#emc | Logs for 2010-10-27

[00:03:48] <theorb> theorb is now known as theorbtwo
[01:22:55] <pcncx> wholly smokes I was concentrating on this diff -y here so hard I had an out of body experience!
[04:02:03] <LawrenceG> fest in in Wichita?
[04:02:25] <LawrenceG> is in
[04:04:07] <LawrenceG> thats like 4 days of driving from Vancouver!
[04:10:01] <pcncx> 2nd build and I'm online!
[04:21:35] <Jymmm> LawrenceG: Start driving now =)
[04:24:38] <LawrenceG> hey Jymmm
[04:24:53] <Jymmm> LawrenceG: whats happening
[04:27:06] <LawrenceG> http://imagebin.ca/view/6FPotuS.html
[04:27:16] <LawrenceG> not much
[04:28:50] <Jymmm> LawrenceG: Just working on the air scrubber here (still)
[04:29:31] <pcncx> Jymmm you dive?
[04:29:53] <Jymmm> pcncx: Kinda of a personal question dont you think?
[04:30:07] <pcncx> Jymmm what else would you need an air scrubber for?
[04:30:35] <LawrenceG> you are practically on the way! want me to pick you for fest?
[04:30:46] <Jymmm> pcncx: My laser... http://www.engraversnetwork.com/uls/m300.html
[04:30:56] <Jymmm> LawrenceG: who me?
[04:30:57] <elmo40> only 3200km...
[04:31:09] <LawrenceG> Jymmm, sure
[04:31:24] <pcncx> Jymmm you got sharks too?
[04:31:27] <Jymmm> LawrenceG: Sure, if you go 800 miles south first =)
[04:31:29] <LawrenceG> need to put some miles on the jeep
[04:31:56] <Jymmm> LawrenceG: But, It's Kansas..
[04:32:17] <pcncx> well it is centrally located
[04:32:31] <pcncx> bout as far away as you can get from everywhere!
[04:32:32] <LawrenceG> south might be a good idea this time of year
[04:32:34] <Jymmm> So is SFO =)
[04:33:39] <Jymmm> I hate to say it, but I can't see any purpose for traveling east of Colorado
[04:33:49] <LawrenceG> :}
[04:34:00] <pcncx> my general rule is don't go west of the Mississippi
[04:34:31] <pcncx> world ends long before then anyways
[04:34:38] <Jymmm> The south is just muggy and humid.
[04:34:55] <Jymmm> New England is cold
[04:35:12] <LawrenceG> I wonder if anyone has welded up a set of wheels on a bridgeport and towed it cross country with a wrangler
[04:35:22] <Jymmm> LawrenceG: LOL
[04:35:28] <LawrenceG> (without getting arrested)
[04:35:36] <pcncx> Jymmm upstate Maine is nice
[04:35:58] <Jymmm> pcncx: Maybe for visiting in the fall with the autumn leaves.
[04:36:11] <pcncx> Jymmm oh there's no trees with leaves there
[04:36:26] <pcncx> its all pine
[04:36:50] <Jymmm> I got OR and WA and NorCal for that.
[04:36:59] <pcncx> when you go there you know how an ant feels running around in a lawn
[04:37:10] <Jymmm> heh
[04:37:44] <elmo40> Ontario needs a'fest
[04:37:50] <elmo40> OpenSource Fest
[04:37:52] <pcncx> millions and millions of acres of nothing but pine trees
[04:38:08] <elmo40> silly Linux conference was cancelled :(
[04:38:20] <pcncx> elmo40 telecommute
[04:38:45] <pcncx> I can't even believe 2nd kernel build and i got online I'll have to fix that wit hthe next build I do
[04:38:48] <LawrenceG> Potland would be a good west coast fest location
[04:39:06] <LawrenceG> Portland... although potland could work!
[04:39:23] <Jymmm> heh
[04:39:24] <LawrenceG> isnt that near you Jymmm
[04:39:37] <Jymmm> LawrenceG: Yeah, it's on the ballot this year.
[04:39:45] <LawrenceG> more northern ca
[04:39:51] <pcncx> Wackenhut needs inmates
[04:39:57] <skinnypup> lol
[04:40:05] <Jymmm> LawrenceG: Well, medical MJ, is legal in Calif now.
[04:40:07] <pcncx> pot heads make the best prisoners!
[04:40:29] <elmo40> if a fest was happening on a Canadian long weekend... I might be able to do the drive ;)
[04:40:37] <pcncx> Jymmm no it isn't it violates federal and international laws
[04:40:42] <Jymmm> LawrenceG: $60 gets you legal anywhere in calif.
