvs up
[12:43:04] <alex_joni> did you compile?
[12:43:20] <danielbr> make? yes
[12:43:22] <jepler> you mean this message still appears on the terminal? bad menu entry index "Passo"
[12:43:42] <danielbr> yes jepler
[12:43:58] <jepler> what directory were you in when you ran "cvs -z5 update -Pd"?
[12:44:17] <jepler> did it print a line showing that it updated the file share/axis/tcl/axis.tcl?
[12:44:46] <jepler> P share/axis/tcl/axis.tcl
[12:44:52] <danielbr> my error no jepler
[12:45:25] <jepler> when you run "update" you must be in the top directory, not the src directory
[12:45:40] <danielbr> no the problem that continue is bad number format message in axis window
[12:46:10] <danielbr> updated in top directory
[12:46:37] <jepler> tell me what you do to get the error "bad number format".
[12:47:19] <danielbr> open axis and run axis emc sample code
[12:50:18] <jepler> please pastebin the output of this command: export | egrep LC_\|LANG
[12:51:13] <danielbr> only this:
[12:51:19] <danielbr> declare -x LANG="pt_BR.UTF-8"
[12:51:19] <danielbr> declare -x LANGUAGE="pt_BR:pt:pt_PT"
[12:55:03] <danielbr> in the terminal that message is emc/task/emctask.cc 275: interp_error: Bad number format
[12:55:48] <danielbr> and the cpu continue 100%
[13:00:16] <jepler> in the terminal, execute these commands and then run emc:
[13:00:16] <jepler> declare -x LANG="en_US.UTF-8"
[13:00:16] <jepler> declare -x LANGUAGE="en"
[13:00:28] <jepler> do you still get the problem?
[13:01:07] <danielbr> yes
[13:01:16] <danielbr> in tkemc also
[13:01:29] <danielbr> bad caracter
[13:05:44] <jepler> in brazil, is it the custom to use "," to separate the whole part of the number from the decimal fraction part, so 1/4 would be written as 0,25?
[13:06:49] <danielbr> yes 1/4 is 0,25
[13:07:01] <danielbr> like europe
[13:08:59] <Dallur> jepler: I think techies here in Europe don't bother with language/locale settings, we just use english with our local keyboard, which could explain why none of us has experienced this issue
[13:10:21] <danielbr> this only occur in the trunk, installed and 2.1 branch are ok
[13:11:30] <danielbr> i need use like dallur one timer and i need set pin in float width with pyvcp
[13:11:57] <danielbr> in the new oneshot.comp
[13:14:31] <danielbr> so tried paste the last version in the emc2.1-branch/src and make && sudo make setuid but this do not work
[13:14:33] <jepler> OK, good news -- I managed to get the incorrect behavior
[13:14:41] <danielbr> good
[13:14:47] <jepler> milltask consumes 80% CPU and I get the message "bad number format"
[13:14:58] <danielbr> yes is like me
[13:18:56] <CIA-8> 03jepler 07TRUNK * 10emc2/src/emc/task/emctaskmain.cc: setting LC_NUMERIC led to odd behavior, including 100% CPU use and the error "Bad number format" when running gcode
[13:18:59] <jepler> OK, do the "update" and "make" again, let's see if it's fixed now
[13:20:32] <barum> Can I find doxygen generated pages from emc on somewhere?
[13:20:36] <barum> I saw doxconfig file on CVS but folder docs/html is not generated
[13:23:10] <jepler> barum: very little of the software ever got doxygen documentation. there are manual pages for the public hal and rtapi APIs, though.
