#emc | Logs for 2006-05-24

[00:02:49] <dmessier> Robin_sz... i agree.... burn the ol high end stuff... its dangeous...
[00:05:07] <dmessier> have a party,,, fluff and burn... to some old sony boy williamson
[00:36:11] <fenn> ah.. i feel much better after exercising my consumer prerogative
[01:26:03] <Jymmm> lo robin
[01:29:46] <Jymmm> Gawd this is lame ---> http://www.diynetwork.com/diy/ld_other/article/0,2041,DIY_14021_2274369,00.html
[01:32:56] <fenn> shut up and buy something jymmm
[02:00:10] <jepler> cradek: you mentioned measuring how many segments per second were delivered onto the queue. How did you measure that?
[02:00:31] <jmkasunich> he was watching the queue fill up with halscope
[02:01:28] <jepler> aha
[03:15:50] <Jymmm> * Jymmm whacks fenn up side the wallet!
[03:36:16] <cradek> jepler: hal_whatever->debug_float_0 = tpqLen(something);
[03:37:30] <cradek> fwiw http://timeguy.com/cradek-files/emc/qlen.diff
[03:37:57] <jmkasunich> you got everything put away now?
[03:38:03] <cradek> not really
[03:38:23] <jmkasunich> same here
[03:38:25] <cradek> while I had it all packed up I decided I should unpack it in the optimal place
[03:38:32] <cradek> that requires moving other things
[03:38:43] <jmkasunich> I'm trying to do some cleanup too
[03:38:49] <cradek> the previous arrangement grew organically and was due to no actual thought
[03:38:51] <jmkasunich> just put one KVM switch on ebay
[03:39:14] <Jymmm> jmkasunich have any network KVM's ?
[03:39:16] <jmkasunich> getting ready to test another
[03:39:32] <jmkasunich> you mean like TCP/IP ones? no, sorry
[03:39:43] <Jymmm> jmkasunich Yeah, okey dokey.
[03:39:59] <cradek> goodnight, I'm off
[03:40:05] <jmkasunich> got a couple 4-way, one two way, and one 16 way
[03:40:11] <Jymmm> I have WAY TOO MANY cables under my desk.
[03:40:26] <Jymmm> jmkasunich got the cables too?
[03:40:32] <jmkasunich> some
[03:40:36] <Jymmm> on the 16
[03:40:42] <jmkasunich> 6 cables for that one
[03:41:48] <Jymmm> bah... sorry for asking... I don't need any more shit around here =)
[03:41:59] <jmkasunich> darn
[03:42:13] <fenn> jymmm whats wrong with x forwarding?
[03:42:20] <Jymmm> I have my 4ch kvm, then my backup 2ch kvm
[03:42:35] <Jymmm> fenn nothin, if you use X
[03:42:49] <fenn> vnc then
[03:42:59] <jmkasunich> kvm is handy at times
[03:43:04] <jmkasunich> for doing installs for instance
[03:43:07] <Jymmm> no DOS vnc
[03:43:11] <jmkasunich> messing with bios stuff
[03:43:15] <Jymmm> no bios
[03:43:16] <fenn> dos! pff
[03:43:27] <fenn> isnt there sshd for dos?
[03:43:33] <jmkasunich> dos, and windows if you don't want to fsck around with vnc
[03:43:57] <Jymmm> I use synergy alot
[03:44:07] <Jymmm> I use synergy2.sf.net alot
[03:44:26] <Jymmm> I love that you can copy n paste between machiens
[03:44:34] <fenn> thats cool
[03:45:48] <Jymmm> so between synergy and my KVM, I got it pretty much down.
[03:48:46] <Jymmm> jmkasunich so, what else you got for sale?
[03:49:12] <jmkasunich> right now that it, gotta dig deeper
[03:49:32] <Jymmm> jmkasunich mostly computer stuff or any tools/tooling too?
[03:49:40] <jmkasunich> mostly computer
[03:49:57] <Jymmm> trackballs?
[03:50:07] <jmkasunich> no
[03:51:01] <jmkasunich> really the main thing I want to get rid of is the KVMs
[03:51:14] <Jymmm> Ok, not much I'd be interested in computer wise. Years ago I got rid of 4 pallets of computer stuff each 4' tall
[03:51:16] <jmkasunich> between the switches and the cables I have several largish boxes
[03:53:22] <jmkasunich> got a lot of other stuff, but I'm too disorganized to even know what I have
[03:53:44] <Jymmm> I have gops of crap left... bunch of Zip drives/disks, Jaz drives/disks, scsi cards/cables, and lots of various cables.
[03:54:10] <Jymmm> all work too
[03:54:20] <Jymmm> * Jymmm takes good care of his stuff.
[03:54:24] <jmkasunich> similar here
[03:54:54] <Jymmm> Hell, my 1mb Palm Pilot Pro has zero scratches on it.
[03:55:30] <jmkasunich> most of my stuff didn't start out as my stuff
[03:55:46] <jmkasunich> someone else put the scratches on it
[03:56:05] <Jymmm> I'm in Silicon Valley, we have lots of toys during the dot com days.
[03:59:05] <jmkasunich> drat...
[03:59:27] <jmkasunich> I don't seem to have a power supply for the 16-way KVM
[03:59:58] <Jymmm> RS has em, just sell it w/o it.
[04:00:28] <jmkasunich> in wants 17VAC center tapped
[04:00:39] <jmkasunich> on a mini-din connector
[04:00:55] <Jymmm> wow, that's old =)
[04:01:12] <jmkasunich> I have power supplies that I picked out of the same dumpster, same time, with matching connector
[04:01:26] <jmkasunich> but they are +/-12 and 5, not AC
[04:01:49] <Jymmm> try +12vdc
[04:01:58] <Jymmm> you knwo internally it's DC
[04:02:15] <Jymmm> sometimes the AC ones were cheaper
[04:02:46] <jmkasunich> internally it has a diode bridge, makes +/-12 (approx)
[04:02:54] <jmkasunich> and I assume +5 from that
[04:03:06] <jmkasunich> wonder if the +/-12 would work
[04:03:11] <Jymmm> there ya go... it should
[04:03:18] <Jymmm> and the polarity won't matter
[04:04:11] <jmkasunich> gotta take the switch apart again and check pinout
[04:04:16] <Jymmm> heh, that's about the only thing I save anymore... walworts.... they are jsut too expensive these days.
[04:04:34] <Jymmm> used to be cheap a long time ago.
[04:05:23] <jmkasunich> I hate wall-warts
[04:05:40] <jmkasunich> I'm glad my current 4-way KVM has AC input
[04:05:54] <jmkasunich> (120V AC that is)
[04:06:07] <Jymmm> Huh? My kvm doesn't have any external power
[04:06:14] <fenn> hopefully everything will be usb soon so we can scrap the wallwarts
[04:06:25] <Jymmm> fuck USB damn POS
[04:06:32] <jmkasunich> what he said
[04:06:39] <fenn> yep i know, but at least its all the same
[04:07:10] <Jymmm> My KVM is PS/2 so that might make a difference
[04:07:32] <jmkasunich> my 2-way KVM leeched power from the PC(s)
[04:07:39] <jmkasunich> but the 4-way has its own cord
[04:07:46] <Jymmm> bummer
[04:07:50] <jmkasunich> its a fairly fancy one, with onscreen display, etc
[04:07:55] <Jymmm> ah, ok
[04:08:43] <Jymmm> I just tap <LEFT_CTRL>+1or2or3or4 to switch to whatever machien
[04:09:10] <Jymmm> it's got other funky commands I never use like rotate between machines
[04:09:27] <jmkasunich> http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=9731348795
[04:09:42] <jmkasunich> thats the one I'm selling, my main one I use is almost identical
[04:10:30] <Jymmm> $13 shipping?!
[04:10:42] <Jymmm> =)
[04:11:01] <jmkasunich> UPS ground, 8 lbs
[04:11:13] <jmkasunich> I don't believe in dicking people over on shipping
[04:11:15] <Jymmm> whats the width?
[04:12:09] <jmkasunich> 10.5" wide 9" deep 1.75" hi
[04:12:13] <Jymmm> http://shop.usps.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=10152&storeId=10001&categoryId=13359&productId=11697&langId=-1
[04:12:26] <jmkasunich> plus the cables of course, they take a lot more space
[04:12:30] <Jymmm> oh damn almost will fit.
[04:12:52] <Jymmm> USPS has flat rate now... up to 75 lbs for $9
[04:13:08] <Jymmm> they have a couple different size boxes too
[04:13:12] <jmkasunich> hmm, maybe have to make that an option
[04:13:25] <Jymmm> jmkasunich you can't, it won't fit in the box
[04:13:41] <jmkasunich> not for this item
[04:13:42] <Jymmm> you MUST use usps flat rate boxes to get the flat rate pricing
[04:14:06] <Jymmm> you said 9" wide, the box is 8.5"
[04:14:29] <fenn> i got a flat-rate box completely packed with aluminum blocks
[04:14:36] <fenn> it was sorta funny
[04:14:45] <Jymmm> the OTHER flat rate box is bigger, but shallower
[04:15:18] <Jymmm> fenn yeah, it's a great deal of dense stuff
[04:15:33] <Jymmm> if you can fit it in the boxes that is.
