#emc | Logs for 2008-01-28

[00:01:34] <cxix> alex - do i just select one of the sample configuration files and configure it from there?
[00:04:38] <alex_joni> cxix: right
[00:04:46] <alex_joni> stepper_mm or stepper_inch are probably closest
[00:05:05] <alex_joni> there's also a program called StepConf which helps you configure a simple stepping machine
[00:05:46] <cxix> i might try that first
[00:06:00] <alex_joni> it's on the menu
[00:07:00] <cxix> i have four motors on three axes, under axis configuration, do i select XYZ or XYZA ?
[00:07:22] <cxix> A is slaved to Y
[00:09:08] <alex_joni> A is a rotary axis
[00:09:37] <cxix> so i can set up the fourth motor as a slave later?
[00:11:46] <cxix> found it, n,
[00:11:47] <cxix> nm
[00:13:34] <Jymmm> alex_joni: Hmmm on newegg, there are a LOT of "open box" VIA EPIA boards for sale.
[00:14:03] <alex_joni> Jymmm: nice
[00:14:24] <Jymmm> alex_joni: Well, I'm thinking open box == returns
[00:15:32] <eric_U> wonder what that means?
[00:15:42] <eric_U> people don't like the performance?
[00:16:51] <Jymmm> I really dont know, but returns never sound good when you get a whole page of them
[00:16:57] <Jymmm> various models
[00:23:40] <eric_U> maybe they just never sell the returns
[00:32:37] <dave_1> cxix
[00:32:47] <dave_1> how did you come out?
[00:33:53] <cxix> the software is working on my server computer
[00:34:03] <cxix> i'm trying to set it up now
[00:34:09] <dave_1> that is good. ... progress
[00:34:35] <cxix> but i went to set up the axes, and i have a 4th motor slaved to the y, and i'm not real sure where to set that up
[00:34:41] <dave_1> I got good index pulses ... 2 ms wide at 4 ipm
[00:35:15] <dave_1> 4 axis or 4 dac?
[00:35:30] <cxix> 3 axis, two motors on the gantry
[00:35:42] <alex_joni> cxix: stepconf won't produce a good config for that
[00:35:44] <dave_1> ohhhhh way beyond me
[00:35:48] <alex_joni> it's only intended for simple machines
[00:35:56] <alex_joni> you need to read/understand the docs for that
[00:36:02] <alex_joni> Integrators Manual is what you need
[00:36:02] <cxix> ok
[00:36:17] <dave_1> or find a ini that is close and then fix it
[00:36:31] <cxix> ok, i did see a sample that was similar
[00:36:51] <cxix> but i have to go see someone... so i may be back after a while.
[00:36:54] <cxix> i appreciate all the help!
[00:37:01] <dave_1> good luck
[00:37:06] <cxix> thanks!
[00:37:07] <cxix> ttyl
[00:37:53] <dave_1> alex ... I've got nice wide index pulses at search speeds ... now to get them into emc
[00:38:09] <alex_joni> dave_1: sounds good
[00:39:15] <dave_1> bbl ... got to do something about food ;-)
[00:52:22] <Jymmm> Was just watching a show on tv about bbq.... I want to make my own smoker this spring =)
[01:39:44] <gezar> howdy
[01:40:21] <gezar> oh this is a nice irc client
[01:48:27] <owhite> hello folks. I have some questions about the mesa M5I20 card. Any takers?
[01:48:57] <SWPadnos> ask, and ye may receive
[01:49:32] <owhite> I get the impression that they do not ship the .BIT files , I have to download the xilinx compiler and make them myself?
[01:49:46] <SWPadnos> bzzzzzzt!
[01:50:03] <SWPadnos> actually, the mesa driver for EMC2 has the bitfile built in, and it will be loaded at driver load time
[01:50:04] <owhite> great. I asked the wrong question already?
[01:50:07] <SWPadnos> heh
[01:50:16] <SWPadnos> no, you had the wrong answer (luckily) :)
[01:50:25] <owhite> okay, how to test if I have the correct version of emc2?
[01:50:33] <SWPadnos> what do you mean?
[01:50:54] <owhite> well like how do I know emc2 has the bitfile?
[01:51:03] <owhite> was wondering if that's a recent version or something.
[01:51:04] <jepler> the mesa driver (including the bit file) is included with every precompiled version of emc2
[01:51:12] <SWPadnos> ummm - yeah - what he said
[01:51:36] <SWPadnos> if you load the driver (loadrt m5i20, and there are no load errors, then the FPGA should be configured
[01:51:42] <SWPadnos> are you having a particular problem?
[01:51:46] <owhite> okay. so here's a question....it looks like when I put in the M5I20 it prevents my wireless card from getting an IP address.
[01:51:49] <jepler> what led you to believe you might have to get the bitfile somewhere else?
[01:52:15] <owhite> without the linux machine on the network, makes it hard to do much.
[01:52:20] <SWPadnos> hmmm. there could be a change to the PCI enumeration in the kernel, though the wireless driver shouldn't care about that
[01:52:33] <jepler> the only thing you change is to put the 5i20 card in a PCI slot, and your wireless stops working?
[01:52:45] <owhite> jepler -- because I started by reading the M5I20 docs in the abscence of other documentation.
[01:53:00] <jepler> 19:52:33 <jepler> the only thing you change is to put the 5i20 card in a PCI
[01:53:00] <jepler> slot, and your wireless stops working?
[01:53:13] <owhite> yes. Shutdown machine, stick in card, power up. No IP address. Shutdown machine again, remove card, get an IP address.
[01:53:16] <jepler> oops
[01:53:55] <skunkworks> did you try a differnt slot?
[01:53:56] <SWPadnos> is this a desktop with a PCI PCMCIA adapter+wireless card, a PCI wireless card, or what?
[01:54:09] <owhite> PCI wireless card.
[01:54:29] <owhite> you guys think it'd help to try other slots?
[01:54:46] <SWPadnos> ok - as skunkworks said, a different slot may be a good idea - you could try putting the mesa card where the wireless card is, and put the wireless card where you were putting hte mesa
[01:55:11] <SWPadnos> the other thing is that the wireless antenna would be next to any cables that would come from the mesa card, so that could affect reception
[01:55:51] <owhite> okay. When you guys move around cards do you reboot each time?
[01:55:54] <jepler> is the wireless card *detected* but just doesn't get a connection? you can use the graphical network manager to find out if it's there or not
[01:56:02] <SWPadnos> of course, you have to turn off the PC to move the cards
[01:56:08] <jepler> or run "ifconfig -a" in the terminal and see if there is an entry for the device in the stuff that prints
[01:56:19] <jepler> yes, always power off before inserting or removing cards in a PC
[01:56:25] <owhite> Jepler: The wireless card is detected. but ifconfig -a shows no IP address.
[01:56:28] <SWPadnos> and that means fully off - unplug it and press the power switch (assuming it's ATX)
[01:56:57] <jepler> you might also look in the output of "dmesg" to see if problems are reported with the wireless card
[01:57:07] <SWPadnos> (pressing the power switch can make the computer attempt to power on, which will drain the caps for the standby supply)
[01:58:22] <owhite> okay. so I will mess around with that for a while. So there's a loadrt m5i20, and I can try launching that to test the thing?
[01:58:40] <SWPadnos> you should be able to load the 5i20 sample config
[01:59:19] <jepler> the HAL module name is hal_m5i20.
[01:59:37] <jepler> but the prefix on the pins and so on is just m5i20
[02:00:46] <jepler> bbl
[02:01:53] <owhite> so to launch it would be "emc configs/m5i20/m5i20" ?
[02:02:16] <jepler> .ini
[02:02:16] <SWPadnos> do you have a GUI? (or are you running remotely via ssh or something)?
[02:02:45] <owhite> well I can run the GUI.
[02:03:15] <jepler> bbl
[02:03:20] <SWPadnos> either way. running from a terminal is probably better so you can see any error messages
[02:03:28] <SWPadnos> see you jepler
[02:03:46] <owhite> righto. okay I'll do some rebooting and fooling around. Thanks folks.
