#emc | Logs for 2008-03-19

[00:05:22] <SkinnYPupp1> Here are the g8x codes http://linuxcnc.org/docs/2.2/html/gcode_main.html#sub:G81-to-G89:
[00:10:48] <ken> thanks you brb
[00:15:06] <fenn> what's two sided milling?
[00:17:12] <Guest486> work on the first side then bring up the backside
[00:17:56] <Guest486> like in deskproto
[00:18:27] <Guest486> consumerfriendly
[00:20:14] <ken> thank you Skinn ...that helped
[00:21:58] <ken> nite all
[00:39:23] <renesis> http://darkertechnologies.com/image/ttmustang_goodside.jpg
[00:39:37] <renesis> before that conical finally broke
[00:40:02] <renesis> it wouldnt do .010 traces like that before
[00:40:35] <renesis> (further confirmation of taig wtf backlash thing)
[00:40:53] <renesis> oh thats not my board
[00:41:07] <renesis> so dont tell me it sucks or anything
[00:41:10] <renesis> i dont care
[01:15:01] <Twingy> your 30 broke?
[01:16:55] <Twingy> those're some pretty small thruholes
[01:28:21] <fenn> big fat traces are caused by runout, not backlash
[02:08:40] <toastydeath> i bought an abacus on ebay
[02:10:19] <fenn> why?
[02:10:30] <toastydeath> i've been trying to learn to use one
[02:10:32] <toastydeath> using java applets
[02:10:37] <eric_U> because you can't add on a slide rule
[02:10:43] <fenn> yes you can
[02:10:55] <toastydeath> you can add and multiply on a slide rule!
[02:10:57] <toastydeath> er
[02:10:57] <fenn> just move the ruler over by the number you're adding
[02:10:58] <toastydeath> abacus!
[02:11:00] <toastydeath> one of those things
[02:11:01] <eric_U> I don't want to hear about any stupid tricks
[02:11:38] <fenn> slide rules are dumb anyway
[02:11:54] <eric_U> the alternative wasn't pleasant
[02:13:01] <toastydeath> but an abacus is cool!
[02:13:11] <toastydeath> it's quite a mental workout to use at the moment
[02:13:14] <toastydeath> i'm hoping that goes away a little
[02:13:59] <fenn> now i've never used a slide rule, but why didnt they make them out of tape wound on reels?
[02:14:54] <toastydeath> because what
[02:15:26] <fenn> for more accuracy (or, to reduce cost resulting from high precision)
[02:16:15] <eric_U> there were round sliderules
[02:17:05] <fenn> you'd have two cranks, one for each tape, and the pairs of tape reels would be held together by a light spring/friction fit
[02:17:36] <eric_U> can't see how that would be as portable as a slide rule
[02:17:48] <fenn> a slide rule is a stick about 18" long right?
[02:18:05] <fenn> this would look like a small camera
[02:18:06] <eric_U> that would have been a large sliderule
[02:18:41] <jmkasunich> fenn: typical slide rules were 10-12", there were 6" ones
[02:19:58] <eric_U> I was thinking about selling mine, maybe I should keep it for when the revolution comes
[02:20:19] <fenn> maybe you should stockpile calculators
[02:20:30] <eric_U> emp
[02:20:41] <fenn> abachi?
[02:20:46] <jmkasunich> keep the stockpile in a faraday cage
[02:20:51] <eric_U> toasty and I will be ruling the world
[02:21:45] <fenn> ooo "The plural of abacus has a been a matter of heated debate for some time."
[02:22:40] <eric_U> is the debate among people that should know, or idiots on the internet?
[02:24:17] <fenn> between people who speak latin and the OED
[02:25:06] <fenn> oo between OED and Webster
[02:25:17] <fenn> oh the drama, the passion
[02:27:36] <toastydeath> i have two abacuses, i have two abacii, i have two abacus?
[02:27:53] <toastydeath> things toasty does not care about, part three
[02:27:55] <fenn> toastydeath: why didnt you make one instead?
[02:28:07] <toastydeath> i tried, out of paper
[02:28:24] <toastydeath> i got the tray and everything, i just didn't want to invest the time to come up with some sort of bead
[02:28:27] <fenn> no no no, you must use the skulls of your enemies
[02:28:36] <toastydeath> CLEARLY
[02:29:17] <fenn> it just doesn't have the same reassuring 'clunk' otherwise
[02:29:55] <toastydeath> well basically i got tired of dicking around in a java applet
[02:30:10] <toastydeath> because the reflex memory wasn't developing with, you know, a mouse
[02:31:25] <toastydeath> i'm hoping it will give me a better memory with numbers, help me memorize some multiplication tables, and some other stuff
[02:31:49] <fenn> just play number muncher
[02:32:04] <toastydeath> lol
[02:32:36] <toastydeath> well "other stuff" includes a better sense of engineering precision, spacial relationships, and lower calculation error as compared to a calculator
[02:32:53] <toastydeath> i'm not sure if number muncher can help with that, but if it can i'd be willing to try
[02:32:56] <toastydeath> :D
[02:36:35] <fenn> lower calculation error?
[02:37:15] <fenn> i find that using the 'units' program helps immensely in understanding the relationships between parameters
[02:37:29] <toastydeath> abacus users have a lower rate of "miskeying" than calculator users
[02:37:51] <fenn> instant gratification, vs having to dig around in tables of conversion factors
[02:37:59] <toastydeath> ?
[02:38:29] <fenn> GNU units, a command line tool that should be installed on every engineer's computer
[02:39:01] <toastydeath> right
[02:39:11] <toastydeath> i do all unit analysis by hand
[02:39:36] <toastydeath> it helps score higher on tests etc
[02:39:41] <toastydeath> at least in my limited experience
[02:39:56] <fenn> compared to what?
[02:40:05] <fenn> not doing any unit analysis?
[02:40:09] <toastydeath> yeah
[02:40:34] <toastydeath> "i seem to have invented a new unit, the inch meter second^2 volt"
[02:40:48] <toastydeath> "it is a unit of error in my equations"
[02:40:54] <toastydeath> "anywhere it appears, i am wrong"
[02:40:58] <fenn> hmm
[02:41:18] <SWPadnos> I can vouch for that
[02:41:25] <toastydeath> like when i'm doing pure algebra, i do the unit analysis at the same time
[02:41:46] <toastydeath> when i have everything rearranged, i know the final answer is right and that the rounding error is minimal
[02:42:45] <fenn> sure, but what i'm saying is that having GNU units at your fingertips all the time builds a sense of proportion so that when you do make an error it should give you an immediate gut reaction
[02:42:58] <fenn> and the rounding error is nonexistent
[02:43:55] <toastydeath> i think we are saying the same thing but we use different tools to accomplish it
[02:44:14] <eric_U> when I went to engineering school, the units were most of the points
[02:44:38] <toastydeath> yes, well times have changed, and most of my classmates can't do basic statics by themselves
[02:45:02] <toastydeath> if they graded on units, three people per year would graduate
[02:45:11] <eric_U> what school?
[02:45:19] <toastydeath> delaware tech
[02:45:47] <toastydeath> it's a community college, but it's still ABET accredited to teach engineering
[02:45:58] <toastydeath> some folks from UD attend classes there
[02:46:11] <eric_U> I've heard the difference is that they now take student evaluations of teachers seriously
[02:46:27] <toastydeath> that might be part of it
[02:46:44] <toastydeath> my thought is how it's taught at our little school is way different from what the people are used to
[02:46:45] <eric_U> they didn't even do those when I was an undergrad
[02:47:09] <eric_U> I don't recall anyone being used to statics when I took it
[02:47:12] <toastydeath> all our teachers come from industry and are like "we are going to solve real problems, not just boradwork"
[02:47:37] <toastydeath> so part of the problem is the people are being asked to actually engineer a device and they've never done it before, so they screw up simple stuff
[02:48:58] <fenn> better than being asked to do fake stuff and still screwing it up
[02:49:04] <toastydeath> tru
[02:49:27] <toastydeath> like if you give them raw math and say "solve this" they can do it
[02:49:35] <fenn> or even getting a contrived problem right - what's the point of that?
[02:49:37] <toastydeath> but if you start asking them questions about projectile motion, they flip out
[02:50:02] <eric_U> really, that was my favorite part of statics
[02:50:06] <fenn> reminds me of feynman bitching about the brazilian education system
[02:50:14] <toastydeath> the question yesterday was, "you have a tomcat, it is landing on the USS Enterprise. Find the force necessary to stop it."
[02:50:23] <Twingy> ok, just finished gcam debian package
[02:50:29] <toastydeath> that was it, the class had to look up the weight/thrust/etc of the tomcat and carrier
[02:50:33] <fenn> students could solve all the problems in the book, but if you asked them about something in the real world they were clueless
[02:50:56] <toastydeath> and then give an assesment of how much force the arrester cable had to hold
[02:50:57] <eric_U> that's a cool problem
[02:51:19] <fenn> you have to know the landing speed, mass of the tomcat, and length of runway
[02:51:20] <toastydeath> it was, it's a really fun program provided you have really good problem solving skills
[02:51:37] <toastydeath> fenn: we also had to factor in the jet lands at full throttle
[02:51:45] <fenn> huh?
[02:51:56] <fenn> who lands at full throttle?
[02:51:58] <eric_U> mil power probably
[02:52:00] <toastydeath> apparently jets throttle up before landing in case they miss the cable
[02:52:04] <toastydeath> ?