[04:15:40] <Jymmm> ceramic tile, etc
[04:16:30] <jmkasunich> fsckit... looks like this power supply won't work
[04:16:39] <jmkasunich> pinout doesn't match
[04:16:43] <jmkasunich> that could be fixed I guess
[04:16:59] <jmkasunich> but its also only rated at 0.5A on the +/- 12V
[04:17:10] <jmkasunich> the switch looks like it needs more than that
[04:19:55] <jmkasunich> tricky trash... they throw out a power supply in the same place as the switch, and the connector physically fits the switch, but its actually a completely unrelated item... and the real power supply is nowhere to be found
[04:24:05] <Jymmm> jmkasunich GND should be common to the case, so just a matter of finding +v
[04:24:15] <jmkasunich> I figured out the pinout
[04:24:21] <Jymmm> =)
[04:24:21] <jmkasunich> theres gnd, +, and -
[04:24:27] <Jymmm> ewwwwww
[04:24:34] <jmkasunich> but they don't match the supply I have
[04:24:47] <jmkasunich> and I suspect it draws more than 1/2 amp
[04:25:16] <Jymmm> that sucks
[04:25:52] <jmkasunich> yep
[04:26:44] <jmkasunich> I can either sell it as untested with a starting bid of $1 (and maybe get no buyers anyway) or put it back on the shelf until I feel more like rigging up something to test it with
[04:28:44] <Jymmm> jmkasunich: Sell it now... I'm looking at the closed auction on ebay and most 3-4 port KVM's are goin for ~$10
[04:29:11] <Jymmm> unless it's SUN/MAC
[04:29:33] <jmkasunich> this is a 16 port with 6 cables
[04:29:47] <jmkasunich> hopefully if working it would be a bit more than $10
[04:30:19] <Jymmm> looking
[04:30:36] <Jymmm> 16p over tcpip $450
[04:30:50] <Jymmm> 16p $188
[04:31:13] <Jymmm> you have 16 cables for it?
[04:31:28] <jmkasunich> this is a 16 port with 6 cables
[04:32:25] <Jymmm> Ok, get the PS goin for it =)
[04:33:05] <A-L-P-H-A> jmkasunich, what's this?
[04:33:14] <Jymmm> 16 port KVM A-L-P-H-A
[04:33:18] <A-L-P-H-A> wow.
[04:33:26] <A-L-P-H-A> i don't think i have 16 computers...
[04:33:56] <A-L-P-H-A> EMC, workstation, laptop, spare, 2, 1, 3.... I only have 9 computers in the house.
[04:34:50] <A-L-P-H-A> I would have absolutely no need for a 16 port kvm.
[04:35:00] <jmkasunich> most people don't
[04:35:28] <jmkasunich> I have a 4, never used more than three ports tho
[04:35:30] <A-L-P-H-A> honestly... why would anyone? there's always SSH... VNC...
[04:35:39] <jmkasunich> but if you have a server room full of doze boxes
[04:36:06] <A-L-P-H-A> my workstation and laptop are the only things that get turned on... others are used by others... I have a router... so why would I need a server persay.
[04:36:56] <jmkasunich> _you_ don't
[04:37:22] <jmkasunich> the big KVMs are usually used in a commercial setting
[04:38:21] <A-L-P-H-A> I understand, a server setting.
[04:38:30] <A-L-P-H-A> but even then... SSH + VNC should do.
[04:38:51] <jmkasunich> sometimes
[04:39:01] <Jymmm> A-L-P-H-A install an OS via VNC =)
[04:39:02] <A-L-P-H-A> only time I see a KVM being critical, is in a rendering farm...
[04:39:17] <Jymmm> A-L-P-H-A upgrade a BIOS via vnc =)
[04:39:20] <A-L-P-H-A> Jymmm, it's call enterprise pushes.
[04:39:33] <Jymmm> A-L-P-H-A it's called KISS
[04:39:46] <A-L-P-H-A> Jymmm, then plug in a screen at the server.
[04:40:26] <jmkasunich> A-L-P-H-A: somebody must use em, BlackBox asks $5K for some of the big ones
[04:40:41] <A-L-P-H-A> jmkasunich, Jesus.
[04:40:59] <A-L-P-H-A> 16 port you say?
[04:41:09] <jmkasunich> yeah
[04:41:10] <A-L-P-H-A> Hmm... doesn't ATMEL do a KVM thing??
[04:41:21] <jmkasunich> the $5K ones are 16 computers and 4 users
[04:41:27] <jmkasunich> any user can access any computer
[04:41:33] <A-L-P-H-A> INTERESTING.
[04:41:48] <jmkasunich> a 16 computer to 1 user one is less, maybe $1000-1500
[04:41:51] <A-L-P-H-A> wait wait...
[04:41:58] <A-L-P-H-A> can 4 users use one computer?
[04:42:04] <jmkasunich> dunno
[04:42:20] <jmkasunich> probably only one user at a time per computer
[04:42:22] <Jymmm> oh gawd.... myspace.com/burgerking
[04:42:27] <Jymmm> just saw it on tv
[04:42:37] <Jymmm> the world *IS* coming to an end
[04:42:49] <A-L-P-H-A> Cause 4 users using 16 different systems, that's not hard... I could do make one... but simultaneous usage of one computer, that's more tricky to me.
[04:43:13] <jmkasunich> I actually had one of the 16x4 ones
[04:43:17] <jmkasunich> sold it to cradek'
[04:43:29] <jmkasunich> sold it to cradek's company (he's a sysadmin)
[04:43:39] <jmkasunich> he has one console in the server room and another in his office
[04:43:45] <jmkasunich> and can access any system from either place
[04:44:05] <jmkasunich> (they paid me a heck of a lot less than $5K for it)
[04:44:34] <A-L-P-H-A> jmkasunich, so you got ripped.
[04:44:35] <A-L-P-H-A> keke
[04:44:46] <jmkasunich> no, because I got it from the dumpster
[04:44:54] <jmkasunich> so both cradek and I got a great deal
[04:44:59] <A-L-P-H-A> did you get $1K+ for it?? :)
[04:45:13] <jmkasunich> bout half that
[04:47:27] <A-L-P-H-A> you know, it wouldn't be hard at all... how many pins are used in the KB? how many pins are used on VGA video?
[04:47:48] <jmkasunich> video switching is non-trivial because of the bandwidth
[04:48:06] <A-L-P-H-A> PIC/atmel catures/relays KB signals... looks for the toggle stroke to increment up/down/#.
[04:48:27] <A-L-P-H-A> jmkasunich, should be a atmel chip that does it... or does it well enough.
[04:48:32] <jmkasunich> keyboard/mouse switching is non-trivial because you have to fake out the computers that don't have the kb/mouse (some will lock up if they think the mouse or kb went away)
[04:49:04] <jmkasunich> I agree that $5K is rediculously high priced tho
[04:49:23] <A-L-P-H-A> jmkasunich, that's actually not to hard... as long as you maintain a power signal, you're fine.
[04:49:46] <jmkasunich> thats not my understanding, but who knows...
[04:49:53] <A-L-P-H-A> jmkasunich, you just don't send data back/forth btwn compy + kvm box.
[04:50:03] <A-L-P-H-A> jmkasunich, I actually believe it is that. USB would be different.
[04:51:10] <Jymmm> USB sucks, damn POS (echo echo)
[04:51:29] <A-L-P-H-A> i love usb.
[04:51:33] <A-L-P-H-A> it's so freak'n convenient.
[04:52:02] <Jymmm> yeah, you try to get real time data using USB
[04:52:21] <Jymmm> I'll stick with my UARTS
[04:52:44] <A-L-P-H-A> hmm... that I don't know about... there are USB to Serial. USB to Parallel stuff... wonder how that stuff would do.
[04:52:54] <A-L-P-H-A> both are $20.00 CDN at local computer store.
[04:52:55] <Jymmm> USB2SERIAL don't work
[04:53:09] <A-L-P-H-A> how does it not work?
[04:53:17] <Jymmm> RT
[04:53:18] <jmkasunich> they work for some things
[04:53:20] <A-L-P-H-A> oh.
[04:53:22] <jmkasunich> not for RT
[04:53:37] <jmkasunich> same with USB->parallel
[04:53:39] <A-L-P-H-A> I never tried RT stuff with USB<-> (serial/parallel)
[04:53:51] <jmkasunich> it will run a printer just fine, but you can't run steppers with it
[04:54:05] <Jymmm> And ironically they would only work on one version of an OS, and not the next (could be different now though)
[04:54:06] <A-L-P-H-A> jmkasunich, have you played with the G101, or G100 from gecko?
[04:54:16] <jmkasunich> not yet
[04:54:20] <jmkasunich> I actually have one
[04:54:24] <A-L-P-H-A> who's played with it?
[04:54:26] <Jymmm> PCMCIA to scsi/serial/parallel are fine though.
[04:54:47] <A-L-P-H-A> jmkasunich, I'm looking for one... cause I want to make my system work.
[04:55:01] <A-L-P-H-A> I'll have to do some tinkering tomorrow... I'll get my hands dirty with the electronics again.
[04:56:29] <A-L-P-H-A> wow. torrent transfer at 280 to 400 kBps... love the modem.