[02:03:49] <jepler> I think SWPadnos asks about gui because it's just as easy to type "emc" then scroll down to the item for m5i20 and less error-prone (you don't have to worry about typing the .ini extension, which is required at the commandline)
[02:03:59] <SWPadnos> yep
[02:04:54] <jepler> anyway .. bbl
[02:05:00] <SWPadnos> heh - see you :)
[02:22:50] <cxix> hi again all
[02:33:26] <jepler> hi cxix
[02:37:21] <cxix> what's happening?
[02:38:21] <jepler> I'm working on the final packaging for the next bugfix release, emc 2.2.3
[02:39:10] <cxix> cool, i'm just trying to configure an ini file
[02:39:44] <jepler> if you have some questions, just ask 'em -- I'm distracted at the moment but maybe somebody else will be able to help you out
[02:40:26] <cxix> just going through the manual step by step for now, i'll throw 'em at you guys if i run into something
[02:40:59] <jepler> sounds good
[02:53:30] <cxix> so, if i have a fourth motor slaved (gantry has two motors), is that in the settings somewhere in the program? I don't see where I would edit the ini file
[02:54:42] <owhite> can someone suggest how I find the slot address for my parallel port card? It used to be 0x9800, but I had to move it, dont know what the new slot address is.
[02:55:02] <SWPadnos> cxix, that's a hardware configuration, so it goes in the hal files (in addition to some things in the ini)
[02:55:06] <SWPadnos> owhite, lspci
[02:55:06] <cxix> * cxix could tell you on windows...
[02:55:19] <owhite> thanks swpadnos
[02:55:22] <SWPadnos> sure
[02:55:49] <SWPadnos> lspci will tell you several addresses, one of which will work. I'm not sure which one it is, but it's easy enough to try them all :)
[02:55:51] <cxix> swp, which 'some things in the ini' ?
[02:56:41] <SWPadnos> cxix, I think you tell the system that there are 4 axes in the ini
[02:56:57] <SWPadnos> but you configure the fourth as a slave in the hal files
[02:57:18] <SWPadnos> I'm not sure exactly how that stuff works though, so I can't help you much in the specifics
[02:57:49] <cxix> oh ok. i have all four axes setup as linear in the ini file, i thought something had to be negative or something, but i set a sample file to "reversed" and it didn't show up, so i was confused.
[02:57:51] <cxix> thanks.
[02:58:26] <SWPadnos> oh - do you have two motors on the same screw, or two screws (to prevent racking)?
[02:58:53] <cxix> i have a rack and pinion setup to drive the gantry, each with its own motor
[02:59:00] <cxix> two racks, two pinions
[02:59:13] <SWPadnos> ok
[02:59:28] <SWPadnos> steppers or servos?
[02:59:34] <cxix> steppers
[02:59:37] <owhite> swpadnos: can you confirm that lspci -vv shows any addresses for you? I aint seeing nuffin.
[02:59:51] <skunkworks> don't you use gantry kin?
[03:00:06] <SWPadnos> ok - you could just swap the phases, or swap the polarity of either set of phase wires, to change the direction of the motor
[03:00:12] <SWPadnos> or change the sign of the scale in the ini
[03:00:19] <SWPadnos> (or invert the direction pin or ... :) )
[03:01:37] <SWPadnos> owhite, it should - I can't check because the only Linux machine I have booted at the moment actually has no parport (new laptop)
[03:01:56] <SWPadnos> you may need to use sudo though, root gets a little more info (though it shouldn't be needed for parport addresses)
[03:02:20] <owhite> trying...
[03:02:35] <jepler> lspci typically won't show the built-in parport, because it's not a PCI device. it should show add-on PCI parport cards though.
[03:02:40] <jepler> the lines you're looking for would say something like
[03:02:41] <jepler> Region 1: I/O ports at 9000 [size=128]
[03:02:54] <owhite> yah this is a parallel card in the PCI.
[03:02:56] <cradek> does anyone have a fishtail gauge for acme, or do you use something else?
[03:03:14] <cradek> oh sorry, you're talking about emc :-)
[03:03:17] <SWPadnos> if you klnow the name (like NetMos), you can use lspci | grep -C 10 "the_name"
[03:03:18] <Jymmm> If anyone cares, Target has all their heaters on clearance right now
[03:03:44] <SWPadnos> Jymmm, it got above freezing today - who needs a heater?
[03:04:03] <Jymmm> SWPadnos: Me, in the garage
[03:04:11] <SWPadnos> heh - ok, me too, in the garage
[03:04:12] <twice2> hiya, congrats on 2.2.2!
[03:04:20] <SWPadnos> I'd need two or three though
[03:04:35] <owhite> does cc00 sound like a reasonable address?
[03:04:40] <jmkasunich> cradek: you mean the flat piece of steel that you use to check the angles on the toolbit, and to set it square to the work?
[03:04:40] <SWPadnos> owhite, yep
[03:04:42] <Jymmm> SWPadnos: Did you see me earlier msg that there's not enough room the the chassis for everything?
[03:04:47] <cradek> jmkasunich: yes
[03:04:51] <SWPadnos> Jymmm, yes, I did notice it
[03:04:56] <SWPadnos> what did you expect? :)
[03:05:01] <jmkasunich> sorry, don't have one (got regular threads, but not acme)
[03:05:05] <jmkasunich> what are you making?
[03:05:28] <cradek> just something half-baked of course
[03:05:40] <GNieport1> Hi guys. Are yesterday's chat logs available online?
[03:05:43] <jmkasunich> of course
[03:05:44] <Jymmm> SWPadnos: Damn good double cheeseburger and home made fries
[03:05:44] <cradek> thinking about those strange antibacklash nuts again
[03:05:52] <Jymmm> cradek: acme?
[03:05:56] <jmkasunich> which strange ones?
[03:06:08] <cradek> http://www.hobbycnc.hu/CNC/Otletek/Otletek.htm
[03:06:19] <jmkasunich> oh
[03:06:20] <cradek> I still need to do something about max's Z
[03:06:31] <jmkasunich> those don't really need acme do they?
[03:06:44] <cradek> probably not, but why not use it
[03:07:29] <toastydeath> actually they do
[03:07:31] <toastydeath> er
[03:07:34] <toastydeath> ballscrews?
[03:07:38] <toastydeath> i will shut up nevermind
[03:07:50] <jmkasunich> no, rollerscrews - follow that link, a pic is worth 1000 words
[03:07:52] <cradek> I now realized the nuts/rollers can't be threads
[03:08:09] <cradek> they have to be ... uh what would it be called
[03:08:14] <cradek> just parallel troughs
[03:08:14] <jmkasunich> they can't?
[03:08:20] <jmkasunich> are you sure?
[03:08:31] <cradek> no because the whole nut can move along the screw without the screw itself turning
[03:08:51] <owhite> insmod: error inserting '/home/owhite/emc2-head/rtlib/hal_parport.ko': -1 Device or resource busy
[03:08:53] <cradek> it won't repeat unless the rolling is perfect and doesn't slip at all
[03:08:57] <owhite> any suggestions?
[03:09:52] <cradek> jmkasunich: imagine the axis is stationary and you reach in there and turn each of the rollers a bit - the position changes
[03:10:07] <jmkasunich> yes
[03:10:27] <SWPadnos> owhite, make sure that parport_pc and lp (and anything else lsmod tells you is using parport_pc) are unloaded
[03:10:41] <jmkasunich> if you use straight "threads" on the rollers though, you'll either need a non-standard thread form, or you'll need to incline the roller axes
[03:10:41] <SWPadnos> also, look at dmesg for the real error
[03:10:47] <SWPadnos> dmesg | tail -30
[03:11:34] <twice2> in a former life I seem to remember farand or optical feedback mounted on the work being the best it gets
[03:12:06] <cradek> jmkasunich: yeah I can't quite picture what it should be
[03:12:06] <jmkasunich> that would be another approach - use a scale on the work for exact position, to offset any creep in the rollers
[03:12:38] <jmkasunich> I bet there will be rubbing instead of rolling too
[03:14:15] <jmkasunich> btw, it would be better to call those things "anti-friction" instead of "anti-backlash"
[03:14:45] <jmkasunich> with a single set of rollers, you'll still have lash, cause the rollers would be able to move a short distance axially on their shafts
[03:15:01] <jmkasunich> (unless you preload them axially as well as toward the screw)
[03:15:12] <Jymmm> SWPadnos: I wonder if that cnc mill at TechShop works?
