#emc | Logs for 2007-04-24

[00:00:41] <petev> jmkasunich, have you tried these guys?
[00:00:41] <petev> http://www.directron.com/bx80557e6600.html
[00:00:41] <petev> they have been good for me in the past
[00:01:13] <jmkasunich> ordered it from http://www.antaresdigital.com/customer/product.php?productid=17250&cat=419&page=1
[00:01:20] <jmkasunich> it shipped this afternoon
[00:01:33] <jmkasunich> I'll keep that link for future use tho, thanks
[00:04:46] <jmkasunich> hmm, they have nice ram prices - I could have gotten everything in one place I bet
[00:19:45] <lerman_> lerman_ is now known as lerman
[01:22:14] <skunkworks> and the mach people show up. http://www.cnczone.com/forums/showthread.php?p=289404#post289404
[01:22:35] <skunkworks> this is the ncpod http://ncpod.oemtech.com/
[01:24:20] <eric_u> is the ncpod supposed to work with emc?
[01:25:54] <skunkworks> No
[01:25:56] <skunkworks> it is usb
[01:27:23] <skunkworks> says it includes the emc trojectoy planner.
[01:27:41] <skunkworks> Trajectory
[01:32:06] <eric_u> I see
[01:32:46] <skunkworks> think it is used as a pre-proccessor - then that info is sent to the pod
[01:35:21] <eric_u> that post has been there a while
[01:37:19] <eric_u> I like the idea that kiwichris had to keep the tape drive on your k&t
[01:38:45] <skunkworks> :)
[01:38:56] <skunkworks> I don't think so.
[01:39:20] <eric_u> I think the ncpod runs the emc code natively, they have done that on the grex
[01:39:43] <eric_u> the rest of it is pretty scary though
[01:39:54] <eric_u> you programmed it using number wheels?
[01:40:42] <skunkworks> No - it used the tape drive to send the programs to it. We hacked into it to make a vitual tape drive with a 386 computer.
[01:40:52] <eric_u> were the tape drive motors big enough to drive a small mill?
[01:41:11] <skunkworks> the number wheels where for mdi input and offsets.
[01:41:33] <skunkworks> I don't think they where anything good.. I would have remembered ;)
[01:42:25] <eric_u> what was the hydraulic unit for?
[01:43:38] <eric_u> woah: http://www.electronicsam.com/images/KandT/conversion/rearcontrller.JPG
[01:44:47] <skunkworks> that is what used to make it go - until about a year ago.
[01:45:11] <eric_u> that thing has wiring in the most unheard-of places
[01:45:22] <skunkworks> The machine had hydaulic servos. - also lots of things like counter balance and tons of cylinders.
[01:46:00] <eric_u> like the moog mills
[01:46:22] <skunkworks> (used to have a hydraulic spindle also - but we replaced it with a 5hp vfd)
[01:47:13] <eric_u> pretty ambitious to retrofit a hydraulic powered mill
[01:47:37] <eric_u> with that toolchanger, it looks like tool storage shouldn't be a problem
[01:49:47] <skunkworks> :) 60 tool change
[01:49:51] <skunkworks> changer
[01:50:05] <eric_u> that one tool on there looks very lonely
[01:52:25] <eric_u> learned something on cnczone, I was blaming my ISP for poor performance browsing, it's the ipv6 setting though
[01:53:13] <skunkworks> I heard. I have never touched it in dapper.
[01:53:23] <renesis_> heh
[01:53:34] <renesis_> i deleted all the ipv6 stuff
[01:53:36] <eric_u> it's probably set wrong though
[01:53:49] <renesis_> from hosts at least
[01:54:10] <eric_u> look in about:config in firefox
[01:55:11] <eric_u> network.dns.disableIPv6 should be true
[01:55:23] <eric_u> probably
[02:48:57] <Jymmm> Well, running RT test alone for the last hoaur and still going, this is without running glxgears and a APG video card w/ 128mb video ram
[02:49:08] <Jymmm> But running them together DID crash X
[02:49:30] <Jymmm> I'll try glxgears solo in about 20 minutes
[02:53:26] <skunkworks> jymmm: http://www.electronicsam.com/images/pent3shot.png
[02:53:54] <Jymmm> how did you get the fps at the bottom?
[02:54:39] <skunkworks> /usr/lib/xscreensaver/gears -window -fps
[02:54:47] <skunkworks> cradek figured it out.
[02:54:48] <skunkworks> pretty cool
[02:55:02] <Jymmm> hmmmmm
[02:55:32] <Jymmm> that's not the same as glxgears though, is it?
[02:56:13] <skunkworks> cradek: seemed to think it was ok. :)
[02:56:46] <cradek> glxgears doesn't seem to draw right (it jerks around), every other gl program does
[02:56:51] <Jymmm> But, here X is crashing on rt test and glxgears, not screensaver
[02:57:18] <cradek> what's in the X log when it crashes? a sig 11?
[02:57:20] <Jymmm> three different boxes and now two different video cards in box #1
[02:57:34] <Jymmm> cradek iirc 15
[02:57:38] <Jymmm> 15 or 11
[02:58:28] <cradek> 15 would be very interesting to say the least
[02:58:53] <Jymmm> I had both, but I think I fixed that one.
[03:02:35] <Jymmm> ok, trying gear screensaver now. Will let run for 40 minutes
[03:03:17] <Jymmm> how do you get the screensaver to dispaly fps
[03:03:23] <Jymmm> automagically
[03:03:32] <skunkworks> jymmm: http://www.electronicsam.com/images/pentIII2.JPG
[03:03:46] <jmkasunich> /usr/lib/xscreensaver/gears -window -fps
[03:03:50] <skunkworks> jymmm -fps
[03:03:52] <jmkasunich> see the -fps at the end?
[03:04:07] <Jymmm> oh so not a setting in the menu
[03:04:45] <jmkasunich> I had a "really cool, sort of" find in the dumpster today
[03:04:54] <Jymmm> skunkworks dual cpu?
[03:05:10] <jmkasunich> really nice 16 AWG 6 conductor cables with heavy duty connectors
[03:05:15] <jmkasunich> great for connecting steppers, etc
[03:05:38] <jmkasunich> about 50 mated pairs of connectors, probably cost a couple grand originally
[03:05:56] <jmkasunich> the "sort of" part is that each connector has about 2 feet of cable on it
[03:06:13] <jmkasunich> so unless your runs are 4 feet or less.... not so usefull
[03:06:27] <eric_u> what kind of connector?
[03:06:45] <jmkasunich> circular, 6 pins, gold, screw lock, molded onto the cable
[03:07:00] <jmkasunich> sort of like mil style circular connectors, but mostly rubber instead of metal
[03:07:09] <jmkasunich> totally waterproof I think
[03:07:20] <eric_u> bayonet style?
[03:07:26] <jmkasunich> thread
[03:07:35] <Jymmm> oh those... I gat zillions of 20ft cables with those connectors on the.
[03:07:37] <Jymmm> them
[03:07:46] <jmkasunich> 16AWG?
[03:07:48] <eric_u> so they don't have separate pins?
[03:08:21] <jmkasunich> separate as in removable?
[03:08:43] <eric_u> right
[03:08:54] <jmkasunich> nope - everything is overmolded with rubber
[03:09:03] <eric_u> no fun
[03:09:14] <jmkasunich> I cut one apart, trying to see if it other wire could be attached
[03:09:35] <jmkasunich> they're not just molded - they're impregnated - every nook and cranny
[03:09:39] <eric_u> I think those are commercial grade parts
[03:09:51] <Jymmm> mil grade
[03:10:10] <eric_u> the ones that you mold the backshell aren't mil are they?
[03:10:21] <eric_u> digikey used to sell something like that
[03:10:46] <eric_u> I'm assuming they are 97 series, I'm probably wrong
[03:11:36] <eric_u> I have some cables for my AB servomotors that have molded backshells
[03:11:36] <Jymmm> http://www.ttabconnectors.com/images/products/ukan-connector.jpg
[03:11:45] <Jymmm> toop right
[03:12:21] <eric_u> looks like a standard 97 series jack
[03:12:29] <Jymmm> http://www.parvus.com/uploads/ProductImages/113.jpg
[03:12:41] <Jymmm> but green =)
[03:13:32] <eric_u> forgot about the square threaded ones
[03:13:42] <eric_u> that's not 97 series
[03:14:21] <eric_u> I was going to buy some of usb connectors like the one on the lower left
[03:14:43] <eric_u> even though it wasn't my money, I couldn't justify spending $100+ for a connector
[03:15:12] <eric_u> ethernet shown
[03:16:02] <eric_u> Air Force doesn't use the green connectors any more
[03:17:01] <eric_u> one of my regrets from the Air Force is that I didn't have one of the electricians show me how to remove the pins from the connectors like the ones on the parvus. com site
[03:17:17] <jmkasunich> something like this: http://www.ab.com/sensors/products/connection_systems/TPE_Cable_CatPgs.pdf
[03:17:48] <jmkasunich> except 6 conductors and 16AWG
[03:18:05] <jmkasunich> very high flex
[03:19:17] <jmkasunich> I have 50+ pieces of male and female straight, each with about 2' of cable, and 10-12 sets of 1 right-angle + 1 straight, with maybe 15" of cable
[03:20:24] <eric_u> somewhere in your company, they have a batch of wire that is eventually going to go into the dumpster
[03:20:34] <jmkasunich> I'll probably use some on my machine (where the length allows) and bring the rest to the CNC workshop for the swap table
[03:20:45] <eric_u> too bad they didn't put it there while it was still connected to the connectors
[03:21:14] <jmkasunich> the ends of the cables have been stripped - they were probably attached to terminal blocks
[03:21:42] <eric_u> ok, the rest of the cable went in the dumpster before now
[03:21:47] <jmkasunich> I bet these started out as assemblies and were cut off and stripped for some test
[03:21:57] <jmkasunich> they may not have been all that long in the first place
[03:22:16] <jmkasunich> from that pdf, they sell them in 2, 5, and 10 meter lengths
[03:22:29] <jmkasunich> I bet theses were 2m ones, and they cut them in half
[03:22:36] <jmkasunich> maybe to test the connectors?
[03:22:54] <jmkasunich> found them in a box that looked like it came from a lab cleanup
[03:22:57] <eric_u> don't they sell them as pigtails?