[04:57:46] <Jymmm> jmkasunich oh, how was the fest?
[04:58:09] <jmkasunich> nice
[04:58:14] <jmkasunich> very very busy
[04:58:27] <Jymmm> doing?
[04:58:48] <jmkasunich> coding, talking to folks, working on the mazak, etc etc
[04:59:02] <Jymmm> ah. mazak?
[04:59:13] <jmkasunich> got the toolchanger working (stayed up all night - couldn't concentrate until everybody else went away
[04:59:25] <Jymmm> I know that one.
[05:00:53] <jmkasunich> I think somebody (SWP?) made a movie of it changing tools
[05:01:12] <jmkasunich> we also cut metal on it for the first time
[05:01:16] <A-L-P-H-A> I miss SWP.
[05:01:25] <A-L-P-H-A> I want to talk to him... about the G100/G101
[05:01:26] <jmkasunich> (although it was capable of doing that back in october)
[05:01:53] <Jymmm> mazak is what, a cnc center?
[05:02:05] <jmkasunich> a big vertical mill
[05:02:13] <Jymmm> with a tool changer?
[05:02:19] <jmkasunich> yep
[05:02:23] <jmkasunich> 24 tools
[05:02:25] <Jymmm> wow, MUST be big
[05:02:56] <Jymmm> it kinda sucks that I can't do metal... no cooling and not waterproof
[05:03:33] <jmkasunich> this is from last year: http://www.cardinaleng.com/emc_mazak_project.htm
[05:04:29] <Jymmm> shit... 10,000 lbs... might as well be a cnc center
[05:08:53] <jmkasunich> yeah, its a beefy machine
[05:09:03] <jmkasunich> the toolchanger is all hydraulic
[05:11:01] <jmkasunich> fenn has a lot more pics of the project
[05:11:08] <jmkasunich> trying to find a url
[05:11:39] <Jymmm> dont sweat it, was just trying to get an idea of what a mazak is/was =)
[05:13:15] <fenn> fenn.freeshell.org/retrofest
[05:13:39] <fenn> ghghgh
[05:14:04] <fenn> my daily 10mb quota has been exceeded
[05:15:11] <jmkasunich> bedtime for me
[05:15:12] <A-L-P-H-A> fenn, flickr!
[05:15:32] <A-L-P-H-A> why can't people just use flickr? or someone elses picture hosting.
[05:15:36] <Jymmm> fenn: flickr.com free or use your yahoo id to login.
[05:16:09] <fenn> poop on flicky
[05:16:22] <A-L-P-H-A> flickr is cool.
[05:16:31] <fenn> i dont like it
[05:17:02] <A-L-P-H-A> flickr is pretty cheap if you think about it... 2gig upload/month while... unlimited views... <$50 for two years.
[05:17:34] <fenn> how hard is it to get a plain ugly shell account with apache
[05:17:48] <fenn> * fenn sighs
[05:18:06] <fenn> i better go to bed before anything else randomly breaks
[05:19:43] <fenn> too late
[05:19:53] <fenn> * fenn cries
[05:20:13] <A-L-P-H-A> there there...
[05:20:19] <A-L-P-H-A> we'll take the server to the farm, and put it down.
[05:22:22] <fenn> tomorrow, when i wake up, everything will be magically working again
[05:35:30] <alex_joni> alex_joni has changed the topic to: Welcome to the Enhanced Machine Control forum - a linux based CNC control. | EMC2.0.0 is out .. | Home: www.linuxcnc.org | wiki up @ wiki.linuxcnc.org | EMC usage map: www.frappr.com/emc2/
[05:42:22] <alex_joni> The frappr map has been moved to an easier to remember name: www.frappr.com/emc2
[05:42:32] <alex_joni> unfortunately frappr.com/emc was already taken..
[05:43:04] <Jymmm> ANYTHING is better than what it was.
[05:45:34] <fenn> whoever wanted to see the mazak pictures: http://fenn.dyndns.org/pub/emc/retrofest/default.html
[05:46:37] <fenn> and to all, a good night!
[05:46:50] <alex_joni> bye all
[05:46:58] <Jymmm> g'night alex_joni
[10:50:14] <chinamill> Hello giacus
[10:50:15] <chinamill> How is your robot coming along?
[10:53:04] <giacus> hello chinamill, waiting the bluetooth board to unlock it by the pc with the software
[10:53:59] <giacus> it don't recognize the biometric password.. won't accept any vocal command
[10:54:38] <chinamill> yaeh, you have make it behave
[10:55:55] <giacus> uploading the video of sound follower test ..
[11:01:41] <giacus> http://www.giacus.org/files/mov01491.mpg
[11:02:12] <giacus> But I have to resize it, and add some titles yet ..
[11:07:44] <giacus> I'll reply you only if I recognize the passord, say the passwd:
[11:07:53] <giacus> linux
[11:08:12] <giacus> password incorrect
[11:08:17] <giacus> linus
[11:08:19] <giacus> password incorrect
[11:08:25] <giacus> lainux
[11:08:31] <giacus> you don't know the passwd
[11:08:33] <giacus> :(
[11:11:50] <giacus> so, the only way to reset the passwd is by the BT software
[11:15:40] <chinamill> almost ready?
[11:17:13] <giacus> I should receive the module soon
[11:17:17] <giacus> some week
[11:58:33] <giacus> found the reason of gcc segfaults compiling the kernel on my cely 800 ..
[11:59:04] <giacus> was a ram issue: 128 + 128 bank ):
[11:59:25] <giacus> using only one it work fine
[11:59:51] <giacus> Kernel: arch/i386/boot/bzImage is ready
[11:59:55] <giacus> darn ram
[12:01:37] <giacus> weird, the 2 bank of ram where working fine at all, except for compiling the kernel :/
[12:02:41] <anonimasu> hm, cant you tell the kernel to avoid a bit of ram ;)
[12:03:09] <giacus> wonder if it depend on gcc vers. used
[12:06:35] <anonimasu> probably not
[12:06:39] <anonimasu> gcc is heavy on memory usage..
[12:10:49] <giacus> it trot, until now got no segfault
[12:11:24] <giacus> maybe the 2 banks are a bit different for latency time
[12:11:41] <giacus> but they where working for all the rest .. strange
[12:15:23] <giacus> its a pleasure to see the kernel compiling without segfaults :P
[12:17:36] <giacus> anonimasu: what processor you have and choosed in make menuconfig for the 2.6.16 ?
[12:18:03] <giacus> I seen rtai don't accept 386 right ?
[12:18:13] <giacus> rtai patches
[12:19:46] <giacus> at least, rtai from cvs ..
[12:25:46] <giacus> * giacus play Stadio - Sorprendimi.mp3
[12:27:40] <CIA-8> 03jepler 07HEAD * 10emc2/src/emc/task/emctaskmain.cc:
[12:27:40] <CIA-8> when the interp_list is less than 2/3 full and EMC_TASK_INTERP_READING, run
[12:27:40] <CIA-8> through emcTaskPlanRead and emcTaskPlanExecute up to 1000 times without
[12:27:40] <CIA-8> going back to main to wait. With a non-pathological program, this means
[12:27:40] <CIA-8> the interp_list fills to 667 elements right away and stays about that full
[12:27:40] <CIA-8> through the entire run. With a pathological program (an infinite O- while
[12:27:44] <CIA-8> loop) it's still abortable.
[12:28:52] <skunkworks> are you talking about me again?
[12:36:19] <jepler> mmmmmaybe
[12:36:26] <skunkworks> :)
[12:40:54] <giacus> is mkinitrd /boot/initrd-2.6.10.img 2.6.10 correct at http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?RtaiSteps ?
[12:41:18] <giacus> debian work with mkinitrd -o
[12:41:48] <giacus> maybe it work for Ubuntu
[12:41:53] <giacus> ?
[12:45:25] <giacus> * giacus play Simply_red -track04.mp3
[12:47:34] <cradek> wow, that looks like a much better fix than mine.
[12:50:17] <giacus> ;P
[12:51:18] <jepler> cradek: that keeps interp_list full, but tcqLen() still increases slowly
[12:51:23] <jepler> if at all
[12:51:36] <jepler> too .. many .. levels
[12:57:35] <skunkworks> I have a 117k 2d file made up of short line segments - if you still want it. i thought I had some bigger ones.
[12:58:19] <jepler> sure, I'll take it
[12:58:22] <jepler> that's 117k or 117klines?
[12:58:42] <skunkworks> 117k in size
[12:59:32] <jepler> how do you want to send it to me? e-mail would be fine.
[13:00:00] <skunkworks> I can do that - let me look for a few more also. what is your email?
[13:00:06] <jepler> jepler@unpy.net
[13:00:20] <skunkworks> give me about 10 min
[13:00:38] <jepler> I won't have a chance to look at them before this evening anyway .. I'm late for work as it is
[13:01:06] <skunkworks> that would be good - I will also ask the laser department.
[14:02:50] <giacus> http://www.beppegrillo.it/eng/2006/05/911_without_truth.html
[15:48:22] <jepler> tc.maxvel = tp->ini_maxvel * ACCEL_USAGE;
[15:49:10] <cradek> jmk was offended by my magic number, so he made it a define ... poorly
[15:49:23] <jepler> I guess so
[15:49:28] <jepler> how come maxvel is decrased by ACCEL_USAGE?