[02:52:07] <toastydeath> so they don't die
[02:52:15] <eric_U> not full afterburner
[02:52:21] <toastydeath> nah, just full throttle
[02:52:29] <eric_U> right mil power :)
[02:52:33] <toastydeath> o
[02:52:36] <toastydeath> i don't know the terms
[02:53:03] <eric_U> I used to be the throttle engineer on the f-16, not exactly rocket science
[02:53:15] <eric_U> jet science
[02:53:17] <toastydeath> lol pun
[02:53:18] <fenn> not quite entirely unlike rocket science
[02:53:59] <toastydeath> OR IS IT
[02:54:15] <eric_U> I knew a rocket scientist once
[02:54:15] <tomp2> Twingy: thank you
[02:54:18] <toastydeath> we also sized a transmission the other day, 150 hp motor at 1800 rpm
[02:54:29] <fenn> do they still make rocket scientists?
[02:54:33] <toastydeath> 3 spd
[02:54:41] <eric_U> no, he was a specialist
[02:54:52] <eric_U> he wasn't all that good, but he was the only one in the air force
[02:55:07] <eric_U> so he got promoted like crazy
[02:55:08] <fenn> omfg there are legions of nerds chomping at the bit to do rocket stuff
[02:56:50] <fenn> toastydeath: what was involved in 'sizing' the transmission? find the max torque required? (what about downshifting)
[02:57:05] <eric_U> toasty, ford took a transmission out of a 50hp 4 cylinder, and put it in my car, 220hp v-6
[02:57:14] <toastydeath> eric_u: out
[02:57:14] <toastydeath> fenn:
[02:57:16] <eric_U> then they wondered why the clutch failed
[02:57:18] <toastydeath> eric_u: *ouch
[02:57:21] <toastydeath> fenn:
[02:57:34] <fenn> multitasking is for mainframes
[02:57:44] <toastydeath> looking at the polar shear stress on each shaft in the transmission in each gear
[02:57:49] <dave_1> and ford calls that engineering? found on road dead
[02:57:51] <Twingy> tomp2, I'm seeing what I can do to make it dapper friendly
[02:58:06] <eric_U> it actually works pretty well once they beefed up the clutch
[02:58:36] <toastydeath> we didn't take into account dynamics, this particlar class is purely strength of materials
[02:58:40] <eric_U> it also has a problem in the differential, but that doesn't show up until the 130k mark or so
[02:59:43] <fenn> does 'dynamics' include things like modal analysis?
[02:59:51] <eric_U> yeah, right
[03:00:00] <fenn> * fenn sneers at eric_U
[03:00:08] <toastydeath> fenn: yes
[03:00:16] <eric_U> * eric_U sneers back at fenn
[03:00:28] <toastydeath> dynamics, anything that results from motion
[03:00:41] <toastydeath> we are pretending it's a static torque on the input shaft
[03:01:06] <toastydeath> because the strength of materials is the same no matter how you generate the forces
[03:01:32] <fenn> eh, really?
[03:01:42] <toastydeath> yeah man.
[03:01:43] <eric_U> sorta
[03:01:45] <Twingy> so when I go to install my gcam debian package it pops up an error on dependency "not satisfiable" for libatk-1.0-0... I go into Synaptic and libatk-1.0-0 is installed, how can this be?
[03:01:53] <toastydeath> the loads just change.
[03:01:58] <toastydeath> you still have to reduce it to a stress.
[03:02:23] <eric_U> there is research done into multi-dimensional fatigue, where they pull in two directions
[03:02:39] <eric_U> I said, din't you guys ever hear of Mohr's law?
[03:02:46] <toastydeath> no sir?
[03:03:05] <toastydeath> also fenn vibration and modal analysis are a 4 year program
[03:03:29] <eric_U> I doubt an undergrad would really cover that
[03:03:40] <eric_U> unless you did it on a computer
[03:03:53] <toastydeath> eric_u: regenerative and self-excited vibration are graduate stuff
[03:04:05] <toastydeath> but they cover the basics in a 4 year
[03:04:33] <toastydeath> which is sad because that's the crap i really want to get into
[03:06:22] <eric_U> I had a course in vibes I really liked in grad school
[03:06:34] <eric_U> I hated the undergrad class
[03:06:40] <toastydeath> why?
[03:07:00] <fenn> because undergrad anything is a scam
[03:07:18] <eric_U> they didn't really treat it as a dynamic system
[03:07:29] <toastydeath> lol what
[03:08:00] <eric_U> it was the old voodoo crap from when engineers only had sliderules
[03:08:38] <toastydeath> oh
[03:09:18] <eric_U> my master's thesis is on active control of structural vibrations
[03:09:38] <toastydeath> nice dude!
[03:09:50] <fenn> have you read 'rainbows end'?
[03:09:55] <eric_U> lotta good that did me
[03:10:01] <toastydeath> structural as in civil engineering or structural as in general structures that vibrate
[03:10:08] <eric_U> aerospace
[03:10:09] <toastydeath> air tables etc
[03:10:13] <toastydeath> oh crap.
[03:10:23] <eric_U> no it was a large space structure
[03:10:34] <eric_U> fenn: I think I have
[03:10:42] <fenn> the dancing library
[03:11:17] <eric_U> no, I'm wrong
[03:12:53] <eric_U> I was going through some of my notes from my master's degree, I took a dynamics course that was really great
[03:12:57] <fenn> did robert forward contribute significantly to the field?
[03:13:09] <eric_U> there was no book though :(
[03:13:13] <toastydeath> oh
[03:13:27] <eric_U> I think he was outstanding in his field
[03:13:27] <toastydeath> that sounds weird, how do you have a college class without a 175 dollar book
[03:13:40] <eric_U> the books that covered the material were all out of print
[03:14:09] <toastydeath> you know you're into some obscure shit when all the books for that subject are out of print
[03:14:22] <eric_U> one of the problems was a mass on the end of a spring on the end of two pendulums, so it was like 3 pendulums
[03:14:40] <toastydeath> haha that's awesome
[03:14:46] <toastydeath> i want to do stuff like that =(
[03:14:53] <Twingy> so what's going on if a .deb package says a package (dependency) is not satisfiable and the package happens to already be installed, does it thing the package isn't installed when in reality it is?
[03:14:56] <eric_U> I didn't understand the first thing I did
[03:15:23] <fenn> Twingy: perhaps the installed package isn't providing the functionality your package requires
[03:15:38] <eric_U> version problem?
[03:15:45] <toastydeath> i hate that, "i don't know what this math does, but i know i have to do it"
[03:15:53] <Twingy> well the version (libatk-1.0-0) it's wanting is already installed (libatk-1.0-0)
[03:16:15] <eric_U> I ran into that when there was no devel package?
[03:16:15] <Twingy> so I'm like wtf
[03:16:27] <JymmmEMC> there's some switch you can use, but I dont know what it is
[03:16:27] <Twingy> do I uninstall it and reinstall it?
[03:16:44] <Twingy> I think apt is confused
[03:16:53] <fenn> are you sure it's not libatk1.0-0 (not libatk-1.0-0)
[03:17:06] <JymmmEMC> no, it's the pkg dependancies... read the apt-get man page for options
[03:17:13] <SWPadnos> --force-depends
[03:17:22] <JymmmEMC> what SWPadnos said
[03:17:27] <SWPadnos> there's another more specific one as well
[03:17:29] <fenn> --force is not something the package maintainer should be doing
[03:17:35] <Twingy> well, I'm not trying to bypass it, I'm trying to make it so anyone can double click on my .deb package and sail through the install
[03:17:39] <SWPadnos> oh, well there is that :)
[03:17:52] <JymmmEMC> Twingy: so you made a pkg and fucked it up?
[03:18:04] <eric_U> maybe not in that order?
[03:18:06] <Twingy> JymmmEMC, it installs just fine double clicking on it on my gutsy machine
[03:18:08] <JymmmEMC> =)
[03:18:17] <Twingy> on the dapper machine it's confuzzled
[03:18:30] <JymmmEMC> Twingy: Yes, on *YOUR* machine, but that aint the point of you making a pkg now is it =)
[03:18:34] <eric_U> don't you just hate that?
[03:18:36] <SWPadnos> you can use (package1 or package2) or similar in the dependency info
[03:18:48] <Twingy> let me take it to another gutsy machine
[03:18:53] <Twingy> and see if it complains
[03:19:04] <Twingy> then I know if it's a dapper problem
[03:19:05] <JymmmEMC> Twingy: you probably missed some lil tiny thing when you made the pkg is all.
[03:19:20] <eric_U> dapper machines look good, but what you really want is a gutsy machine, for the guts
[03:19:29] <Twingy> jymm, not sure what though, I have gtk2.0-dev in the dependencies
[03:19:34] <Twingy> and gtk requires atk
[03:19:41] <fenn> Twingy: are you sure the libatk-1.0-0 package install is spelled with a dash?
[03:19:43] <Twingy> so that's probably where it stems from
[03:19:54] <JymmmEMC> Twingy: you've made a pkg that needs a dev dependancy?
[03:20:06] <Twingy> fenn, I'll double check that it matches character for character, one moment, computer is in the garage with the CNC mill
[03:20:10] <Twingy> I'm in the basement
[03:20:18] <Twingy> let me turn on VNC connection
[03:20:31] <Twingy> jymm, I think dependencies are ok, brb
[03:20:34] <JymmmEMC> Twingy: If apt.sources doesn't have dev enabled I can see an issue here
[03:22:00] <Twingy> fenn, the package it's complaining about and the one installed are verbatim
[03:22:05] <Twingy> character for character
[03:22:22] <Twingy> I'll do a screen shot if you want
[03:22:31] <fenn> what happens if you try to install gtk2.0-dev by itself?