[03:23:02] <jmkasunich> dunno
[03:23:16] <jmkasunich> we don't make anything in our building, its all development and testing
[03:23:23] <eric_u> I recognize those connectors from industrial sensors
[03:24:50] <jmkasunich> the ones in the PDF do come in pigtale as well as connector both ends
[03:25:02] <tomp2> nice cables, oil resistant, rt angle, yeah, harsh enviro sensor cnxrs
[03:25:06] <eric_u> trying to remember if I have any in the basement
[03:25:19] <jmkasunich> so far I haven't found any docs that match the exact things I have
[03:25:35] <jmkasunich> everything is 18 or 22 ga, and most are 4 cond or less
[03:26:15] <eric_u> digikey catalog printing is too small for my eyes, woulda thought they would have those
[03:26:28] <jmkasunich> might
[03:26:41] <jmkasunich> probably $10-$20 each
[03:27:15] <jmkasunich> at least
[03:30:22] <tomp2> i just wired up 8 of these 'laser' distance sensors, used to see if tools were in pods. they have a digital output ( yes there's something out there) and also an analog output ( how far away ) Yamatake/Azbil http://www.yamatakeusa.com/LaserSensors.asp. but bare ends, no nice connectors :(
[03:31:37] <skunkworks> jymmm: yes - dual cpu
[03:33:37] <Jymmm> skunkworks did you tweak the Rt kernel to get emc going ?
[03:34:13] <skunkworks> No - cradek did...
[03:34:54] <Jymmm> huh?
[03:35:02] <Jymmm> he shipped you the mobo?
[03:35:42] <skunkworks> No - he made up a smp kernel. smp? (multi processor rt kernel)
[03:36:10] <Jymmm> is it part of the livecd or something?
[03:36:25] <skunkworks> No - not yet. do you have something to test with?
[03:36:32] <eric_u> automation direct sells a cross to those connectors
[03:36:44] <skunkworks> jymmm - do you want to test it?
[03:36:54] <Jymmm> oh about 4 different mobo's.... Dual P3 1Ghz with 2gb ram
[03:37:13] <Jymmm> skunkworks not tonight
[03:37:46] <skunkworks> mine is a dual 600mhz pent III
[03:37:55] <skunkworks> that screen shot is from that.
[03:38:16] <skunkworks> 512 megs
[03:39:24] <eric_u> but they only seem to go up to 5 connectors
[03:39:34] <eric_u> 5 pins
[03:40:22] <jmkasunich> eric_u: I don't have one in my hand at the moment, so I can't be sure, but I think the ones I have are same style but a bit larger
[03:40:34] <skunkworks> it was pretty easy - install 3 deb's and then grab trunk and build it.
[03:40:44] <jmkasunich> I didn't bring them home yet, just piled them in a box and left it in the lab
[03:40:57] <jmkasunich> when I accumulate a good pile, I load up the truck ;-)
[03:41:28] <skunkworks> under your desk? ;)
[03:41:46] <eric_u> there is some indication that they are m12 connectors?
[03:41:47] <jmkasunich> I have a pile building there too
[03:41:53] <jmkasunich> m12?
[03:42:03] <eric_u> metric standard
[03:42:23] <eric_u> I have a batch of motors with m23 connectors
[03:42:52] <eric_u> there have to be sizes in-between
[03:43:02] <jmkasunich> these are similar to http://www.ab.com/sensors/products/connection_systems/TPE_Cable_CatPgs.pdf
[03:43:20] <jmkasunich> but the pdf says cable OD 0.28 inches - the ones I have are closer to a half-inch OD
[03:43:28] <jmkasunich> connector nut is maybe 3/4"
[03:44:10] <eric_u> don't find the cable size too interesting, but the shoulder on the connector would be good to know
[03:44:33] <jmkasunich> maybe I'll bring one home tomorrow
[03:44:37] <jmkasunich> can post pics and dims
[03:45:31] <jmkasunich> ah, so thats why they are so heavily overmolded: "NEMA 6P, IP 67, 1200psi (8270kPa) washdown"
[03:45:38] <jmkasunich> 1200 psi washdown!
[03:46:15] <eric_u> http://www.phoenixcontact.com/global/news/222_6865.htm
[03:46:51] <eric_u> phoenix always does things differently
[03:47:57] <jmkasunich> push in, eh?
[03:48:10] <skunkworks> 'explosion proof' ;)
[03:48:38] <jmkasunich> explosion proof is easy for connectors, very little volume inside
[03:49:01] <jmkasunich> gets a lot harder as the size increases
[03:50:46] <jmkasunich> IIRC the test for explosion proof involves an explosive gas/air mix inside and out - ignite the mix inside, and the device must not only not come apart, it must not allow escape of enough flame or sparks to ingite the mix outside
[03:51:14] <ds3> can it melt? ;)
[03:51:31] <jmkasunich> only if it keeps the flame inside
[03:52:01] <ds3> that sounds aweful like the NEBS tests
[03:52:32] <eric_u> http://www.bannerengineering.com/products/sublit.php?modelFam=CABLES_PE&sub_id=6
[03:53:48] <jmkasunich> NEBS?
[03:53:55] <eric_u> if you download the banner engineering accessories catalog at that link, they have connectors starting at about page 31
[03:54:15] <eric_u> looks like there is m8, m12 and m16 connectors
[03:54:24] <eric_u> sounds like you might have some m16
[03:54:24] <ds3> NEBS is a telco standard for equipment in a CO
[03:55:28] <ds3> one NEBS test involve filling the inside of the unit with a flamable gas while it is running; it must not eject enough flames to set adjacent equipment on fire and must self extinguish after the gas is removed
[03:55:58] <eric_u> you got nothing on the Air Force
[03:56:02] <jmkasunich> http://www.reliance.com/prodserv/motgen/acmain_explosion.htm
[03:56:17] <Jymmm> O, the system does NOT crash on screensaver gears, but it DOES on glxgears and RT test from the livecd; could someone try that instead?
[03:56:53] <eric_u> there was one test for fuel bearing components we did where the item had to withstand a 2000 degree f flame for 5 minutes.
[03:57:03] <jmkasunich> I ran glxgears and rtai latency test for about 20 mins, on two systems
[03:57:31] <jmkasunich> this one: uniprocessor sempron, "standard" EMC RT kernel
[03:57:45] <jmkasunich> and kayak: 2xP3-666, cradek's new RT SMP kernel
[03:58:16] <Jymmm> jmkasunich crashes after 30 minutes
[03:58:17] <eric_u> the rep from one of the companies started going "come on baby, you can make it" at about the 4:30 point, but it burst into flames with about 10 seconds left
[03:58:29] <jmkasunich> ouch
[03:58:45] <eric_u> they had to go back to the funky expanding paint
[03:59:16] <eric_u> what version kernel is he using?
[03:59:32] <Jymmm> eric_u from the livecd
[03:59:33] <jmkasunich> which he?
[03:59:42] <eric_u> cradek
[04:00:11] <jmk-kayak> Linux kayak 2.6.17-rtai3.5 #2 SMP Fri Apr 20 22:15:13 CDT 2007 i686 GNU/Linux
[04:01:14] <Jymmm> bbiab food
[04:01:33] <jmkasunich> bbt sleep
[04:01:43] <eric_u> g'night
[04:42:49] <Skullworks_PGAB> * Skullworks_PGAB has gone lights out.
[05:45:33] <Jymmm> Ok, scrnsvr gears jsut crashed X after a couple hours.
[05:53:33] <renesis_> pcb-gcode for eagle is neato
[06:10:52] <chr0n1c> i have been wanting to mill my first pcb.. but i need to find a neato project .. ;)
[06:11:19] <chr0n1c> i just ordered some skate bearings from ebay for a proto type milling (slow speed) for my mini mill
[06:12:16] <Jymmm> chr0n1c do you have a plan already, or going to play it by ear
[06:22:10] <chr0n1c> play it by ear
[06:22:24] <chr0n1c> i got 8 bearings so i can screw a few up i guess
[06:22:37] <chr0n1c> and make extras for sale possibly
[06:22:44] <Jymmm> make what?
[06:22:57] <chr0n1c> anyone interested? extra milling spindles
[06:23:22] <chr0n1c> *made with 8mm pro skate bearings
[06:23:38] <Jymmm> maybe, depends on how they turn out, but more or less I'll just steal your idea as I have about 50 bearings of my own =)
[06:24:10] <chr0n1c> oh my
[06:24:26] <Jymmm> I may be a thief, but I'm an honest one
[06:24:32] <chr0n1c> i almost bought 100 of them... i'll find a use for em.. but i went with 8 for now
[06:24:57] <Jymmm> actually, I think I did buy 100 of them ABEC7 rated for 30K rpm iirc
[06:25:18] <chr0n1c> i am thinking two gears, a dc motor... some drill rod... aluminum frame lots of mounting options
[06:25:29] <chr0n1c> vairable speed controller
[06:25:35] <chr0n1c> maybe a speed encoder
[06:25:57] <Jymmm> Well, right off, I'd just be interested in the mechanical part atm
[06:25:58] <chr0n1c> i am gonna shoot around 200 - 1000 rpms
[06:26:16] <Jymmm> lots of torque?
[06:26:23] <chr0n1c> of course
[06:26:28] <anonimasu> ouch
[06:26:28] <chr0n1c> it's for steel ;)
[06:26:47] <chr0n1c> i won't be taking 1/2 cuts or anything like that
[06:26:51] <chr0n1c> with it*
[06:27:05] <chr0n1c> i have a few different motors
[06:27:24] <chr0n1c> a couple greaboxes from a cordless drill ... like planetary gears and stuff
[06:27:39] <chr0n1c> for the prototypes..
[06:28:27] <chr0n1c> i've looked through about 100 different spindle designs on the hobby sites
[06:28:36] <chr0n1c> i should be able to come up with something
[06:29:45] <Jymmm> But, I won't be doing anything if I can't get X to stop crashing.
[06:29:55] <chr0n1c> :o
[06:29:59] <chr0n1c> oh noes
[06:30:51] <chr0n1c> i just had a great idea.. for testing code on a machine.. use a qtip
[06:31:00] <chr0n1c> for the cutter*
[06:31:15] <chr0n1c> x crashing eh?
[06:33:49] <Jymmm> 3 system
[06:45:29] <Jymmm> I'm not sure who to go with this now... I cna't have X crashing on me in the middle of a part
[06:46:03] <Jymmm> I even bought a 128mb nvidia baed agp vard tonight, and it still does it.
[06:46:12] <Jymmm> based
[06:53:03] <chr0n1c> thats a crazy problem
[06:53:12] <chr0n1c> maybe you are pushing it too hard...
[06:53:33] <chr0n1c> will emc push it like glx and the rt tests?
[07:02:27] <alex_joni> chr0n1c: it might
[07:02:36] <alex_joni> but you can't spec a machine for half the load
[07:02:37] <Jymmm> I honestly don't know, but it's just a test to determine stability
[07:02:48] <alex_joni> and it then creeps when you start printing something while cutting
[07:03:12] <alex_joni> Jymmm: any errors on the X crashing?
[07:03:22] <alex_joni> in /var/log/..
[07:03:22] <Jymmm> alex_joni nope
[07:03:31] <Jymmm> signl 11 or 15
[07:03:32] <alex_joni> Jymmm: tried memtest?
[07:03:33] <Jymmm> that's it
[07:03:44] <Jymmm> yeah, and on there different boxes.