[04:41:23] <pcncx> until they can figure out how to outlaw stupidity i say keep the drug laws on the books they're our only protection!
[04:44:01] <Jymmm> pot has never killed anyone. They could tax it for entertainment purposes, and helps those in need for medicinal purposes, and is natural.
[04:44:30] <skinnypup> stupidity has killed on the other hand.....
[04:44:42] <skinnypup> that's natural too
[04:44:52] <LawrenceG> known side effects.... excessive craving for nacho chips
[04:45:03] <elmo40> with dip?
[04:45:04] <pcncx> yeah I'm not seeing people getting any smarter puffing the funny stuff
[04:45:05] <Jymmm> take off the warning labels and let evolution take over
[04:46:32] <pcncx> LawrenceG show somebody stoned a box of Yodels I dara ya
[04:47:45] <LawrenceG> yodels? doesnt translate to canuck
[04:48:03] <pcncx> LawrenceG wow you guys don't have Drakes cakes?
[04:48:47] <LawrenceG> been a while since I have done much shopping, but it doesnt sound familiar
[04:48:48] <pcncx> they were better years ago
[04:50:02] <LawrenceG> pcncx, yea much fresher!
[04:50:17] <pcncx> LawrenceG I think they had more saturated fats in them back then
[04:57:09] <pcncx> anyone ever wonder why the kernel .config file is hidden?
[04:59:56] <LawrenceG> off to bed
[05:00:04] <Jymmm> LawrenceG: G'Night
[05:00:07] <LawrenceG> gn
[05:02:41] <pcncx> I'll tell you this EMC2 stuff really makes you learn about Linux
[05:04:35] <skinnypup> pcncx, you may like this
[05:04:36] <skinnypup> http://wiki.debian.org/DebianLive/HelperTree
[05:05:44] <pcncx> skinnypup instructions on how to make a live image?
[05:06:00] <pcncx> I tried to do that once years ago man it was complicated
[05:06:22] <skinnypup> basically but the amount of customizing you can do there is unlimited
[05:06:37] <pcncx> well its not like we don't have the source code ;)
[05:06:57] <skinnypup> hear ya :)
[05:07:00] <pcncx> so yes it is basically limited only by the hardware you're running on
[05:07:19] <pcncx> and if you're godo with a soldering iron even that can be hacked
[05:07:59] <pcncx> skinnypup it is pure dumb luck I got an internet connection on the second build this try
[05:08:26] <pcncx> I was just about to reboot to do some searching when I said let me do an ifconfig and I saw an IP in it
[05:08:27] <skinnypup> awsome !
[05:08:51] <pcncx> there is still stuff i need to sort out
[05:09:07] <pcncx> like when I'm in menuconfig it locks some stuff on me and i can't deselect it
[05:09:29] <pcncx> I got this junk called hugetlbfs why I don't know!
[05:09:58] <skinnypup> i need to squeeze a new xorg into lenny so i can use www.intellinuxgraphics.org drivers on my atom 330 and 510
[05:10:06] <pcncx> if this machine ever was a 16GB file I think it'd burst into flames
[05:10:21] <skinnypup> will be the next linux hurdle i hop
[05:10:34] <pcncx> well I'm running with no gfx drivers now
[05:10:45] <pcncx> being as the other ones gave me so much troubles
[05:11:09] <pcncx> if i had r128 loaded it's lock my machine up when I started EMC2
[05:11:45] <pcncx> well it was strange i could still move my mouse cursor around but it'd do nothing
[05:11:53] <skinnypup> i saw some of that scanning through earlier
[05:12:16] <pcncx> like i always have focus follows mouse on my windows but they wouldn't even raise
[05:12:26] <pcncx> and you couldn't click on anything
[05:12:36] <pcncx> and the keyboard was dead too
[05:12:54] <pcncx> but I didn't have this module called endev loaded then
[05:13:02] <skinnypup> ive had that a few times when i had 10k browser tabs open ....
[05:13:15] <pcncx> or enedev? I can't remember i have it noted in a file someplace
[05:13:56] <pcncx> so I will give it another shot with that going and see if it clears it up or not
[05:14:16] <pcncx> but right now I'm in X with no drm or r128 drivers
[05:14:49] <pcncx> but I'm here so go figure
[05:15:48] <pcncx> evdev
[05:15:55] <pcncx> just did an lsmod and saw it
[05:17:37] <pcncx> hey do I need ipv6? I never use the stuff
[05:17:56] <pcncx> stuff is horking back a ton of RAM
[05:17:58] <Jymmm> no
[05:20:21] <pcncx> sometimes i see people on IRC using it but I've really no idea how to
[05:22:02] <pcncx> I swear they're going to be chasing after me with butterfly nets if i keep on working on this project
[05:24:19] <Jymmm> * Jymmm pulls out the tranq gun
[05:26:19] <pcncx> a couple more rounds of tweaking this kernel and I'm going to be tweaking
[05:55:05] <Jymmm> HAHAHAHAHA
[07:41:33] <Jymmm> http://www.reuters.com/news/video/story?videoId=163932228&videoChannel=2602&refresh=true
[12:59:51] <Paragon39> Hello Everyone!