[03:15:22] <SWPadnos> the flashcut one, or the big iron?
[03:15:29] <Jymmm> SWPadnos: The 2000+lb one
[03:15:49] <SWPadnos> I bet it "could", but I think it was still partially plastic wrapped, wasn't it?
[03:15:56] <twice2> yep, mechanical issues, to a point would be non issue
[03:16:11] <Jymmm> SWPadnos: No, had a lot of grease and swarf though
[03:16:30] <SWPadnos> ok - was there a lathe or something that was still partly wrapped for shipping?
[03:16:39] <Jymmm> SWPadnos: yeah
[03:17:58] <twice2> cradek: may i ask what the device is glued to the cap in the photo of your pwm mod sherline lathe?
[03:18:26] <cradek> url?
[03:18:50] <twice2> uhm, timeguy something :)
[03:18:53] <cradek> ha
[03:19:14] <twice2> nice clock btw
[03:19:23] <twice2> the neon
[03:19:23] <cradek> thanks (which one?)
[03:19:31] <jepler> jepler has changed the topic to: Welcome! EMC (Enhanced Machine Controller) is a linux-based opensource CNC control. | Latest release: EMC 2.2.3 | http://www.linuxcnc.org | http://wiki.linuxcnc.org
[03:19:34] <cradek> yay!
[03:20:00] <twice2> RAH!
[03:20:25] <jepler> in other exciting news, I finally wired up amplifier enables on my bitty stepper mill
[03:20:35] <Jymmm> and I JUST installed 222, eeeeesh
[03:20:54] <cradek> twice2: http://timeguy.com/cradek-files/cnc/lathe/DSCN6292.JPG
[03:20:56] <cradek> this?
[03:21:17] <Jymmm> I dont know how many years you've guys have been doing that.... Put out a release, then another one right behind it by less than 45 days
[03:21:25] <Jymmm> or so
[03:23:08] <cxix> out of curiosity, what kind of cnc uses 9 axes?
[03:23:15] <cradek> cxix: not many I bet
[03:23:20] <toastydeath> lathes
[03:23:25] <toastydeath> swiss lathes
[03:23:26] <cradek> but lots use more than 3 linear axes
[03:23:35] <cradek> lots use some linear and some rotary
[03:23:37] <toastydeath> and cnc screw machines use 60+ axes
[03:23:38] <cxix> that was the first thing that caught my eye when i read about emc
[03:23:49] <cxix> holy crap
[03:23:57] <jepler> Jymmm: actually, you may not know this -- but we have a "target number" for installed base; when we reach that number, we pretend that the old version was buggy (it wasn't), then release an "upgrade"
[03:24:20] <cradek> twice2: that's just an optoisolator
[03:24:53] <Jymmm> jepler: No, I didn't. What purpose does that serve?
[03:25:11] <cradek> jepler: (?)
[03:25:28] <jepler> whoops, looks like sarcasm-over-tcpip transmission failed again
[03:26:16] <twice2> cradek: sorry, yes
[03:26:29] <cradek> jepler: vim's irritating autoindent chomped up your * list a bit
[03:27:43] <jepler> cradek: argh, I thought I double-checked that
[03:27:56] <jepler> oh well
[03:27:58] <cradek> that gets me all the time
[03:28:05] <cradek> yeah, no big deal
[03:28:42] <cradek> twice2: if I had any photography skills, you could maybe divine the schematic for that... but it's pretty poor isn't it
[03:29:45] <owhite> folks...I keep getting ...
[03:29:51] <owhite> insmod: error inserting '/home/owhite/emc2-head/rtlib/hal_parport.ko': -1 Device or resource busy
[03:30:13] <fenn> owhite: check dmesg
[03:30:13] <cradek> this is an emc you compiled?
[03:30:25] <owhite> which is odd, in that all I did was move my parallel card to another pci slot, found the address, and then changed taht address in a .hal file.
[03:30:29] <cradek> do you have a package also installed, or have you otherwise disabled the loading of parport_pc?
[03:30:37] <cradek> [a packaged emc2]
[03:31:21] <twice2> cradek: no pic is great. so you put the trans side in series with the speed pot and drive the current side with pwm from par port
[03:31:24] <owhite> cradek, sorry cant follow you. how do I disable loading parport_pc? and note, the problem just started when I changed the slot of my parallel card
[03:31:25] <fenn> cradek: the roller nut thing - you can't just turn the screws as it would change the preload
[03:31:52] <cradek> twice2: yes
[03:31:59] <owhite> oh hm....
[03:32:14] <owhite> dmesg reports...
[03:32:19] <owhite> [ 407.863818] config string '0xc400'
[03:32:19] <owhite> [ 407.864122] PARPORT: ERROR: request_region(c400) failed
[03:32:25] <cradek> owhite: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting#Parallel_port_no_longer_works_in_EMC_2_0_1_or_later_hal_parport_Device_or_resource_busy
[03:32:36] <fenn> i think you would need an even number >2 of rollers in order to preload them for anti-backlash
[03:32:50] <twice2> cool beans
[03:33:19] <cradek> twice2: it gives very nice control once I got the pwm scale/offset correct
[03:33:36] <twice2> cradek: right
[03:34:06] <dave_1> =//////////////////////= input from the cat
[03:35:14] <twice2> cradek: but then you must not touch the pot
[03:35:31] <cradek> it just stays up all the way
[03:35:38] <twice2> ah
[03:35:40] <cradek> I could use it as 'spindle override' (decrease only) but I don't
[03:37:48] <twice2> shop torn down, all boxed up and selling house, when sold I'll be shopless and homeless
[03:38:06] <cradek> yikes
[03:38:11] <cradek> hope you have a 'plan B'
[03:38:15] <twice2> and bored
[03:38:32] <dave_1> twice2 ... and planning a bigger house and shop :-)
[03:38:36] <twice2> yep, up nebraska way
[03:38:59] <cradek> why would you want to go there?
[03:39:04] <dave_1> why NB
[03:39:10] <cradek> or NE
[03:39:12] <twice2> beef
[03:39:16] <cradek> haha
[03:39:18] <dave_1> opps
[03:39:35] <twice2> haha
[03:40:11] <owhite> so my system was configured as described in the "Parallel port no longer works" documentation. People have any other suggestions for what I might try?
[03:40:47] <cradek> if 'request region' fails, something has already claimed that region. maybe you have the wrong address?
[03:41:05] <jepler> 'cat /proc/ioports' and look for what linux driver using that address range
[03:41:24] <SWPadnos> maybe CC00 isn't the right address - there should be others in the list
[03:41:38] <twice2> owhite: bios?
[03:41:50] <owhite> yah the address has changed in lspci, to 0xc400
[03:42:18] <owhite> twice2: would that apply to a pci parallel port card?
[03:42:33] <twice2> owhite: pnp?
[03:42:37] <SWPadnos> only for possible enumeration order problems
[03:42:48] <SWPadnos> twice2, PCI - always PnP
[03:42:52] <owhite> proc/ioports reports...
[03:43:09] <owhite> c400-c43f : 0000:02:08.0
[03:43:09] <owhite> c400-c43f : e100
[03:43:41] <SWPadnos> oh. good thing the request_region failed then
[03:43:47] <cradek> interesting
[03:43:53] <SWPadnos> no sense screwing up the ethernet card
[03:44:18] <SWPadnos> I wonder if the BIOS is buggy, which could also explain the problems with wireless vs. 5i20
[03:44:52] <twice2> owhite: is it an i/o flex card?