[07:04:25] <Jymmm> three
[07:05:03] <Jymmm> I eve md5sum /dev/cdrom on the 2nd one I burned. The first one failed the md5sum even thought the boot menu "check cd" passed
[07:05:29] <alex_joni> any thing common between the three boxes?
[07:05:42] <alex_joni> like moving memory or HDDs ?
[07:05:44] <Jymmm> I am really running out of ideas. Just the LiveCD (and electricity)
[07:06:03] <Jymmm> no, each system has it's own stuff in it, no swapping at all.
[07:06:09] <alex_joni> did you try the stock ubuntu?
[07:06:23] <Jymmm> no, but I just got done burning debian 4 cd
[07:06:36] <alex_joni> http://www.bitwizard.nl/sig11/
[07:06:45] <alex_joni> look for "If it happens to be the hardware it can be:"
[07:08:41] <Jymmm> well I can let memtest run again for 3 hours.
[07:09:06] <Jymmm> let me look at the log again...
[07:09:11] <alex_joni> Jymmm: sounds like a real pita
[07:09:24] <alex_joni> and it only crashes when GL is involved?
[07:10:10] <Jymmm> so far, yes. GLXGEARS from console within 40 minutes. screensaver gear after two hours.
[07:10:52] <alex_joni> hmm.. sounds like a burnin problem
[07:10:57] <alex_joni> do you have proper cooling?
[07:11:46] <Jymmm> actually nothing in the xorg.0.log
[07:11:57] <Jymmm> yeah it doens't get hot at all
[07:12:28] <alex_joni> I tihnk it gets rewritten when you restart your box
[07:12:39] <Jymmm> hell the cpu heatsnk is cold
[07:12:54] <alex_joni> how about nothbridge, mem, etc?
[07:14:05] <chr0n1c> what about the rt baseperiod?
[07:14:12] <chr0n1c> have you tried changing that?
[07:15:06] <Jymmm> north is warm but it took me a few seconds to figure that out
[07:15:07] <chr0n1c> my box was sluggish until i changed it.. now a p 233 runs as fast as my 1.2ghz xp machine
[07:15:35] <alex_joni> chr0n1c: not running emc2 at all
[07:15:38] <Jymmm> alex_joni but I've tried three different boxes
[07:15:45] <alex_joni> Jymmm: :/
[07:16:04] <Jymmm> the ONLY thing common between them is the LiveCD
[07:16:23] <Jymmm> And it crashes X when installed too.
[07:16:32] <Jymmm> not just booting off the cd
[07:16:36] <chr0n1c> can you run some sort of debugger and see WHY it crashes?
[07:16:49] <chr0n1c> i dunno because i'mnot a programmer
[07:16:53] <Jymmm> I guess, no clue how/what though.
[07:16:54] <chr0n1c> jsut asking
[07:17:06] <Jymmm> np =)
[07:17:12] <alex_joni> Jymmm: I suggest you get an clean ubuntu live
[07:17:17] <alex_joni> and test that
[07:17:41] <chr0n1c> build a new kernel specifically for that machine with all the right options?
[07:17:43] <Jymmm> alex_joni what about debian 4 ???
[07:17:45] <chr0n1c> then retest it?
[07:18:29] <chr0n1c> it sounds like something is overheating
[07:18:38] <chr0n1c> a chip/ram/cpu or something...
[07:18:49] <alex_joni> Jymmm: you'll have to get a working RT kernel
[07:18:54] <alex_joni> if you plan to use emc2 later..
[07:19:02] <Jymmm> chr0n1c Ive tried three different computers
[07:19:17] <Jymmm> alex_joni fair enough, dl iso now....
[07:19:40] <alex_joni> Jymmm: which may turn out to be a pita.. depending on luck
[07:19:46] <Jymmm> alex_joni 6.06 or ??? 7.04 or ???
[07:19:51] <alex_joni> 6.06
[07:20:01] <alex_joni> as that's what you'll be using for emc2
[07:20:01] <Jymmm> alex_joni no 6.10 ?
[07:20:10] <alex_joni> 6.06 LTS
[07:20:29] <alex_joni> 6.10 is worse than both 6.06 and 7.04
[07:20:30] <chr0n1c> document your build process?
[07:20:41] <alex_joni> chr0n1c: it's not a matter of build process
[07:20:52] <alex_joni> it's a matter of catching the right kernel/rtai combo
[07:20:54] <Jymmm> i have to hit another machine to DL, brb
[07:20:55] <chr0n1c> for the kernel i mean
[07:21:22] <alex_joni> chr0n1c: lots of kernel versions, lots of RTAI versions
[07:21:33] <alex_joni> mixing and matching is _pure_ pleasure
[07:21:35] <chr0n1c> i know it confuses me
[07:21:49] <chr0n1c> which is why i heart the live cd
[07:22:06] <chr0n1c> it saved me hours of pulling my hair out i think
[07:23:38] <alex_joni> make that days
[07:23:58] <alex_joni> especially if it's the first time
[07:24:38] <chr0n1c> :(
[07:26:18] <Jymmm> glad i brought a stack of blanks *sigh*
[07:38:37] <chr0n1c> i found a v9 to vX post convertor for MC... lets hope it works!
[07:57:31] <Jymmm> ok, iso burned. md5sum /dev/cdrom now
[07:58:24] <chr0n1c> different burners?
[07:58:49] <Jymmm> burner drive not same as cdrom drive that's md5'ing now
[07:59:00] <Jymmm> not even the same machine.
[08:01:17] <Jymmm> I'm tempted to buy one of these http://newspaperads.mercurynews.com/ROP/ads.aspx?advid=32664&adid=4428278&subid=15914980&type=
[08:02:34] <Jymmm> ok md5sum of /dev/cdrom PASSED!!!!!
[08:04:08] <anonimasu> :)
[08:04:31] <Jymmm> any thoughts on that system????
[08:04:46] <Jymmm> I know NOTHING AT ALL about AMD
[08:05:02] <Jymmm> It could be the equic of a 486 for all I know
[08:05:06] <Jymmm> equiv
[08:05:53] <Dallur> hmm that link seems to be missing something
[08:06:06] <anonimasu> hi
[08:06:13] <Dallur> hey :D
[08:06:47] <Jymmm> Dallur http://newspaperads.mercurynews.com/ROP/ads.aspx?advid=32664&adid=4428278&subid=15914980&type=
[08:07:16] <Dallur> Jymmm: all I get is a page showing the address of Fry's in San Jose
[08:07:42] <Jymmm> Dallur try another browser maybe?
[08:07:58] <Dallur> Jymmm: firefox is all I got on this machine ....
[08:08:08] <Dallur> Jymmm: could be a country/ip based filter
[08:08:18] <Jymmm> maybe, but doubtful
[08:08:28] <cncman> +
[08:08:54] <Jymmm> http://newspaperads.mercurynews.com/ROP/ads.aspx?adid=4428278&advid=32664&type=
[08:09:02] <Jymmm> then click on the 3rd ad
[08:10:31] <anonimasu> works or me
[08:10:53] <Jymmm> anonimasu which?
[08:11:05] <anonimasu> the page..
[08:11:23] <Dallur> It's possible that there is some flash thingy embedded, running on 64bit ubuntu here at work so no flash and only firefox
[08:11:45] <anonimasu> ok
[08:12:03] <Dallur> Anyways, no matter, Jymmm, is that a Geode or ?
[08:12:25] <Jymmm> http://img.travidia.com/rop-sub/15914980
[08:12:28] <Jymmm> try that
[08:12:46] <Dallur> Jymmm: thx, works
[08:13:28] <Dallur> $449, not bad for complete machine
[08:14:11] <Jymmm> but should it work "well" for emc2?
[08:14:26] <Jymmm> assuming I add memory iif needed
[08:14:49] <Dallur> Jymmm: Do what alex_joni does, go to the store with a livecd preconfigured with a couple of diff. base threads and a simulator :)
[08:16:00] <Jymmm> Dallur alex_joni has WAY TOO MUCH time on his hands... he needs 2 or 5 MORE girlfriends to keep him occupied
[08:16:10] <Dallur> lol
[08:16:41] <Jymmm> Ok, installing from the cd now...
[08:25:50] <Jymmm> alex_joni when the install finished, what do I do next?
[08:27:06] <anonimasu> configure emc
[08:27:17] <Jymmm> with what? apt-get ?
[08:27:36] <anonimasu> no.. just configure it..
[08:27:58] <Jymmm> anonimasu alex_joni suggested to try a ubuntu cd, then install emc from there instead of the livecd
[08:27:58] <anonimasu> star it and let it copy a config file to your home directory..
[08:28:35] <anonimasu> oh..
[08:28:39] <anonimasu> apt-get install emc2
[08:28:40] <Jymmm> since the problem I've been haviing on three seperate machines only has one common thing - the livecd
[08:28:49] <anonimasu> but you need to add the repository for emc..
[08:28:54] <anonimasu> in /etc/apt/sources.list
[08:28:59] <Jymmm> which is?
[08:29:00] <anonimasu> it's described on the wiki
[08:29:12] <anonimasu> 2I'd help you but im kind of busy
[08:29:21] <Jymmm> np
[08:37:26] <alex_joni> Jymmm: try running it really hard
[08:37:30] <alex_joni> and see if you can get it to fail
[08:37:38] <alex_joni> glxgears (a couple of them)
[08:37:41] <alex_joni> etc
[08:38:11] <alex_joni> let them run for a while (same as you did while it barfed before)
[08:38:50] <alex_joni> if that works ok, follow these 7 simple steps for installing emc2 on it: http://www.linuxcnc.org/content/view/21/4/lang,en/
[08:47:38] <Jymmm> alex_joni glxgears alone doens't seem to crash X, just glxgear + RT Test
[08:48:11] <Jymmm> alex_joni: but I'll try, running updates now. a wee bit more bw here than at home =)
[08:48:53] <Jymmm> Gotta love it, we are retiring our OC-48's =)
[08:49:08] <anonimasu> heh
[08:49:17] <Jymmm> too slow =)
[08:50:31] <Jymmm> alex_joni do you know where the apt sources links are in the wiki? I couldn't find them
[08:51:10] <Jymmm> this? http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?DebianInstall
[08:53:14] <Jymmm> That says Debian Etch, isn't that debian 4 ?
[09:12:14] <alex_joni> Jymmm: follow these 7 simple steps for installing
[09:12:14] <alex_joni> emc2 on it:
[09:12:14] <alex_joni> http://www.linuxcnc.org/content/view/21/4/lang,en/
[09:12:36] <alex_joni> there is a script there you need to run, which will set you up with the right repos
[09:12:44] <Jymmm> gotcha
[09:12:47] <alex_joni> but it's for Ubuntu, not Debian
[09:13:03] <Jymmm> I installed 6.06
[09:13:04] <alex_joni> there are no kernel packages for Debian right now
[09:13:09] <Jymmm> ah, ok
[09:13:49] <Jymmm> IF I get this going finally, I'll talk to cradek about SMP on a dual P3 1Ghx w/ 2GB ram I have
[09:39:16] <Jymmm> ok, emc installed... rebooting
[09:50:11] <Jymmm> I have to pack up in a few minutes, but 4 glxgears are going and just started RT test.