[13:00:28] <elmo40> Paragon39: morning
[13:01:43] <elmo40> Jymmm: wtf are they talking about 'labour shortage' ?
[13:02:07] <elmo40> there are more people now then there ever has been.
[13:54:04] <Jymmm> elmo40: what you talking about?
[13:54:38] <elmo40> that reuters video about the robot... the japanese guy in there said (paraphrasing) 'this is good for the labour shortage we will have'.
[13:54:49] <elmo40> and I say what damn shortage? we have a shortage of employment!
[13:55:42] <Jymmm> Oh, in Japan as most workers are in their 60's aka baby boomers
[13:57:19] <elmo40> is there... so japan is way different then every other country? I don't buy it.
[13:57:51] <Jymmm> factory workers
[13:59:51] <elmo40> because kids in the last 15 years have been brought up with the 'tech bubble/bust' and think it is the best (easiest) way to make money. They don't realize that actually making the parts is still required ;)
[14:38:19] <skinnypup> Yeah i halfway wanted to stick the robot in the woodchipper....
[15:50:40] <Paragon39> Has anyone managed to get EMC2 Axis running over Xming (X11) ? Every time the Gui starts i get fatal error popup on windows.
[15:51:13] <SWPadnos> I have used Cygwin, but not Xming
[15:51:45] <Paragon39> SWPadnos: And that displayed the axis gui ok?
[15:51:47] <SWPadnos> try tkemc. If that works, it points somewhat at an improper (or missing) openGL on xMing
[15:51:50] <SWPadnos> yes
[15:52:04] <Paragon39> ok
[15:53:53] <Paragon39> tkemc worked ok.
[15:54:50] <Paragon39> seeing the tkEMC gui has taken me back a few years ;-)
[15:56:40] <Paragon39> Huh! The axis gui has started up after I ran the tkEMC gui but the preview is blank? Weird
[15:57:01] <SWPadnos> opengl ...
[15:57:28] <Paragon39> Yep must be! Any ideas for a solution?
[15:58:41] <psha> google is silent about it?
[15:59:19] <Paragon39> Found this ... Xming uses either [WWW]Mesa or [WWW]Microsoft WGL [WWW]OpenGL for rendering 3D graphics. But...
[16:00:07] <psha> there is large troubleshooting page on xorg
[16:00:07] <atmega> glxgears works with xming
[16:00:16] <psha> http://www.straightrunning.com/XmingNotes/trouble.php
[16:00:22] <psha> oops not xorg
[16:00:26] <atmega> (for a semi-random data point)
[16:01:20] <Paragon39> glxgears starts but is so so slow. Not surprised really.
[16:02:08] <Paragon39> psha: That the same page I took the above quote from ;-)
[16:06:36] <psha> have you tried various environ quirks described there?
[16:07:25] <psha> btw small poll: what gladevcp widgets are missing and needed?
[16:16:50] <cradek> psha: none - you already did the one I needed
[16:23:51] <Jymmm> I have (lets call it) a bed of nails, and there is airflow horizontally thru the bed. I'd like some way to incorporate a vacuum holddown when placing a piece of material on the nail. any suggestions?
[16:24:19] <skunkworks> hollow nails ?
[16:24:40] <Jymmm> No, they are 0.093" pins.
[16:24:55] <cradek> sounds contrary to nature
[16:25:00] <cradek> I mean, impossible
[16:25:13] <cradek> you can't hold something with vacuum unless you have two matching surfaces that make a seal
[16:27:03] <Jymmm> It doens't have to be perfect, mostly a downward draft
[16:27:29] <Jymmm> a strong draft preffered.
[16:27:47] <mrsunshine> hmm, is there a special instruction to tell emc to do a tool set?