[15:49:46] <cradek> it's my "solution" for the stepgen headroom problem
[15:50:04] <cradek> but until I can convince jmk to do the same for jogs, it's useless
[16:00:50] <Jymmm> ~~~~~ Ah, finally found it... http://sawdustmaking.com/Sheet%20Rack/rack.htm
[16:03:17] <giacus> hello Jymmm
[16:03:41] <giacus> hey guys, I compiled a new kernel, how you see the rtai test ? http://rafb.net/paste/results/3htlPG23.html
[16:03:53] <giacus> houls I compare it somewhere ?
[16:03:57] <giacus> should*
[16:04:41] <Jymmm> overall max 20,000 or less
[16:05:16] <Jymmm> so you're doing pretty good giacus
[16:05:29] <giacus> ok, thanks
[16:05:44] <Jymmm> If you want to do better, disable secondary IDE, unused serial ports, etc in bios
[16:05:50] <jepler> make sure you do things like switch windows in X, cause disk activity, etc
[16:05:57] <jepler> to see if any of those cause latency
[16:06:05] <giacus> K
[16:10:14] <cradek> giacus: it's good to let the test run while you surf and do other normal things
[16:10:31] <giacus> ok, running X now ..
[16:10:56] <cradek> also be sure to run some opengl programs if you are using an accelerated driver/card
[16:11:11] <giacus> I used the patches from magma cvs on 2.6.16
[16:11:33] <giacus> video card is nvidia tnt 16 mb, let's see..
[16:11:57] <cradek> you will see overruns if you are using the nvidia closed-source driver
[16:13:26] <giacus> ok, testing it
[16:14:21] <Jymmmmm> Jymmmmm is now known as Jymmm
[16:16:19] <Jymmm> with such the popularity of nvidia, I always thought it was open source
[16:16:41] <cradek> no they have a binary driver for linux
[16:16:56] <cradek> there is also some open source support, which works fine with rtai
[16:17:19] <Jymmm> ah, ok
[16:18:33] <Jymmm> Oh this sucks... http://absolutelyfreeplans.com/fraud/fraud.html
[16:19:39] <jepler> ha ha ha
[16:19:49] <jepler> seems like ebay sellers get away with murder
[16:20:47] <Jymmm> I wonder if "mercinary for hire" is againest ebay's TOS ?
[16:21:17] <Jymmm> outside the US of course =)
[16:22:21] <skunkworks> did you guys guess that the video card was causing the latency in the mazak?
[16:23:13] <cradek> onboard video has a long history of breaking rtai - it wasn't a hard guess to make
[16:23:18] <skunkworks> "guess" kinda sounds bad - assumed from past experience
[16:23:46] <cradek> also the overruns corresponded to big video activity
[16:23:52] <jepler> g0x0 / g1x1f1 should take 1 minute?
[16:24:01] <skunkworks> yes
[16:24:05] <skunkworks> if in ipm
[16:24:12] <skunkworks> english
[16:24:44] <cradek> unless your machine can't go that fast :-)
[16:25:00] <jepler> it's going faster than that
[16:25:01] <jepler> I wonder why
[16:25:08] <cradek> how much?
[16:25:17] <skunkworks> (guess it doesn't matter if it is in english) it should take 1 minuit either way
[16:25:17] <jepler> I've been working on merging the moves with the velocity and accel messages
[16:25:43] <jepler> when I use ., axis says the speed is about 2.5
[16:26:16] <jepler> please don't tell me I have a units bug
[16:26:20] <cradek> without your changes?
[16:26:27] <jepler> no, with all these changes
[16:27:10] <jepler> er, when I use F.1, axis says the speed is about 2.5
[16:27:19] <jepler> Motion id -5 took 23.816572 seconds.
[16:27:24] <jepler> 1 inch move at should-be-1ipm
[16:27:30] <jepler> 1 inch move at should-be.-1ipm
[16:27:33] <jepler> argh
[16:29:44] <alex_joni> hi guys
[16:29:54] <cradek> hi alex
[16:30:14] <skunkworks> Hi alex
[16:30:43] <alex_joni> boring day today ;)
[16:31:51] <skunkworks> did you take pictures of the robot? ;)
[16:32:32] <alex_joni> skunkworks: yeah, I did.. wanna see?
[16:32:46] <skunkworks> sure
[16:33:31] <alex_joni> hang on.. need to download from the cam
[16:39:42] <giacus> hi alex_joni
[16:40:23] <giacus> well, seems ok, running mplayer with a divx, firefox and others, overruns at 0
[16:41:16] <giacus> should be ok
[16:42:48] <giacus> here's the last dmesg http://rafb.net/paste/results/NX8Wz859.html
[16:43:08] <giacus> can't see issues
[16:45:21] <skunkworks> jepler: I sent you 2 files - one is just a 100k or so - (looks like deer) the other is a test out of a image-gcode program. it is 228k - Let me know how it looks in axis. I have not run either of them in emc. the test.ncg file should look pretty neat I would think.
[16:45:46] <alex_joni> skunkworks: http://www.robcon.ro/emc/rIMG_4290.JPG
[16:46:07] <alex_joni> http://www.robcon.ro/emc/rIMG_4291.JPG
[16:46:26] <alex_joni> http://www.robcon.ro/emc/rIMG_4296.JPG
[16:46:38] <skunkworks> neat - is that 4 or 5 joints?
[16:46:48] <alex_joni> http://www.robcon.ro/emc/rIMG_4299.JPG
[16:46:54] <alex_joni> 6 on the bot, 2 on the table
[16:47:01] <alex_joni> so 8 synchronized here
[16:47:10] <alex_joni> the control can do 18 ;)
[16:47:34] <alex_joni> http://www.robcon.ro/emc/rIMG_4303.JPG
[16:47:51] <alex_joni> http://www.robcon.ro/emc/rIMG_4307.JPG
[16:47:58] <skunkworks> nice. is it programmed most of the time or do you teach it?
[16:48:33] <alex_joni> teach
[16:48:54] <alex_joni> but only once, afterwards it#ll do the same thing over and over again :D
[16:49:02] <alex_joni> it's got a pascal like language
[16:49:09] <skunkworks> I would hope so :)
[16:49:30] <alex_joni> depends on the customer.. some have very few parts of the same size/shape
[16:49:36] <alex_joni> so they do lots of programming..
[16:49:45] <skunkworks> we have a sankio robot - it is programed in sankio c
[16:49:54] <alex_joni> coo
[16:51:06] <skunkworks> buzz or something like that
[16:53:29] <alex_joni> later guys..
[17:13:50] <giacus> msgfmt -o po/de_rs274_err.mo po/de_rs274_err.po
[17:13:50] <giacus> make: msgfmt: Command not found
[17:13:50] <giacus> make: *** [po/de_rs274_err.mo] Error 127
[17:13:58] <giacus> mmhhh :-))
[17:15:20] <giacus> emc2 seems don't like to be translated :P
[17:36:32] <giacus_2> Starting emc...
[17:36:32] <giacus_2> /home/jack/emc2/bin/emc_module_helper: Invalid usage with args: insert /usr/realtime/modules/rtai_sched.ko
[17:36:35] <giacus_2> Usage: /home/jack/emc2/bin/emc_module_helper insert /path/to/module.ext [param1=value1 ...]
[17:36:51] <giacus_2> shoud be manully fixed the script I suppose.. ?
[17:37:28] <giacus_2> it seems disagree the sintax
[17:43:02] <giacus_2> or just forgot the module somewhere ..
[17:43:26] <chinamill> giacus_X are you in the other dimension?
[17:43:38] <giacus_2> there's no /rtai_sched.ko
[17:43:50] <giacus_2> hi chinamill :)
[17:46:20] <giacus_2> instead there's rtai_ksched.ko
[17:46:36] <giacus_2> are they brothers ? :D
[17:46:45] <giacus_2> can I use it ?
[17:47:31] <chinamill> cosmic twins!
[17:48:08] <giacus_2> :))
[18:05:02] <jepler> skunkworks: I got the ngc files
[18:05:19] <jepler> skunkworks: bigdeer looks like it needs to be reordered to decrease the distance of G0 moves
[18:05:37] <jepler> skunkworks: and for some reason, test shows the preview plot in a different location than the backplot
[18:09:22] <skunkworks> jepler: g92 issue?
[18:09:41] <jepler> I don't know yet
[18:09:58] <skunkworks> I used to use g92 a lot until emc ;)
[18:11:39] <jepler> I guess different people learned different offset-related tricks
[18:11:44] <jepler> I use G54 exclusively
[18:11:49] <jepler> but that doesn't excuse axis from showing the wrong thing
[18:13:11] <jepler> ugh. "g92 x0 y0 z0" makes the current point be (0,0,0)?
[18:13:21] <jepler> there's no way axis will get that right, without reloading after each jog
[18:13:39] <skunkworks> yes that is now normal fanuc g92 works
[18:14:07] <skunkworks> g92x9y0z0 sets the current posision to that.
[18:16:05] <jepler> I have an internal debate between trying to find out why in the world you'd want to do that inside a ngc file, and just forgetting about it and getting along with my own life.