[09:57:17] <Jymmm> G'Night Folks, thanks alex_joni
[10:22:10] <jepler> I'm home!
[10:24:33] <Dallur> jepler: No more Europe ?
[10:25:22] <jepler> Dallur: I assume it's still there...
[10:26:23] <Dallur> jepler: let me check, hmm can't see that far
[10:26:39] <Dallur> jepler: I assume since I can't see it that it's not there
[10:29:56] <jepler> good point
[11:04:29] <alex_joni> yay.. hi jeff
[11:04:41] <alex_joni> it's there.. at least here
[11:19:02] <jepler> hi alex
[11:21:20] <alex_joni> up so early? or still getting adjusted to timezones?
[11:22:43] <jepler> at 5AM local time I just didn't feel like I could sleep anymore
[11:22:57] <jepler> that was after 6-7 hours of sleep, though, so it's not too bad
[11:24:25] <cradek> I bet you have a little email to read
[11:26:17] <jepler> not all that much
[11:26:20] <cradek> and I kept at least one hard bug for you...
[11:26:26] <jepler> I skipped over the latest "emc for usb and no realtime" thread, for instance
[11:26:55] <jepler> speaking of that, it looks like I got a segfault, but it looks a bit different from the traceback you mailed me
[11:26:55] <cradek> yeah I can understand that
[11:27:01] <jepler> Program received signal SIGSEGV, Segmentation fault.
[11:27:01] <jepler> [Switching to Thread -1231410256 (LWP 16601)]
[11:27:01] <jepler> 0xb6a1163c in Logger_start (s=0xb7d30d20, o=0xb49c5fe8)
[11:27:01] <jepler> at emc/usr_intf/axis/extensions/emcmodule.cc:1527
[11:27:01] <jepler> 1527 np.x = op.x; np.y = op.y; np.z = op.z;
[11:27:13] <cradek> oh, yours is better
[11:27:36] <cradek> I only got that one the first try, and then never again
[11:27:44] <cradek> I'll try some more I guess
[11:27:56] <jepler> but yours was in wait_complete (but without debug information, who knows if it was the same or if it was different)
[11:27:57] <cradek> it seemed a strange time for a clone()
[11:28:28] <cradek> bbl, have to get ready
[11:28:31] <jepler> oh -- that's like how main() always shows at the top of a traceback (of the main thread)
[11:28:54] <jepler> (gdb) print op
[11:28:54] <jepler> $2 = (logger_point &) @0x9b7faf0: Cannot access memory at address 0x9b7faf0
[11:29:52] <jepler> aha
[11:30:10] <jepler> you have to have a program that changes color at the same logger cycle as it resizes the buffer
[11:30:12] <jepler> struct logger_point &op = s->p[s->npts-1];
[11:30:15] <jepler> [... realloc]
[11:30:21] <jepler> 1527I np.x = op.x; np.y = op.y; np.z = op.z;
[11:30:43] <jepler> ..but op is a reference to memory that has been freed by the realloc..
[11:46:49] <CIA-19> 03jepler 07v2_1_branch * 10emc2/src/emc/usr_intf/axis/extensions/emcmodule.cc:
[11:46:49] <CIA-19> Fix a bug reported at http://www.pastebin.ca/454207 -- axis would sometimes
[11:46:49] <CIA-19> segfault, and this problem made it happen somewhat repeatedly.
[11:47:23] <CIA-19> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/extensions/emcmodule.cc: merge rev fix segfault in axis backplot on http://www.pastebin.ca/454207
[11:59:05] <CIA-19> 03compile-farm 07BDI-4.51 ( * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot6_log.txt
[12:03:19] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot1_log.txt
[12:04:18] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot2_log.txt
[12:04:36] <cradek> jepler: wow, good catch, his program constantly switches between arcs and lines
[12:04:57] <cradek> compile-farm: we know, we know
[12:05:02] <skunkworks> Is jepler back?
[12:05:08] <cradek> yes and he fixed the hard bug already
[12:05:13] <skunkworks> Nice
[12:05:14] <cradek> (well, mostly)
[12:05:40] <skunkworks> the wrong velocity one?
[12:05:52] <cradek> no, the axis-blows-up one
[12:06:01] <skunkworks> ah - didn't know about that one ;)
[12:06:34] <cradek> it was very rare, but someone had a program that did it
[12:07:28] <cradek> bbl
[12:07:43] <skunkworks> He just gets back - and you make him work ;)
[12:09:11] <skunkworks> logger_emc: bookmark
[12:09:11] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-04-24.txt
[12:15:46] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2.1branch/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2.1branch_slot1_log.txt
[12:16:40] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2.1branch/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2.1branch_slot2_log.txt
[12:17:30] <CIA-19> 03compile-farm 07BDI-4.51 ( * 10emc2.1branch/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2.1branch_slot6_log.txt
[12:19:18] <CIA-19> 03compile-farm 07Ubuntu 6.06 LTS (dapper) realtime (2.6.15-magma) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot7_log.txt
[12:24:11] <CIA-19> 03compile-farm 07Ubuntu 6.06 LTS (dapper) realtime (2.6.15-magma) * 10emc2.1branch/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2.1branch_slot7_log.txt
[12:46:42] <jepler> whoops, what did I do to the build?
[12:46:50] <jepler> hmph
[12:46:50] <alex_joni> missed a &
[12:47:05] <jepler> but it built for me
[12:48:32] <jepler> * jepler notices that on this machine he's had an emc/axis running for about 3 1/2 weeks now
[12:49:43] <alex_joni> heh.. nice :)
[12:50:29] <jepler> it's always reassuring that software doesn't crash just because it's been running a long time
[12:50:39] <alex_joni> yeah..
[12:50:42] <jepler> jepler 15128 0.3 0.2 4052 2236 pts/16 S+ Mar29 134:12 /net/filesrv1/sd3a/users/jepler/emc2-src/bin/rtapi_app load pumakins
[12:50:49] <alex_joni> and that there are no memory leaks
[12:51:17] <jepler> looks like the last thing I ran was that "cloos" logo file you sent me
[12:51:27] <alex_joni> heh, cool
[12:52:06] <skunkworks> I tried the multi-processor rt kernel last night. Painless and worked. :)
[12:52:17] <CIA-19> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/extensions/emcmodule.cc: fix up the earlier compile errors
[12:53:50] <CIA-19> 03jepler 07v2_1_branch * 10emc2/src/emc/usr_intf/axis/extensions/emcmodule.cc: fix up the earlier compile errors
[12:58:03] <alex_joni> http://maps.google.com/
[12:58:13] <alex_joni> search for "from new york to london"
[12:58:20] <alex_joni> and look at the instructions step 24
[13:00:35] <CIA-19> 03compile-farm 07BDI-4.51 ( * 10emc2head/: build PASSED
[13:02:19] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2head/: build PASSED
[13:02:53] <skunkworks> http://www.cnczone.com/forums/showpost.php?p=289605&postcount=48
[13:04:51] <skunkworks> He was the one asking about why the spindle didn't shut off when a limit was hit.
[13:05:10] <skunkworks> Another one praising the virtues of emc ;)
[13:08:06] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) non-realtime (2.6.12-10-386) * 10emc2.1branch/: build PASSED
[13:09:58] <CIA-19> 03compile-farm 07BDI-4.51 ( * 10emc2.1branch/: build PASSED
[13:10:07] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2head/: build PASSED
[13:13:54] <CIA-19> 03compile-farm 07Ubuntu 6.06 LTS (dapper) realtime (2.6.15-magma) * 10emc2head/: build PASSED
[13:14:24] <alex_joni> skunkworks: cool
[13:14:29] <alex_joni> can't see the pics though
[13:15:23] <skunkworks> crap
[13:16:15] <skunkworks> it just shows his machine.. mdf style (wood)
[13:16:31] <alex_joni> ah, ok..
[13:16:31] <skunkworks> he also has a bunch of screen shots of emc/axis.
[13:16:40] <alex_joni> nice
[13:16:41] <skunkworks> very
[13:17:02] <CIA-19> 03compile-farm 07Ubuntu 5.10 (breezy) realtime (2.6.12-magma) * 10emc2.1branch/: build PASSED
[13:17:14] <skunkworks> * skunkworks wonders why there are no pictures up on jeff's site... ;)
[13:17:39] <skunkworks> yet
[13:18:34] <skunkworks> quote
[13:18:35] <skunkworks> It was interesting trying Mach3, as a lot of people rave about it on here. It's easy to configure, and seems very powerful, but the bang bang acceleration curve and being based on Windows are detractors for me.. Also the user interface is too busy for my liking, although 1000% better than Mach2, which I also tested out of curiosity...
[13:18:40] <skunkworks> /qoute
[13:19:34] <CIA-19> 03compile-farm 07Ubuntu 6.06 LTS (dapper) realtime (2.6.15-magma) * 10emc2.1branch/: build PASSED
[13:37:52] <cradek> a nice video of it
[13:40:06] <jepler> skunkworks: it'll be at least a few days before I get my trip photos up
[13:44:10] <alex_joni> cradek: what video?
[13:44:42] <cradek> there's a link in skunkworks's link
[13:45:05] <alex_joni> oh, missed that one
[13:45:47] <skunkworks> he has a few videos on google
[13:46:39] <skunkworks> http://video.google.com/videoplay?docid=5594809606355326574
[13:49:04] <skunkworks> boy - that thing really runs nice
[13:49:49] <cradek> yeah it looks snappy
[13:52:59] <alex_joni> bbl
[13:57:52] <skunkworks> jepler: no hurry ;)
[14:02:53] <skunkworks> hmm - I have a motherboard that smoked :)
[14:03:11] <skunkworks> PII 450 I think
[14:03:46] <skunkworks> thought the power supply went bad - but one of the regulators on the motherboard burned up.
[14:04:09] <cradek> that sounds familiar
[14:06:10] <Dallur> skunkworks: cartridge cpu ?
[14:07:57] <cradek> ah, and the cat just came in. Hi bonnie. [meooorw!]
[14:08:42] <Dallur> does bonnie speak tcp/ip ?
[14:08:58] <cradek> no it sounds like she just speaks cat
[14:09:34] <Dallur> cat5 or cat6 ?
[14:10:26] <cradek> just cat I think
[14:11:08] <skunkworks> Dallur: yes
[14:11:11] <skunkworks> slot 1
[14:11:48] <Dallur> skunkworks: just be happy it's gone then :=
[14:13:38] <Dallur> skunkworks: I was a sysadmin for 300ppl software company when those were started showing up, dell and micron machines had about 50% failure rate within the first year,, horrible,, horrible idea to move the cache off the die
[14:13:39] <skunkworks> but.. but.. it would run emc2 ;)
[14:14:26] <skunkworks> hmm - never had that problem.. we had a bunch of ecs motherboard that the caps exploded on/
[14:17:01] <skunkworks> and a bunch of gateway monitors that would smoke after a year or so.