[16:27:57] <mrsunshine> that is when you using a tool setter to set the tool offset
[16:28:15] <Jymmm> It's the DOWN part I can't figure out. I though of maybe diaganal slats to direct the airflow
[16:28:57] <skunkworks> mrsunshine: does this help? http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=nc_files/tool-length-probe.ngc;h=8b41233d2afe19ef6b9f17706108d0eb356f066b;hb=HEAD
[16:32:44] <mrsunshine> skunkworks, cant realy figure how to use it :/
[16:38:33] <skinnypup> mrsunshine, http://linuxcnc.org/docs/2.1/html/gcode.html http://linuxcnc.org/docs/html/gcode_main.html
[16:43:40] <Paragon39__> Paragon39__ is now known as Paragon39
[16:58:56] <IchGuckLive> hi all here from germany
[17:01:56] <JT-Work> has there been any issues with Test This Axis in Stepconf in the past?
[17:03:30] <archivist> I think the problems I had were due to mouse over the button problems
[17:03:46] <JT-Work> ok thanks
[17:04:14] <archivist> or yes but never quantified what it was
[17:35:33] <SWPadnos> JT-Work, I seem to recall some people saying that things worked in the test but not in emc, or the other way around
[17:35:47] <SWPadnos> by that I mean that I think I've seen both claims
[17:36:16] <IchGuckLive> are you talking on 10.04 RTI
[17:36:35] <JT-Work> I'm not sure which version he has
[17:36:55] <JT-Work> http://www.linuxcnc.org/component/option,com_kunena/Itemid,20/func,view/catid,16/id,4878/lang,english/#4907
[17:38:11] <IchGuckLive> ok by for me its dark here
[17:40:09] <SWPadnos> estop out on pin 1 - I don't know that the test will activate/deactivate that
[17:40:17] <SWPadnos> (if it's needed for the driver)
[17:40:48] <SWPadnos> he didn't say that he had inverted the estop output (only the input)
[17:40:58] <JT-Work> I'm thinking that it only puts out step and direction
[17:41:06] <SWPadnos> if that isn't it, it may be time for him to post his .stepcond file
[17:41:12] <SWPadnos> s/d/f
[17:41:14] <SWPadnos> /
[17:42:11] <JT-Work> I bet it is the e-stop pin
[17:43:30] <SWPadnos> maybe there should be an option in the test window to turn on/off estop and/or to enable a charge pump
[17:43:47] <SWPadnos> if either is selected as an output
[17:44:04] <JT-Work> that would make sense
[17:51:31] <psha> a.y i.tel D510 board users here?
[17:51:51] <psha> sorry for dots so.e letters are faulty :)
[19:08:07] <Paragon39> What manual explains the .ini directives in detail. I have noticed that the new manuals that are released with the version 2.4 Ubuntu 10.04 release are more geared around the graphical configurators?
[19:09:04] <skunkworks> http://www.linuxcnc.org/docs/2.4/html/config_ini_config.html
[19:09:20] <Paragon39> Thanks skunkworks!
[19:11:24] <skunkworks> cradek: btw - so far the 'only shift when the S word changes seems to work well. So - as you move the spindle overrid up and down - the gears don't shift/
[19:12:26] <skunkworks> it is odd though pressing the +/- spindle buttons in axis (which changes the s word) because each time you press the buttons more likely than not the gear will change. :)
[19:15:25] <skunkworks> I do have it setup so if you do change the s word while the spindle override is 0 - it will shift to the correct gear when the overide goes above 0.
[19:30:55] <Jymmm> SOB... drill press table has been unsquare with quill all this time!!! ARGH
[19:57:31] <cradek> skunkworks: sounds perfect. how long for the locking rotary?
[19:58:08] <skunkworks> dad is working on mounting the b axis servo. we are using a floor scrubber motor. (low rpm - high torque)
[19:58:25] <cradek> cool
[19:58:43] <cradek> is the unlock hydraulic?
[20:01:07] <skunkworks> yes
[20:01:18] <skunkworks> it actually lifts the table off of the coupling/
[20:01:41] <skunkworks> so it raises about 1/2 inch or so.
[20:01:47] <cradek> a switch says it's succeeded?
[20:01:51] <skunkworks> yes
[20:02:02] <skunkworks> easy peesy ;)
[20:02:04] <cradek> cool, that's what emc wants.
[20:03:01] <skunkworks> cradek: http://electronicsam.com/images/KandT/conversion/tablerotation.JPG
[20:03:13] <skunkworks> that is the 5 degree increment coupling
[20:04:58] <cradek> I don't understand what that does
[20:05:11] <cradek> it decouples?
[20:05:24] <cradek> loses encoder position?
[20:05:28] <skunkworks> heh
[20:06:15] <cradek> really, not seeing it
[20:06:20] <skunkworks> no - the table can only be set down at 5 degree increments.
[20:06:38] <cradek> ok, that I get
[20:06:54] <skunkworks> so the table lifts off the coupling (big ring that has teeth in it (not the ring gear))
[20:06:57] <cradek> but the servo stays coupled, or no?