[03:22:40] <fenn> i.e. leaving your package out of the whole process
[03:22:47] <Twingy> fenn, one moment
[03:23:16] <JymmmEMC> apt-cache show <pkg_name>
[03:23:23] <Twingy> installing libgtk2.0-dev
[03:23:34] <eric_U> I think my dog was playing with my embedded computer
[03:23:44] <JymmmEMC> apt-cache depends <pkg_name>
[03:23:48] <Twingy> ever since I installed 802.11n over the weekend my connection to the garage is smoking fast
[03:24:01] <eric_U> I need to upgrade
[03:24:03] <fenn> libgtk2.0-dev Depends: libatk1.0-dev
[03:24:30] <eric_U> obviously
[03:25:00] <Twingy> ok, libgtk2.0-dev was installed by hand, but same problem "Error: Dependency is not satisfiable: libatk1.0-0"
[03:25:15] <Twingy> and Synaptic reports that "libatk1.0-dev" is installed
[03:25:18] <Twingy> err
[03:25:19] <fenn> libgtk2.0-0 Depends: libatk1.0-0
[03:25:25] <Twingy> libatk1.0-0
[03:25:27] <fenn> so what happens if you install libgtk2.0-0
[03:25:32] <Twingy> sec
[03:25:45] <fenn> twingy why dont you just ssh to the other machine?
[03:25:50] <Twingy> fenn, it's already installed
[03:26:06] <fenn> ok, i pictured you running out to the garage and back
[03:26:09] <Twingy> no
[03:26:12] <Twingy> I'm vnc'd in
[03:26:17] <Twingy> remote desktop
[03:26:27] <eric_U> this is why it's good to be working on a remote machine, can't punch it
[03:26:46] <Twingy> I just think dpkg is wigging out
[03:26:56] <eric_U> but sometimes it's fun to go kick the robots
[03:27:03] <Twingy> it's like yo, this package isn't there, and synaptic is like yo, the package is right here f00
[03:27:41] <Twingy> this is a vanilla EMC install too
[03:27:47] <Twingy> haven't really installed anything
[03:27:54] <Twingy> EMC live CD thing
[03:28:01] <Twingy> installed to HD
[03:28:05] <fenn> libgtk2.0-0 installs fine?
[03:28:11] <Twingy> fenn, yea, it was already there
[03:28:47] <Twingy> libgtk2.0-0, libgtk2.0-bin, libgtk2.0-common, libgtk2.0-dev are all installed
[03:28:55] <Twingy> yet the problem remains
[03:29:11] <Twingy> libatk1.0-0 and libatk1.0-dev are also installed
[03:29:32] <Twingy> can I give you the .deb package?
[03:29:49] <SWPadnos> can you pastebin the control file?
[03:29:53] <Twingy> yea, sec
[03:29:56] <SWPadnos> (I think that's teh one anyway)
[03:30:29] <fenn> i notice the emc package Depends: libatk1.0-0 (>= 1.9.0)
[03:30:33] <Twingy> http://pastebin.com/m7d372a87
[03:31:17] <fenn> those ${shlib} should turn into real package names after compiling
[03:31:43] <fenn> if you unpack the deb you can get the real control file
[03:32:13] <Twingy> umm, dpkg something
[03:32:43] <SWPadnos> dpkg -e
[03:33:04] <SWPadnos> according to man on my 7.10 machine
[03:33:31] <Twingy> http://pastebin.com/m3bc3a34a
[03:33:53] <Twingy> 1.13.2 perhaps?
[03:34:15] <SWPadnos> what's installed now?
[03:34:27] <fenn> now we're getting somewhere eh
[03:34:43] <fenn> dapper provides 1.11.4-0ubuntu1
[03:34:44] <Twingy> 1.11.4-0
[03:34:47] <Twingy> right
[03:34:53] <SWPadnos> we havea weiner!
[03:34:57] <Twingy> yep
[03:35:01] <Twingy> ok, hrm
[03:35:08] <Twingy> so maybe I need to build the package on dapper?
[03:35:16] <Twingy> and then it'll work on gutsy
[03:35:27] <Twingy> instead of build on gutsy to work on dapper
[03:35:30] <fenn> or figure out how to use autoconf or whatever is making that ${shlibs} stuff
[03:35:30] <SWPadnos> there may be anotehr way, but that seems like the easiest thing to think of at the moment
[03:35:54] <Twingy> I'm just going through this debian package tutorial
[03:36:04] <Twingy> I haven't been doing this but for an hour
[03:36:14] <Twingy> so I'll dig into it
[03:36:15] <SWPadnos> the package info available on guty/feisty will have the versions needed for that - there's no way to know on that system how far back you can go
[03:36:19] <Twingy> should be easy enough
[03:36:47] <Twingy> well, if I can support dapper, edgy, feisty, gutsy, I'm good
[03:37:08] <tomp2> Twingy: please post the url to this channel when you get it done (gotta go now )
[03:37:09] <SWPadnos> don't forget Hardy! :)
[03:37:22] <Twingy> well hardy heron will just work
[03:37:25] <fenn> really though, i dont think the package should specify all the dependencies recursively down the tree
[03:37:36] <SWPadnos> no, it shouldn't
[03:39:01] <Twingy> who's working on migrating the EMC live CD to heron?
[03:39:19] <SWPadnos> Alex, and jepler has been doing some kernel bulding as well
[03:39:40] <Twingy> think they would mind tossing gcam onto the image?
[03:39:55] <SWPadnos> if there's room, I bet we can work that out :)
[03:40:05] <Twingy> it's like 780kB
[03:40:23] <Twingy> I could probably trim it down some
[03:40:39] <SWPadnos> there's probably room for that, though the CD is pretty packed
[03:40:41] <fenn> then there's the dependencies
[03:40:46] <Twingy> yea
[03:40:51] <Twingy> gtkglext and gtk
[03:40:55] <fenn> i wouldnt worry about <1MB
[03:40:57] <SWPadnos> I think gtk is needed by EMC anyway due to halscope/halmeter and the like
[03:41:06] <Twingy> so gtkglext
[03:41:10] <Twingy> which is fairly tiny
[03:41:17] <SWPadnos> and we use gl too, so one never knows
[03:41:23] <Twingy> right
[03:42:04] <Twingy> do you guys keep the debian build stuff in the repository
[03:42:11] <SWPadnos> yep
[03:42:17] <Twingy> I'm thinking about comitting my debian setup into my subversion repo
[03:42:27] <Twingy> kk
[03:42:32] <SWPadnos> it's probably helpful - you can track changes and stuff ;)
[03:42:35] <Twingy> makes life easier
[03:42:40] <SWPadnos> http://cvs.linuxcnc.org/cvs/emc2/debian/control.in?rev=1.23
[03:43:12] <SWPadnos> I don't see gtkglext specifically listed in there
[03:43:48] <fenn> not that emc necessarily does it the "right" way
[03:44:15] <SWPadnos> well, we are idiots after all - no way to know
[03:44:15] <fenn> i still cant figure out how to get it to accept libpth20
[03:44:28] <Twingy> SWPadnos, hrm
[03:44:31] <SWPadnos> heh
[03:45:08] <SWPadnos> fenn, vs. what?
[03:45:15] <Twingy> that sucks
[03:45:26] <Twingy> I guess libgtkglext didn't get added until edgy?
[03:45:44] <fenn> eric_U: scroll down to "smart structures" for an interesting read: http://www.robertforward.com/Fast_Forward_Fifty_Years.htm
[03:45:48] <Twingy> maybe I should wait until hardy is EMC standard
[03:46:00] <SWPadnos> hmmm
[03:46:19] <Twingy> april 2008 release
[03:46:22] <fenn> SWPadnos: it wants libpth2
[03:46:32] <fenn> somehow it gets that package name from libpth-dev
[03:46:33] <Twingy> so that's only like 1-2 months away
[03:47:06] <Twingy> I'll just wait until the LIVE CD is built on hardy
[03:47:17] <SWPadnos> yep - about 6 weeks, and an unspecified amount of added time for us (mostly Alex) to make sure all the RT stuff works :)
[03:47:49] <SWPadnos> http://packages.ubuntu.com/search?suite=dapper&searchon=names&keywords=gtkglext
[03:47:55] <SWPadnos> looks like there's something for dapper
[03:48:09] <SWPadnos> fenn, hmmm. where's pth specified there?
[03:48:48] <fenn> SWPadnos: in debian/configure: KERNEL_HEADERS=libpth-dev
[03:48:54] <SWPadnos> ah
[03:49:55] <fenn> only for -sim packages
[03:49:59] <fenn> (sry)
[03:50:28] <fenn> but it makes sense to try to support sim packages on more distro's
[03:51:09] <SWPadnos> I don't see a sim package in that file ...
[03:51:29] <SWPadnos> the control file that is
[03:51:40] <fenn> case $TARGET in sim)
[03:52:00] <SWPadnos> I mean in the debian/control file
[03:52:07] <SWPadnos> err - control.in I guess
[03:52:28] <fenn> uh, yes, and now you can see why i flounder so much
[03:52:48] <SWPadnos> actually, I'm not sure what the problem is :)
[03:53:19] <SWPadnos> libpth-dev depends on libpth20 (on my Feisty machine)
[03:53:38] <fenn> but emc2-sim depends on libpth2
[03:54:12] <fenn> (which presumably doesn't exist on feisty, and certainly doesnt exist on debian lenny)
[03:54:21] <SWPadnos> ok - I don't see any sim package listed at all in the control.in file, so I'm not sure where to look
[03:54:52] <fenn> i think it gets put in a new control file that gets created at package build time
[03:56:00] <fenn> when you run debian/configure it modifies control.in to make the control file (for both sim and rt packages)
[03:57:06] <SWPadnos> ok. I guess someone needs to do that on an older system - all I see referenced is libpth-dev, so libpth2 vs libpth20 is coming from the depends on the build machine
[03:57:49] <fenn> right
[03:58:04] <SWPadnos> well, get cracking - yer the one with the problem! :)
[03:58:16] <fenn> ideally the package would say Depends: (libpth2 | libpth20)
[03:58:40] <fenn> i dont know where the libpth2 comes from though (maybe the 'subst control.in > control' does it?)