[13:23:49] <danielbr> stepper sample config run ok now but axis xyza do not start
[13:24:59] <danielbr> /usr/bin/xgettext -k_ -o po/rs274_err.pot emc/task/emctaskmain.cc emc/rs274ngc/rs274ngc_errors.cc && touch po/rs274_err.pot
[13:27:44] <alex_joni> barum: and you can generate it on a local checkout
[13:33:06] <danielbr> when i MAKE after that oneshot change in emc2.1-branch i get this error:
[13:33:28] <danielbr> ../bin/comp --document -o ../docs/man/man9/oneshot.9 hal/components/oneshot.comp
[13:33:29] <danielbr> hal/components/oneshot.comp:35:1: Trying to find one of "$", "component", "pin", "param", "function", "option"
[13:45:49] <jepler> danielbr: probably the error compiling oneshot.comp is that the "license" line is not recognized. reverse this change, then it may work with 2.1. http://cvs.linuxcnc.org/cgi-bin/cvsweb.cgi/emc2/src/hal/components/oneshot.comp.diff?r1=1.2;r2=1.3
[13:47:09] <danielbr> i'll try
[13:56:50] <danielbr> jepler: new pin oneshot working in emc2.1-branch, Thank YOU
[16:20:58] <xemet> hi
[16:21:31] <xemet> EMC2 installed, not run in place, anyone knows where is placed the file scope.cfg used to save the halscope configuration?
[16:21:43] <xemet> I can't find it
[16:25:44] <jepler> it is placed in the working directory of halcmd
[16:25:48] <jepler> er, of halscope
[16:25:50] <SWPadnos> I think it's in the "current directory", and it's called ".scope.cfg" -the leading dot will cause it to not show up in directory listings by default
[16:26:17] <jepler> if you are starting halscope from the axis menu, the directory is the one with the .ini, because the emc script changes to that directory
[16:26:56] <xemet> so, it is created in when I start halscope
[16:27:15] <jepler> actually I think it is created when halscope exits, but I am not sure of the details
[16:27:20] <SWPadnos> I think it's created when you exit halscope, and read when you start
[16:27:40] <cradek> the behavior of saving settings across runs of halscope is supposed to be transparent. are you having trouble with this feature?
[16:27:49] <xemet> ok...I start halscope from halcmd...
[16:27:55] <xemet> 1) I open a terminal
[16:27:59] <xemet> 2) start realtime
[16:28:03] <xemet> 3) halcmd -kf
[16:28:09] <xemet> 4) loadusr haslcope
[16:28:45] <SWPadnos> in that case, I believe that .scope.cfg will be in the directory in which you started halcmd
[16:29:20] <xemet> ok
[16:29:24] <xemet> it is in the home
[16:29:31] <xemet> now, question...
[16:30:02] <xemet> I've a hal program that I launch from a simple bash scripts that contains only one line: halcmd -f esempio.hal
[16:30:33] <xemet> I would like that this program open a halscope ready and configured
[16:31:07] <xemet> so, if I add in the script a line that copy a scope.cfg file from I given directory and put it in the working directory, would it work?
[16:31:19] <SWPadnos> it should work
[16:31:48] <SWPadnos> also, you can put the scope config file contents in the bash script, and cat it out to a config file before running halcmd
[16:32:11] <xemet> so, the process is: I use my program first time, launch halscope, configure it, find the scope.cfg file and save it
[16:32:23] <SWPadnos> something like cat > .scope.cfg <<EOF
[16:32:26] <SWPadnos> my
[16:32:27] <SWPadnos> scope
[16:32:28] <SWPadnos> config
[16:32:31] <SWPadnos> data
[16:32:32] <SWPadnos> EOF
[16:32:34] <SWPadnos> halcmd ...
[16:32:41] <SWPadnos> or something like that
[16:33:18] <xemet> uhm...understand...what's EOF? End Of File?