[03:44:57] <owhite> yah the wireless and 5i20 started to cooperate in a sense -- I moved them around, and they work together, but only when I removed the parallel card.
[03:45:09] <owhite> twice2: not sure.
[03:45:24] <dave_1> :wq
[03:45:34] <SWPadnos> heh - ok, is there a BIOS option something like "reset ESCD data"?
[03:45:34] <SWPadnos> if so, select it and reboot
[03:45:35] <SWPadnos> then reboot again
[03:45:48] <SWPadnos> (of course ytou have to reboot to get to the BIOS setup too)
[03:46:12] <owhite> so, reboot. go into the "fixxy the bios by hitting F12" mode, look at the bios options, and the reboot?
[03:46:19] <owhite> ah, got it.
[03:46:32] <SWPadnos> yes, and select "Reset ESCD Data"
[03:46:36] <owhite> "reset ESCD data". I'll check.
[03:46:42] <SWPadnos> with all cards installed
[03:47:09] <owhite> anything else to check in the bios?
[03:47:37] <SWPadnos> hmmm. make sure "Plug n Play OS" is set to "no"
[03:47:54] <owhite> yeah I've done taht before I'm pretty sure.
[03:48:01] <SWPadnos> ok
[03:51:03] <cxix> in a hal file, what is the difference between 'linksp' and 'net' ?
[03:51:59] <fenn> you have to create the signal before you can use linksp
[03:52:00] <twice2> uhmm, you can make multiple connections with net?
[03:52:26] <fenn> net uses the first argument as the signal name and finds the type automatically
[03:52:57] <cxix> in standard_pinout.hal and xylotex_pinout.hal, they're the exact same up until they "connect the pins to the signals" xylotex uses net, standard uses linksp
[03:53:10] <cradek> net can do the newsig + all of the link[sp] for that new signal at once
[03:53:38] <cradek> net is a newer addition to halcmd
[03:54:12] <cxix> so, stick with net?
[03:54:47] <SWPadnos> net is shorter for sure
[03:54:56] <twice2> stick with emc2, when I see mach it's like i'm looking a AOL.com :)
[03:55:05] <cxix> lol
[03:55:48] <cxix> see, but i've used mach, and all it took was about an hour to get a cnc running when i've never even seen one before. now i try to switch to emc2, and i've been at this for two days straight
[03:56:29] <SWPadnos> "with great power comes great responsibility" :)
[03:56:40] <fenn> if your machine is so standard then why are you messing with hal?
[03:56:46] <cxix> and the need for more knowledge
[03:56:50] <SWPadnos> dual-X gantry
[03:56:53] <cxix> it is apparently not standard
[03:57:14] <fenn> hm. how does mach handle homing for dual gantry machines?
[03:57:29] <fenn> or is that a dumb question?
[03:57:45] <cxix> homing? i duno. but to set it up, you click "slave a axis" and then "reverse a axis"
[03:58:14] <SWPadnos> fenn, I'm reasonably sure it doesn't
[03:58:16] <owhite> >>grumble<<
[03:58:26] <SWPadnos> but people use it with slaved axes, so there must be a way
[03:59:11] <owhite> why would this dumb parallel port card start screwing up just cause moved it to another slot?
[03:59:23] <cxix> well, there is a sample file the "stepper-gantry" uses two motors on the gantry, but it has all sorts of extra crap in there i tried to sort through
[03:59:24] <owhite> er, cause -I- moved it...
[03:59:25] <SWPadnos> BIOS and/or kernel issues
[03:59:54] <owhite> okay. well I guess all I can do is move it back.
[04:00:09] <owhite> or, try another one. :-)
[04:00:17] <jepler> 'night all
[04:00:22] <cxix> i've moved on to trying to modify the "stepper-xyza" and just set the a from angular to linear
[04:00:31] <SWPadnos> it sounds like you'll have problems with 3 PCI cards in that machine - I'm betting it won't matter where they are
[04:00:39] <owhite> right.
[04:00:51] <SWPadnos> cxix, what extra crap was in the stepper-gantry config?
[04:01:04] <SWPadnos> night jepler
[04:01:08] <SWPadnos> thanks for 2.2.3
[04:01:31] <owhite> irony is taht if I can get the M5i20 working then I wont need the parallel card. Guess I just have to move de-commisioning the card ahead in my schedule. :-)
[04:01:42] <SWPadnos> heh - you have a 7i37?
[04:02:00] <owhite> no. M5i20.
[04:02:00] <cxix> SWP - it kept saying stuff about servos, and has something like four extra files thrown in there, several hal files, i'm not sure what to do with them all and have been looking through the manuals.. i dont want my cnc to get up and dance... i just want it to cut things
[04:02:21] <SWPadnos> that's strange for a stepper config
[04:02:26] <SWPadnos> (to mention servos)
[04:02:35] <cxix> stepper-xyza has less stuff to sort through ... hoping it goes faster
[04:02:38] <cxix> yeah i thought so to
[04:02:39] <cxix> too.
[04:02:39] <twice2> night jep
[04:03:14] <fenn> # Hook functions to servo thread <- that?
[04:03:39] <cxix> yeah, what is that exactly?
[04:03:57] <fenn> it means emc recalculates all the velocities and such every millisecond
[04:03:57] <SWPadnos> there are multiple threads of execution in EMC, the one that runs at about 1 KHz is called the Servo thread
[04:04:19] <SWPadnos> even if you use steppers, that thread also does things like trajectory planning and such
[04:04:48] <cxix> i see i see. so what are all those other files in there for?
[04:05:08] <SWPadnos> you can split up the config info into multiple files
[04:05:14] <fenn> it was based on dallur's plasma table
[04:05:14] <SWPadnos> jarl apparently did that ;)
[04:05:33] <fenn> i agree it's too much random crap for a demonstration config
[04:06:07] <SWPadnos> note that you'd need to change one line from this:
[04:06:09] <SWPadnos> setp gantrykins.joint-3 1
[04:06:10] <SWPadnos> to this:
[04:06:12] <SWPadnos> setp gantrykins.joint-3 0
[04:06:17] <SWPadnos> in kinematics.hal
[04:06:38] <SWPadnos> you also have the option of just concatenating all those hal files into one single file (in the order they appear in the ini)
[04:06:48] <cxix> what if i skip that and use the stepper-xyza? it doesn't have a kinematics.hal
[04:06:59] <GNieport1> http://imagebin.ca/view/W6ajjMYO.html
[04:07:01] <SWPadnos> you need the connections in the kinematics.hal file
[04:07:12] <SWPadnos> that's what makes the fourth axis mimic the first axis
[04:07:16] <GNieport1> I think the tuning is getting better :)
[04:07:27] <fenn> you can include the kinematics.hal in stepper-xyza if it's easier
[04:07:39] <SWPadnos> yep, that'll work
[04:07:41] <cxix> i'll try that.
[04:07:52] <SWPadnos> but then you need to merge the two - you need gantrykins instead of trivkins
[04:08:18] <fenn> right - change 'loadrt trivkins' to 'loadrt gantrykins'
[04:08:43] <cxix> where is this text? the hal file in xyza?
[04:08:56] <cxix> found it
[04:11:18] <twice2> during times of universal deceit, telling the truth becomes a revolutionary act
[04:11:51] <twice2> orwell
[04:12:03] <SWPadnos> how apropos
[04:13:45] <twice2> Obsolete. by the way.
[04:14:42] <twice2> websters
[04:15:59] <twice2> I've been looking at Jon's pico stuff, I like the idea a true emc dro
[04:16:21] <SWPadnos> any servo system will get you that due to the encoder feedback
[04:16:31] <cxix> bbiab
[04:16:35] <SWPadnos> see ya
[04:16:45] <twice2> yeppers
[04:19:00] <twice2> if the encoders stay on
[04:24:28] <SWPadnos> that's a requirement for any feedback system
[04:26:38] <maddash> is it possible to get halcmd to display all the pins that are connected to parport.*?
[04:26:53] <SWPadnos> yes
[04:26:56] <maddash> and only those pins that are related to parport?