[03:58:49] <SWPadnos> I wonder if adding that explicitly would override the automagic dependency from libpth-dev
[03:59:16] <SWPadnos> dunno - it does seem weird that all those would be listed - that's dpkg's job
[04:03:45] <fenn> hmm.. "Regardless of what you put in the Depends: line,
[04:03:46] <fenn> the built binary package does depend on the environment in which it is
[04:03:47] <fenn> built, because of the way dynamic linking works."
[04:13:34] <fenn> but i think libpth2 provides the file /usr/lib/libpth.so.20 so that should not matter
[04:22:40] <fenn> oh, right, now i remember why i gave up on this earlier.. all the package build dependencies
[04:25:44] <eric_U> shouldn't this all be automated?
[04:26:27] <fenn> the problem is it's too automated
[04:26:53] <fenn> in order to build the emc2-sim package you have to build all the docs, and recompile emc2 from scratch
[04:27:00] <fenn> that takes forever on this machine
[04:34:19] <fenn> most likely i am not understanding how to do something
[04:39:26] <toastydeath> i had a performance review at work today
[04:39:41] <toastydeath> and my boss goes "i don't know what i'm supposed to tell you in this, because we work together all day"
[04:39:55] <toastydeath> so we just bullshitted for an hour and called it good
[04:39:59] <toastydeath> a+ review
[07:22:22] <micges> hello everyone
[07:38:52] <Unit41> micges your ugly im good looking
[07:52:12] <fenn> tape slide rule, just as i imagined: http://www.pat2pdf.org/patents/pat2826361.pdf
[07:53:31] <fenn> i bet it would be smaller with plastic tape (tighter bend radius = smaller spools)
[08:08:42] <alex_joni> now I know why I like wireless http://www.techamok.com/?pid=4244
[08:13:08] <fenn> maybe they had just watched 'fight club'
[12:12:37] <Guest532> Guest532 is now known as skunkworks_
[13:02:02] <cradek_> cradek_ is now known as cradek
[13:05:18] <jepler> fenn: you are talking about the install-time dependency of emc2-sim on libpth2? that is created by ${shlib:Depends} which is created by dh_shlibdeps. It finds this dependency by seeing that bin/rtapi_app is linked to /usr/lib/libpth.so.20, and that /usr/lib/libpth.so.20 is provided by libpth2 (on ubuntu dapper systems). If you build on a system that uses a different package name for the package that provides libpth.so.20, the dependency will be differ
[13:05:52] <jepler> if you're trying to install a dapper .deb on lenny, I think you should just give up; that's not intended to work.
[13:27:14] <alex_joni> http://www.techamok.com/?pid=4243
[13:32:59] <skunkworks_> Cool :)
[13:46:56] <archivist_win> .join #php-support
[13:54:30] <archivist> very good robot, the reaction to a kick in the side and slipping on ice is good
[14:03:20] <lerman> I just took a look at the EMC requested feature list on the Wiki. http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?EmcFeatures.
[14:04:19] <lerman> Number 18 requests a macro language. I believe that the recent named o-word feature completes the implementation of this requirement and would like to move this to the done list.
[14:04:31] <lerman> Does anyone disagree?
[14:09:09] <alex_joni> hmm.. it's been a long time since I added that
[14:09:31] <alex_joni> but I think there are a couple of things that a general macro language could do, and g-code isn't suitable for that
[14:09:44] <alex_joni> like.. check current axis position, and do something if ..
[14:11:32] <cradek> so many features seem to want interpreter introspection
[14:12:36] <lerman> Indeed. That should be added to the list. :-)
[14:14:02] <skunkworks_> couldn't a 'simple' (yes I know) kins module do simple coordinate rotation? taking a point a degree of rotation?
[14:14:24] <skunkworks_> taking a point and degree of rotation
[14:15:23] <cradek> sort of. the problem is that when you rotate, you get a jump in position
[14:15:52] <cradek> so this rotation actually needs to be before the interpolator [motion controller] so you can make a nice path from the old point to the new (rotated) point
[14:16:06] <skunkworks_> ah
[14:16:12] <cradek> it could easily be at the canon level (right at the output of the interpreter)
[14:16:42] <cradek> you'd just make the interpreter call a new canon call ROTATE_Z(degrees)
[14:17:00] <cradek> or at least that's how I would do it.
[14:17:20] <lerman> How would the center of rotation be specified?
[14:17:27] <cradek> I don't know
[14:17:57] <cradek> rotating around the machine origin would be the easiest to implement...
[14:19:48] <cradek> a gcode programmer could "easily" (ha) move his g54 origin to compensate
[14:26:01] <cradek> I decided that I was wrong
[14:26:15] <lerman> About what?
[14:26:40] <cradek> there should not be a canon ROTATE call. Like the various translations, rotation should be handled in the interpreter, and the canon calls should retain their simplicity of being only in machine coordinates
[14:27:15] <lerman> Makes sense.
[14:27:49] <lerman> Didn't someone post some gcode that did rotations of g0 g1 g2 g3?
[14:28:03] <lerman> Gcode subroutines, I believe.
[14:30:51] <cradek> lerman: could be. it doesn't seem hard, but I think it would be somewhat nasty when done
[14:37:36] <lerman> That would be one neat application of the feature I've considered to allow overriding of existing gcodes. If we would replace g1 by an integrator supplied g1 that worked with rotated coordinates that would provide for a neat user interface.
[14:37:57] <lerman> Of course the integrator supplied g1 would have to be able to invoke the original g1.
[16:00:23] <The_Ball> Any eagle experts around? I'd like to connect some header pins to some nets, but not the standard supply nets, just arbitrary nets, any hits?
[16:07:55] <cradek> nets become connected, even if they aren't connected by a line on the screen, if their names are set the same
[16:24:03] <micges> hello all
[16:45:36] <The_Ball> cradek, thank, that's great
[17:31:35] <BigJohnT> hi micges
[18:07:12] <beege> Hi everyone, I have a router with Softservo on it, runs on Win2000. Going to see what it would take for EMC to run it
[18:07:37] <skunkworks_> sounds like fun.. what kind of hardware are the motors and drives?
[18:07:48] <SWPadnos> probably just the liveCD, unless there's some proprietary interface to the motors
[18:08:10] <beege> Motors are .75kW brushless servos X4
[18:08:33] <SWPadnos> hmmm. is there an interface card of some sort?
[18:08:39] <beege> Drives are Ormec servowire dirves
[18:08:47] <SWPadnos> oh. well then you'll need more :)
[18:08:54] <beege> more what?
[18:09:01] <SWPadnos> more than just the liveCD :)
[18:09:22] <beege> these run thru firewire rght now
[18:09:39] <skunkworks_> That is not going to work with emc...
[18:10:11] <beege> I've talked to Ormec, Softservo
[18:10:24] <beege> no help ther
[18:10:25] <skunkworks_> Do you know how the softservo whatever comunicates with the ormec drives? step/dir,+/-10v...
[18:10:27] <beege> there
[18:10:35] <SWPadnos> firewire
[18:10:47] <skunkworks_> oh - really?
[18:10:55] <beege> really
[18:11:02] <skunkworks_> heh
[18:11:34] <beege> drives were obsloete the day after they were made
[18:12:38] <beege> is there any way of finding out how to talk to the drives, esp. tuning?
[18:12:54] <SWPadnos> sadly, it's not always easy to match drives to motors when they're brushless
[18:13:08] <SWPadnos> not really. you have (at least) two problems:
[18:13:22] <SWPadnos> 1) if the softservo folks don't tell you the protocol, then there's nothing you can do
[18:13:23] <beege> at least
[18:13:37] <SWPadnos> 2) there is no realtime firewire driver for EMC2
[18:14:15] <SWPadnos> oh, and 3) EMC2 doesn't usually play well with "smart" drives - EMC2 likes to do the PID itself, using velocity / torque more servos
[18:14:28] <beege> ok. i've sort of come to the conclusion that new drives are needed
[18:14:42] <beege> How to match to servos?
[18:15:02] <SWPadnos> which brings a different problem: you need to be careful to be sure that any drive you ge tcan drive your servos
[18:15:10] <SWPadnos> or you need to replace those too
[18:15:18] <SWPadnos> ... get can ...
[18:15:44] <beege> what d oI need to know?
[18:16:22] <beege> ...do I...
[18:16:28] <beege> :)
[18:18:06] <SWPadnos> well, if you have the torque/speed/power ratings of the servos you have now, you can match that with some other system :)
[18:19:21] <beege> I'm at home right now, specs are at the shop. Do I call someone with those specs to select?
[18:20:03] <beege> Whom might you suggest?
[18:20:03] <skunkworks_> who are some ac servo amp suppliers?
[18:20:55] <SWPadnos> well, it really depends on your budget
[18:21:19] <SWPadnos> I only have experience with Yaskawa, and that's in the $1500 range per axis (for servo and drive)
[18:21:40] <SWPadnos> I've seen a lot of stuff on eBay, but it also depends on your power needs
[18:21:43] <beege> Ok... spindle drive is Yaskawa
[18:22:09] <SWPadnos> do you know for sure that your drives have no input other than firewire?
[18:22:46] <beege> There's an I/O plug on the side, maybe 4 sockets per
[18:22:56] <SWPadnos> SD230 or SD460?
[18:23:20] <beege> Don't know what that means
[18:23:37] <SWPadnos> those are some series of Ormec ServoWire drives: http://www.ormec.com/smlc/smlc_products/SD_drives_smlc.htm
[18:24:29] <skunkworks_> hast to be the 230 I would guess as he said .75kw..