[14:17:36] <Dallur> skunkworks: that sweet smell of burned circuits
[14:17:49] <cradek> I fixed dozens of PS/2 VGA monitors back in the day - they all had the same problems
[14:18:38] <cradek> Dallur: I read the cache yield was so slow, moving it outside the processor was the only way to make the processors affordable
[14:19:55] <Dallur> cradek: yeahh but a pentium pro would outperform a P2/P3 cartridge double it's clockspeed ....
[14:20:20] <cradek> no..... really?
[14:20:38] <Dallur> cradek: yeah, really sad, only thing that saved the P3 was mmx
[14:21:18] <Dallur> At that time we were working really closely with Intel, got to spend a whole lot of time down in San Jose with em ,,
[14:21:31] <cradek> hmm somewhere I still have a dual PPro 200 board with processors
[14:22:10] <Dallur> cradek: Try firing it up, I'm betting your will be pleasantly supprised
[14:22:21] <cradek> I might do that if I can find it
[14:23:45] <jepler> I know p3 outperformed p4 of the same speed, I didn't know the same was true of the ppro -> p2 switch
[14:24:09] <skunkworks> I have a dual proliant server (but only one processor)
[14:24:14] <skunkworks> p-pro
[14:28:04] <jepler> "Likely Pentium Pro's most noticeable addition was its on-package L2 cache. At the time, manufacturing technology did not feasibly allow L2 cache to be integrated into the processor core. Intel instead placed the L2 die separately in the package which still allowed it to run at the same clock speed as the CPU core." -- wikipedia
[14:33:04] <skunkworks> jepler: http://www.electronicsam.com/images/pentIII2.JPG
[14:33:22] <skunkworks> http://www.electronicsam.com/images/pent3shot.png
[14:35:43] <jepler> looks like the PII was not a bad per-MHz competitor to the ppro: http://www.tomshardware.com/1997/04/30/the_empire_strikes_back/
[14:36:44] <jepler> "The Pentium II doesn't include any second level cache anymore as its predecessor Pentium Pro. This is due to production costs, since in case only one die in the Pentium Pro is not functioning in the post-bonding CPU tests, the whole CPU has to be thrown away. The chips can only be tested after the bonding and at this stage it's too late to use one of the two components anymore. "
[14:38:33] <jepler> wow that was 10 years ago
[14:41:56] <skunkworks> wow. the proliant pentium pro was the first server I bought when I took this position
[15:00:37] <cradek> if I have a board that runs two PII, can I plug in two PIII instead?
[15:00:48] <skunkworks> cool - the burnt motherboard has a agp ati rage 128
[15:36:58] <jepler> I forget whether it's that card or the matrox that chris has gotten to work with accelerated opengl and hard realtime..
[15:41:18] <skunkworks> hmm - cool
[15:41:26] <cradek> matrox
[15:41:30] <skunkworks> aw
[15:41:52] <cradek> pci g400 or g450 I think
[15:42:10] <skunkworks> cradek: I have a matrox myst/mga
[15:43:23] <skunkworks> powerd by MGA 64-bit graphics.
[15:45:41] <cradek> you may get direct rendering with that - not sure
[15:55:10] <cradek> (it really doesn't make much difference except for screensavers and stuff like that)
[15:59:38] <Guest980> damn internet
[16:00:00] <Guest980> Guest980 is now known as skunkworks_
[16:03:04] <skunkworks_> skunkworks_ is now known as skunkworks
[16:17:49] <CIA-19> 03jepler 07v2_1_branch * 10emc2/docs/src/gui/image-to-gcode.lyx: give section name for additional inifile lines
[16:18:34] <CIA-19> 03jepler 07TRUNK * 10emc2/docs/src/gui/image-to-gcode.lyx: merge rev give section name for additional inifile lines
[17:07:58] <feoc> moo
[17:30:35] <skunkworks> http://www.tuxcnc.org/pivot/entry.php?id=16#body
[17:35:13] <skunkworks> Hi paul ;)
[17:37:45] <feoc> * feoc wonders who paul is
[17:38:36] <skunkworks> He does the original bdi emc1 stuff
[17:39:26] <feoc> cool
[18:19:25] <CIA-19> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: do not change the text of the 'touch off' menu item, it causes troubles for i18n and string-based menu enable/disable
[18:19:51] <cradek> ack I didn't think of that
[18:20:21] <cradek> I'm afraid its operation is mysterious without it
[18:21:06] <jepler> * jepler hmms
[18:21:17] <cradek> can i18n just have %c in a fmt string?
[18:21:27] <cradek> or, maybe the design sucks
[18:21:46] <jepler> yes, it's more the second thing that is important: I'm changing the enable/disable so that it will use menu entry names instead of menu entry indices
[18:22:07] <jepler> state {$interp_state == $INTERP_IDLE && $taskfile != ""} \
[18:22:07] <jepler> {.menu.file "_Reload"}
[18:22:09] <jepler> e.g., ^^^
[18:22:14] <cradek> ok that's good, it sucks now
[18:22:41] <cradek> it probably took me an hour to get the File menu right
[18:23:27] <cradek> although most of that was figuring out how to hook together the tcl and python variable and only remove those entries once
[18:25:38] <jepler> indeed
[18:25:46] <jepler> I think the code I just wrote will grey out Edit all the time if no editor is defined
[18:26:00] <cradek> is that better than removing it?
[18:26:19] <jepler> no
[18:26:23] <cradek> (I wish I would have thought of that - it's much easier)
[18:27:14] <jepler> I am sure my first try will get the menu enables wrong but it'll be easier to fix than before
[18:27:39] <cradek> sure
[18:27:48] <cradek> I tested it a LOT before I was sure
[18:28:17] <jepler> +setup_menu_accel .menu.machine end [_ "Touch-O_ff current axis in system..."]
[18:28:39] <jepler> can you help me understand what will be "mysterious" about this?
[18:29:30] <cradek> well it's not your wording, it's the whole design where the thing that happens when you pick the menu entry changes based on something that's not even in the menu
[18:30:06] <cradek> for instance, to home an axis, you don't have to pick it on the tabs before going to the menu
[18:30:07] <C_Kode> Hey all. I'm using Navisphere and wanting to connect one of our Linux servers to the CX300. Does anyone know the correct connection setting in Navisphere?
[18:30:08] <jepler> maybe there should be Touch Off > X > G54 (one more level)
[18:30:24] <jepler> C_Kode: wrong emc
[18:30:25] <C_Kode> whoops
[18:30:25] <cradek> C_Kode: I think you have the wrong channel...
[18:30:31] <C_Kode> yeah I just saw that! Sorry
[18:30:35] <jepler> C_Kode: s'okay
[18:30:37] <cradek> good luck
[18:30:40] <C_Kode> thanks
[18:31:03] <cradek> jepler: or the touch off window (that comes from poking the button) should let you select the system on radiobuttons.
[18:31:31] <cradek> then at least it's all run from the tab
[18:32:06] <cradek> but, since this is fairly "advanced" usage, I did it this way to "hide" it from the usual operation where you want to not worry about systems
[18:39:55] <jepler> I suppose I should actually look at how the menu is organized
[18:40:36] <jepler> * jepler sucks in air through is teeth: P1 G5_4
[18:40:55] <jepler> 'P1' is the code used with G10 L2 ?
[18:41:00] <cradek> yes
[18:41:17] <cradek> and the spec calls it coordinate system 1, not 54
[18:41:27] <jepler> * jepler groans
[18:41:39] <cradek> (please make it not suck)
[19:28:27] <Jymmmmmmm> Hey gene, ltns
[19:45:46] <CIA-19> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: fix bug reported by cradek: pressing - or = in the mdi field would display a traceback on the terminal
[19:55:11] <Jymmmmmmm> Just for the record (if anyone cares).... Under ubuntu 6.06 (at least) at the boot menu when you have the "Check CD" option, it can report "GOOD", but when you md5sum /dev/cdrom can still FAIL
[19:55:41] <cradek> often md5sum /dev/cdrom doesn't work right
[19:56:01] <cradek> it reads too much or too little from the CD
[19:56:04] <erdizz> right. just as cat /dev/cdrom no longer produces a correct image
[19:56:14] <cradek> if you dd the right number of bytes from /dev/cdrom, it might work
[19:56:44] <cradek> (I have read that scsi cdrom drives are better)
[19:56:51] <erdizz> me too
[19:57:18] <cradek> the built-in test is the one to trust
[19:57:33] <erdizz> the scsi me is, I mean...
[20:00:58] <Jymmmmmmm> the built in one is the one that lied.
[20:02:12] <Jymmmmmmm> Buffer I/O failures whne I did md5sum the first time, then dl iso and reburned, and md5sum was good
[20:03:07] <Jymmmmmmm> Ok, have 8 glxgears running right now, hasn't crashed X yet (but did crash earlier when I run rt test
[20:03:19] <erdizz> Jymmmmmmm, see this post at lkml: http://lkml.org/lkml/2006/4/12/46
[20:03:32] <Jymmmmmmm> cradek: someone said you have a smp kernel?
[20:05:03] <cradek> it's only experimental so far
[20:05:18] <Jymmmmmmm> cradek: easy enough to install configure?
[20:05:34] <Jymmmmmmm> I'm running out of options
[20:05:35] <cradek> yes, but it comes with no support
[20:06:46] <cradek> linuxcnc.org/experimental/
[20:07:02] <cradek> you will have to build emc under it of course
[20:07:09] <cradek> no packages
[20:07:46] <Jymmmmmmm> ah, I have no idea how todo that, so nm
[20:09:31] <Jymmmmmmm> ok, 8 glxgears have been running for 20+ minutes, now starting RT test
[20:13:23] <Jymmmmmmm> is there anyway to timestamp and log everything I do? Maybe I cna find a pattern
[20:17:57] <Jymmmmmmm> Just crashed
[20:22:02] <Jymmmmmmm> Ok running RT Test solo
[20:28:10] <skunkworks> So it restarts X?
[20:28:44] <Jymmmmmmm> yes
[20:30:19] <jymmmmm> There are no timestamps, but there are errors in .xsession-errors mostly non-existant sound card, but here are a couple that aren't sound card related:
[20:30:30] <jymmmmm> (gnome-panel:4503): GdkPixbuf-CRITICAL **: gdk_pixbuf_scale_simple: assertion `dest_width > 0' failed
[20:30:30] <jymmmmm> Gnome-Message: gnome_execute_async_with_env_fds: returning -1
[20:30:42] <jymmmmm> (gnome-panel:4503): Gtk-WARNING **: gtk_widget_size_allocate(): attempt to allocate widget with width -9 and height 24
[20:30:59] <jepler> most gnome and gtk programs spew warnings like that; it's not clear to me which ones matter, if any.