[04:27:01] <SWPadnos> yes
[04:27:12] <twice2> SWPandos: yes, should be
[04:27:13] <SWPadnos> halcmd show pin "*parport*"
[04:27:29] <SWPadnos> quoting not needed (and probably bad) if you're at a halcmd prompt
[04:27:44] <maddash> cool, thanks
[04:28:06] <SWPadnos> if you only want signals that are connected to parport pins, and you want to show the other connections, then it's a little harder
[04:28:09] <SWPadnos> sure
[05:25:42] <GNieport1> For a standard Bridgeport-size vertical mill, what would be a typical minimum following error setting? 0.005? 0.001?
[05:26:13] <SWPadnos> depends on a few things, including your encoder resolution and driver/motor specs
[05:26:50] <SWPadnos> MIN_FERROR is the allowable error at very low speeds, so that should be able to be pretty small
[05:32:16] <GNieport1> 8k counts/rev; 750W brushless
[05:32:16] <GNieport1> I have it at 0.002 right now, I have to quit EMC to recover from a following error?!
[05:32:16] <GNieport1> No errors with 300 IPM G0 moves, 60 IPM G1 moves
[05:32:16] <GNieport1> :)
[05:32:29] <SWPadnos> you shouldn't need to quit
[05:32:59] <SWPadnos> you should just be able to hit F2 or F1 then F2 to set "machine on" again
[05:33:11] <SWPadnos> cool onthe moves though :)
[05:33:30] <SWPadnos> 750W is pretty good power
[05:34:44] <maddash> rawwwr!!!!!! dot rules!
[05:35:03] <SWPadnos> um - yeah
[05:35:10] <SWPadnos> you may need to lay off the Jolt at night
[05:36:28] <tomp> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Following_Error (the graphic is broken, and that makes the explanation hard to follow)
[05:48:21] <GNieport1> SWPadnos when I hit machine on, the motor jumps as if still wound up and then EMC faults again.
[05:48:38] <GNieport1> thanks tomp
[05:49:29] <GNieport1> anyhow, if I change my INPUT_SCALE from -40000 to 40000, do I need to swap my encoder A and B
[05:49:39] <SWPadnos> no
[05:50:03] <SWPadnos> is the encoder connected to the drive?
[05:50:46] <SWPadnos> err - wait, does the axis travel in the wrong direction now, or do you just want the number to be positive?
[05:51:05] <GNieport1> okay. I jogged my table in y to the extreme front edge, closest to operator, and apparently my input scale has the wrong sign
[05:51:35] <GNieport1> Jogging back to the machine column would be an increasing negative move
[05:52:03] <SWPadnos> this is a BP-style machine - a knee mill?
[05:52:17] <GNieport1> the encoder is connected to the drive, but feeds out of the drive and into EMC
[05:52:23] <GNieport1> yes, knee mill
[05:52:37] <SWPadnos> ok, with the encoder fed to the drive, you don't want to move wires around :)
[05:52:57] <GNieport1> Well, I can swap them at the 5i20, but n/m
[05:52:58] <SWPadnos> on a BP-style knee mill, the table moving toward the operator is a positive Y move
[05:53:20] <SWPadnos> if you change the sign and the encoder wires, you wiill have done two negations, and changed nothing
[05:53:28] <SWPadnos> change one or the other if the direction is wrong
[05:53:37] <GNieport1> I thought 0,0 was supposed to be with the spindle over the front, left corner of the table
[05:54:14] <GNieport1> anyhow, my scaling is correct then
[05:54:32] <SWPadnos> depends on how you define "front" :)
[05:54:43] <GNieport1> the problem is, i hit "home" at the front edge, and EMC will not allow me to jog more negative
[05:54:58] <SWPadnos> it also depends on where you want the zero reference - middle of table or one end of travel
[05:55:06] <GNieport1> one end of travel
[05:55:35] <SWPadnos> ok, if you define home as 0, have a soft limit of 0, and that's the extent of travel, you shouldn't be allowed to jog any further negative
[05:55:47] <GNieport1> I broke it :)
[05:56:15] <SWPadnos> well, don't do that :)
[05:56:31] <GNieport1> The Y axis DRO has a little "center of gravity" symbol showing, and I think that is why I cannot jog
[05:57:12] <GNieport1> ah, i'll just estop it
[05:57:22] <GNieport1> that will 'break' home
[05:57:58] <SWPadnos> I think that means "homed"
[05:58:50] <GNieport1> hrm, it didn't clear the Home indicator
[05:58:59] <maddash> jmkasunich: are you there? in src/hal/utils/hal_commands.c, aren't data_arrow{1,2} identical? why are both needed?
[05:59:49] <GNieport1> any way to unhome without quitting EMC?
[06:01:17] <SWPadnos> hmmm. there's some way to override soft limits, but I don't know the keystroke
[06:01:36] <SWPadnos> is the Y scale the wrong sign?
[06:01:45] <SWPadnos> if so, wuit and fix it - you'll be happier :)
[06:01:49] <SWPadnos> s/wuit/quit/
[06:04:05] <GNieport1> thanks
[06:04:12] <SWPadnos> aure
[06:04:14] <SWPadnos> argh
[06:04:16] <SWPadnos> sure
[06:09:12] <tomp> GNieport1 http://pastebin.ca/875660 you had some probs and found some more :) heres some thoughts on the motor jump at startup & I'm off to bed
[06:09:35] <SWPadnos> indeed. bedtime here too (past, actually)
[06:32:46] <maddash> jepler: can I submit a patch for halcmd?
[06:35:43] <C_Morley> anyone here?
[06:36:07] <maddash> doesn't seem like it
[07:35:42] <cxix> swp - you still around?
[07:45:13] <cxix> what does joint 2 following error mean?
[07:55:07] <Jymmm> cxix: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Following_Error
[07:56:50] <cxix> jymmm - im also trying to setup this kinematics.hal file, but if i put it in the hal section of the ini, i get an error on loading the program, what did i do wrong?
[07:59:48] <toastydeath> toastydeath is now known as toast_sleep
[13:24:00] <toast_sleep> toast_sleep is now known as toast_work
[13:43:27] <acemi> probably this is a xorg bug in Debian Lenny but EMC2 puma interface resets X in my machine. Caught signal 11. Server abaroding...
[13:44:11] <eric_U> you sure it's not a kernel prob?
[13:45:03] <acemi> I'm not sure
[13:45:40] <eric_U> did you build it yourself?
[13:45:46] <acemi> yes
[13:57:25] <acemi> scara interface can't remove rtai modules when closing too. ERROR: Can't remove RTAI modules, kill the following process(es) first
[13:57:30] <jepler> you could boot a non-rt system, build emc with --enable-simulator, and see whether pumagui still crashes the X server. If so, I'd report a bug against the X server
[13:57:57] <acemi> ok jepler
[13:59:54] <acemi> but I can do that tomorrow. this a USB stick and there is no space for another kernel
[14:01:01] <jepler> I can duplicate your pumagui unload problem about one time out of three with halrun / loadusr pumagui / ctrl-d
[14:01:21] <jepler> ERROR: Can't remove RTAI modules, kill the following process(es) first
[14:03:57] <acemi> do I need to start realtime before this command
[14:04:08] <acemi> ./script/realtime start
[14:04:19] <jepler> halrun is a convenience script which sets up the realtime environment,
[14:04:19] <jepler> executes halcmd with the given arguments, optionally runs an interac-
[14:04:19] <jepler> tive halcmd -kf if -I is given, then tears down the realtime environ-
[14:04:19] <jepler> ment.
[14:04:25] <jepler> ^^^ from 'man halrun'
[14:04:31] <jepler> it's very convenient for me
[14:05:22] <acemi> yes it's dublicated
[14:06:47] <acemi> this is 2.2.3
[14:09:15] <jepler> I was using TRUNK but I am not surprised the problem exists in 2.2.3 as well
[14:14:17] <acemi> bbl, I want to try puma interface with vesa driver
[14:40:19] <acemi> the same problem exists for the vesa driver too. I think that X resets if I get Unexpected realtime error
[14:41:34] <acemi> normally my latency results are normal. no overrun, no jitter greater that 20000
[14:56:27] <skunkworks_> Great job on the 2.2.3 release!!
