Back
[00:00:25] <MattyMatt_> or maybe you do, if you pick the wrong brand of CF card. they use a similar wear leveling system to HDD, and may implement it badly enough to cause random slowdown
[00:03:07] <MattyMatt_> emc doesn't use x-windows does it?
[00:04:04] <MattyMatt_> I really ought to try emc sometime :)
[00:05:40] <andypugh> I am not sure what the windowing system is.
[00:07:10] <andypugh> And 'tis time I were asleep.
[00:10:31] <jymm> NEVER use CF/SD, etc in a predominantly RW system. Good for tings like a LiveCD installation to CF, where you are only saving the config once in a great while.
[00:12:35] <jymm> Most OS'es have far to many write cycles for that type of media.
[00:19:39] <MattyMatt_> most OSes, but to be a RT os it would have to forego a lot of its chatter or at least use a ramdisk
[00:22:58] <MattyMatt_> I'll be treating my emc machine as an embedded one, no other progs
[00:32:15] <MattyMatt_> a killer NIC would be nice, with the leisure time to transfer the whole IP duties to it
[00:34:37] <tom3p> interviewing cnc's in taiwan: lots used cf boot. saw lots of 'sorry, we need a new cf chip' (limited write life), near the end, we expected the demos to have this problem.
[00:35:02] <tom3p> and i carry a 1G and a 256M cf daily
[00:35:14] <tom3p> jumpstart
[00:42:58] <klick0> I have good luck with transcend CF cards, they seem to last well in RW environments
[00:49:26] <klick0> I just upgraded from the old ubuntu version (6.x ?) to the new 8.04 live cd, I copied over my configs and everythign seems alright, i've machined probably 2-3 hours worth, so it works. The problem is though, when I go into Axis, it won't let me run anythign unless all the axis's are homed, is there a way to disable this? In the old version I would enter back in, it would remember it's position, and I just just run things again.
[00:49:26] <klick0> When I start it up now, it remembers it's position (which is good), but then I can't run anything.
[00:49:48] <klick0> I think this is more of an axis problem then emc..
[00:50:20] <klick0> Regardless, how do I make it so that I don't have to re-home when I start back up?
[01:01:28] <tom3p> i dont know how. you may have moved a tiny bit.
[01:01:38] <tom3p> there is a way to not move and set 0 where you wake up.
[01:01:44] <tom3p>
http://linuxcnc.org/docs/2.2/html/config_ini_homing.html#Homing Configuration sez set 3 parms to 0, then...
[01:01:45] <tom3p> do 'touch off'
http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?CoordinateSystems#Touch_Off
[01:01:58] <tom3p> as far as can tell
[01:02:03] <tom3p> I
[01:03:23] <klick0> yea, i would agree, but that's not the issue, Axis simply won't let me run any g-code until i've homed the axis's... not touch off, I generally don't home but once every 6 months, my machine doesn't move when it's turned off...
[01:03:38] <klick0> when I home it, a little bullseye comes up next to each axis
[01:03:47] <klick0> once I have them all homed, it then lets me run g-code
[01:04:12] <klick0> but before I upgraded to 8.04 (which includes some new version of axis, it looks a litlte different) I can't run anything until i've homed
[01:04:26] <Valen> theres a thing in the ini file to let you run without homing
[01:04:27] <klick0> i was just hoping there was an axis flag that says "allow me to run g-code without beign homed"
[01:04:32] <klick0> cause it wasn't like that before
[01:04:35] <klick0> ok
[01:04:36] <tom3p> 3 things
[01:04:43] <klick0> i'm searching the documentation and can't find that
[01:04:44] <tom3p> as above
[01:04:46] <Valen> its not axis specifically its in the ini somewhere
[01:05:22] <tom3p> oh, if theres a way to not move and not home and not say 'here's 0'... i dontknow, so i'll shut up
[01:05:44] <klick0> hehe, it's alright, i appreciate any help
[01:07:41] <Valen> home_search_vel
[01:07:49] <Valen> home_search_vel' is a member of the joint structure (as defined in motion.h). The default value is zero. A value of zero causes EMC to assume that there is no home switch. The search and latch stages of homing are skipped, EMC declares the current position to be “home_offset”, and does a rapid to “home” if “home” is not equal to “home_offset”.