[20:32:07] <Jymmmmmmm> jepler: : Ok, I'm just grasping at straws now. Three different systems, two differnent livecd's, and a solo ubuntu+emc script installs - all crash X when I run glxgears and RT test.
[20:32:37] <cradek> what happens if you don't do that?
[20:32:53] <Jymmmmmmm> "that" ?
[20:33:29] <Jymmmmmmm> I ran 8 iterations of glxgears solo for 20+ minutes, np. Running RT Test solo right now
[20:33:43] <anonimasu> are you sure you didnt run out of memory?
[20:33:54] <Jymmmmmmm> video or system?
[20:33:56] <anonimasu> that makes X die hard..
[20:33:56] <anonimasu> system
[20:34:01] <Jymmmmmmm> 768MB
[20:34:07] <anonimasu> hm, should be enough
[20:34:39] <Jymmmmmmm> I bought a APG nvidia based video card last night with 128MB video ram, still crashed X
[20:34:56] <Jymmmmmmm> went back to my ATI Mach64 AGP card this monring
[20:35:04] <anonimasu> this is really odd..
[20:35:15] <anonimasu> got any usb stuff plugged in?
[20:35:37] <ds3> Jymmm: what driver are you using?
[20:35:38] <Jymmmmmmm> None at all as all these systems are USB1.1, using PS/2 kybd and mouse
[20:35:52] <skunkworks> usb stuff for me just gives me overuns on certain hardware.
[20:36:00] <anonimasu> yep
[20:36:00] <anonimasu> me too
[20:36:49] <jymmmmm> I removed all the items related to watcom table from xorg.conf as it was giving errors in the log "not found"
[20:36:57] <jymmmmm> err wacom
[20:37:41] <Jymmmmmmm> serial ports are disabled in the bios
[20:38:15] <jymmmmm> http://pastebin.ca/456449
[20:38:51] <skunkworks> Is there a way to look at system resources in linux.. like memory and such. Might be nice to have up while running everything.
[20:39:00] <anonimasu> yeah
[20:39:01] <anonimasu> top
[20:39:04] <ds3> free
[20:39:15] <anonimasu> or free
[20:39:21] <jepler> there's some gnome program, look around in the menus
[20:39:26] <ds3> cat /proc/meminfo
[20:39:28] <ds3> procinfo
[20:39:31] <ds3> ;)
[20:39:53] <Jymmmmmmm> but does it log? when X crashes I can't see crap
[20:40:09] <ds3> Look at /var/log/Xorg.*
[20:41:29] <Jymmmmmmm> ds3: nothing in there significant
[20:43:19] <jymmmmm> I wish there was
[20:44:00] <anonimasu> jymmmmm: try "dmesg"
[20:44:05] <anonimasu> and paste the output
[20:44:06] <Jymmmmmmm> Ok, RT Test has been running solo for 20 minutes, will let it run for a while longer to see if it crahses
[20:44:32] <anonimasu> there might be something in there..
[20:44:35] <anonimasu> or in /var/log/messages
[20:45:36] <jymmmmm> dmsg http://pastebin.ca/456458
[20:46:54] <jymmmmm> messages http://pastebin.ca/456460
[20:47:58] <alex_joni> obie shutdown[3956]: shutting down for system halt
[20:48:59] <anonimasu> hm
[20:49:00] <chr0n1c> say a retarded person sent me a part to quote... and it's a couple pictures in an undecipherable pdf.. with not much information
[20:49:12] <chr0n1c> should i.. 1: work with the guy or 2: ignore it
[20:49:42] <anonimasu> well, charge him for drawing the parts..
[20:49:59] <anonimasu> ;)
[20:50:03] <chr0n1c> hmm..
[20:50:31] <anonimasu> I guess it depends if it's worth it..
[20:50:31] <chr0n1c> i guess he just wants letters and stuff for a door sign cut out by cnc
[20:50:43] <chr0n1c> then glued together to make a 8.5" x 11" sign?
[20:50:49] <chr0n1c> that's about all i can make out of it
[20:51:26] <chr0n1c> i was dumpster diving a few weeks ago and came across over 300 dvd cases
[20:51:30] <chr0n1c> all black plastic...
[20:51:39] <anonimasu> if it's a nice
[20:51:41] <anonimasu> nice.. :)
[20:52:04] <chr0n1c> and before the y guy emailed me i had the idea of cutting out some funky letters and numbers from the flat half of the cases
[20:52:14] <chr0n1c> for making house signs...
[20:52:26] <chr0n1c> i got like 3000 fonts i think
[20:52:35] <chr0n1c> it would be a nifty way to get rid of the dvd cases...
[20:52:50] <chr0n1c> unless someone needs some? they are from like uhhh blocky busters
[20:52:56] <anonimasu> hehe
[20:52:58] <anonimasu> melt them
[20:53:08] <alex_joni> fire !!!
[20:53:08] <anonimasu> and make blocks for prototyping
[20:53:11] <chr0n1c> big fire in the backyard huh?
[20:53:20] <anonimasu> I guess in a oven..
[20:53:21] <chr0n1c> if i only had a furnace
[20:53:26] <chr0n1c> or my oven!
[20:53:36] <chr0n1c> shew that would stink up the house for a few days
[20:53:38] <alex_joni> mom's gonna get upset
[20:53:39] <alex_joni> :P
[20:53:47] <chr0n1c> i own my own house yo ;)
[20:53:56] <chr0n1c> i haven't talked to my mom in about ... 4 years
[20:53:57] <Jymmmmmmm> chr0n1c: toaster oven
[20:54:09] <chr0n1c> toaster oven out back!
[20:54:13] <chr0n1c> great idea
[20:54:41] <chr0n1c> i can chunk them up i gess and make a square mold
[20:54:47] <chr0n1c> or i could use a biscut pan, lol
[20:54:52] <chr0n1c> plastic muffins
[20:55:37] <Jymmmmmmm> vaseline + xylene as a mold release, but wait till it evaporates
[20:55:49] <chr0n1c> hmm.. think the brand new crock pot i never used still in the packaging... would get hot enough?
[20:56:05] <Jymmmmmmm> chr0n1c: not hot enough
[20:56:25] <Jymmmmmmm> will just become gooey
[20:56:33] <chr0n1c> yeah...
[20:56:35] <Jymmmmmmm> then you'll run out of heat
[20:56:45] <chr0n1c> unless i tim taylor the crock pot .. :D
[20:56:53] <Jymmmmmmm> Ug UG UG
[20:56:58] <chr0n1c> lol
[20:57:17] <Jymmmmmmm> iirc you need about 280F+
[20:57:36] <chr0n1c> we used to use a crock pot at the shop for coating badass parts with that gooey rubbery endmill stuff
[20:57:41] <Jymmmmmmm> coffe can + burner
[20:57:45] <chr0n1c> *before we shipped them
[20:58:04] <Jymmmmmmm> chr0n1c: but tha'ts not a thermoplastic
[20:58:13] <chr0n1c> ya...
[20:58:40] <Jymmmmmmm> Now THAT you can dilote with xylene
[20:59:09] <Jymmmmmmm> chr0n1c: PORTABLE burner, like camping burner
[20:59:17] <chr0n1c> yeah...
[20:59:30] <Jymmmmmmm> but camping burner can't control the heat very well, unless propane
[20:59:39] <chr0n1c> i have a gas grill outback... that should work ok
[20:59:48] <skunkworks> jymmmmm: does it crash faster - the more glxgears are running?
[20:59:50] <Jymmmmmmm> ew stinky steaks
[20:59:50] <chr0n1c> i can do a double boiler thing with the coffe can and the grill
[20:59:59] <Jymmmmmmm> skunkworks: No, not at all.
[21:00:17] <Jymmmmmmm> chr0n1c: do you wnat it induce moisture to the platic?
[21:00:32] <chr0n1c> you got the source code for glxgears and for the rt test?
[21:00:48] <chr0n1c> can someone who knows programming... find the clash?
[21:01:04] <chr0n1c> good point jymm
[21:01:04] <anonimasu> there is no clash..
[21:01:04] <Jymmmmmmm> Can'tfind it till we know what's causing it.
[21:01:05] <anonimasu> usually
[21:01:26] <Jymmmmmmm> RT Test solo is still going
[21:01:48] <anonimasu> Jymmmmmmm: try it with one instance of glxgears
[21:01:50] <anonimasu> later
[21:02:07] <Jymmmmmmm> Ok, RT Test has been running for 50 minutes now
[21:02:43] <Jymmmmmmm> I'll give it a full hour, then start one glxgears
[21:02:56] <anonimasu> ok
[21:02:57] <anonimasu> :)
[21:03:34] <Jymmmmmmm> Whats ironic is that if I do scrnsaver gears, it takes two hours, then crashes X
[21:06:40] <skunkworks> do you have a ups you could plug it into? (do you have power problems?)
[21:07:03] <skunkworks> * skunkworks is just throwing stuff out there.
[21:07:03] <Jymmmmmmm> skunkworks: I did this at work last bight, same problem
[21:07:08] <Jymmmmmmm> night
[21:07:24] <Jymmmmmmm> We have enough power at work to power a small city for a week
[21:08:00] <Jymmmmmmm> all conditioned
[21:08:23] <chr0n1c> i was gonna mention that. maybe a small power spike...
[21:08:28] <skunkworks> do you have an excpetionally electric personallity?
[21:08:29] <chr0n1c> not much to notice it
[21:08:36] <chr0n1c> but enough for the computer to hiccup
[21:09:06] <Jymmmmmmm> skunkworks: heh, testing at work completely eliminates any power issues, that's for sure.
[21:09:25] <skunkworks> Yes
[21:09:35] <anonimasu> Jymmmmmmm: did you replace your keyboard and mouse?
[21:09:38] <anonimasu> *grins*
[21:09:45] <anonimasu> just kidding
[21:09:49] <alex_joni> CMOS battery?
[21:10:00] <Jymmmmmmm> 119.6VAC @ 1.87 apms 59.9 Hz
[21:10:01] <skunkworks> Pretty cool - some one actually comparing emc2 to mach. (lord knows I haven't) ;) http://www.cnczone.com/forums/showthread.php?p=289808#post289808
[21:10:02] <anonimasu> did you take ram out of the other box?
[21:10:07] <alex_joni> * alex_joni is feeling silly..
[21:10:15] <chr0n1c> is someone flushing the toilet while the test are running?
[21:10:28] <Jymmmmmmm> anonimasu: no, each system has it's own mobo, ram, hdd, video card, etc
[21:10:33] <anonimasu> Jymmmmmmm: ok
[21:10:35] <skunkworks> * skunkworks wonders about the tin-foil hat
[21:10:36] <Jymmmmmmm> chr0n1c: LOL, I know that one =)
[21:10:37] <anonimasu> it's really odd..