[16:15:36] <archivist> * archivist would like a bunch of different gerber files to test prototype parser produced last night/this morning www.archivist.info/search/index.php/Gerber
[16:17:17] <SWPadnos> archivist, does that program deal with RS274-X (tool data embedded in the file)?
[16:17:44] <archivist> as in embeded apertures yes
[16:17:52] <SWPadnos> I can make you some test files - simple shapes with just traces, some with holes/pads, then some with various SMT pad shapes
[16:17:54] <SWPadnos> ok
[16:18:24] <SWPadnos> I'm not sure if I can still run Protel, but I can do that with Altium Designer
[16:18:25] <archivist> format has changed a bit over the years
[16:18:42] <archivist> I have old PCAD and Kicad
[16:18:49] <SWPadnos> what language(s) is it written in?
[16:18:54] <SWPadnos> (your converter)
[16:19:00] <SWPadnos> uh - reader :)
[16:19:09] <archivist> erm some wont like the answer to that
[16:19:14] <archivist> php
[16:19:16] <SWPadnos> VB?
[16:19:20] <SWPadnos> oh, phew!
[16:19:31] <SWPadnos> I'm sure a C version could be done ;)
[16:19:37] <archivist> VB never heard of it
[16:19:50] <SWPadnos> good good, let's keep it that way
[16:20:01] <archivist> yes C to php to C is easy enough
[16:20:34] <SWPadnos> I'll try to make some files for you sometime, but I'm preparing for a trip at the moment (as usual)
[16:20:46] <archivist> php make it web'able as well
[16:20:57] <archivist> yes no rush
[16:21:33] <SWPadnos> it's funny - I'm not a fan of web apps in the slightest
[16:22:04] <archivist> damned easy way to make server apps though
[16:22:36] <SWPadnos> there is one advantage that I can see ("anywhere" accessibility), and some other things that could be advantages (possibly better maintained/protected data store), but I still don't like them
[16:22:36] <SWPadnos> sure
[16:26:17] <archivist> must have a google for the RS274-X spec later/tomorrow
[16:27:07] <SWPadnos> yeah - I'm not sure of the exact format - never bothered to open those big files
[16:27:58] <archivist> Im used to reverse engineering so its not too hard to read
[18:46:31] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[20:01:48] <alex_joni> archivist: hmm, then we need http filters for AXIS :)
[20:17:02] <alex_joni> see you guys
[20:20:54] <skunkworks> emc2 - try it because it's free - use it because it works.
[20:27:35] <alex_joni> skunkworks: did you do marketing in another life?
[20:28:42] <skunkworks> heh - not yeet
[20:28:43] <skunkworks> het
[20:28:45] <skunkworks> yet
[20:28:46] <skunkworks> oops
[20:29:01] <skunkworks> My spelling would get me in trouble. ;)
[20:31:51] <alex_joni> skunkworks: obviously :)
[20:32:13] <alex_joni> I think your subconscience was looking for 'njet'
[20:36:38] <jensor> Hello
[20:40:53] <jensor> I'm trying to file share between my EMC box and my desktop XP box. Information I have gleaned indicates I need Samba installed in order to do this. However when I start looking at that I see things indicating that I need to recompile the kernal it scares me to death thinking that I will mess up my 2.6.15-magna
[20:41:26] <jensor> Does anyone have any suggestions?
[20:42:50] <jensor> I want to do all my gcode creation on the desk top unit5
[20:44:27] <jensor> The ethernet card is installed and working - can now go on line with the EMC box
[20:47:41] <cradek> you do not need to compile anything.
[20:47:51] <cradek> go to places/home folder
[20:47:59] <cradek> right click the folder you want to share
[20:48:38] <skunkworks> cradek: on the microsoft box?
[20:48:48] <cradek> no on the linux
[20:48:52] <skunkworks> ok
[20:49:20] <cradek> share folder, samba, etc etc
[20:49:32] <cradek> it may install software (samba) if necessary
[20:50:15] <cradek> http://www.google.com/search?q=dapper+sharing+files+with+windows
[20:50:24] <cradek> the second hit here looks good
[20:50:46] <jensor> it's installing samba now
[20:50:54] <cradek> 4th message of the 2nd hit of that search
[20:52:57] <jensor> My desktop is already sharing with my laptop okay
[20:56:53] <jensor> oh oh- error msg: The following problems were found on your system E:Samba:subprocess post-installation script returned error exit status 102
[20:57:17] <SWPadnos> you shouldn't have to install any Samba packages to be able to access Windows shares
[20:57:56] <alex_joni> SWPadnos: only smbclient
[20:58:02] <SWPadnos> in the Places menu, there's an entry something like "Network"
[20:58:05] <alex_joni> or smbfs which is part of smbclient
[20:58:14] <alex_joni> good night all
[20:58:22] <SWPadnos> click that, and you should see something like the Windows "all networks" icon
[20:58:32] <SWPadnos> alex_joni, yes, but that's there by default on 6.06, I think
[20:59:08] <cradek> ok, jensor did not say which way he wanted to do the sharing, but since he talked about samba I assumed that direction
[20:59:37] <SWPadnos> yeah - I'd expect him to want to access CAM files posted from a Windows program
[20:59:42] <SWPadnos> so the other way around
[21:00:10] <cradek> your guess is as good as mine (or maybe even better)
[21:00:26] <SWPadnos> roughly equivalent ("conditionally true" :) )
[21:04:32] <jensor> now have a window - under WINS server 3 choices - (1) Do not use WINS server (2) This computer is a WINS server and (3) Use WINS server - for this choice it wants a name
[21:04:49] <SWPadnos> do not use WINS server
[21:04:56] <jensor> ok
[21:04:58] <SWPadnos> (same as on the Windows machines, I bet)
[21:06:49] <jensor> Now have file browser open on EMC box don't see my desktop files any where
[21:07:52] <SWPadnos> "desktop files"?
[21:07:57] <jensor> Looking at desktop windows explorer -don't see any EMC stuff
[21:08:12] <SWPadnos> ok, which way are you trying to share files?
[21:08:33] <SWPadnos> reading Windows shares from the EMC machine or reading an EMC share from the windows machine(s)?
[21:08:43] <jensor> MY XP madchine I mean to say doesn't show on the EMNC box
[21:08:52] <jensor> either way
[21:08:56] <SWPadnos> ok
[21:09:05] <SWPadnos> one sec
[21:10:22] <toast_work> toast_work is now known as toastydeath
[21:10:34] <SWPadnos> where are you looking (for the windows machine), and how did you get there?
[21:10:55] <jensor> I am using the linux file browser in ubuntu
[21:11:54] <jensor> I have already setup the XP box to share the folder named MY Documents
[21:12:55] <Gamma-X> how do i look up the email address of the poster for msg#00085?
[21:13:21] <SWPadnos> jensor, what is in the window when you go to Places -> Network Servers?
[21:16:09] <SWPadnos> Gamma-X, whatever mailing list it is should have tools for that. if it doesn't, then your best bet is a web search for some archinve that may show the address
[21:16:13] <jepler> Gamma-X: there are several places online where the emc mailing lists are archived. They each differ in their rules and practices concerning e-mail addresses. The only dependable way to find the address of people who have written on the mailing lists is to be a subscriber via e-mail
[21:16:58] <jensor> By Golly - it's all there Thanks Now how do I share my linux home folder with the XP box
[21:17:05] <Gamma-X> thanks
[21:17:25] <SWPadnos> jensor, that's a little harder, and requires that Samba be working, and there are some permissions to set as well
[21:17:38] <SWPadnos> but you should be able to right-click and share, similar to Windows
[21:18:11] <jensor> I see . Thats not as important to me although it might be handy
[21:23:35] <jensor> Ok I rt click see window - shows path to my emc2 folder, share with: SMB under share properties see NAme I named it EMC and a button to click entitled General windows sharing settings
[21:25:18] <jensor> I guess it's not going to work since I got that error during the Samba install
[21:27:57] <maddash> jepler: are you there? I'd like to submit a patch that adds a feature to halcmd
[21:29:04] <SWPadnos> jensor, the error may or may not be a problem
[21:29:24] <SWPadnos> maddash, halcmd changes could be sent to jmkasunich or myself as well as jepler
[21:29:34] <SWPadnos> what's the feature, out of curiosity?