[18:16:18] <giacus_2> guys, here's the log http://rafb.net/paste/results/5Bv83B54.html
[18:16:37] <giacus_2> it sound to me like an old issue .. but I can't recall
[18:17:03] <skunkworks> that is why I stopped using it and everyone on here calls it the devil. I come from a fanuc controller running a laser. - make the program - jog machine to where you want to start - run program with g92x0y0 at the beginging
[18:17:34] <skunkworks> :)
[18:17:58] <robin__sz> yeuw
[18:17:58] <skunkworks> I have been getting along without it though. It was just a supprise when it didn't work the way I was used to.
[18:18:35] <jepler> skunkworks: do you want me to condescendingly explain to you how I use g54 offsets, or shall we just skip that step?
[18:18:56] <robin__sz> not quite the same thing
[18:19:17] <robin__sz> although, thats how we run our laser
[18:19:49] <robin__sz> put on sheet, use tape measure to locate the offset, enter it as a G54, all programs run G54
[18:20:54] <jepler> you can still use the "jog to" method with g64 (or g92 for that matter) but set the offset before running the program. in tkemc, set the g92 offset by right-clicking the axis position; in axis, set the g54 offset by moving to the desired "0" and pressing shift-home.
[18:21:49] <jepler> it's having the .ngc file contain the instruction to create the offset that baffles me most, g92 vs g54 vs g5x is nearly unrelated
[18:22:05] <skunkworks> I understand - that program though that I sent you - test.ncg - was spit out of a program exactly the way I sent it to you. So I must not be the only one.
[18:22:43] <fenn> the way i understand it, g92 sets the machine<->world offset, g54 sets the part<->machine offset
[18:23:46] <jepler> I'm not sure what you mean by "world"
[18:23:56] <cradek> I was also just thinking that
[18:24:17] <skunkworks> (program was called profiler6)
[18:24:59] <fenn> in emc the "world" is fixed with respect to the machine.. theres some #defines in the code that set the relationship to zero
[18:25:03] <fenn> er crap
[18:25:20] <fenn> in emc the "world" is fixed with respect to the bed of the machine tool..
[18:25:43] <Jymmm> the world is flat damnit!!!
[18:26:21] <cradek> I think the only confusion/disarray caused by g92 is that the g92 offset is cleared at the end of the program (m2).
[18:26:32] <jepler> not cleared .. just .. unapplied
[18:26:44] <fenn> wtf
[18:26:50] <fenn> /emc/motion/command.c:EmcPose worldHome = { {0.0, 0.0, 0.0}
[18:26:58] <fenn> ah the leading slash was giving me problems
[18:27:24] <cradek> jepler: right (like g92.2)
[18:30:01] <cradek> so if you want your offsets unapplied at m2, use g92; otherwise use g54
[18:31:29] <jepler> running a program twice should cut in the exact same place again. it seems like this is harder to achieve with g92 than g54, because of the way it interacts with m2. putting g92x0y0z0 in a program just compounds the problem.
[18:31:45] <cradek> I agree
[18:32:25] <skunkworks> I am learning. I was just used to lining something up - entering g92x0y0 in mdi and then being able to run a program. or having the g92x0y0 at the begining of the program. No biggie - it just happend to be one of the first things I tried in emc - and it suprised me it didn't work the way I was used to.
[18:32:32] <cradek> most people want the offsets they set by right-clicking in the gui to persist, so it's probably a gui bug that it uses g92. if that were fixed, I bet nobody would complain about g92 anymore.
[18:32:54] <cradek> skunkworks: it seems to me like that would work. do you remember what didn't work about it?
[18:34:16] <skunkworks> the mdi one seemed to get reset when the program started. But It has been a while since I had played with it. when I came on here explaining my problem I got the " don't use g92" ;)
[18:34:24] <cradek> and does anyone know what "exchange pallet shuttles" means? it's all over the m2 section of the spec
[18:34:33] <jepler> cradek: nfc
[18:34:56] <fenn> you know what a pallet shuttle is right?
[18:35:02] <cradek> skunkworks: it will be reset on m2, but I think it would be a bug if it reset at program start
[18:35:13] <cradek> uh, that thing that you exchange? (no)
[18:35:19] <jepler> skunkworks: as an experiment, try putting g92.3 (no xyz) in your program, then see if setting g92 x0y0z0 before a run works.
[18:35:53] <skunkworks> jepler: I think you may be right - I think that worked.
[18:36:12] <jepler> skunkworks: if that fixes it, then when starting a program something happens similar to m2 to unapply the g92 offset
[18:36:23] <cradek> somewhere was a many-page document/study about g92 tests, and iirc it concluded that it works as the spec says, even though some people don't like it, but it's been a while since I've seen that.
[18:36:29] <fenn> on the k&t 800 at nist there were interchangeable fixtures you could bolt stuff to called pallets, and they sit on top of "shuttles" that move around on tracks
[18:36:46] <skunkworks> http://www.electronicsam.com/images/KandT/DSCCurrent.JPG
[18:36:48] <fenn> so you could work on setup for one part while it was machining another
[18:36:56] <cradek> ok
[18:37:21] <skunkworks> cradek: look at the picture - there are 2 tables - one in front of the spindle and one to the left.
[18:38:11] <fenn> skunkworks: do they actually move around on tracks or did i make that up?
[18:38:59] <fenn> this shows the pallets (the thing with two circles) http://www.electronicsam.com/images/KandT/oldkandt.JPG
[18:39:05] <skunkworks> the current one would slide onto a station to the right of the machine - the left pallet would be pulled then onto the table.
[18:39:19] <skunkworks> fenn: exactly ;)
[18:39:30] <skunkworks> forgot about that one
[18:39:52] <robin__sz> I'll tell you wahtr I want, what I really really want ...
[18:40:03] <robin__sz> a nice little welding robot :)
[18:41:22] <fenn> you wanna really really wanna zigazig ha?
[18:41:58] <robin__sz> hmmm .. you might be right ... but only with the cute blonde one
[18:42:10] <skunkworks> robin__sz: http://www.robcon.ro/emc/rIMG_4290.JPG alex posted a few pictures earlier
[18:43:22] <fenn> oh crap now i've got that song in my head
[18:44:04] <skunkworks> ditto
[18:44:09] <robin__sz> nice
[18:44:26] <robin__sz> there are a few things i didnt like about the Cloos wire feed
[18:44:38] <robin__sz> very long torch assembly
[18:45:03] <robin__sz> the ABB has the wire feed mounted a lot closer to the business end, on the back of the arm
[18:45:10] <skunkworks> jepler: did the test at least look cool in axis? :)
[18:45:35] <robin__sz> on the other hand ...
[18:45:43] <robin__sz> (4 fingers and a thumb ...)
[18:45:58] <robin__sz> on the other hand ... the Cloos has a more compact wrist joint
[18:48:12] <jepler> skunkworks: yeah, and the trajectory planner seemed to have no problem running them at top speed too
[18:48:28] <skunkworks> jepler: cool
[18:48:44] <jepler> do you want a screenshot or anything?
[18:49:07] <skunkworks> if it isn't much trouble - I dont have an emc machine here running.
[18:50:41] <alex_joni> robin__sz: hi there ;)
[18:50:51] <robin__sz> Mr Joni, are you well?
[18:50:54] <alex_joni> there is a wirefeed that goes on the arm aswell
[18:51:00] <alex_joni> ;-)
[18:51:03] <robin__sz> yeah?
[18:51:08] <alex_joni> yup, just not that usual
[18:51:10] <robin__sz> thats the way to go ... for sure
[18:51:27] <alex_joni> we usually go for wire drum (250-300kg)
[18:51:32] <robin__sz> wait .. I have photos of the old ABB and cell Im getting
[18:51:37] <alex_joni> and about 10-15m of length to the drum
[18:51:41] <alex_joni> I think I've seen some
[18:51:43] <alex_joni> on redpoint
[18:52:07] <robin__sz> yeah, but there are ways to do that nicely, especially with 250Kg packs of wire
[18:52:15] <robin__sz> yeah, those
[18:52:21] <jepler> skunkworks: coming right up
[18:52:46] <jepler> http://emergent.unpy.net/index.cgi-files/sandbox/axis-bigdeer.png http://emergent.unpy.net/index.cgi-files/sandbox/axis-test.png
[18:53:00] <robin__sz> http://www.redpoint.org.uk/photos/upload/robot/
[18:53:04] <robin__sz> ^^ htat one
[18:53:59] <robin__sz> hmmm , come to think of it .. I dont see the wire feed ....
[18:54:03] <skunkworks> jepler: I can increase the lines on the program to make the test.ncg 2.9m - do you want that?
[18:54:03] <alex_joni> hi jeff
[18:54:06] <jepler> hi alex
[18:54:17] <jepler> skunkworks: No, I don't want anything 3 meters big
[18:54:25] <robin__sz> ah well, have a better look when I go to see ti on Friday
[18:54:29] <skunkworks> same physical size
[18:54:38] <alex_joni> jepler: what are you doing with the deer?