[18:24:40] <SWPadnos> yeah - I was thinking that
[18:24:44] <beege> Yes, runs on 208V
[18:24:52] <SWPadnos> but this is one of those sites that requires registration to get their friggin manuals
[18:25:16] <beege> Had a hard time with registration
[18:25:49] <SWPadnos> hmmm. there's a
[18:25:55] <SWPadnos> an SM line as well.
[18:26:10] <SWPadnos> similar power ranges, also 115V / 230V
[18:28:48] <beege> I think they are SM something
[18:30:55] <beege> This was supposed to be a commercial machine, not a hobby tool. So maybe I contact Yaskawa? Will drive they recommend work with EMC?
[18:32:12] <SWPadnos> their SGDH drives can accept analog or step/dir inputs
[18:32:41] <SWPadnos> they have add-ons for their "smart" interface "Mechatrolink", which you don't need
[18:38:23] <beege> OK.... once I find drives that will work, what kind of card do I need in the PC?
[18:39:01] <SWPadnos> you're unlikely to find drives that will run your motors, if they are actually brushless DC motors (vs. AC servos)
[18:39:31] <SWPadnos> the Yaskawa drives are pretty much matched to their motors, I'm not sure if you can use them on other motors easily
[18:39:38] <beege> I'll double check that and get back to you
[18:40:37] <beege> Thanks, and I have to be going... more to follow
[18:41:17] <SWPadnos> ok
[18:42:35] <skunkworks_> as far as the card for the pc... http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?EMC2_Supported_Hardware
[18:42:58] <skunkworks_> mesa is well liked
[18:52:31] <gene> Ok, I'm out here and playing. To get rid of that initial RTAPI error, I have to set a base_period of 140000ns, but then nothing will run without following errors.
[18:53:34] <SWPadnos> I wouldn't worry about it
[18:53:47] <SWPadnos> it's annoying, but I'm pretty sure it's nothing more than that
[18:54:08] <cradek> gene: did you say you are using the nvidia video driver?
[18:54:24] <gene> At 30000ns, I get both errors at startup, but motion.servo.overruns never shows anything but 0.
[18:54:29] <gene> yes
[18:55:02] <cradek> the fact that you only have a problem when displaying to that screen, combined with my previous bad experience with the nvidia driver, makes me think that's probably the culprit
[18:55:07] <gene> nv won't run any nvidia card i own at more than 640x400
[18:55:20] <cradek> switching to the vesa driver, or a non-nvidia card, would be interesting to try
[18:55:24] <SWPadnos> I think the RTAI (or RTAPI)-produced error is a "zero-tolerance" threshold - if the timing is one clock too late, the error gets flagges (the deadline was missed)
[18:55:42] <cradek> SWPadnos: what makes you think that?
[18:55:43] <SWPadnos> the other RT error has a 20% margin before it gets reported
[18:55:51] <SWPadnos> I looked at the code at some point :)
[18:55:56] <SWPadnos> (but I could be remembering wrong)
[18:56:08] <cradek> SWPadnos: (I thought it was the same thing that caused an overrun report in the original rtai latency test)
[18:56:33] <SWPadnos> I'm not sure. I believe there are two places where an error can be reported
[18:56:35] <cradek> the problem with the test in the servo thread is that it's the slowest thread
[18:56:57] <SWPadnos> one is pretty low-level in RTAI/RTAPI, before any HAL threads are executed
[18:57:03] <SWPadnos> the other is in the motion controller
[18:57:12] <cradek> yes
[18:57:18] <gene> yes, motion.servo.last.preiod is dancing about 134xxxxx way slow.
[18:57:26] <SWPadnos> I don't think they use the same tolerance
[18:57:58] <SWPadnos> at the low level, a missed deadline is a pretty critical error. in motion, we know the absolute timing isn't as important
[18:58:06] <gene> yes, motion.servo.last.periodns is in the 900000's range is dancing about 134xxxxx way slow.
[18:58:37] <gene> damn up arrow
[18:58:50] <SWPadnos> heh
[18:58:56] <gene> ignore the last half of that
[18:59:25] <gene> I have an old ATI card, would that be worth a try?
[18:59:27] <SWPadnos> probably in the high 900000's
[18:59:41] <SWPadnos> it shouldn't matter much with the VESA driver
[18:59:59] <cradek> yeah just switch to vesa, I think the nvidia cards work fine with vesa
[19:00:03] <cradek> they are slow of course
[19:00:15] <gene> yes, in the 96xxxxx area
[19:00:31] <gene> How do I do that?
[19:01:02] <gene> in xorg.conf, or etc/modprobe
[19:01:12] <SWPadnos> xorg
[19:01:28] <gene> lemme get that in an editor
[19:01:42] <SWPadnos> if you run AXIS remotely and get no error at startup, it's very very likely that the video driver is the problem
[19:01:50] <SWPadnos> or at least one problem :)
[19:02:27] <gene> got it, next?
[19:02:52] <SWPadnos> err - continue correcting problems until all of them have been eliminated?
[19:03:13] <gene> no, how to make it use the vesa driver?
[19:03:31] <SWPadnos> oh :)
[19:04:12] <SWPadnos> dunno - you can change the driver to vesa instead of nv/nvidia
[19:04:25] <skunkworks_> you need to be sudo to do this..
[19:04:27] <SWPadnos> there may also be changes needed to load mesa instead of nvidia glx
[19:04:36] <gene> the last time I tried to muck with that file on this machine, I wound up re-installing, so how to I get a non x startup? Thats first...
[19:04:37] <skunkworks_> vesa ;)
[19:05:39] <gene> what runlevel in /etc/inittab? Remember I'm fedora everyplace else.
[19:09:53] <gene> I reset it to a 1, lets go for effect...
[19:10:01] <SWPadnos> runlevel 2 is usually networking/multiuser
[19:10:06] <SWPadnos> but no X
[19:10:23] <gene> yes but it starts x, and if x can't run, there is no rescue
[19:10:30] <SWPadnos> ?
[19:10:37] <SWPadnos> runlevel 1 is single, no networking
[19:10:49] <gene> my login screen is an x screen I think.
[19:10:54] <SWPadnos> 2 is usually (on non-Debian) multiuser/netwroked, so you can ssh into it
[19:10:58] <SWPadnos> not in runlevel 2 :)
[19:11:15] <SWPadnos> on Debian/Ubuntu, runlevel 2 is also graphical
[19:11:16] <cradek> why would you mess with your inittab? just change the video driver
[19:11:26] <gene> ok, how do I get a non x terminal in runlevel 2 then?
[19:11:42] <cradek> Ctrl-Alt-F1
[19:11:45] <SWPadnos> locally, the console is a terminal
[19:11:52] <SWPadnos> remotely, just SSH like normal
[19:11:53] <cradek> or, change it and then log out
[19:12:28] <SWPadnos> or /etc/init.d/gdm restart if you're rmote
[19:13:47] <SWPadnos> hmmm. or kdm since that machine is kubuntu
[19:14:05] <gene> I set it to vesa, and a ctl-alt-backspace is 5 seconds away
[19:14:13] <cradek> there you go, that's all it takes
[19:15:42] <SWPadnos> I always get a kick out of your sign-on message
[19:15:49] <SWPadnos> "gene@pool" ...
[19:17:35] <gene> Crazy, I'm back to a 1600x1200 screen, 'pool'? NDI where that is set, this is konversation
[19:18:20] <skunkworks_> gene: have you ever run latency-test from terminal?
[19:19:28] <gene> now what? Look at this:
[19:19:47] <gene> Traceback (most recent call last):
[19:19:51] <gene> File "/usr/bin/axis", line 3369, in ?
[19:19:55] <gene> o = MyOpengl(widgets.preview_frame, width=400, height=300, double=1, depth=1)
[19:19:59] <gene> File "/usr/bin/axis", line 255, in __init__
[19:20:03] <gene> Opengl.__init__(self, *args, **kw)
[19:20:07] <gene> File "debian/tmp/usr/lib/python2.4/site-packages/rs274/OpenGLTk.py", line 218, in __init__
[19:20:11] <gene> File "debian/tmp/usr/lib/python2.4/site-packages/rs274/OpenGLTk.py", line 167, in __init__
[19:20:15] <gene> File "debian/tmp/usr/lib/python2.4/site-packages/rs274/OpenGLTk.py", line 93, in __init__
[19:20:19] <gene> File "/usr/lib/python2.4/lib-tk/Tkinter.py", line 1862, in __init__
[19:20:19] <cradek> probably nvidia has porked up your mesa opengl libraries
[19:20:23] <gene> self.tk.call(
[19:20:27] <gene> _tkinter.TclError: Togl: X server has no OpenGL GLX extension
[19:20:28] <cradek> ok stop
[19:20:31] <gene> Shutting down and cleaning up EMC2...
[19:20:35] <gene> Cleanup done
[19:20:39] <gene> EMC terminated with an error. You can find more information in the log files
[19:20:43] <gene> /home/gene/emc_debug.txt
[19:20:47] <gene> and
[19:20:48] <gene> /home/gene/emc_print.txt
[19:20:48] <gene> as well as in the output of the shell command 'dmesg' and in the terminal
[19:20:48] <gene> is that the openGL.so being only the nvidia vesion?
[19:20:49] <gene> because that is all thats installed.
[19:21:18] <gene> or should I say, all I can find with sudo locate
[19:21:41] <gene> I need the mesa packages now?
[19:22:09] <cradek> other than going back in time and not installing that accursed .run file, I don't know how to fix it
[19:22:45] <gene> I think if I put in the mesa.* stuff it will overwrite the nvidia stuffs.