[16:33:21] <SWPadnos> it's just a text file I believe, so you should be able to stick the contents into the script. that's a little safer
[16:33:28] <SWPadnos> it's just a name you give bash
[16:33:36] <SWPadnos> you can do cat <<FRODO .... FRODO
[16:33:55] <SWPadnos> it counts anything between the two occurrences of the name as a file (called a "here file")
[16:34:04] <SWPadnos> err - not name, tags
[16:36:23] <xemet> understood
[16:36:30] <xemet> thanks
[16:36:36] <xemet> I will try
[17:08:47] <acemi_> acemi_ is now known as acemi
[17:10:21] <xemet> hi
[17:10:35] <xemet> regarding the scope.cfg question I posted few minutes ago
[17:11:17] <xemet> I trid copying the file .scope.cfg, it worked for channels, positions in the screen etc.
[17:12:10] <xemet> the only thing doesn't work is that it says hal error: scope.sample function not found
[17:12:32] <xemet> could not find thread: THREAD ciclo
[17:12:41] <xemet> and I've to select the thread...
[17:12:46] <xemet> after taht it works
[18:07:54] <xemet> what is the bash command to say that a script has not to print commands on the terminal?
[18:08:18] <xemet> (if any)
[18:09:10] <ds2> print commands or just stdout?
[18:09:28] <xemet> I think stdout
[18:09:34] <ds2> exec > /dev/null
[18:09:52] <ds2> or for specific commands, just appead a "> /dev/null"
[18:14:54] <xemet> uhm...i prints all the commands result like before
[18:15:02] <xemet> it
[18:20:37] <ds2> are you sure it is stdout, not stderr?
[18:20:54] <ds2> try adding both 'exec > /dev/null' and 'exec 2> /dev/null'
[18:21:09] <SWPadnos> command name 2>&1 > /dev/null
[18:21:23] <xemet> I'm not sure....
[18:21:33] <SWPadnos> what kind of stuff is printing?
[18:21:49] <ds2> that's per command and itgets old fast
[18:21:51] <xemet> I don't even knowwhat stdout is...
[18:22:23] <SWPadnos> when a program does something like "printf"Hello, World\n"), that goes to stdout (standard output)
[18:22:39] <SWPadnos> err - printf("Hello, World.\n") ...
[18:23:00] <SWPadnos> if there's a problem, many programs print to stderr - the "standard error output channel"
[18:23:32] <SWPadnos> that generally goes to the console as well, and has to be separately redirected to /dev/null, which is basically a black hole for data
[18:24:35] <ds2> unless he's using insmod and those are printk's ;)
[18:24:40] <SWPadnos> heh
[18:24:56] <SWPadnos> though those would go to the syslog, regardless of redirection
[18:25:05] <SWPadnos> and probably not to the console at all
[18:25:24] <ds2> I donno how ubuntu sets it up but by default it does the console
[18:25:37] <ds2> console as in /dev/console
[18:25:49] <SWPadnos> ok
[18:27:11] <xemet> sorry, my oruter disconnected
[18:27:46] <xemet> so, I've a script that launch a file hal with pyvcp widgets inside
[18:28:21] <xemet> when I launch it some things are printed in the terminal...like waiting for component etc etc
[18:28:33] <xemet> is there a way to tell the script to not prin anything?
[18:28:43] <xemet> I don't know what is stout or sterr
[18:28:48] <xemet> std
[18:28:54] <ds2> do what SWPadnos said
[18:29:01] <ds2> or at the top put the 2 exec
[18:29:06] <ds2> should cover both stderr and stdout
[18:30:06] <xemet> now I try
[18:32:23] <xemet> if I add exec 2> /dev/null at the beginning of the file, the terminal closes and nothing more happens
[18:33:22] <xemet> ok, now it works
[18:33:27] <SWPadnos> try using halcmd -q
[18:33:29] <xemet> maybe there was a space
[18:33:41] <SWPadnos> that means "quiet", and it will only print errors (which you probably want to see anyway)
[18:34:10] <SWPadnos> or use -Q, if you want it to not print anything
[18:34:56] <SWPadnos> that's for halcmd, so if your entire script is halcmd -kf your_file.hal, just change it to halcmd -Dkf your_file.hal
[18:35:01] <SWPadnos> argh
[18:35:11] <SWPadnos> if your entire script is halcmd -kf your_file.hal, just change it to halcmd -Qkf your_file.hal
[18:35:23] <xemet> will try
[18:35:45] <xemet> now, I'm using a bash script with a launcher with the option "run in terminal"
[18:35:48] <SWPadnos> and you should be able to remove all the > /dev/null stuff too
[18:36:07] <xemet> is there a way to let it run in terminal without opening a terminal window? :)
[18:36:28] <SWPadnos> is there a "start minimized" option?