[18:54:47] <skunkworks> 2.9 mb file size
[18:55:42] <robin__sz> * robin__sz goes out to pull a motorbike apart
[18:56:20] <skunkworks> 114401 lines of g-code
[18:58:29] <jepler> skunkworks: all it'll succeed in doing is slowing axis down to a crawl
[18:58:42] <jepler> alex_joni: nothing really, I'm just showing skunkworks what it looks like in axis
[18:58:58] <jepler> If I keep messing with the interpreter and planner these files may come in handy as real-world examples though
[18:59:19] <skunkworks> it the buffer keeping up?
[18:59:44] <skunkworks> * skunkworks is trying to sound like he knows what he is talking about.
[18:59:59] <jepler> skunkworks: yeah it looks that way
[19:00:42] <Jymmm> skunkworks: "If you can't dazzle them with brillance, baffle them with bullshit"
[19:01:02] <jepler> skunkworks: if I understood things any better I'd be happy to explain them to you
[19:01:08] <jepler> but unfortunately I'm sumbling around in the darkness too
[19:01:29] <cradek> I attack the darkness!
[19:01:38] <jepler> <nasal laugh>
[19:01:57] <alex_joni> get a flashlight
[19:02:16] <skunkworks> I read the comments of it being many layers of crap to get from the gcode file to the tp - (at least that is what I got out of it)
[19:02:17] <Jymmm> "cradek" must mean "Lord of darkness and all that is evil" in some ancient and forgotten language
[19:02:48] <Jymmm> ... compleltely foiled by the invention of light
[19:03:24] <alex_joni> I think they don't have a word for artificial light
[19:03:27] <skunkworks> onions have layers
[19:03:44] <Jymmm> alex_joni: who said anything about "artificial" ? =)
[19:04:07] <alex_joni> well.. they probably know daylight
[19:04:14] <alex_joni> anything besides that is artificial
[19:04:25] <Jymmm> alex_joni: underground cave ppl
[19:08:57] <cradek> skunkworks: is there any advantage to your big mill being horizontal except for the ability to fit it under a reasonable ceiling?
[19:10:24] <skunkworks> Cradek: it is like ford and chevy for the most part. I have found out chip removal is a lot better as it can fall out of the way amoung other things.
[19:10:45] <cradek> that's a good point, I didn't think of it
[19:10:59] <cradek> however any coolant will be much harder to control
[19:11:33] <skunkworks> most coolant flows down into skirt around the table to be reclaimed.
[19:12:10] <alex_joni> hi ray
[19:12:21] <cradek> hi
[19:12:29] <skunkworks> home?
[19:14:11] <cradek> skunkworks: what's in the spindle in this photo?
[19:16:06] <skunkworks> it was a die grinder - we where using it for 1/8 carbide mill
[19:17:28] <skunkworks> poor mans high speed spindle
[19:24:35] <rayh> Hi guys
[19:24:49] <rayh> Yep. Made it home yesterday evening.
[19:28:38] <skunkworks> did you have a good time?
[19:29:19] <Jymmm> skunkworks: Sure he did, just ask his bail bond man!
[19:29:37] <rayh> I had so much fun it took a day of sleeping to recover.
[19:29:55] <cradek> oh man, me too
[19:30:01] <rayh> yup.
[19:30:30] <rayh> A lot of very good stuff came about there.
[19:30:30] <alex_joni> rayh: nice to see you back online ;)
[19:30:43] <rayh> Thanks alex. It has been a while.
[19:30:51] <alex_joni> how were your lectures?
[19:31:18] <rayh> I think that most of the guys got the ideas about how to use emc2
[19:31:53] <rayh> I need to draw up a graphic of my thinking about it and start a wiki page.
[19:32:21] <rayh> thinking about how to explain HAL to new users.
[19:32:51] <rayh> halui worked for the mazak.
[19:33:14] <alex_joni> it did? YAY
[19:33:15] <cradek> did you hook up the jogwheel?
[19:33:30] <rayh> Not in any perminent way.
[19:33:48] <rayh> I kinda went brain dead by Sunday noon.
[19:34:12] <rayh> Roland doesn't plan to use it for a couple months.
[19:34:26] <rayh> By then I'll visit him again and add several features.
[19:34:35] <alex_joni> sounds like fun..
[19:34:41] <alex_joni> maybe we'll extend halui by then
[19:34:51] <alex_joni> * alex_joni has seen lots of buttons on the old panel ;)
[19:35:02] <rayh> You bet. I think that will help us a lot.
[19:35:04] <cradek> I'm going to set up a jogwheel on my own mill
[19:35:13] <cradek> maybe the code will get polished up that way
[19:35:16] <rayh> Fantastic. I'll be asking for help.
[19:35:19] <alex_joni> cradek: that's great
[19:35:33] <alex_joni> maybe we'll go on and add it into motion (where it belongs)
[19:35:47] <cradek> yeah
[19:35:48] <alex_joni> I mean RT code
[19:35:52] <giacus> hi ray !
[19:35:54] <cradek> it would be more responsive that way, but it does already work
[19:35:56] <cradek> brb
[19:36:12] <rayh> Hi giacus.
[19:36:36] <anonimasu> cradek: that sounds like a nice thing
[19:39:11] <cradek> back
[19:41:47] <jepler> cradek: looks like timing belt pulleys are anywhere from $8 to $20 on mcmaster-carr
[19:41:53] <jepler> depending on the specifics
[19:42:19] <cradek> I wonder if I could confidently match the ones that are already on the motors
[19:42:20] <Jymmm> I know theres a online tool for calculating deflection of a rod/bar, but what about a flat surface, like a table as example?
[19:43:35] <jepler> Jymmm: It would be tempting to calculate deflection as for a rod but using the diagonal as the rod length, but I have no idea how valid that would be
[19:44:21] <cradek> to me it sounds like a complex question that you would have to ask a mechanical engineer
[19:44:32] <cradek> I'm sure it depends a lot on the specifics of the table
[19:48:18] <fenn> hmm
[19:48:27] <Jymmm> Ok, what I'm trying todo is use 1/4" high density fiberboard about 10" x 10" without it bowing in the middle. The problem is that I can't add much in the respect to supports due to what it's being mounted on.
[19:48:39] <fenn> i guess i was dreaming yesterday
[19:48:52] <fenn> all the things that randomly broke for no reason have now fixed themselves
[19:49:02] <Jymmm> fenn PEBKAC
[19:49:08] <Jymmm> fenn PEBTAC
[19:49:22] <fenn> problem exists between eyes and brain
[19:49:25] <Jymmm> fenn PEBTAO
[19:49:36] <Jymmm> Proble exist between tool and operator =)
[19:49:55] <Jymmm> hmmm, that was kind lame. nm
[19:50:26] <Jymmm> (on my part that is)
[19:50:58] <Jymmm> anyhow... I'm just trying to get an idea on how much deflection I might have to deal with
[19:51:21] <fenn> the calculation should be about the same as for a beam
[19:51:42] <Jymmm> 10"x10"x.25"
[19:51:45] <fenn> constrained at both ends
[19:52:14] <Jymmm> will be supported on all four edges, just not sure how much I'll be able to towards the middle
[19:52:28] <fenn> does the phrase "preventative overkill" mean anything to you? perhaps it should
[19:52:53] <giacus> any idea about this issue ? http://rafb.net/paste/results/5Bv83B54.html
[19:53:07] <Jymmm> fenn Like I said, I won't be able todo a lot of support in the middle, there are ibsturctions preventing that.
[19:53:14] <Jymmm> obstructions
[19:53:41] <fenn> giacus: what is the output of uname -a
[19:54:09] <alex_joni> giacus: is adeos loaded?
[19:54:17] <giacus> Linux emc2 2.6.16 #5 PREEMPT Wed May 24 16:04:52 CEST 2006 i686 GNU/Linux
[19:54:19] <alex_joni> put a dmesg in there
[19:54:39] <giacus> a minuts .. done the compilation of emc2
[19:55:01] <jepler> giacus: maybe you need to add rtai_sched to the list of modules allowed by emc_module_helper. the list is in emc2/src/module_helper/
[19:55:04] <jepler> giacus: in the .c file
[19:55:20] <giacus> oh.. ok looking
[19:55:21] <jepler> my system doesn't have an rtai_sched.ko
[19:55:32] <jepler> (ubuntu)
[19:56:52] <alex_joni> it's ksched and should be a symlink
[19:56:56] <alex_joni> to rtai_up usually
[19:57:01] <giacus> uhm.. there are a lot of rtai_shm: Unknown symbol rt_get_registry_slot in dmesg ..
[19:57:03] <giacus> a moment
[19:57:59] <jepler> /home/jack/emc2/bin/emc_module_helper: Invalid usage with args: insert /usr/realtime/modules/rtai_sched.ko
[19:58:07] <jepler> the error message is about rtai_sched though .. not ksched
[19:58:30] <jepler> ./scripts/rtapi.conf.in: MODULES="adeos rtai rt_mem_mgr rtai_sched rtai_fifos rtai_shm rtai_libm"
[19:58:42] <jepler> maybe this needs to be changed?
[19:59:29] <fenn> is it rtai_ksched on his system instead of rtai_sched?
[19:59:46] <alex_joni> I might remember it wrong..