[19:23:00] <gene> lemme take a look with adept, nbrb
[19:30:48] <gene> and I can't find it in adcepts listings
[19:30:56] <gene> still loking though
[19:31:41] <cradek> libgl1-mesa?
[19:31:47] <cradek> libglu1-mesa?
[19:33:11] <jepler> there are also other xorg.conf changes made by the nvidia driver -- after trying to fix all the files that the nvidia installer renamed or (hopefully not!) removed, you might need to run dpkg-reconfigure on the X server to get a working opengl again
[19:33:55] <gene> ok, I'll give that a shot, can't hurt at this point
[19:36:58] <gene> reconfigure what, nothing I've enetred has been valid
[19:37:16] <SWPadnos> dpkg-reconfigure xserver-xorg
[19:38:05] <fenn> isnt there a command that inspects every package on the system to see if its files are installed?
[19:38:25] <skunkworks_> don't auto-detect. pick visa and the defaults for everything else. (atleast that is what I do)
[19:43:52] <gene> Still no GLX extension
[19:46:41] <gene> And the xorg log says it loading the nvidia version, bahh
[19:46:51] <gene> blow it away.
[19:46:54] <SWPadnos> wait
[19:47:07] <SWPadnos> look at the xorg.conf file for "nvidia-glx"
[19:47:14] <SWPadnos> or fglrx
[19:47:25] <SWPadnos> and remove (or comment) those references
[19:47:40] <SWPadnos> you amy also be able to use dpkg to reinstall mesa
[19:48:07] <SWPadnos> and possibly to remove the nvidia restricted drivers (though you'll need to be careful there, since there may be unintended dependencies)
[19:55:18] <gene> adept is now removing everything, damn
[19:56:04] <gene> the whole kubuntru desktop, I was going to remove, both nvidia and mesa and re-install mesa only
[19:56:28] <gene> if it survives on in memory long enough to do that...
[19:58:27] <fenn> unix wont delete files that are open by a program (i.e. running)
[19:58:43] <cradek> depends what you mean by delete
[20:00:38] <gene> its survived so far, but the kubuntu reinstall is busy pulling 400 megs of language packs...
[20:08:04] <gene> FWIW, there wasn't any rference to nvidia-glx in the xorg.conf, just plain glx, which then loaded the nvidia verswion silently.
[20:10:21] <cradek> didn't apt warn you that it was going to delete 400MB of stuff you still wanted?
[20:12:09] <gene> adept, and no, it just did it
[20:12:26] <cradek> amazing
[20:12:39] <cradek> the commandline tools warn you, of course
[20:12:46] <fenn> never give your keys to a drunk programmer
[20:13:06] <gene> well, i think it might be ready bfor a restart, here goes
[20:13:36] <gene> Hey, I'm sober, darnit!
[20:15:08] <gene> back now to see what the Xorg log says
[20:16:34] <gene> (WW) Warning, couldn't open module glx
[20:16:57] <gene> so which package is it in?
[20:21:54] <gene> adept gives me quite an extensive list for opengl as search term
[20:22:36] <fenn> search for glx
[20:22:46] <fenn> but i think that's not what you want to do
[20:22:57] <fenn> grep glx xorg.conf
[20:26:37] <gene> Load "glx"
[20:28:07] <gene> and the last install pulled in a linux-image file of some sort, wtf?
[20:29:56] <gene> I think I'd better check the menu.lst
[20:31:20] <gene> its ok yet
[20:34:36] <gene> still getting this:
[20:34:40] <gene> II) LoadModule: "glx"
[20:34:40] <gene> (WW) Warning, couldn't open module glx
[20:34:41] <gene> (II) UnloadModule: "glx"
[20:34:41] <gene> (EE) Failed to load module "glx" (module does not exist, 0)
[20:35:06] <gene> so what package is this module in?
[20:40:26] <fenn> holy crap mfg.com is what i've been yammering about for years, but it's already done and doing hundreds of millions of dollars worth of transactions
[20:42:10] <gene> installed the packages an error message said, no effect
[20:42:13] <gene> II) LoadModule: "glx"
[20:42:14] <gene> (WW) Warning, couldn't open module glx
[20:42:14] <gene> (II) UnloadModule: "glx"
[20:42:14] <gene> (EE) Failed to load module "glx" (module does not exist, 0)
[20:45:14] <fenn> gene: you need to edit your xorg.conf file (remove the LoadModule: "glx")
[20:48:30] <gene> don't I need that with vesa?
[20:49:02] <gene> and emc claims it can't run because its not available
[20:49:23] <dimas> hello all
[20:49:26] <fenn> uh, hrm. try installing libgl1-mesa-swx11 or whatever it's called
[20:50:10] <dimas> anyone could recommend schematics for diy stepmotor driver?
[20:50:28] <fenn> dimas: there are several on pminmo.com
[20:50:33] <cradek> gene, fenn: this is from the readme that comes in the accursed nvidia .run file: http://pastebin.ca/949463
[20:50:41] <cradek> you will have to reverse/undo these changes
[20:50:59] <cradek> or, like many have suggested, dpkg-reconfigure xserver-xorg and generate a new config
[20:51:13] <fenn> i thought he did that already.. meh
[20:51:29] <cradek> oh maybe he did
[20:52:34] <gene> and to top that off, now:
[20:52:34] <fenn> gene: you will have to reverse/undo these changes: http://pastebin.ca/949463
[20:52:47] <dimas> fenn: thanks, i think a have it in saved links just hard to decide which one to choose, checking now
[20:52:50] <fenn> either manually or by doing dpkg-reconfigure xserver-xorg
[20:52:52] <gene> ene@shop:~/emc2/nc_files$ emc
[20:52:52] <gene> bash: emc: command not found
[20:52:52] <gene> gene@shop:~/emc2/nc_files$
[20:53:02] <fenn> emc is not installed
[20:53:24] <gene> geezuss h krist
[20:53:47] <fenn> dimas: i suggest the l298 one, as it's common, simple, rugged, and modular
[20:54:00] <fenn> L297-8
[20:55:18] <gene> has anyone ever commented that this package manager sucks dead toads through soda straws?
[20:55:30] <fenn> i dont think anyone else uses adept
[20:55:41] <fenn> i use apt-get, most users use synaptic
[20:56:14] <gene> ok, to get glx, whats the apt-get linew?
[20:56:20] <gene> ok, to get glx, whats the apt-get line?
[20:56:23] <fenn> you dont want glx
[20:56:40] <fenn> glx is provided by the nvidia module which you are trying to get rid of
[20:56:42] <gene> emc says it needs it???
[20:57:10] <fenn> emc needs some kind of opengl infrastructure. right now you probably have libgl1-mesa-glx installed, which is what emc calls (?)
[20:57:11] <gene> oh hell, I nuked all that both my hand and witth adept an hour ago.
[20:58:24] <fenn> gene: sudo apt-get install libgl1-mesa-swx11
[20:58:57] <dimas> fenn: and i need to do 3 unit for each axis, something else?
[20:59:09] <fenn> dimas: right, one for each axis
[20:59:20] <dimas> yes correct
[21:00:03] <gene> gene@shop:~$ sudo apt-get install libgl1-mesa-swx11
[21:00:03] <gene> Reading package lists... Done
[21:00:04] <gene> Building dependency tree... Done
[21:00:04] <gene> E: Couldn't find package libgl1-mesa-swx11
[21:00:06] <fenn> i never understood why he removed the eagle files from his site
[21:00:27] <dimas> fenn: thanks, now will ping local supplier for the parts
[21:00:27] <fenn> gene: eh, well, sorry
[21:00:39] <gene> in oter words, that file doesn't excst
[21:00:48] <fenn> right, it must be named someting else in ubuntu
[21:02:08] <fenn> libgl1-mesa-swrast
[21:05:31] <ken_> can someone help, i need to cut a 525rad in emc2 ?
[21:06:08] <cradek> what the heck is a 525rad?
[21:06:27] <ken_> just over a 1/2"
[21:06:39] <tomp> 0.525" radius i think he means
[21:06:45] <ken_> that's what they want
[21:06:50] <ken_> ooppsss yes
[21:06:51] <cradek> a .525 radius what?
[21:06:54] <ken_> .525
[21:07:02] <tomp> mm or inch
[21:07:06] <ken_> inch
[21:07:11] <fenn> a hole? a sphere?
[21:07:17] <tomp> from where to where
[21:07:24] <ken_> hole
[21:07:34] <tomp> thats a 0.525" circle
[21:07:36] <ken_> 2.25" x 1.5"
[21:07:59] <tomp> lost again, please explain or sketch and post
[21:08:01] <cradek> are you asking how to move a tool in a circular path?
[21:08:08] <ken_> yes
[21:08:21] <cradek> you use a G2 or G3 arc command to make a circular path
[21:08:35] <ken_> and that's a 225" arc
[21:08:38] <ken_> g2
[21:09:04] <ken_> i need a .225" rad in the conners
[21:09:47] <fenn> ken_: draw a picture, and upload it to imagebin.org
[21:10:27] <ken_> k
[21:10:37] <OoBIGeye> * OoBIGeye is going totaly crazy with the inches and feet of americans
[21:10:39] <OoBIGeye> :O)
[21:10:53] <fenn> OoBIGeye: hey at least you dont have to live with it
[21:10:59] <tomp> include where the rounded rectangle is located
[21:12:50] <tomp> and if it's a pocket in a larger plate or a piece that measures 2.25" x 1.5" on it's outside
[21:13:03] <ken_> k
[21:14:53] <ken_> aluminum 3" wide 1.25" deep 15" long
[21:15:05] <ken_> i got to cut 8 pockets in this
[21:15:14] <ken_> br
[21:15:16] <ken_> br
[21:15:21] <ken_> brb
[21:15:26] <tomp> pocket depth?