[18:36:43] <xemet> I launch the script and it opens a terminal window that I don't need...what I've to use is all outside terminal
[18:36:53] <xemet> uhm...not seen that option
[18:37:36] <SWPadnos> you don't have to run it in a terminal, since you're trying to get rid of all output anyway, the terminal seems useless here
[18:37:58] <xemet> ok, but if I remove run in terminal...I think it doesn't work...
[18:38:00] <xemet> now I try
[18:38:05] <SWPadnos> just take out all the -Q / -q / ... > /dev/null, and uncheck the "run in terminal" option
[18:38:16] <SWPadnos> it shouldn't need a terminal (I think)
[18:38:42] <xemet> wow, it works...
[18:39:10] <xemet> The first time I tried it didn't...maybe I made something wrong
[18:39:58] <SWPadnos> that's the problem with not running in a terminal - when something goes wrong, there's no way to find out what it was ;)
[18:40:02] <ds2> exec 2 > /dev/null is bad and means something very different ;)
[18:42:36] <xemet> ok...now it works...If I find I need terminal again for that I will use it
[18:42:54] <xemet> ds2, yes, infact it didn't work
[18:43:23] <xemet> last question, I think with no answer...why in ubuntu I can create a launcher only in the Desktop?
[18:43:44] <xemet> if I'm in a directory I don't have the option create a launcher
[18:43:45] <ds2> it worked just fine... just wasn't what you wanted... 'exec xyz > /dev/null' means replace your current program with xyz and send the stdout to /dev/null; in your case, it was trying to replace your script with something called '2' =)
[18:44:37] <SWPadnos> if you have an executable script in a directory listing, you can automatically launch it by double-clicking
[18:45:16] <xemet> yes...BUT (i know, this is trivial) how do I put an icon?
[18:45:25] <SWPadnos> oh, that I don't know ;)
[18:45:31] <xemet> when I use create launcher, it let me to choose an icon
[18:45:35] <xemet> eheh
[18:46:11] <xemet> that's the difference :)...I know, you programmers do not like icons...but I like those stupid graphic things...
[18:46:19] <SWPadnos> hmmm. try copying/dragging the launcher icon from the desktop into a window ...
[18:46:32] <xemet> I made it
[18:46:34] <xemet> it works
[18:46:48] <xemet> I make the launcher in the desktop and after I copy it in a directory
[18:47:12] <xemet> but would be easier if I could right click and create the launcher directly where I want...
[18:47:29] <SWPadnos> true
[18:48:48] <xemet> well, now I go for dinner
[18:49:00] <xemet> thank you very much for your help!
[18:49:17] <SWPadnos> sure. bon appetit
[18:49:28] <xemet> thanks
[18:49:33] <xemet> see you
[18:49:41] <JymmmEMC> * JymmmEMC waves a zepplin under SWPadnos' nose
[18:49:57] <SWPadnos> I'm not a big fan of Zeppelin
[18:50:14] <JymmmEMC> lower case z
[18:50:27] <SWPadnos> ah - must be a small zeppelin
[18:51:11] <JymmmEMC> yeah, prime rib powered zepplins
[18:51:19] <SWPadnos> mmmmmm
[18:51:29] <SWPadnos> hey - I should have breakfast. bbiab
[18:51:34] <JymmmEMC> k
[18:56:58] <gene_> I found a minor quibble in the manual...