[19:59:52] <alex_joni> might still be sched not ksched
[20:00:03] <alex_joni> but it definately is a symlink to one of the schedulers chosen
[20:00:07] <alex_joni> I suspect up
[20:00:14] <jepler> lrwxrwxrwx 1 root root 10 2006-02-19 16:14 /usr/realtime-2.6.12-magma/modules/rtai_ksched.ko -> rtai_up.ko
[20:00:20] <jepler> that's on my ubuntu
[20:00:28] <jepler> so .. I'm confused
[20:00:28] <fenn> same here
[20:00:50] <fenn> ditto
[20:01:13] <giacus> lrwxrwxrwx 1 root root 13 2006-05-24 18:02 /usr/realtime/modules/rtai_ksched.ko -> rtai_sched.ko
[20:01:29] <giacus> -rw-r--r-- 1 root root 80859 2006-05-24 18:02 /usr/realtime/modules/rtai_sched.ko
[20:01:50] <giacus> modules are all there
[20:01:54] <alex_joni> what rtai is that?
[20:01:56] <alex_joni> 3.3?
[20:02:11] <giacus> yes, the latest patch
[20:02:25] <fenn> and the latency test works?
[20:02:32] <giacus> yes
[20:02:33] <jepler> scripts/realtime resolves a level of symlinks, so that "rtai_sched" is what it actually passes to module_helper when it tries to load/unload rtai_ksched
[20:02:38] <jepler> on giacus machine
[20:02:45] <jepler> on my machine it ends up trying to load rtai_up, I think
[20:03:01] <jepler> presumably this dance is necessary so that the module can be unloaded by the right name?
[20:03:38] <jepler> so I'm back to thinking it's right to add "rtai_sched" to module_whitelist, not that I ever really stopped
[20:04:14] <giacus> /usr/src/magma/base/arch/i386/patches/hal-linux-2.6.16-i386-1.3-00.patch is what I used
[20:04:25] <giacus> from cvs
[20:05:50] <giacus> latency test work fine RTH| lat min| ovl min| lat avg| lat max| ovl max| overruns
[20:05:51] <giacus> RTD| -1480| -1480| -361| 7621| 7621| 0
[20:06:04] <jepler> giacus: have you tried changing module_helper yet?
[20:06:11] <giacus> not
[20:06:24] <giacus> I can try
[20:06:44] <jepler> please do
[20:06:47] <CIA-8> 03jepler 07HEAD * 10emc2/src/module_helper/module_helper.c: if not setuid root, complain about it right away instead of leaving users scratching their heads
[20:06:52] <giacus> working on TESTING , I also tried HEAD, same issue
[20:07:05] <giacus> ok give me few minuts
[20:07:46] <jepler> * jepler waits patiently
[20:08:06] <Jymmm> jepler: Hot or cold beverage while you wait?
[20:08:15] <jepler> Jymmm: no thanks
[20:08:54] <Jymmm> jepler bag or two of airline peanuts?
[20:10:25] <giacus> compiling..
[20:13:52] <jepler> Jymmm: did you give up the CNC and get a new job as an airline steward now?
[20:14:02] <jepler> you're doing very good in that role
[20:14:12] <giacus> sorry.. few lines:
[20:14:14] <giacus> Compiling module_helper/module_helper.c -Wall -Werror -DMODULE_EXT=".ko" -DRIP_MODULEDIR="/home/jack/emc2/rtlib"
[20:14:15] <giacus> module_helper/module_helper.c:33: error: called object '"rtai_sched"' is not a function
[20:14:16] <giacus> make: *** [objects/module_helper/module_helper.o] Error 1
[20:15:10] <giacus> I added it here: "stepgen", "supply", "threads", "timedelay", "timertask", "hal_vti","rtai_sched"
[20:16:10] <giacus> in module_helper.c
[20:16:14] <jepler> giacus: I don't understand the error message. you do need to have a comma after "rtai_sched", but I'm not sure how it leads to that message.
[20:16:37] <giacus> a moment..
[20:17:48] <giacus> http://rafb.net/paste/results/671Wp577.html
[20:17:50] <fenn> i like how classicladderr shows you the latency in the corner
[20:17:54] <giacus> what's wrong ?
[20:18:04] <jepler> giacus: add a comma after "rtai_sched"
[20:18:23] <skunkworks> jepler: at what feed rate was the test running at. (says 64.06 actual)
[20:18:34] <jepler> http://rafb.net/paste/results/hgcF0f65.html
[20:18:36] <skunkworks> forgot to say - looks cool in axis.
[20:19:09] <jepler> skunkworks: well it was running in my CVS checkout where I've screwed up the feed rates. I believe it was trying to run at the machine maximum 72ipm.
[20:20:04] <jepler> skunkworks: I'm also not entirely sure how accurate the speed display is, I kina just hacked it on
[20:20:18] <skunkworks> ok
[20:20:31] <jepler> it varies a lot as the program runs
[20:20:48] <Jymmm> Just curious... does anyone have, or know of anyone that still has the blade guard on their table saw by chance?
[20:21:30] <Jymmm> jepler: Nah, just being hospitable while you "wait patiently" for giacus.
[20:22:10] <skunkworks> When I was looking over cradeks shoulder when he was running the spiral on the mazak - I asked if he had thought any about having actual feedrate on the screen. he promptly pulls down a menu and activates it. :)
[20:23:55] <jepler> skunkworks: I'm not sure what prompted me to add it, it was something I did last week thought
[20:23:58] <jepler> though
[20:24:00] <giacus> some warning ..
[20:24:04] <giacus> going on
[20:24:07] <Jymmm> skunkworks: You should of asked him about beer and wine dispensing option =)
[20:24:33] <giacus> tryng to run EMC2
[20:24:51] <skunkworks> neat
[20:25:22] <giacus> jepler: it definitively work
[20:25:28] <giacus> wher you looked at ? O_O
[20:25:31] <skunkworks> used to be I would ask about something being added and it would be added. Now it gets added before I ask for it ;)
[20:25:35] <giacus> the cristal sphere ? :P
[20:25:50] <Jymmm> Magic eight ball =)
[20:26:03] <giacus> nice thanks
[20:26:09] <jepler> giacus: emc2 starts now? good. I'll add "rtai_sched" to module_helper in TESTING and HEAD, so anyone else using rtai 3.3 will not have the same trouble.
[20:26:30] <giacus> jepler: yeah, running tkemc
[20:26:39] <giacus> it seems work fine
[20:26:52] <jepler> giacus: great
[20:26:55] <giacus> going to compile Axis now
[20:28:02] <jepler> CIA-8: hey, I just checked something in. aren't you going to tell the channel about it?
[20:28:25] <Jymmm> jepler (only if it fails compiling)
[20:30:11] <Lerneaen_Hydra> hi
[20:30:15] <jepler> hi Lerneaen_Hydra
[20:30:27] <CIA-8> 03jepler 07v2_0_branch * 10emc2/src/module_helper/module_helper.c: backport: permit rtai_sched to be loaded and unloaded
[20:30:43] <Jymmm> my bad
[20:53:48] <giacus> exit
[20:53:50] <giacus> ops
[21:03:50] <jepler> 4082 N May 24 EMC CVS server ( 0) [Emc-commit] emc2/src/module_helper module_helper.c,1.20,1.21
[21:03:53] <jepler> 4083 N May 24 EMC CVS server ( 0) [Emc-commit] emc2/src/module_helper module_helper.c,,
[21:03:58] <jepler> cradek: any idea why the two subject liens are spaced differently? seems odd
[21:05:19] <cradek> echo "Subject: $1"
[21:05:27] <cradek> all I can say is it comes from cvs like that
[21:05:37] <jepler> huh
[21:08:12] <cradek> even more puzzling, I don't think the script would work with spaces there
[21:10:31] <cradek> Jymmm: I think those guards might make a table saw less safe, not more
[21:11:22] <Jymmm> cradek: Me too, but usually I think it's a CYA for the dumbass operator sorta thing. But I thought I'd ask
[21:11:57] <cradek> for instance they get in the way of using a push stick
[21:12:09] <cradek> I doubt they are dumbass-proof at all
[21:12:26] <cradek> I bet "dumbass" is hard to fix with any kind of guard
[21:12:32] <Jymmm> probably an "ansi" safety issue/regulation sorta thing.
[21:13:06] <cradek> did you ever see the videos of the meat-detector addons for table saws etc?
[21:13:21] <Jymmm> lol, no.
[21:13:33] <cradek> I wonder where those were - it was extremely cool
[21:13:41] <cradek> they had slow motion videos
[21:14:32] <cradek> they put a hot dog on top of a board and sawed along - the sensor somehow senses the blade touching the "meat" and throws a brake into the blade's teeth, and also the blade instantly retracted
[21:14:50] <cradek> and it left only a nick on the hot dog
[21:15:09] <Jymmm> wow, I'd liek to see that
[21:16:10] <anonimasu> :)
[21:16:24] <cradek> http://www.sawstop.com/
[21:16:38] <skunkworks> fires a spring loaded almuninum chunk into the blade
[21:17:18] <cradek> yeah I recall it stops rather destructively
[21:17:35] <cradek> but if you fall onto the blade, you would probably be ok with buying a new one
[21:18:08] <anonimasu> yeah
[21:19:19] <cradek> jepler: the maillog doesn't contain the subjects so I can't be sure what added the spaces
[21:22:21] <cradek> wow, their video says you just have to replace the blade and metal brake
[21:22:53] <alex_joni> that looks awesome
[21:23:48] <alex_joni> hmm.. it looks at conductive material?