[21:11:32] <Jymmmmmmm> I went from LiveCD (as that was the only thing commmon between them) to ubuntu + emc install script last night. verified iso md5 as well as md5sum /dev/cdrom too
[21:12:23] <Jymmmmmmm> The thing is has anyone else tried a simular setup for long term (2+hours) as see if they get the same results.
[21:12:56] <Jymmmmmmm> boot from the livecd, run glxgears, then start rt test as mentioned in the wiki
[21:13:07] <Jymmmmmmm> not scrnsaver gears or anything like that
[21:13:36] <skunkworks> hmm - interesting. Have you tried to install ubuntu? doesn't take that long
[21:14:00] <Jymmmmmmm> skunkworks: it is installed on the hdd, then ran updates, them dl the script and installed emc
[21:14:07] <Jymmmmmmm> then ran update agian
[21:14:25] <skunkworks> ah
[21:14:43] <Jymmmmmmm> just to eliminate the possibility thatthe livecd might have issues
[21:17:10] <Jymmmmmmm> So we cna eliminate the live vs stand alone install beingthe issue. The only thing "common" is the distro and of course emc
[21:18:13] <Jymmmmmmm> RT test has been running solo for 55 minutes, will start glxgears in 5 minutes
[21:20:31] <Jymmmmmmm> I wonder.... maybe I should try breezy instead of dapper???
[21:21:36] <Jymmmmmmm> anonimasu: as far the kybd/mouse, yes I did try a different set (at work)
[21:21:47] <Jymmmmmmm> even a differnt monitor too
[21:22:09] <skunkworks> power cord?
[21:22:12] <skunkworks> ;)
[21:22:13] <Jymmmmmmm> yep
[21:22:23] <anonimasu> *sigh* formatting 600gb takes time
[21:22:38] <chr0n1c> format /q
[21:22:46] <anonimasu> it's unformatted
[21:22:50] <chr0n1c> oh.. lol, we are talkin linux
[21:23:18] <chr0n1c> biggest drive i ever formatted was a 120 gig
[21:23:20] <Jymmmmmmm> ok, I'll give it a couple more minutes then start X
[21:23:22] <anonimasu> mkfs.ext3
[21:23:24] <anonimasu> ;)
[21:23:25] <Jymmmmmmm> err glxgears
[21:23:31] <anonimasu> or whatever you write nowdays
[21:24:44] <Jymmmmmmm> starting glxgears................................now!
[21:24:58] <chr0n1c> oh noes!
[21:25:04] <skunkworks> * skunkworks hides
[21:25:10] <chr0n1c> * chr0n1c ducks
[21:25:43] <chr0n1c> * chr0n1c peeks from under the bench to see if everything is okay
[21:26:40] <Jymmmmmmm> isnt there an xtop or something liek that?
[21:27:13] <chr0n1c> i used apachetop for my webserver through ssh
[21:27:23] <chr0n1c> to check on it from work and such
[21:27:41] <chr0n1c> and there is emctop and plain old top
[21:27:46] <chr0n1c> so i bet there is xtop!
[21:28:16] <skunkworks> bbl
[21:28:32] <Jymmmmmmm> used memory: 134.8MB of 758.3
[21:28:40] <Jymmmmmmm> cpu 100%
[21:29:48] <Jymmmmmmm> load avg 7.04
[21:30:14] <chr0n1c> did ya found an xtop?
[21:30:51] <Jymmmmmmm> no
[21:31:15] <chr0n1c> achooo!
[21:45:25] <Jymmmmmmm> -
[21:45:56] <Jymmmmmmm> ok 20 minutes later, started anohter 4 iterations of glxgears - 5 total now running
[21:46:22] <Jymmmmmmm> oh hell, lets go for 9 total
[21:47:00] <chr0n1c> the gears turn at 1 tooth advance every 3 or 4 seconds on my emc box
[21:47:25] <chr0n1c> with nothing else but a terminal running... with the live cd HD install
[21:47:38] <Jymmmmmmm> yeah same here
[21:47:52] <Jymmmmmmm> but just running glxgears should put a load on the system
[21:47:57] <chr0n1c> p2-233mhz/383 mb ram
[21:48:12] <Jymmmmmmm> 768 total, 445mb used
[21:48:23] <Jymmmmmmm> 95% cpu
[22:03:20] <Jymmmmmmm> X crashed, but I was out of the room. but from the other me leaivng I'd suspect almost 30 miuntes later exactly
[22:03:51] <Jymmmmmmm> (14:24:42) Jymmmmmmm: starting glxgears................................now!
[22:04:00] <Jymmmmmmm> (14:54:10) jymmmmm left the room (quit: Remote closed the connection).
[22:04:48] <Jymmmmmmm> If anyone has an ideas, I'm open
[22:07:35] <ds3> setup a getty on a serial port
[22:07:49] <ds3> login via the serial from a vt100, laptop, etc
[22:08:01] <ds3> start X from there. wait for it to crash. look at the reason for the crash
[22:08:38] <ds3> probally can also do a ssh session if that is easiser
[22:09:01] <toastydeath> how are you going to get the details from an X server crash
[22:09:03] <toastydeath> over ssh
[22:09:11] <ds3> if that does not turn up anything, make sure you are not running out of memory - reboot with the console on the serial port
[22:09:21] <toastydeath> i suppose you could dump the output
[22:09:22] <ds3> X usually dumps on stdio; so you start the X server on the ssh session
[22:09:26] <cradek> are you getting sig 11? what's in the X log?
[22:09:28] <ds3> and keep the ssh session open
[22:10:41] <ds3> and if you are really paranoid, start X in the background with the stdio attach to the ssh session, i.e. "startx&"
[22:10:55] <ds3> then you can do periodic 'dmesg; free' to see if anything odd is going on
[22:13:03] <Jymmmmmmm> cradek: cheking now...
[22:17:28] <jymmmmm> http://pastebin.ca/456576
[22:18:15] <cradek> is that after a crash, or with X running?
[22:19:00] <Jymmmmmmm> cradek: That is whats in there currently, no timestamps so I cant tell, but my other irc client left irc at 14:54:10
[22:19:18] <cradek> is X running?
[22:19:39] <Jymmmmmmm> right now, yes. it went back to the login screen (gui) after the crash
[22:19:54] <cradek> then this is the wrong log file, look for the older one
[22:19:58] <Jymmmmmmm> k
[22:21:37] <jymmmmm> http://pastebin.ca/456582
[22:21:50] <Jymmmmmmm> Jymmmmmmm is now known as Jymmm
[22:22:08] <cradek> look, a backtrace
[22:22:26] <cradek> pastebin the output of "glxinfo" on that machine please
[22:22:59] <cradek> you haven't tried to install the ATI binary driver have you?
[22:24:11] <jymmmmm> http://pastebin.ca/456591
[22:24:30] <Jymmm> No, just whatever was install when I installed ubuntu
[22:25:21] <cradek> huh, looks like a mesa bug
[22:25:23] <Jymmm> The ONLY thing I did in respect to xorg conf was tell it my monitor and max res of 1600x1200, plus set the vram to 8192
[22:26:05] <Jymmm> oh and remove all the wacom table entries sections
[22:26:44] <cradek> you're sure it's an 8 meg card?
[22:27:20] <cradek> just for fun, you could try the vesa video driver instead of ati
[22:27:20] <jymmmmm> xorg.conf http://pastebin.ca/456597
[22:28:05] <cradek> might try taking out UseFBDev
[22:28:06] <Jymmm> cradek: sure, not sure what I need todo for that
[22:28:16] <cradek> change "ati" to "vesa"
[22:28:18] <Jymmm> the cesa driver thing that is
[22:28:20] <Jymmm> ah, ok
[22:28:30] <cradek> (but it looks like a mesa bug)
[22:28:49] <cradek> this isn't going to affect running emc - I bet if you stop running 9 glxgears, it won't crash anymore
[22:29:21] <cradek> also if you change your DefaultDepth to 16 instead of 24, you'll have some video ram free and might get direct rendering
[22:29:21] <Jymmm> it did it with running just one glxgears, I was just trying to get it to crash asap if it was going to
[22:31:48] <jymmmmm> restarting X. screw it rebooting instead
[22:32:42] <Jymmm> will repast my new xorg.conf when the box comes back up
[22:34:08] <Jymmm> I just wish I could get it to crash faster
[22:34:35] <cradek> do you get this same crash while running emc, verified by seeing this same backtrace in the log?
[22:35:05] <cradek> I think it's a hard-to-trigger mesa bug, and maybe it's not a big deal if you don't try to trigger it
[22:35:24] <Jymmm> cradek: I haven't gotten that far, just trying to "test" with just glxgears and RT test
[22:35:38] <Jymmm> since it's unpredictable with it's going to crash
[22:35:46] <Jymmm> s/with/when/
[22:35:49] <skunkworks> umm - running the latency test and glxgears makes my portable restart x also
[22:36:02] <cradek> skunkworks: it looks like a Mesa bug
[22:36:14] <skunkworks> mesa?
[22:36:26] <skunkworks> the i/o card?
[22:36:34] <jymmmmm> no OpenGL
[22:36:36] <cradek> no, the software opengl rendering library with the same name
[22:36:46] <skunkworks> ah - confusing
[22:37:00] <skunkworks> logger_emc: bookmark
[22:37:00] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-04-24.txt
[22:37:22] <jymmmmm> ok, Rt test going
[22:37:48] <jymmmmm> 4 glxgears going
[22:38:36] <Jymmm> X just crashed
[22:38:37] <cradek> http://bugs.freedesktop.org/attachment.cgi?id=5893
[22:38:44] <cradek> ^^ doesn't require realtime
[22:41:48] <jymmmmm> http://pastebin.ca/456620
[22:42:10] <Jymmm> any fix mentioned?
[22:42:12] <cradek> don't bother - it won't help - it's a known mesa bug
[22:42:29] <cradek> hard to tell - sf is so slow I can't read the mesa list archives
[22:42:43] <jepler> I can't figure out what bug that traceback is attached to, either
[22:43:12] <cradek> http://bugs.freedesktop.org/show_bug.cgi?id=7205
[22:46:15] <skunkworks> so could that be why I also get random overruns?
[22:46:23] <skunkworks> when running glxgears
[22:46:36] <cradek> skunkworks: probably not related
[22:46:43] <skunkworks> darn
[22:49:31] <skunkworks> so is there anything else that I could use to test video?
[22:49:38] <cradek> wow, that was a long read...
[22:50:13] <cradek> looks like they decided at the end that it's fixed...
[22:51:31] <cradek> so the remaining (hard!) question is how does Jymm get the fixed code
[22:52:38] <ds3> prehaps, the process of being able to get the fix code is a bug in itself? ;)
[22:52:56] <cradek> maybe bring it to the attention of the dapper team (without mentioning realtime), including a pointer to this report we found
[22:53:13] <cradek> it's only been 20 days. maybe they are already planning to update mesa, hard to say
[22:53:58] <ds3> couldn't he use the binary nVidia drivers? think they include their own version of OpenGL
[22:54:24] <cradek> jymm doesn't have nvidia hardware, and also their driver breaks the realtime latency
[22:55:52] <cradek> Jymmm: if you change the DefaultDepth to 16 and turn off UseFBDev, but leave Driver=ati, do you get direct rendering?