[18:57:42] <gene_> on page 153, where it gives the example of a subroutine call
[18:58:38] <gene_> it says O100 call [1] [2] [3] etc but it needs to be O100 call [#1] [#2] [#3] etc
[18:59:22] <SWPadnos> nope - the example is demonstrating something else
[18:59:34] <gene_> huh?
[18:59:55] <SWPadnos> Because ?1 2 3? is parsed as the number 123, the parameters must be enclosed in square brackets. The following calls a subroutine with 3 arguments:
[19:00:20] <gene_> if the subroutine copys all passed vars to #31+ numbers, it doesn't get any data unless the # signs are in there
[19:00:23] <SWPadnos> that's the line just above the sample ccode
[19:00:54] <gene_> Yes, I see that, but without the # signs, it doesn't work
[19:01:01] <SWPadnos> I understand - the ezxample is there to demonstrate how to pass literal numbers to the subroutine, without the parser taking out all the spaces and turning them into one large number ...
[19:01:23] <gene_> ahh, ok
[19:01:28] <SWPadnos> sure it does, that is a subroutine call with the parameters one, two, and three, not the values of variables numbered onw, two, and three
[19:02:00] <gene_> maybe another line needs to be added demoing how to pass #numbers?
[19:02:17] <SWPadnos> there probably should be an example with variables as well, since there's not really a good separation between the variable concept and the no-spaces concept
[19:02:26] <gene_> agreed
[19:02:47] <gene_> who's the math whiz here?
[19:02:58] <SWPadnos> that depends on who else is here ;)
[19:03:04] <gene_> chuckle
[19:03:09] <JymmmEMC> gene_ is the math wiz
[19:03:21] <skunkworks> * skunkworks points at SWPadnos
[19:04:03] <gene_> I need to do a least square fit on this data I scanned, so that I can move the locus to the true center, and substitute perfect data in place of the rough scan
[19:04:30] <SWPadnos> OOCalc!
[19:04:39] <JymmmEMC> MOOCalc!
[19:04:43] <gene_> me, a math whiz, with an 8th grade education? You gotta be kidding...
[19:04:59] <SWPadnos> but back in your day, that meant something ;)
[19:05:02] <JymmmEMC> gene_: You *ARE* smarter than a 5th grader!
[19:05:31] <gene_> OOCalc, humm, that never crossed my mind, I'll have to look at it.
[19:05:45] <JymmmEMC> Object Oriented Calc?
[19:06:10] <JymmmEMC> Bovine Calc is funnier though
[19:06:14] <SWPadnos> heh
[19:06:17] <SWPadnos> apt-get moo
[19:06:35] <gene_> witht he help of another friend I'm learning awk, and its able to pick the data out and reformat it into gcode so I can send it back to emc2 and dup what I scanned, but the accurcy leaves a bit to be desired...
[19:06:39] <JymmmEMC> ROTFLMAO... BovineCALC == BCMATH
[19:07:24] <SWPadnos> gene_, you can probably write it to a comma-delimited spreadsheet just as easily (even easier, I bet)
[19:07:35] <gene_> I'd hope so jymmmm, after all I have been thre CE at a tv station for the last 23 years.
[19:07:48] <SWPadnos> import into OOCalc, and I'd bet there's a linear regression function
[19:07:48] <JymmmEMC> SWPadnos: I don't know if you care, but I found a BCMATH lib for JS
[19:08:02] <SWPadnos> BCMATH?
[19:08:15] <JymmmEMC> gene_: all you old timers just lick your fingers to measure voltages anyway
[19:08:20] <SWPadnos> I know of bc, the command-line calculator ,,,
[19:08:33] <gene_> is that one of those unlimited precision things as long as you have hte cpu time available?
[19:08:42] <JymmmEMC> SWPadnos: yeah, just a sec...