[21:29:52] <maddash> DOTTY!!!!!!!!!!!!!!!
[21:29:58] <maddash> rawr!
[21:30:01] <SWPadnos> in english?
[21:30:51] <maddash> how about a thousand words, instead?
[21:31:00] <SWPadnos> no, pictures are fine
[21:31:01] <maddash> I ran tkemc.sim
[21:31:07] <maddash> http://img165.imageshack.us/my.php?image=testmi5.png
[21:31:11] <SWPadnos> are you talking about support for the AT&T DOT/DOTTY software?
[21:31:41] <maddash> so, may I submit and will it commit?
[21:32:00] <SWPadnos> what does the patch do, exactly?
[21:32:11] <SWPadnos> is it a new form of "show" output?
[21:32:14] <maddash> er, didn't you see the link?
[21:32:17] <maddash> yes
[21:32:27] <SWPadnos> I see the graph, I'm wondering how it relates to halcmd
[21:32:28] <SWPadnos> ok
[21:32:32] <maddash> ...
[21:32:58] <maddash> er, the patch gets halcmd to generate those automatically
[21:32:59] <SWPadnos> I think the arrows are backeards
[21:33:07] <maddash> hm,
[21:33:09] <maddash> A => B
[21:33:13] <maddash> who's input?
[21:33:17] <SWPadnos> look at ypos
[21:33:31] <SWPadnos> you can't have two writers to a signal, and there are two arrows pointing at ypos
[21:33:45] <maddash> argh, so A's output?
[21:33:52] <SWPadnos> in A => B, A writes to B
[21:33:55] <SWPadnos> yes
[21:34:06] <maddash> hm, it was vice versa in print_pin_info
[21:34:11] <SWPadnos> when you say "can generate those automatically", what do you mean?
[21:34:17] <SWPadnos> shouldn't be
[21:34:33] <maddash> look at "data_arrow2() " inside halcmd_commands.c
[21:34:37] <maddash> or dataarrow1
[21:35:06] <maddash> by automatically, I mean that I generated that using "halcmd -k dotty all"
[21:36:13] <jensor> SWPadns:Under System > Administration > Shared folders , I see the selected folder listed okay, but still can't see anything from the windows box.
[21:36:44] <jensor> You said something about permissions need to be set
[21:36:49] <jensor> how do I do hta t?
[21:37:08] <maddash> SWPadnos: you're right, they're backwards
[21:37:44] <maddash> SWPadnos: actually, since you're familiar with halcmd, can you tell me why there are two data_arrow functions?
[21:37:58] <SWPadnos> jensor, I don't know specifically, but I bet you can ask google or the ubuntu wiki and find out - that's a typical Windows networking question
[21:38:18] <SWPadnos> maddash, dunno off the top of my head
[21:38:21] <jensor> Thks - I'll give it a try!
[21:38:38] <SWPadnos> jensor, sure - good luck
[21:39:07] <maddash> er, so do I submit or not?
[21:39:57] <SWPadnos> maddash, I don't know yet. you mentioned that halcmd can do this "automatically" - what do you mean by that?
[21:39:59] <maddash> and if I do, should I submit to http://sourceforge.net/tracker/?group_id=6744&atid=356744 (feature request page)?
[21:40:17] <maddash> 16:35:06 <maddash> by automatically, I mean that I generated that using "halcmd -k dotty all"
[21:40:25] <SWPadnos> ah - ok
[21:40:38] <maddash> :)
[21:40:56] <SWPadnos> and the dotty command takes the same arguments as the show command?
[21:41:10] <maddash> well, only those that make logical sense
[21:41:34] <maddash> so, if you tried, "halcmd dotty function" or s/function/param/, it wouldn't work
[21:42:02] <maddash> because a digraph isn't appropriate to displaying a list of things that aren't related to each other
[21:42:05] <SWPadnos> ok. does thread work?
[21:42:30] <maddash> I could add it in. how would threads be related to each other?
[21:42:37] <SWPadnos> well, functions are related - you could do a central node with the component name and children with all the functions (or params, for that matter)
[21:42:50] <SWPadnos> you'd have a single chain for each thread, showing the order
[21:43:06] <maddash> ah, I see what you mean
[21:43:12] <maddash> sure, hold on
[21:43:30] <SWPadnos> that stuff would be more helpful if it also had data printed (like show would do) - thread period/frequency, etc.
[21:43:39] <maddash> while I'm at it, are there any other types besides "pin, sig, thread"" that you'd like to include?
[21:44:21] <SWPadnos> as for accepting the patch, I think it's cool, but I'm not sure it fits with the idea od halcmd, which is "an interface for HAL manipulation"
[21:44:30] <SWPadnos> s/od/of/
[21:44:44] <SWPadnos> that may be a jmk call more than anyone else
[21:44:47] <maddash> hm, well the dotty command is essentially the show command, on steroids
[21:45:12] <SWPadnos> yep, but only useful if you have dotty installed - I'm sure a patch that has halcmd create the bitmap wouldn't be accepted ;)
[21:45:59] <maddash> I don't understand the second part of that remark
[21:46:42] <SWPadnos> if the output of halcmd is a bitmap, then that's not appropriate
[21:46:57] <SWPadnos> it if's a script that dotty uses to create a bitmap, that's more appropriate
[21:47:01] <SWPadnos> arg
[21:47:08] <SWPadnos> if it's ...
[21:47:22] <maddash> clearly not. but even the raw digraph output is useful by itself.
[21:47:39] <maddash> er, in response to halcmd/bitmap
[21:47:40] <SWPadnos> but it still requires either some dotty-lib to be linked in, or another package installed, to be useful
[21:47:52] <maddash> sure, so use a "Recommends"
[21:48:14] <SWPadnos> could do that - as I say, it's questionable whether it fits with the halcmd premise
[21:48:16] <maddash> and adding dotty in wouldn't subtract anything from the existing halcmd
[21:48:50] <SWPadnos> can a utility be written that takes the output of existing halcmd functions and creates the same digraph, or do you need HAL internal data
[21:49:00] <SWPadnos> I think it could be done as a separate program
[21:49:08] <maddash> hm, that'd be a bit more work
[21:49:17] <maddash> the halcmd output isn't exactly script friendly
[21:49:23] <SWPadnos> just pointing out where there may be resistance
[21:49:31] <SWPadnos> have you used script mode? (halcmd -s)
[21:49:53] <SWPadnos> no headers, things like signals printed on one line, for example
[21:49:59] <maddash> and it wouldn't be very flexible either, because if the halcmd output changed tomorrow, the haldotty util would need to change as well
[21:50:07] <maddash> let me go see
[21:51:21] <maddash> i'll wait until jmkasunich returns
[21:51:39] <SWPadnos> that's a good plan. I"m not so great at deciding what should / shouldn't be added ;)
[21:53:00] <maddash> btw, if I bought an AVR, i would most certainly need to fabricate my own PCB before I could begin to program it, right?
[21:53:15] <jepler> I take the viewpoint that programs can already parse halcmd's output. http://axis.unpy.net/01174426278
[21:53:21] <bill2or3> nah, you could use a breadboard.
[21:53:25] <cradek> you can program one easily on a proto board
[21:53:38] <bill2or3> assuming you have a programmer of some sort.
[21:53:41] <SWPadnos> maddash, depends on whether you get a through-hole one or surface mount
[21:53:44] <maddash> yeah, but the bottom line is that some extra work needs to be done
[21:53:50] <bill2or3> a little.