[22:56:26] <skunkworks> cradek: what do I change in the x config to try to see if direct rendering works?
[22:56:39] <cradek> skunkworks: some hardware does it, most doesn't
[22:56:51] <cradek> skunkworks: you have to have some video ram free, you can't use it all for the framebuffer
[22:57:00] <cradek> (that's why jymm's reducing the depth might enable it)
[22:57:06] <skunkworks> ah
[22:57:16] <cradek> he only has 8M, but he can get 1600x1200x16 with 4M
[22:57:27] <cradek> so it *might* do direct on his hardware
[22:58:33] <ds3> <Jymmm> I even bought a 128mb nvidia baed agp vard tonight, and it still does
[22:58:33] <ds3> +it.
[22:58:57] <ds3> but the RT problem is a bummer
[22:58:58] <cradek> nvidia will not have direct rendering with the free driver, this ati might
[23:00:02] <cradek> Jymmm: https://launchpad.net/ubuntu/dapper/+bugs?field.tag=xorg
[23:00:31] <cradek> I think you ought to submit a bug report. Include a link to the freedesktop.org bug we found: http://bugs.freedesktop.org/show_bug.cgi?id=7205
[23:03:46] <cradek> actually I'm filing it for you
[23:06:24] <Jymmm> cradek: okey, cool. as I have no clue what the bug actually is
[23:07:07] <skunkworks> So running 16 bit color on the dual processor - I go from 10-15fps to 15-20fpx
[23:07:38] <skunkworks> instead of 24
[23:08:28] <Jymmm> Any idea WHY I found it, and nobody else has?
[23:08:56] <skunkworks> no one runs both for that long.
[23:08:59] <Jymmm> skunkworks: well you said you were able to repeat it iirc
[23:08:59] <cradek> you were looking for it
[23:09:06] <Jymmm> oh, heh...
[23:09:16] <skunkworks> It took about 1/2 hour for mine
[23:09:29] <Jymmm> skunkworks: almost EXACTLY 30 minutes on the nose too
[23:09:36] <skunkworks> but normally I would just run the latency test by itself
[23:10:07] <Jymmm> well, alex_joni suggested to test the system hard RT Test + glxgears to see if the HW had latency issues.
[23:10:08] <skunkworks> anyone remember the command to redetect a video card? xconfig...
[23:10:27] <skunkworks> ah - it is alex's fault ;)
[23:10:29] <Jymmm> sudo dpkg-reconfigure xserver-xorg
[23:10:34] <skunkworks> thanks
[23:10:56] <Jymmm> skunkworks: the only way I know that is because someone told me =)
[23:11:04] <Jymmm> and I wrote it down =)
[23:11:15] <Jymmm> DCC on this crt don't work
[23:11:34] <Jymmm> ok, getting ready for work. bbiab
[23:14:36] <alex_joni> good night all
[23:17:28] <skunkworks> night alex
[23:21:20] <cradek> https://bugs.launchpad.net/ubuntu/+source/xorg-server/+bug/109784
[23:25:00] <skunkworks> with an ati rage 128 agp - I get 35fps full screen :)
[23:26:22] <skunkworks> Cradek: nice work.
[23:27:26] <cradek> I can't quite figure out if I have everything right in the report - it's kind of confusing
[23:27:51] <chr0n1c> anybody want gcode for a nascar logo? ...about 4 inches long by 5/8" tall
[23:27:56] <cradek> there's a place to link to another tracker, but it wants to know that tracker's distribution, and freedesktop is used by many of them
[23:29:24] <chr0n1c> so you guys solved jymmmm's prob?
[23:31:37] <Jymmm> they FOUND it at least
[23:31:56] <skunkworks> cradek: I am running a 20us period and axis runs pretty smooth. ;)
[23:32:12] <eric_u> isn't there a Russian site with all sorts of gcode for logos?
[23:32:27] <Jymmm> Im not sure whats up with the almost EXACTYL 30 mnute timeframe though
[23:34:19] <chr0n1c> maybe something clogs up after that many calcs
[23:34:30] <chr0n1c> a memory leak?
[23:34:31] <Jymmm> oh wth... I can't even run sim axis
[23:35:14] <chr0n1c> i never saw the russiansite
[23:35:20] <chr0n1c> russian site*
[23:35:43] <chr0n1c> but i have done a ford logo... anda nascar logo so far
[23:35:54] <chr0n1c> the ford logo is getting turned into a steel die
[23:36:01] <chr0n1c> for stamping stainless
[23:36:25] <skunkworks> Jymmm: What does it do? you can't have the latency test running and also run emc.
[23:36:26] <Jymmm> Just remember that logos are copyright/tradmarked
[23:36:36] <chr0n1c> yeah
[23:36:37] <Jymmm> skunkworks: I dind't have the test running
[23:36:54] <chr0n1c> it's more jsut to show off my programming skillz
[23:37:05] <Jymmm> like my HD stuff
[23:37:05] <chr0n1c> but the ford is for a reproduction parts company
[23:38:09] <chr0n1c> i cut the nascar thing into the flat side of a dvd case
[23:38:20] <chr0n1c> i didn't realize they were 1/16th thick
[23:38:26] <chr0n1c> i was guessing 1/32
[23:38:38] <chr0n1c> so it didn't go all the way through and looks like booboo
[23:38:57] <chr0n1c> but if it wa slexan with a .020 endmill...
[23:39:14] <skunkworks> picture?
[23:39:25] <chr0n1c> yeah...i took a movie too
[23:39:31] <chr0n1c> most likely the movie sux
[23:39:35] <skunkworks> well?
[23:39:37] <chr0n1c> i'm loadingthem onmy puter now
[23:39:38] <skunkworks> ah
[23:39:40] <skunkworks> :)
[23:40:20] <Jymmm> ok, a reboot recovered and running axis sim now
[23:40:36] <Jymmm> actually running whatever gcode it loads with right now
[23:41:09] <skunkworks> Jymmm: If x restarted - it is still running the latency test.
[23:41:24] <skunkworks> it was sill running
[23:41:26] <chr0n1c> lol, they were amazed i actually vut the axis splash code.... lol
[23:41:27] <skunkworks> still
[23:41:38] <Jymmm> skunkworks: no shit? In the bkgnd?
[23:41:48] <skunkworks> yes
[23:41:50] <chr0n1c> cut*
[23:42:11] <Jymmm> skunkworks: wasn't aware of that since I started it from a term within X
[23:42:25] <skunkworks> right
[23:42:40] <chr0n1c> the dishsoap and water with 70/30 more dishsoap... is awesome cutting fluid for plastics
[23:42:45] <chr0n1c> i haven't tried steel yet
[23:43:08] <chr0n1c> thanks for that suggestion, i forget who it was
[23:43:41] <skunkworks> we have used windex in a pinch ;)
[23:43:42] <ds3> is rust a problem ?
[23:43:54] <chr0n1c> with the soap?
[23:43:57] <ds3> yeah
[23:44:00] <chr0n1c> i dunno...
[23:44:04] <chr0n1c> not on plastic
[23:44:08] <ds3> are your ways rusted?
[23:44:10] <ds3> =)
[23:44:12] <chr0n1c> but my slides are all alum
[23:44:16] <chr0n1c> and chrome ballscrews
[23:44:19] <ds3> oh
[23:44:33] <eric_u> will the dishsoap work on acrylic?
[23:44:39] <chr0n1c> sooo.. maybe on a different machine
[23:44:42] <chr0n1c> yeah
[23:44:46] <chr0n1c> i used it on the dvd case
[23:44:49] <chr0n1c> and on lexan
[23:44:56] <ds3> the standard blue stuff is great on plastics
[23:45:02] <eric_u> I occasionally cut ultrasonic wedges out of acrylic
[23:45:05] <chr0n1c> just make sure the cutter gets lubed with it as it cuts
[23:45:24] <chr0n1c> ultrasonic wedges?
[23:45:26] <ds3> chr0n1c: are you setup for flood?
[23:45:57] <chr0n1c> i am using an old welch's grape 20 oz bottle with a hole drilled in the lid and acid brush
[23:46:09] <chr0n1c> an*
[23:46:24] <chr0n1c> and a rag to catch the runoff
[23:46:33] <ds3> how fast of a spindle are you running?
[23:46:44] <chr0n1c> 8000 - 32, 000
[23:46:57] <chr0n1c> i usually keep it on 8000
[23:47:01] <ds3> and what kind of rain coat do you use? =)
[23:47:01] <chr0n1c> cuz stuff melts!
[23:47:05] <chr0n1c> hah
[23:47:21] <chr0n1c> if you mix it thick.. it stays put
[23:47:31] <chr0n1c> makes tiny bubbles and holds the chips
[23:47:41] <ds3> Oh I see
[23:48:16] <toastydeath> is this an engraving cutter
[23:48:28] <eric_u> wedges: http://www.olympusndt.com/en/panametrics-ndt-ultrasonic/cds_wedges/
[23:49:35] <toastydeath> also i know i sound like a broken record
[23:49:56] <toastydeath> but acrylic is one of those plastics that takes to kerosene
[23:50:09] <chr0n1c> it's a 1/8 carbide center drill
[23:50:14] <chr0n1c> i scarfed it from work
[23:50:49] <chr0n1c> the closest i got to flammable cutting fluid was wd-40
[23:51:05] <chr0n1c> i don't wanna burn to death with chips in my eye from an explosion
[23:51:06] <ds3> wd40 is mostly kero
[23:51:10] <toastydeath> flood kerosene keeps the acrylic very cool and the cutter very lubricated
[23:51:24] <chr0n1c> someone told me to use alcohol on glass
[23:51:30] <chr0n1c> but that sounds like a fire
[23:51:31] <toastydeath> you don't cut glass
[23:51:44] <chr0n1c> with a diamond bit...
[23:51:48] <toastydeath> that's abrasion
[23:51:53] <toastydeath> very different
[23:51:55] <chr0n1c> well... still cutting
[23:52:00] <toastydeath> no, it's not
[23:52:06] <ds3> Mmm diamond dust + way oil
[23:52:11] <chr0n1c> i have a few ball diamond bits that came with the dremel... i need to try some glass
[23:52:46] <chr0n1c> i should maybe get my table 100% flat first
[23:52:47] <toastydeath> you can cut glass and other ceramics with a ductile cutting setup
[23:53:21] <chr0n1c> nascar pics are uploading to the webserver now
[23:53:21] <ds3> your next mission is to cut tempered glass, if you so choose to accept it
[23:53:23] <ds3> =)
[23:53:42] <chr0n1c> this irc message will self destructo in 30 secondos
[23:55:58] <Jymmmmmm> cradek nice writeup, hope they read/act upon it