[19:09:28] <JymmmEMC> http://php.net/bc
[19:09:36] <JymmmEMC> not sure of the REAL term generic
[19:09:49] <JymmmEMC> gene_: Nah, no real limitations
[19:09:51] <gene_> gawk uses C syntax for the most part, so I walked right in and made good code pretty fast.
[19:10:27] <JymmmEMC> gene_: I've done some obscene cals using it
[19:10:29] <gene_> The script to make the gcode is ,aybe 60 lines
[19:10:31] <JymmmEMC> calcs
[19:10:39] <gene_> maybe
[19:11:04] <SWPadnos> the awk script to output 3 numbers as a comma-delimited spreadsheet should be about 20 lines long, including header and format setting
[19:11:29] <gene_> I was skimming for data too...
[19:11:41] <JymmmEMC> SWPadnos: http://en.wikipedia.org/wiki/Arbitrary_precision
[19:11:45] <SWPadnos> ah
[19:11:48] <gene_> to separate output files.
[19:12:15] <SWPadnos> JymmmEMC, I've heard of arbitrary precision calculators, I just didn't know the specific one you were talking about
[19:12:30] <SWPadnos> err - I didn't know that what you were talking about was one ... :)
[19:12:31] <JymmmEMC> SWPadnos: Sorry, couldn't remember the REAL name
[19:13:35] <gene_> I'll take a look at OOCalc then. It can be loaded into gnumeric, but I don't know if gnumeric has the tools to do this
[19:13:54] <JymmmEMC> SWPadnos: We just got tlaking about math, reminded me is all. I've been working on a lil project for a number of years and found a JS variant.
[19:13:56] <SWPadnos> it probably does, a linear regression is a pretty basic analysis function
[19:14:02] <SWPadnos> heh
[19:14:56] <gene_> Is the std deviation how one would find the slope of a noisy straight line?
[19:15:05] <JymmmEMC> SWPadnos: I menat to ask you, what are you programming Atmel's in? C ASM ?
[19:15:14] <SWPadnos> no, that tells you how far away from the fitted line the points are
[19:15:41] <gene_> what would give me the average slop per inch of run then?
[19:15:47] <gene_> slop/slope
[19:16:52] <SWPadnos> argh
[19:17:01] <SWPadnos> assembly, in case you missed that
[19:17:29] <JymmmEMC> SWPadnos: Awwww =(
[19:17:35] <SWPadnos> I have one project that I think would be better in C now, because it's 6 years old and has around 800k of assembly source
[19:17:47] <JymmmEMC> SWPadnos: under nix or win?
[19:17:49] <SWPadnos> it would probably be a lot less C
[19:17:52] <gene_> assembly, as in machine code nemonic src?
[19:17:53] <SWPadnos> AVR ...
[19:17:56] <SWPadnos> yes
[19:18:08] <SWPadnos> lpm r8, Z+ ...
[19:18:17] <JymmmEMC> SWPadnos: you developing under nix or win?
[19:18:38] <gene_> gah, I only do that on a 6x09 cpu for the last 20 years.
[19:18:40] <SWPadnos> that's under Windows, since I'm using a specific assembler that isn't available for Linux (IAR)
[19:18:58] <SWPadnos> the AVR isn't much more complex than a 6809
[19:19:06] <JymmmEMC> SWPadnos: And you're using the STK500 and their ISP dongle?
[19:19:12] <SWPadnos> eys
[19:19:13] <SWPadnos> yes
[19:19:40] <gene_> where 6x09=6809 or 6309, 6309 from hitach, much smarter chipi
[19:19:44] <SWPadnos> and everything from the 12k parts up to the 128k parts
[19:19:58] <SWPadnos> err - not 12k, one k
[19:21:34] <JymmmEMC> SWPadnos: Ok, I'm playing with the STK200 and an AT90S2313, Would you mind writing the C code to make a/all LED(s) blink on PORTB? Every time I try I see the chip being programmed but I can't even make an LED light up and I can't find any C code, only ASM.
[19:22:15] <JymmmEMC> I'm not sure if the dev board is fubared or what.