[21:15:38] <tomp> or thru?
[21:15:42] <ken_> .69
[21:15:54] <gene> fenn: did that, ready to try again
[21:16:55] <cradek> ken_: I don't understand exactly what part you're having trouble with. Do you know the basics of gcode? of machining? do you know how to program an arc but just can't get it to work?
[21:17:25] <cradek> it would be silly for us to write your gcode for you, instead of helping you do it
[21:17:36] <cradek> but to help you do it, you need to ask specific questions that we can answer
[21:18:14] <ken_> pic at imagebin.org
[21:18:43] <ken_> don't need you to write it...just tell me how to make a corner
[21:18:55] <ken_> ok
[21:19:07] <gene> fenn: lost the pastebin reference, repeat please?
[21:19:08] <ken_> picture onder ken
[21:19:12] <archivist> http://imagebin.org/15041
[21:19:22] <ken_> under
[21:19:44] <archivist> ken_, the idea is to give a complete link to the picture
[21:19:50] <tomp> equally spaced along length?
[21:20:07] <ken_> yes 1.9 apart
[21:20:15] <tomp> even at ends?
[21:20:28] <tomp> 1.9 from ends AND 1.9 apart?
[21:20:28] <ken_> bu i just need to make the machine go around a corner
[21:20:41] <tomp> study emc manual on G2
[21:20:42] <ken_> 1.9" apart
[21:20:50] <ken_> k
[21:20:52] <ken_> thanks
[21:21:37] <cradek> ken_: applications / cnc / gcode quick reference
[21:21:50] <cradek> then, click G2 or G3, it will take you right to the detailed information
[21:22:29] <ken_> thanks
[21:22:30] <fenn> also you should study work coordinate systems (g54 and g10)
[21:22:37] <fenn> blah
[21:22:39] <archivist> heh
[21:22:41] <cradek> you tried
[21:24:13] <archivist> * archivist wonders how far to step back from the flying bits of endmill
[21:24:24] <tomp> i looked for the MOPS editor, thought it had a rect pocket rotuine
[21:24:34] <tomp> routine
[21:32:27] <gene> And with all this screwing aro0und, I've lost the 1600x1200 screen mode again.
[21:35:47] <alex_joni> 've lost the 1600x1200
[21:35:47] <alex_joni> screen mode again.
[21:35:53] <alex_joni> sorry :(
[21:39:09] <gene> Ok, what do I do, just reinstall the whoile system?
[21:42:36] <SkinnYPuppY> Here's a trick I have used a few times, boot the live cd and find it's xorg.conf file. If it has the screen mode you want mount your sys partition and replace or reedit xorg.conf accordingly
[21:43:43] <gene> The problem is that a reconfigure always puts in the Load glx, emc insists on it, and x can't find it,.
[21:44:01] <gene> I have probably removed and re-installed a gigabyte of stuff right now.
[21:48:58] <tomp> the cp1 rect pocket works, but doesnt allow for a corner radius different from cutter radius, so the tcl would need hacking http://www.linuxcnc.org/wiki/uploads/cp1.tgz
[21:49:24] <tomp> it even has the pocket location, so he could use it for all the pockets
[21:49:51] <tomp> beware 'llx' 'lly' was Jon's name for the offset location
[21:56:34] <gene> and there is no answer to the emc needs glx connundrum
[21:56:56] <gene> I tried the cd, its xrog.conf quits at 1024x768
[21:57:16] <gene> and has a lot of other stuff I don't think I need.
[21:57:34] <archivist> edit and set what you want
[21:57:43] <gene> Then I have to demolish the cd tray to keep it from mrebooting from the cd forever
[21:58:07] <gene> the bios grabs it before the drive recognizes it.
[21:58:43] <alex_joni> how about removing the CD before the PC reboots?
[21:58:50] <gene> well, it seems I need glx, but its not avauilable at x start
[21:58:51] <alex_joni> there is a prompt from ubuntu to do just that
[21:58:57] <tomp> there's a small hole in face of tray that forces eject, use a resistor wire or paper clip
[21:59:04] <alex_joni> gene: you have to add one line to have glx
[21:59:21] <gene> its in there right now, unless I have a typu
[21:59:28] <tomp> ;)
[22:02:14] <gene> Took the comment out, restarted, get this:
[22:02:28] <gene> II) LoadModule: "glx"
[22:02:28] <gene> (WW) Warning, couldn't open module glx
[22:02:29] <gene> (II) UnloadModule: "glx"
[22:02:29] <gene> (EE) Failed to load module "glx" (module does not exist, 0)
[22:03:12] <gene> and I have asked what package this module named gkx is in several times, with no answer
[22:03:21] <gene> and I have asked what package this module named glx is in several times, with no answer
[22:05:05] <gene> Does anyone know where this module is supposed to exist?
[22:05:06] <alex_joni> it's usually part of the graphics card driver
[22:05:20] <gene> you asked me to use vesa
[22:05:24] <alex_joni> what driver do you currently use?
[22:05:29] <gene> vesa
[22:05:38] <alex_joni> I think you need to install some mesa-foo
[22:05:44] <alex_joni> can't remember the exact name
[22:06:09] <gene> all the mesa stuff is except the .dev
[22:06:49] <alex_joni> libgl-mesa-glx maybe?
[22:18:21] <tomp> cp1 accounts for tool dia, ( AXIS shows path is smaller than desired pocket by 1/2 cutter per side) but doesnt allow for corner radii other than cutter radius :( maybe python or o-words could help (sort of a CAM system)
[22:18:44] <tomp> short of a CAM system
[22:26:52] <gene> <alex_joni> libgl-mesa-glx maybe? doessn't exist
[22:27:37] <SWPadnos> libgl1-mesa-glx
[22:28:00] <SWPadnos> hmmm. that's after dapper though
[22:28:51] <SWPadnos> ah - libgl1-mesa, libglu1-mesa
[22:32:21] <gene> I've reinsstalled the mesa-glx kit but /usr/lib/X11 is still empty of glx stuffs, and both of thoise are installed but let me check the last one
[22:34:23] <maddash> weeeeeeeeeee
[22:39:28] <gene> I just used synaptic to re-install it all, but there still isn't anythying with a glx in its name anywhere on the system
[22:40:07] <gene> again that is
[22:40:18] <SWPadnos> that's probably OK. the issue may be that the nvidia installer links or replaces some file to point at the nvidia version of the library
[22:40:27] <SWPadnos> I don't recall exactly which file though
[22:40:53] <gene> I have nuked every reference I can find of that, using mc, which shows broken link in red
[22:41:59] <SWPadnos> hmmm
[22:42:07] <gene> One thing that disappoints me is that ubuntu has no concewpt of a working 'locate' directive, no error, but returns instantly with no results
[22:42:28] <SWPadnos> there's find though
[22:42:43] <SWPadnos> even if you don't have a locate / slocate database
[22:43:01] <archivist> locate uses a database, it needs updating
[22:43:04] <alex_joni> I think you can install locate though
[22:43:36] <gene> I did that, sudo updatedb, came back in 1 second flat
[22:43:58] <gene> syntax for find?
[22:44:08] <maddash> gene: find / -iname \*glx\*
[22:44:20] <maddash> archivist: of course it does, how else could it return instantly?
[22:44:26] <SWPadnos> or use quotes around the name: find / -name "*glx*"
[22:44:40] <maddash> or use single quotes, if you love your shift key
[22:44:47] <maddash> :P
[22:44:59] <SWPadnos> but not backquotes :)
[22:45:05] <maddash> rofl
[22:47:22] <maddash> * maddash can't program anymore
[22:47:24] <gene> that needs a sudo in front of it, and then locked up the shell after I entered the pw.
[22:47:32] <alex_joni> maddash: I was just wondering the other day what happened to your stepper driver
[22:47:53] <maddash> alex_joni: 'stepper driver'?
[22:47:56] <alex_joni> you started on a multi-*bazillion*-discrete chip driver (with a whole lot of transisters iirc)
[22:47:58] <maddash> alex_joni: what do you mean?
[22:48:12] <alex_joni> that was a couple years ago
[22:48:23] <alex_joni> might be confusing you with someone else though..
[22:48:31] <maddash> er, i've only been around for 1 year
[22:48:55] <alex_joni> ok, prolly someone else
[22:50:10] <alex_joni> there was a guy in here a while ago, that wanted to build a stepper driver from a lot of resisters and transisters
[22:50:26] <tomp> dimas?
[22:51:15] <alex_joni> dunno :)
[22:51:21] <gene> And it doesn't find a 'glx' except some .c and .h files & some mesademo stuff
[22:51:23] <alex_joni> not important ..
[22:52:09] <SWPadnos> gene, is it trying to load the nvidia files from xorg.conf at the moment?
[22:52:10] <maddash> gene: ^C
[22:52:23] <gene> no, they've been blown away
[22:52:47] <gene> I finally used the menu to kill the session
[22:53:50] <gene> Q: are the names of the video card a factor in the xorg.conf, I thought just the drivers were
[22:54:04] <SWPadnos> you don't need sudo for find. you'll just get some messages that you couldn't open certain files/dirs if you're not root
[22:54:05] <alex_joni> gene: how about pastebin ?
[22:54:39] <SWPadnos> there can be multiple card and monitor definitions, you have to make sure the rigth one is selected in the server layout section
[22:55:50] <gene> for the xorg.conf file
[22:55:55] <gene> ?
[22:56:03] <alex_joni> gene: sure
[22:57:50] <gene> http://www.pastebin.ca/949646
[22:59:40] <alex_joni> gene: can you also pastebin the output from dpkg -l | grep mesa ?