[21:54:22] <maddash> jepler: bwahaha, that thing does the exact same thing as mine
[21:55:10] <maddash> jepler: where's the timestamp on that post?
[21:55:43] <cradek> jepler: is that layout autogenerated? it looks great
[21:55:55] <jepler> maddash: 20 March 2007, 21:31 UTC
[21:56:10] <jepler> cradek: yes, my recollection is that it's the automatic layout from 'dot'
[21:56:13] <maddash> er, where'd you get that?
[21:56:19] <jepler> maddash: er, I wrote it
[21:56:53] <maddash> so, where's net2dot?
[21:56:54] <jepler> maddash: oh, the date? it's shown on the blog index, or you can compute it from the URL (the number is seconds since unix epoch)
[21:57:09] <maddash> haha, nice
[21:57:14] <jepler> there's a link to it on that page in the "Files currently attached to this page:" section
[21:57:29] <maddash> oh,so it's not included in the official emc?
[21:58:03] <jepler> uh no
[21:58:44] <jepler> there are lots of things that work with emc that don't make sense to include with emc
[21:58:48] <jepler> IMO
[21:58:53] <jepler> that includes a lot of the things I do
[21:59:15] <jepler> one reason not to include them is that relatively few people will care about them; another is when I have no intention of maintaining them over time and I don't know anybody else who does
[21:59:21] <jepler> I think both apply to this net2dot program
[22:00:45] <maddash> hah, net2dot is way more versatile than my patch
[22:01:21] <maddash> hm, or not
[22:01:22] <SWPadnos> uh - I guess that means it's unlikely to be accepted ;)
[22:01:31] <maddash> why doesn't circo work in place of dot?
[22:01:40] <maddash> SWPadnos: I'm already past that
[22:01:43] <SWPadnos> heh
[22:12:33] <GNieport1> Hi all. I finally got the knack of PID tuning after several swift kicks from you all. Thanks for all of the suggestions :)
[22:13:35] <SWPadnos> heh - any time :)
[22:13:55] <GNieport1> Now to get the spindle speed to obey speed commands from my unused analog out channel...
[22:14:14] <SWPadnos> same instructions, different drive :)
[22:14:48] <GNieport1> No feedback on the spindle drive (yet)
[22:14:56] <SWPadnos> it's a VFD?
[22:15:00] <GNieport1> Just need to figure out the scale factor
[22:15:03] <GNieport1> yes.
[22:15:24] <SWPadnos> it'll probably be fine just running that open-loop (for speed control at least)
[22:15:38] <SWPadnos> once you get feedback hooked up, you'd be able to do threading and the like
[22:15:42] <GNieport1> 0-10V , with a digital signal to select fowrad or reverse
[22:16:08] <GNieport1> any idea how many pulses per rev will be needed for rigid tapping?
[22:16:39] <SWPadnos> depends on the speed, but "the more the merrier"
[22:16:49] <SWPadnos> since you have a Mesa card, count speed shouldn't be an issuw
[22:16:51] <SWPadnos> issue
[22:17:17] <cradek> whatever you have that's a few hundreds or more
[22:17:35] <cradek> mine is 1024, the mazak is 360 I think
[22:17:40] <GNieport1> I think I have a 1048 line encoder somewhere
[22:17:42] <cradek> (cycles)
[22:17:48] <cradek> that'd be fine
[22:17:58] <GNieport1> Sweet
[22:17:58] <cradek> (what a strange number)
[22:18:12] <GNieport1> 2^x
[22:18:13] <SWPadnos> I'm betting that's a typo
[22:18:18] <cradek> I have some 504 line!
[22:18:21] <SWPadnos> 1024 is 2^10
[22:18:23] <cradek> oh, 2048
[22:18:27] <GNieport1> haha, must be that then
[22:18:28] <SWPadnos> that too
[22:18:46] <cradek> SWPadnos: now figure out what 504 is!
[22:18:52] <cradek> (and no, I don't know)
[22:18:55] <SWPadnos> 2^9-8
[22:18:59] <SWPadnos> :)
[22:19:01] <cradek> uh of course
[22:19:04] <cradek> that must be it :-)
[22:19:42] <cradek> 10^3/2 + oops
[22:20:17] <SWPadnos> yeah - 508 I could imagine, but 504 doesn't make much sense
[22:20:41] <cradek> I was so skeptical I opened one, to find '504' stamped on the wheel
[22:21:06] <SWPadnos> it's 126 per quadrant, but I don't know where that would be useful
[22:21:10] <SWPadnos> (or 63 per octant)
[22:21:16] <cradek> they're from printers. probably something to do with some dpi
[22:21:24] <cradek> combined with the number of teeth on the pulley
[22:21:35] <SWPadnos> yep - could be
[22:28:45] <alex_joni> 504 using 4x ?
[22:28:48] <GNieport1> So would I be setting up the spindle VFD the same as a servo for threading? Or does EMC just slave/electronically gear the actual servos to the spindle?
[22:29:27] <SWPadnos> the axes are slaved to the spindle
[22:29:41] <GNieport1> Great!
[22:40:35] <toastydeath> toastydeath is now known as toast_school
[22:52:20] <micges> hi all
[22:54:42] <jepler> 504 counts gives 7 counts per 1|72 inch
[22:55:02] <jepler> 1|72 inch is 1 "point" in postscript
[22:55:24] <alex_joni> 1:7 gearing?
[22:55:47] <jepler> er I'm stating it badly
[22:56:11] <jepler> but I think it's a number one would choose because it's a multiple of 72
[22:56:29] <jepler> obviously it depends on the screw pitch as well
[22:56:36] <micges> jepler: can you tell me name of tool you used to find uninitialized memory regions in emc ?
[22:56:43] <alex_joni> valgrind?
[22:56:50] <jepler> yes, valgrind
[22:57:04] <alex_joni> http://valgrind.org/
[22:57:04] <micges> thanks
[23:01:29] <micges> must check my converter :P
[23:03:01] <alex_joni> fenn: this is cool: http://fenn.dyndns.org/pub/emc/screenshots/3d-spiro.png
[23:09:49] <jepler> bbl .. time to make dinner
[23:30:54] <jensor> hello- I'm in extremely bad trouble now. In the process of setting up file sharingbetween my EMC box and my XP box I can no longer get any response out of Ubuntu. can't get terminal , text editor, or emc2 to run. What do I do now?
[23:32:05] <jensor> It's all because I am so ignorant on the system that I goofed it up - I now plead for help to try to recover
[23:32:40] <jensor> Do I dare try to shut down and reboot
[23:32:47] <jensor> Will that fix it?
[23:38:17] <jensor> The name of the computer is Rex and I think somehow that got altered when try to get file sharing going may have brought this about.
[23:41:57] <jepler> I doubt you'll make things worse by restarting. If it doesn't restart normally, then you'll have to enter the boot menu, choose "recovery mode", and figure out how to put back whatever you changed using a text-mode root login.
[23:42:32] <jepler> entering the boot loader, choosing "e" to edit the commandline, and removing "quiet" and "splash" will show you additional information during startup which may be helpful
[23:42:34] <jensor> ok
[23:42:45] <jensor> we'll see what happens
[23:50:53] <jensor> ok , I'm up called up the terminal and I see that the command prompt is now different. It reads: Jj-nu:`$ Jj-nu is the name of nmy xp box
[23:53:19] <jensor> That implies that somehow my xp box is in the path
[23:54:58] <cradek> looks like you changed the hostname somehow
[23:55:11] <cradek> that would be consistent with not being able to open new programs
[23:56:04] <jensor> how do I correct that?
[23:56:28] <cradek> system/administration/networking, general, hostname
[23:56:44] <cradek> I suspect that's where you changed it shortly before you started having trouble
[23:58:59] <jensor> Okay I entered what I thought should be there and I guessed wrong cuz now it behaves as before.
[23:59:13] <jensor> How do I determine what the host name should be?
[23:59:29] <cradek> you can call it whatever you want
[23:59:53] <cradek> just call it emc or something