[19:23:38] <SWPadnos> I'm not sure what code gets executed at startup when using C
[19:23:45] <SWPadnos> are you trying to use gcc / winAVR?
[19:24:20] <JymmmEMC> Heh, I was trying BASCOM (BASIC) but their examples are crap
[19:25:02] <JymmmEMC> Was going to grba AVR Studio tonight
[19:27:25] <JymmmEMC> is winAVR GUI ?
[19:28:39] <SWPadnos> winavr works with avr studio, though I'm not exactly sure how
[19:29:02] <SWPadnos> winavr has an editor (adtually a very nice one, which I use for most text stuff on Windows)
[19:29:47] <SWPadnos> other than tha, there doesn't seem to be any GUI (in my older version)
[19:29:51] <JymmmEMC> Cool, there's a PortableWinAVR too =)
[19:30:28] <JymmmEMC> Ok, I'll give Studio a shot tonight.
[19:30:33] <SWPadnos> winavr is just a WIndows version of gcc for the AVR, so the code should be portable anyway
[19:30:53] <JymmmEMC> why does one need winavr AND studio?
[19:31:12] <SWPadnos> studio lets you program chips and do debugging
[19:31:27] <SWPadnos> winavr lets you edit (with programmers notepad) and compile (with gcc)
[19:31:43] <JymmmEMC> you can't compile in Studio?
[19:32:12] <SWPadnos> I think you can set up studio to run gcc for you, but AVR studio doesn't come with a compiler
[19:32:20] <SWPadnos> it has an assembler, but not a C compiler
[19:32:23] <JymmmEMC> Ah, gotcha.
[19:32:58] <JymmmEMC> OH! On your STK500. the connector for the CLD backback, is it a single row of pins or dual row?
[19:33:05] <JymmmEMC> LCD
[19:33:44] <SWPadnos> there's no LCD connector on the STK500
[19:34:14] <SWPadnos> have you connected the cable between PORTB and the LEDs?
[19:34:21] <SWPadnos> without that, you won't see any flashing LEDs
[19:34:44] <JymmmEMC> The LED's light when it's being programmed
[19:35:29] <SWPadnos> ok. I haven't seen an stk200 in a long time
[19:35:49] <JymmmEMC> I read the docs and the jumpers are in the correct position.
[19:37:11] <JymmmEMC> I have RevB - REALLY old
[19:38:19] <JymmmEMC> But, I'm only playing around, just cheap entertainment if you will
[19:39:58] <SWPadnos_> argh
[19:40:00] <SWPadnos_> when you wrote your programs to flash the lights, did you set up the port as output?
[19:40:20] <JymmmEMC> config portb = output
[19:40:25] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[19:41:15] <JymmmEMC> I even set the port to 128 just in case I wasn't sure if they were tied ho or lo
[19:41:20] <JymmmEMC> hi
[19:41:46] <SWPadnos> hi
[19:41:59] <JymmmEMC> s/ho/high/
[19:42:03] <SWPadnos> heh
[19:42:50] <JymmmEMC> I unplugged the ISP cable too, just in case it was interfering with the program
[19:43:31] <JymmmEMC> Whenever I read/write to the chip, I see at least 3 LED's light up
[19:43:36] <SWPadnos> well, I don't know the setup stuff that you need to do with winavr or BASCOM, so I can't help you there
[19:44:07] <JymmmEMC> Yeah, I'll try winAVR and/or studio tonight.
[19:44:37] <JymmmEMC> When I glanced at it all I saw was ASM - that aint gonna happen.... homey dont play that!
[19:56:40] <SWPadnos> I find assembly much easier for low-level stuff, and of course C is better for high-level stuff (like multibyte math, print formatting ...)
[19:57:31] <JymmmEMC> I am clueless in respect to ASM. At least if I have to (re)learn C, I can use it elsewhere.