[20:07:04] <skunkworks> servo stays coupled
[20:07:16] <cradek> ok
[20:07:50] <cradek> good - that would be srazy otherwise
[20:07:53] <cradek> c
[20:09:10] <skunkworks> the servo rotates the table (has to rotate it an amount exactly divisable by 5 degrees) and then the hydraulics set the table back down on the coupling.
[20:09:23] <cradek> cool.
[20:09:24] <skunkworks> I have a hard time explaining it well.
[20:09:51] <cradek> so it's not really a coupling - just a lock
[20:11:24] <skunkworks> they call it a (and I have the spelling wrong I am sure) cervic coupling (and I don't know if it is meant to be the same as womans cervic or I am spelling it wrong)
[20:11:49] <cradek> weird
[20:14:35] <cradek> bbl
[20:17:33] <NTU_live> does anybody get this error when running ./usr/realtime/test-suite/user/latency/run: CANNOT INIT MASTER TASK ???
[20:18:05] <NTU_live> and is that normal? also if i replace user with kern, i get all blanks. no 0s, or anything
[20:18:22] <andypugh> Normally you would run the /kernel/latency test
[20:18:38] <NTU_live> that doesnt work. it worked once but thats it
[20:18:55] <andypugh> What do you get if you just type "latency-test" in the terminal?
[20:19:01] <NTU_live> all 0s
[20:19:18] <NTU_live> pastebinning example. 1 sec
[20:20:10] <andypugh> You can't pastebin the output of latency-test
[20:20:17] <NTU_live> i know
[20:20:23] <NTU_live> im not. :)
[20:20:24] <psha> ascii-art!
[20:20:29] <NTU_live> hahah!!
[20:21:49] <NTU_live> http://pastebin.com/5PQ2Sc4g
[20:24:59] <NTU_live> im just popping back in this channel to re-ask something. why does latency-test show all 0s on my dads motherboard but works on my own board? both AMD 785 chipset.
[20:26:48] <NTU_live> oh and EMC doesnt work. USRMOT: ERROR: command timeout USRMOT: ERROR: command timeout USRMOT: ERROR: command timeout /usr/bin/emc: line 358: 4993 Killed $EMCTASK -ini "$INIFILE"
[20:27:01] <psha> dmesg?
[20:27:22] <NTU_live> 1 sec :)
[20:28:57] <andypugh> Whats the kernel version on the working machine and the non-working machine?
[20:29:07] <NTU_live> same version
[20:29:22] <NTU_live> btw same problem on ubuntu live cd
[20:29:33] <NTU_live> RTAI kernel patch and userspace from cvs magma tree
[20:29:40] <andypugh> Line 45 onwards looks like a non-rtai kernel
[20:30:02] <andypugh> (Or something else which badly breaks rtai)
[20:30:15] <NTU_live> http://pastebin.com/JNXSAJ96
[20:30:35] <NTU_live> it is RTAI kernel
[20:30:43] <NTU_live> because it all works on my board
[20:30:49] <NTU_live> this is mobo specific i think
[20:31:10] <andypugh> Could be a BIOS setting
[20:32:11] <psha> too cryptic for me :(
[20:32:30] <NTU_live> oh also, acpi=off causes hard lock on system when loading rtai modules
[20:32:37] <psha> i'll better go to bed... :)
[20:32:44] <NTU_live> good night! ;D
[20:32:47] <psha> thanks
[20:33:49] <andypugh> Well, I can tell you that RTAI is loading then quitting without botherig to say why/
[20:34:02] <andypugh> I guess "halrun" fails too?
[20:34:54] <NTU_live> halrun -V says some stuff
[20:35:13] <NTU_live> what do you want me to run in halrun?
[20:35:39] <andypugh> just type halrun all by itself
[20:35:49] <NTU_live> halcmd:
[20:35:54] <NTU_live> it wants something
[20:36:01] <andypugh> That sounds happy enoigh
[20:36:10] <NTU_live> oh ok
[20:36:22] <andypugh> loadrt threads period1=1000000 name1=slow
[20:36:46] <andypugh> show threads
[20:36:56] <andypugh> (possibly "show thread")
[20:38:28] <NTU_live> ah
[20:38:43] <NTU_live> http://pastebin.com/zpvs1Uqs
[20:39:11] <andypugh> start
[20:39:14] <andypugh> show thread
[20:39:54] <NTU_live> i type in start, show doesnt say anything new
[20:41:30] <NTU_live> http://pastebin.com/A7nfGMgA
[20:41:32] <andypugh> Ah, yes. Because the thread is empty
[20:41:38] <andypugh> loadrt sum2
[20:41:54] <andypugh> addf sum2.0 slow
[20:41:57] <andypugh> show thread
[20:43:38] <NTU_live> http://pastebin.com/TPBYxvfa
[20:44:14] <andypugh> Odd.