[21:24:41] <anonimasu> cool machines
[21:25:22] <alex_joni> probably worth the <3k$
[21:25:27] <anonimasu> yeah
[21:33:41] <dmessier> hi all
[21:37:55] <fenn> i can think of lots of uses for a meat detector
[21:38:29] <anonimasu> hehe
[21:39:05] <dmessier> hmmm
[21:39:18] <fenn> only works with wood though, so you're on your own if sawing aluminum or meat
[21:40:23] <anonimasu> they should make motorsaws like that
[21:40:45] <anonimasu> err chainsaw's..
[21:41:06] <fenn> yeah it'd stop if you hit a nail or something too
[21:41:22] <alex_joni> fenn: only if you're touching the nail I think
[21:41:41] <fenn> yeah moisture and sap would probably mess it up
[21:42:03] <dmessier> so the saw sTOPS if it hits your arm???
[21:42:38] <fenn> it detects the change in capacitance like a switch-less lamp
[21:42:49] <Jymmm> "The bypass switch allows you to cut conductive material"
[21:43:44] <Jymmm> (it's literally a key lock switch too)
[21:44:44] <dmessier> this is in a chainsaw??
[21:44:55] <Jymmm> dmessier: http://www.sawstop.com/
[21:44:58] <anonimasu> no
[21:45:01] <anonimasu> I wish it were :)
[21:45:04] <anonimasu> that's be sweet ;)
[21:45:12] <robin__sz> I would NOT buy one ...
[21:45:28] <anonimasu> why not?
[21:45:45] <robin__sz> the company obviously does not believe in itsown products, so why should I?
[21:46:25] <anonimasu> robin__sz: so you think they should have tried it for real?
[21:46:29] <robin__sz> if they really beleived in them, the CEO should be there, demonstrating with his own finger
[21:46:32] <robin__sz> * robin__sz nods
[21:46:34] <anonimasu> lol
[21:46:52] <dmessier> i drove a buddy to the hospital after a BAD wrestling fight with a chainsaw... got him RIGHT BETWEEN THE EYES on a 45 deg angle
[21:46:54] <cradek> or neck - he could lay on the board and someone could push it through
[21:47:05] <robin__sz> exactly
[21:47:33] <fenn> well at least it nicks you to keep you from doing stupid things like that
[21:47:34] <robin__sz> chainsaws do that ... a helmet can help a lot
[21:48:15] <dmessier> had 1... but a full face bike helmet would have been trashed too
[21:48:27] <robin__sz> I did a chainsaw safety course once ... learnt a lot that day
[21:48:32] <fenn> robin__sz: i'm pretty sure there are laws against live safety testing
[21:49:00] <robin__sz> fenn, subjecting employees to it, sure, but ceo's are immune
[21:49:08] <fenn> a ceo is an employee too
[21:49:33] <robin__sz> dmessier, best bit of chainsaw advice I ever got was "dont stand behind it, stand to the side"
[21:49:38] <dmessier> he who holds the patent..
[21:50:43] <dmessier> twiggy crap kicked it up... he had sprained his wrist the week befor water skiing... it flipped into his face...
[21:51:08] <robin__sz> dmessier, imagine instead of standing at 90 degrees to a log, you stand along side it ... with the chainsaw at 90 degrees to you ...
[21:51:25] <dmessier> ok
[21:51:26] <anonimasu> that way it'll chop your leg off and not a arm.
[21:51:28] <anonimasu> :D
[21:52:03] <robin__sz> if it kicks up, it just harlessly pivots in front of you ... maybe :)
[21:52:17] <anonimasu> aw, how boring
[21:52:19] <robin__sz> the ballistic nylon trouser will save you
[21:52:26] <dmessier> you arent in any type aergomomic type of position to work properly AND safely like that
[21:52:27] <anonimasu> *just talking crap*
[21:53:22] <robin__sz> dmessier, well, its a technique taught over here ... my little sister does a lot of saw work,
[21:53:30] <dmessier> they work... this was an old saw no e-brake stuff at all
[21:53:36] <robin__sz> ick
[21:54:19] <robin__sz> a friend has a big old 48" bar one ... no brakes or anything .. handle on both ends of the bar
[21:54:23] <dmessier> it was 1988 after all...
[21:54:43] <dmessier> and a mittload i'll bet...
[21:54:49] <robin__sz> paintstripper++ # great for cleaning up old carburetors
[21:56:05] <robin__sz> trouble is ... people hire a littel 12 or 14" bar Stihl saw for a weekend to chop a tree and think "hey, I can handle a saw" ... they should try a 30" monster, in winter, on a steep slope in the rain
[21:56:21] <dmessier> my adopted dad was a master of the saw...drop em where you wanted em..
[21:56:37] <robin__sz> yep ...
[21:56:41] <robin__sz> true skill that
[21:56:58] <anonimasu> chopping trees is hard work
[21:57:00] <robin__sz> thats what my mate with the 48" monster does for a living
[21:57:07] <dmessier> i hear ya... big Stihl's & skills needed
[21:57:14] <Jymmm> $2800 USD for the saw, not that bad for what it does.
[21:57:52] <dmessier> good investment once you put it to real paying work..
[21:58:09] <robin__sz> hey, I saw a GREAT plasma tube cutter, 3 axis CNC
[21:58:20] <robin__sz> very neat
[21:58:31] <anonimasu> just 3 ? aw
[21:58:39] <robin__sz> the Z axis was floating free, rested a bar on the pipe
[21:58:43] <dmessier> i could selectively harvest 5000 dollars a week in 8-10 trees
[21:58:47] <robin__sz> so 3 and a half really
[21:59:02] <anonimasu> i got a ad for some weird stuff like that
[21:59:06] <anonimasu> with a \ / axis
[21:59:14] <anonimasu> so you could have a tube -------\--------
[21:59:23] <robin__sz> yeah
[22:01:28] <robin__sz> this did box section too
[22:01:28] <anonimasu> yep
[22:01:28] <robin__sz> so it coul do holes in box section
[22:01:28] <robin__sz> you got a left/right axis ... a forward, back axis, and a rotational axis
[22:01:28] <anonimasu> yep
[22:01:28] <anonimasu> cool :)
[22:01:28] <dmessier> should do it...
[22:04:36] <alex_joni> how's the weather over there?
[22:04:50] <alex_joni> over here it's pretty crazy.. we had about 35C today..
[22:05:42] <alex_joni> going down to 23 tomorrow
[22:05:49] <jepler> 32C here
[22:06:53] <alex_joni> it was damn hot
[22:07:05] <alex_joni> but 12 degrees drop from one day to the next.. weird
[22:07:06] <giacus> same here
[22:07:54] <alex_joni> anyways.. hitting the sheets
[22:07:56] <alex_joni> night all
[22:10:40] <robin__sz> night
[22:13:11] <giacus_2> night alex_joni
[22:42:55] <giacus> http://www.giacus.org/files/axis.png
[22:43:22] <giacus> Axis running on Ion, pretty nice
[22:43:30] <Jymmm> Jymmm is now known as Jymmmmm
[22:43:49] <giacus> I can't resize it under that size, but looks ok
[22:44:25] <giacus> halscope looks very bad: http://www.giacus.org/files/halscope.png
[22:44:43] <giacus> also resizing the frame
[22:45:58] <Jymmmmm> Jymmmmm is now known as Jymm
[22:47:07] <giacus> in the float workspace are ok
[22:47:25] <giacus> act as a 'normal wm'
[22:48:43] <giacus> well, I'll use the WFloatWS for Emc
[22:50:42] <giacus> finally a shot of my desktop http://www.giacus.org/files/shot3.png :)
[22:50:42] <fenn> giacus: you broke it!
[22:50:48] <giacus> hehe
[22:51:38] <giacus> fenn: there's no way to get it working on a WIonWS :(
[22:51:49] <giacus> workspace
[22:52:02] <giacus> looks horrible
[23:16:00] <dmessier> robin__sz... ever done any risk manegment SPECIFIC trianing
[23:18:11] <dmessier> we have a protocol in our club thats called the RCR's Robertsons Charts of Reliability..
[23:19:00] <dmessier> evaluates Wing ... Wind.. and Windividual..
[23:20:22] <dmessier> i've re-written them to be Machine.. Set-up.... Monkey... for our shop
[23:43:21] <jepler> oh no! vista has been pushed back again! http://www.forbes.com/2006/05/24/vista-rollback-microsoft-cx_ck_0524autofacescan11.html
[23:43:52] <rayh> horrors what a blow to mankind.
[23:44:41] <jepler> at least it gives us a few more weeks to work on our vista port of emc2
[23:45:56] <cradek> oh god what will we do?
[23:46:43] <rayh> Do you think the apic virus will be able to run under vista?
[23:47:44] <jepler> gee, I don't know. vista is touted as being the most secure version of windows, after all
[23:48:27] <dmessier> probably will/// it spawned from there after all
[23:48:57] <dmessier> gotta check it some how???
[23:56:06] <giacus> what ?
[23:59:58] <giacus> oh.. windows vista