[19:58:04] <SWPadnos> yep - there is that
[20:34:01] <a-l-p-h-a2> a-l-p-h-a2 is now known as a-l-p-h-a
[21:38:55] <ohiopctechDOTcom> * ohiopctechDOTcom <-should be outside enjoying the sun
[21:39:23] <ds2> you are not a vampire and you will not go up in smoke?
[21:39:45] <ohiopctechDOTcom> i thought so but i went to work today and i am still alive!
[21:39:48] <ds2> (see today's dilbert)
[21:40:02] <ohiopctechDOTcom> oh.. i miss dilbert i don't get the paper anymore
[21:40:17] <ds2> see it online
[21:40:40] <ohiopctechDOTcom> my favorite was the one where the punchline was "get over it."
[21:41:00] <ohiopctechDOTcom> i forget the whole strip but i had it on the timeclock board at work for a while...
[21:41:12] <ohiopctechDOTcom> something about cows
[21:41:53] <ohiopctechDOTcom> ha.. dilbert.com
[21:43:02] <ohiopctechDOTcom> i'm uploading one of my cnc vids to my(lame)space now..
[21:43:24] <ohiopctechDOTcom> maybe give all those emo kids a new hobby!
[21:43:37] <ohiopctechDOTcom> something beside being too emotional
[21:47:23] <ohiopctechDOTcom> :( they say i can't upload porn... but music videos with hoochies shakin it all round is ok...
[22:58:01] <jepler> bahahaha http://catless.ncl.ac.uk/Risks/24.66.html#subj6
[22:59:58] <cradek> nawww
[23:00:05] <cradek> I don't buy it
[23:01:21] <cradek> visited a neurologist before finding that it was only in IE?
[23:01:51] <cradek> my brain would have to be falling out before I'd go see a neurologist about it
[23:04:24] <JymmmEMC> s/neurologist/mortitian/
[23:05:27] <eric_u> my daughter was having problems passing out when she did a particularly contorted routine in gymnastics
[23:05:40] <eric_u> I said "I have a clever idea, stop doing that routine"
[23:06:26] <eric_u> but no, she got lots of tests, and found out she was missing a blood vessel and that caused her to have blackouts if she over-extended her neck
[23:06:40] <JymmmEMC> I'd feel sorry for her if she was allergic to water.
[23:06:46] <eric_u> the neurologist said "stop doing that routine"
[23:06:49] <toastydeath> or allergic to gymnastics
[23:06:55] <toastydeath> nice.
[23:07:26] <JymmmEMC> Gawd help her if she ever got into a car accident.
[23:07:29] <eric_u> they aren't going to do anything about a missing artery, now are they?
[23:08:02] <eric_u> it's a little scary
[23:08:31] <JymmmEMC> They could, but what are the statics behind such a thing causing lifelong complications?
[23:08:59] <eric_u> I really don't think it's practical
[23:09:15] <JymmmEMC> Free granite table and stand =(
[23:09:23] <JymmmEMC> it's HUGE
[23:09:38] <toastydeath> link
[23:09:44] <toastydeath> is it a surface plate or a table
[23:10:01] <eric_u> the guys in the lab next door got two 4x6x2 foot granite tables on air isolators from ibm
[23:10:07] <JymmmEMC> inspection plate... 55" x 34" x 8"
[23:10:19] <eric_u> and another that has to be 4x8x2'
[23:10:31] <JymmmEMC> http://sfbay.craigslist.org/sby/zip/330831554.html
[23:10:42] <JymmmEMC> brand new
[23:10:57] <toastydeath> it doesn't say what grade
[23:11:23] <ds2> get you truck and go pick it up =)
[23:11:36] <eric_u> "This will be a first come first serve basis and the original customer is aware of the free for all takers situation"
[23:11:43] <eric_u> ?
[23:12:04] <eric_u> but does he have the ----s to come pick it up is the question
[23:44:51] <Guest164> Guest164 is now known as skunkworks