[20:44:22] <NTU_live> whats supposed to happen?
[20:44:24] <andypugh> try start again?
[20:44:35] <andypugh> Period FP Name ( Time, Max-Time )
[20:44:36] <andypugh> 999841 YES slow ( 100, 1010 )
[20:44:36] <andypugh> 1 sum2.0
[20:44:46] <NTU_live> still 0s
[20:44:50] <andypugh> thread running, 100nS
[20:45:16] <andypugh> sorry, no ideas
[20:45:41] <andypugh> Is dmesg saying anything now?
[20:45:58] <NTU_live> nothing special
[20:46:19] <NTU_live> im going try running halrun -V loadrt <all RT mods here>
[20:46:31] <NTU_live> hopefully something shows up
[20:47:08] <andypugh> you can use "source filename.hal"
[20:49:12] <NTU_live> http://pastebin.com/GWCgas5T
[20:49:23] <NTU_live> EMCMOT not found.. normal?
[20:51:14] <andypugh> I don't know for sure that halrun can pull values out of INI files.
[20:51:49] <NTU_live> what do you recommend for a hal file?
[20:52:17] <andypugh> Depends what you are hoping to do.
[20:52:53] <andypugh> I don't think anyhting is going to actually work.
[20:53:05] <NTU_live> oh ok
[20:53:29] <andypugh> If you can't make a thread run with a sum2 in it, then loading a more complex HAL file is just going to fail in more ways
[20:54:59] <andypugh> I would be tempted to look very hard at the BIOS settings of the machine that does work in comparison to the one that doesn't. (You did say they were the same CPU / mobo combination?)
[20:55:14] <NTU_live> very similar
[20:55:41] <NTU_live> PHENOM II X4 955 vs PHENOM II X3 720 BE @ 3.2 Ghz + 1 unlocked core
[20:55:44] <NTU_live> same speed
[20:56:15] <andypugh> If one works with a livecd boot and one doesn't (identical kernels and software and settings) then it has to be hardware or BIOS settings I would have thought.
[20:56:37] <NTU_live> what is USRMOT? any idea?
[20:56:53] <NTU_live> http://pastebin.com/A0yjtrhg
[20:59:14] <ries_> ries_ is now known as ries
[20:59:43] <andypugh> if you put "DISPLAY = usrmot" in the [DISPLAY] section you get a command-line interface to emc2.
[20:59:58] <andypugh> (which I learned approximately 30 seconds ago)
[21:02:28] <skunkworks> the errors that come in axis in the right lower corner get stored anywhere?
[21:02:33] <skunkworks> like a history?
[21:03:28] <andypugh> I think they also go to stdio, but you only get to see that if you start emc from the command line
[21:05:32] <NTU_live> so emc uses printf?
[21:05:43] <NTU_live> i dont really look at emc source..
[21:05:52] <andypugh> It uses rtapi_snprintf I think.
[21:07:55] <NTU_live> i replaced tkemc with usrmot in servo_sim.ini and get http://pastebin.com/VbctS75k
[21:09:17] <andypugh> so it seems that usrmot and USRMOT are different things?
[21:09:47] <NTU_live> i dont know, tbh i just want this to work :P
[21:11:02] <andypugh> That error message only appears in the usrmotinf.cc file.
[21:11:39] <andypugh> You are using the standard servo_sim.ini file?
[21:11:44] <NTU_live> #usrmot: can't find emcmot ini file emc.ini can't load ini file emc.ini
[21:12:24] <NTU_live> andypugh, yes
[21:13:13] <NTU_live> is there a way i can get real back trace or since emc is #!/bin/bash you cant get any?
[21:13:39] <NTU_live> id love to run latency-test via gdb but.. :(
[21:13:50] <andypugh> the servo_sim works for me, so I can't help
[21:26:11] <NTU_live> oh cool it randomly worked this time
[21:26:28] <andypugh> That's only semi-cool
[21:26:38] <NTU_live> http://pastebin.com/gGnmVS80
[21:27:23] <andypugh> That's pretty shocking latency.
[21:27:31] <NTU_live> heh
[21:29:19] <paragon_ws> Chaps, I just wanted to ask what I could expect from the 7i43 driving msd542 stepper drives which have the capability to step x128 (25600) per rev?