[22:59:56] <SWPadnos> that's an ell, not a numeral 1 :)
[23:00:18] <alex_joni> lowercase L
[23:00:45] <gene> ahh...
[23:01:24] <alex_joni> SWPadnos: I wonder how many people actually have a 'wacom' hardware
[23:01:31] <SWPadnos> heh - not many
[23:01:42] <SWPadnos> but those who do would be disappointed if it didn't work
[23:01:50] <alex_joni> wonder why they turn the stuff on by default..
[23:02:00] <gene> http://www.pastebin.ca/949648
[23:02:07] <SWPadnos> (it's too bad - we've got a reasonably good calcomp here, but I think the batteries exploded in the pen)
[23:02:07] <tomp> 'they' had wacoms ;)
[23:02:22] <beege> Back fom the shop
[23:02:43] <gene> and you can't remove it from xorg.conf, been there, done that, had to reinstall!
[23:03:00] <alex_joni> gene: sure you can remove, you just need to remove all references
[23:03:14] <alex_joni> and reinstalling because of a busted xorg.conf is just plain wrong
[23:03:21] <gene> I spent several hours trying that
[23:03:31] <alex_joni> obviously before changing any configuration file, you make a copy, don't you?
[23:03:42] <alex_joni> if things fail, just restore it..
[23:03:46] <gene> there are older ones there, yes
[23:04:18] <alex_joni> no, I mean.. a known working copy .. which you save in /home/gene/..
[23:04:32] <alex_joni> anyways, I don't spot any apparent error
[23:04:35] <gene> ahh, no, sorry :(
[23:04:55] <alex_joni> can you paste the logfile for X?
[23:05:11] <gene> except the system doesn't actually have a glx on it. Next
[23:05:16] <alex_joni> Xorg.log or something like that
[23:05:23] <SWPadnos> Xorg.0.log
[23:06:39] <gene> http://www.pastebin.ca/949653
[23:09:04] <toastydeath> shoop da woooop
[23:09:11] <toastydeath> i think i aced another test
[23:09:14] <toastydeath> at skewl
[23:09:39] <gene> Gotta recycle some coffee and check dinner progress.. Kewl!
[23:09:44] <SWPadnos> heh
[23:10:24] <archivist> skewl is a long lost memory
[23:10:33] <SWPadnos> kewl
[23:10:46] <alex_joni> gene: the module should be in /usr/lib/xorg/modules/ something
[23:11:44] <beege> EMC / Linux - touch screen friendly?
[23:11:46] <SWPadnos> hmmm - can you reconfigure the mesa packages, or just --force-reinstall to get those libs put back??
[23:11:51] <SWPadnos> beege, sure
[23:12:13] <SWPadnos> depends on the manufacturer a bit - some have drivers included, others you may have to install on your own
[23:12:27] <SWPadnos> the main problem with touchscreens is that there is no right-click
[23:12:30] <SWPadnos> or middle ...
[23:13:03] <rharpham> hi
[23:13:09] <SWPadnos> hi
[23:13:46] <alex_joni> gene: also look in /usr/X11R6/lib/modules/extensions
[23:14:03] <alex_joni> there might be a libglx.so there, but not libglx.a (or the other way around)
[23:14:39] <rharpham> just playing with emc 2.2.4, with simulation in axis, just taken a program been using to day from our fanuc based control to see how emc handels it. cant understand why its not liking the G80 comand
[23:15:14] <rharpham> i have G80 Z30 M09, and its saying cannot use axis values with G80
[23:15:25] <gene> http://www.pastebin.ca/949660 ios ls -lR of that tree
[23:15:50] <alex_joni> rharpham: http://www.linuxcnc.org/docview/2.2/html//gcode_main.html#sec:G80:-Cancel-Modal
[23:16:11] <rharpham> read manual on gcodes
[23:16:58] <SWPadnos> G80 cancels motion modes, so giving an endpoint is kind of senseless
[23:17:07] <gene> No modules/extensions dir, modules contains only:libglide3x.so -> ../../../lib/libglide3.so.3
[23:17:32] <rharpham> so need G80 g0
[23:17:51] <SWPadnos> that's not how I read the code
[23:18:17] <SWPadnos> G80 is meant to tell the interpreter explicitly that a new motion mode is needed before moving again
[23:18:26] <SWPadnos> "I'm done with the old mode now"
[23:18:30] <toastydeath> swpadnos is so right it burns
[23:18:34] <toastydeath> burns like fire
[23:19:00] <SWPadnos> heh
[23:19:02] <tomp> j.cash
[23:19:08] <toastydeath> ?
[23:19:13] <rharpham> aaah sorted it
[23:19:15] <alex_joni> gene: maybe try "apt-file search libglx.a" or .so
[23:19:24] <tomp> like a burning ring of... johnny cash
[23:19:27] <rharpham> is there a highspeed peck? G73
[23:19:45] <toastydeath> high speed peck?
[23:20:03] <toastydeath> run the drill faster?
[23:20:13] <SWPadnos> gene, can you try apt-get install --reinstall libglu1-mesa libgl1-mesa ?
[23:20:17] <alex_joni> no G73 in emc2
[23:20:36] <SWPadnos> (I'm not sure that's the exact syntax - check apt-get manpage first)
[23:20:56] <gene> Apparenty do not have and apt-file
[23:20:57] <rharpham> but its too slow in G83
[23:21:12] <rharpham> now u doubled my drill times
[23:21:39] <SWPadnos> ah yes, it is --reinstall
[23:21:44] <toastydeath> write a macro i guess
[23:21:55] <toastydeath> if you really want g73
[23:22:04] <tomp> rharpham: many of the g7x g8x can be custom codes from your 'var' not exactly 'standard'. you can write code yourself tho
[23:22:11] <toastydeath> even though i hate that answer
[23:22:39] <gene> No error, does it quik, and doesn't change anything, don't you just love it?
[23:24:06] <rharpham> was stadnard on Fanuc 6m and up lol 99% same as G83 code
[23:24:49] <gene> I just heard the dinner bell. back later
[23:24:57] <SWPadnos> see ya
[23:25:18] <tomp> standard to FANUC ( and thier buddies ), thats not standard to all machine tools
[23:25:32] <toastydeath> no, even okuma has a g73-esque cycle
[23:25:35] <tomp> not wrong, just not standard
[23:25:56] <toastydeath> and okuma is the poster child for "we don't copy fanuc"
[23:26:17] <tomp> no, 'we dont pay fanuc but want thier customers'
[23:26:22] <tomp> ;)
[23:26:25] <toastydeath> no, actually
[23:26:34] <toastydeath> you don't have to pay royalties to use the same formatting
[23:26:48] <toastydeath> okuma just decided they wanted to do things differently and felt they could do better, so that's what they did
[23:26:50] <rharpham> and even mitibichi copyed fanuc on 73
[23:27:45] <tomp> i meant that many japanese companies share technology. some do some dont, some all, some some... but WONT share with german or american control companies
[23:28:01] <toastydeath> right but who cares, because g73 is common amongst all controls
[23:28:22] <toastydeath> and if they don't use g73 exactly, they still have the function represented by g73
[23:29:22] <toastydeath> just like lathe roughing cycles, for example.
[23:30:33] <rharpham> so now whats this bussness with retyping the co-ord on g03 g02 seems to want end point even tho we want full circule from start to finish from where we are currnet
[23:32:14] <toastydeath> a clear majority of controls don't support that, so i can't defend the g02/3 issue
[23:33:12] <rharpham> G90 G54 G00 X52.5 Y-50.8; G43 Z2. H05 M08; G01 Z-.5 F1000.; G17 X54.505 F500.; G03 I-2.005 J0 F800.; G01 X60.505;
[23:33:34] <rharpham> ; ^ denotes line brake
[23:33:37] <toastydeath> rite, but that really does vary from control to control
[23:33:54] <toastydeath> so it's not like a basic function that's not supported, it's just a semantics thing
[23:34:28] <rharpham> ok this being like our simens control on our hardinge lath with a B instead of R on radius lol
[23:34:47] <toastydeath> yup
[23:35:36] <rharpham> better write a new post for camworks :)
[23:35:46] <toastydeath> =)
[23:40:18] <rharpham> is there a rapid over ride in EMC now?
[23:42:22] <rharpham> can you do G52 data shift? nice usefull for sub programing parts
[23:42:44] <toastydeath> work shifts?
[23:42:56] <toastydeath> look at g92
[23:42:58] <toastydeath> in the docs
[23:43:37] <rharpham> not quite same as G52
[23:43:49] <toastydeath> never used g52 before
[23:43:58] <rharpham> 52=local co-ord shift 92=co-ord shift
[23:44:44] <rharpham> 52 will shift the 54,55 etc by ammount defind on 52
[23:45:17] <rharpham> always canceld by g28 when go bk home
[23:45:35] <toastydeath> oh that's pretty sweet
[23:46:17] <rharpham> use it all the time in macro programing, so can have unlimted data shifts
[23:47:40] <beege> We used to call G52 "child coordinate system"
[23:48:53] <rharpham> lol
[23:49:00] <beege> Was also always cancelled by G52X0Y0Z0
[23:49:06] <rharpham> indeed
[23:49:34] <rharpham> guess as we have a massive hard drive program size should not b issue :P
[23:50:40] <rharpham> my program was 12k in macro, now its 35k to half work on emc tho
[23:54:15] <gene> ok, back to bug ya'll again
[23:54:38] <gene> any more ideas short of a re-insstall?
[23:55:17] <gene> This thing defaults to a path to #kubuntu too, I'll ask there
[23:57:30] <rharpham> doing a G91 g28 z0 is not working here :S what am i doing wrong!
[23:57:49] <rharpham> its doing it as if a G90 crashing down into the job
[23:59:18] <SWPadnos> what are you expecting that to do?