[21:29:26] <NTU_live> i suppose i could forward this to the rtai mailing list
[21:29:42] <NTU_live> and ask why not even the built in latency-test reports /dev/null
[21:30:03] <andypugh> http://pastebin.com/ZDw6YjsQ
[21:30:27] <NTU_live> its fake output
[21:30:41] <NTU_live> this mobo cant have that bad of latency
[21:31:30] <andypugh> paragon_ws: I should drive them OK, but it is very debatable whether there is any point in that level of microstepping
[21:31:58] <andypugh> I seem to recall that very high microstepping multipliers lead to wierdness
[21:33:27] <paragon_ws> I was originally running them at 400 steps/rev direct from the paraport but was seeing following errors when attempting to up the speed. Also the motors seemed to run rough. What would be a good u step to go with?
[21:34:52] <paragon_ws> I think I was also running EMC V2.2 or something (from vague memory).
[21:37:55] <NTU_live> im off to make a 64-bit EMC + RTAI live cd now :D hope it works! i dont expect anything new to change though...
[21:39:34] <andypugh> paragon_ws: Following errors with p-port stepgen are a bit "fake"
[21:40:13] <andypugh> What a stepgen following error means is that the base thread is not fast enough to keep up with the step rate you are demanding.
[21:41:30] <andypugh> I would experiment and see what works best. Some folk have reported better speed at 8x than 4x, and some the reverse. I don't know anyone using more than 8x though.
[21:43:16] <paragon_ws> andypugh: I remember playing around with the base thread to try and improve it but even then I think I maxed it so I purchasing a pluto-p but this had issues so have now purchased the 7i43. Hence the enquiry with regards to the 7i43 and steps/rev.
[21:44:38] <andypugh> steps/sec is possibly more relevant, but what can the drivers manage (ie what is the step time and space time ?)
[21:46:05] <paragon_ws> The pulse input is rated from 0-300Khz according to documentation of the driver.
[21:47:19] <DaViruz> anyone know of a 3ph induction motor manufacturer that has 3d models of their motors?
[21:48:08] <paragon_ws> DaViruz: Afraid not.
[22:20:55] <NTU_live> heh it was a BIOS setting..
[22:22:10] <NTU_live> ACC (Advanced Clock Calibration) i noticed that if you have set on "All cores" you get misc issues, setting it to "auto" unlocks cores and what not and also adds a small performance boost
[22:22:42] <NTU_live> 28,000 ns though :/
[22:24:09] <NTU_live> better than nothing so im happy :)
[23:29:10] <MattyMatt> DaViruz, would you be more likely to purchase their motors if they did?
[23:30:05] <skunkworks> cradek: it is a curvic coupling
[23:30:30] <skunkworks> supposed to be accurate to +/-10 arc sec
[23:31:12] <DaViruz> MattyMatt: no, i have already selected a motor
[23:31:48] <DaViruz> well, for this application at least. in the future, sure
[23:31:56] <MattyMatt> I've put a nema17 and 23 in blender if you want them :)
[23:32:14] <DaViruz> i tend to design with linear components now because i like their spice program so much
[23:32:17] <DaViruz> :)
[23:32:20] <DaViruz> (electornics)
[23:33:32] <MattyMatt> their program, or just their component database?
[23:34:01] <DaViruz> MattyMatt: both
[23:34:38] <DaViruz> (their program has a fairly complete library of their devices)
[23:35:18] <MattyMatt> I want something more generic for mechanisms
[23:35:48] <MattyMatt> generic yet with specialised wizards, like ordering 8020 by part number etc
[23:37:09] <DaViruz> speaking of things i like, i'm kind of getting fond of the siemens logo! pseudo-plc's
[23:40:02] <DaViruz> now if i can just find a small vfd that can push DC through the windings when the motor is stationary to keep it above ambient i will be happy
[23:40:05] <pcncx> check this out: http://img87.imageshack.us/img87/2250/wedoct27bootchart.png
[23:40:15] <MattyMatt> which logo is that? on their corporate site they have the traditional one
[23:40:46] <pcncx> its the instant on logo
[23:40:59] <DaViruz> MattyMatt: http://www.automation.siemens.com/mcms/programmable-logic-controller/en/logic-module-logo/Pages/Default.aspx
[23:41:10] <pcncx> this machine boots up so fast it is almost laughable
[23:42:17] <MattyMatt> pcncx for years I dreamed of sth like that to deconstruct the Windows boot sequence
[23:42:18] <DaViruz> i think mine can give it a run for it's money
[23:42:42] <DaViruz> the windows boot sequence is a mystery to me
[23:42:42] <pcncx> MattyMatt aptitude install bootchart
[23:43:26] <pcncx> MattyMatt for years I've been trying to figure out why anyone would want to run Windows
[23:43:30] <MattyMatt> pcncx I never reboot my linux boxes :)
[23:43:46] <pcncx> MattyMatt wiyh a 16 second boot I shut this box off
[23:43:53] <MattyMatt> I do turn off my emc2 machine at night to save power
[23:43:58] <DaViruz> 16 second from the button?
[23:44:00] <DaViruz> +s
[23:44:06] <pcncx> well after POST
[23:44:12] <pcncx> app can't run then
[23:44:28] <MattyMatt> coreboot :)
[23:44:32] <DaViruz> didn't look at the time thingy there
[23:44:53] <MattyMatt> oh wow, I thought that was just the first 15s :)
[23:44:58] <pcncx> still I can hit the button and by the time I'm sat down its ready
[23:45:02] <DaViruz> when i tested windows xp on my ssd it was actually crazy fast
[23:45:07] <DaViruz> (boot-wise)
[23:45:12] <DaViruz> win7 is a fair bit slower
[23:45:31] <MattyMatt> yeah v tempted by 64GB ssd for T30
[23:45:58] <pcncx> this machine is such an old POS I'm sure a modern machine would be a lot faster
[23:46:14] <pcncx> this is only a Pentium 3
[23:46:23] <DaViruz> i need to time my new emc2 box. quickboot bios and i've optimized the OS boot like crazy
[23:46:37] <MattyMatt> hard drive speed and memory size are more important than 1Ghz or 2
[23:46:51] <pcncx> DaViruz that is what I am in the process of doing now optimizing the OS
[23:47:11] <DaViruz> i'm running crux on it, so it's fairly fast by default
[23:47:28] <DaViruz> the ubunto boot is, in a word, awful :/
[23:47:29] <pcncx> MattyMatt I'm in X now with my browser running and I'm using 83 MB RAM
[23:48:00] <pcncx> up just went to 84
[23:48:19] <DaViruz> pcncx: where are you after those 16 seconds? X? terminal login?
[23:48:23] <pcncx> DaViruz this is a fairly minimal Debian Lenny
[23:48:34] <MattyMatt> bleh. I have 4GB and Iceweasel+flash amd64 has issues
[23:48:38] <pcncx> DaViruz just console login X comes up fast though
[23:48:53] <pcncx> blink of an eye I run fluxbox
[23:49:39] <pcncx> this box 512 MB total Linux says 511 MB usable lousy AMI BIOS
[23:49:47] <DaViruz> often graphic drivers slow things down
[23:50:00] <pcncx> DaViruz oddly I am running no gfx drivers
[23:50:15] <pcncx> because EMC had issues with the r128 kernel driver
[23:50:21] <DaViruz> i see, vesa mode?
[23:50:28] <pcncx> I guess it is frame buffer?
[23:50:32] <pcncx> I'm not sure
[23:50:35] <MattyMatt> I need to fix my latency errors and check my base thread timing. I'm getting a 2 second lag on jogging with keys atm, with a complex gcode file
[23:50:41] <pcncx> honestly I don't know how it is running
[23:50:51] <DaViruz> framebuffer should be fast, since it's already online by the kernel
[23:50:57] <pcncx> this build i haven't gotten that far yet that i can check latency
[23:51:06] <pcncx> I'm not totally done tweaking the kernel yet
[23:51:50] <pcncx> the build before i was spiking to 18,000 ns occasionally I am hoping to better that
[23:52:25] <pcncx> but it was running around 6,000 ns a lot just not enough to call it 6,000 ns
[23:52:50] <pcncx> you got to call it by the highest spike you get
[23:53:15] <pcncx> I'd be thrilled ig i could get double 6,000 ns out of this machine
[23:53:28] <pcncx> 12K ns latency would rock my socks
[23:54:46] <MattyMatt> was it you linked that LW article on RTAI? that was interesting
[23:55:09] <pcncx> MattyMatt might have been
[23:55:25] <pcncx> this is my hmm link for today: http://www.tux.org/~mayer/linux/bmark.html
[23:57:15] <pcncx> well that and this: http://cateee.net/autokernconf/
[23:57:53] <pcncx> my big thing lately has been grepping and diffing kernel config files
[23:58:02] <MattyMatt> yeesh. the thing that stands out in the nbench list is the effect icc has on xeons
[23:58:42] <pcncx> MattyMatt I would like to find a better benchmark program but it doesn't seem to be something people are very interested in today
[23:59:01] <pcncx> MattyMatt I think today if people want their computer faster they just go out and buy a faster computer
[23:59:19] <MattyMatt> there's the SpecMark that THG uses