[01:08:21] <Valen> you shouldn't need homing then you just "touch off" and it'll let you start running gcode
[01:08:43] <klick0> hmm, alright, well i did have home_search_val set on X axis, so that could be that
[01:09:07] <Valen> if its set you must home before you can do anything
[01:09:15] <Valen> set to anything other than 0 that is
[01:12:08] <klick0> "Can't run a program when not homed"
[01:13:02] <cradek> yes you must home - if the velocities are zero you can just jog to where you want zero to be and home there - it will do nothing except change to zero
[01:13:03] <klick0> I found a chat log of someone saying this back in january
[01:13:56] <klick0> yes I realize I can just home and then run it, but it wasn't like this before I upgraded. If i'm just faking the home basically, wha'ts the difference, there should be a setting to just turn off this requirement
[01:14:24] <cradek> yes you can turn it off - see the docs
[01:14:39] <cradek> but you should home - you get the benefit of soft limits then.
[01:15:55] <klick0> yes, but i do have the benefits of soft limits
[01:16:01] <klick0> it remembers my position
[01:16:04] <klick0> and i don't move my machine
[01:16:07] <klick0> so they are fine
[01:16:11] <klick0> if i move my machine, i home then
[01:16:18] <cradek> I see
[01:16:27] <klick0> basically, i want to decide when i want to home, i don't want emc to decide...
[01:16:32] <cradek> then turn it off
[01:16:47] <klick0> yes i know, my search velocitys are zero and it still wants it homed
[01:16:51] <klick0> i just tried it
[01:17:00] <klick0> so i'm gonna re-read all the documentation
[01:17:03] <klick0> and attempt to find it
[01:19:27] <cradek> did you use the wiki page UpdatingTo2.3 to see how to update?
[01:19:44] <klick0> nope
[01:19:50] <klick0> i re-installed
[01:19:54] <klick0> but i did use my old configs
[01:19:56] <klick0> i'll read those
[01:20:01] <cradek> oh ok
[01:20:06] <cradek> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UpdatingTo2.3
[01:20:23] <cradek> you'll want to read everything on the UPDATING page it points you to
[01:20:45] <klick0> well hey, it's the first line :)
[01:21:00] <klick0> thank you
[01:21:08] <cradek> imagine that! (depending on what version you had before, other things there may also be important)
[01:21:48] <cradek> welcome :-)
[01:23:06] <klick0> well that's good, glad that won't be an issue anymore
[01:28:50] <tom3p> klick0: does it work for you?
[01:28:53] <klick0> anyone ever try to machine with a 0.15625" diameter end mill? I did today, i broke 2 of them heheh... i was going into aluminum, tried 0.005" depth per pass, 2IPM, it did a bunch of passes, finally died, 13000RPM on the router
[01:28:58] <klick0> tom3p: yep, it works
[01:29:13] <klick0> er, i mean 0.015625"
[01:30:25] <klick0> i don't think i'm going to get into small engravings... I got a 0.03125 i'll try, hopefully that won't snap
[01:30:52] <cradek> I think those are only good for breaking...
[01:31:07] <tom3p> i ran an installed version of 2.3.5 to try for your problem, i set the 3 parms to 0 and the homing type to -1.
[01:31:07] <tom3p> i cant home, what did you do ( i dont need to u/g, i'm there )
[01:31:43] <klick0> in the [TRAG] section, just add the line "NO_FORCE_HOMING = 1"
[01:31:46] <klick0> in your main ini
[01:32:00] <klick0> TRAJ
[01:32:04] <klick0> [TRAJ]
[01:32:21] <tom3p> in [TRAJ], the docs i found all taked about [AXIS_N], i didnt see that anywhere
[01:32:51] <klick0> yea, i was cutting some aluminum parts out with a 0.0625 diameter bit, and they looked so nice, great detail, I figured that maybe the 0.015 bit would be fun...
[01:33:03] <klick0> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UPDATING
[01:33:06] <klick0> read section 1.1
[01:33:17] <klick0> I assume you're running 2.3 now right, you said you upgraded
[01:33:27] <klick0> er, did you
[01:33:43] <tom3p> yeah, thats what i mean, go read the section on upgrading when you dont need to upgrade
[01:34:09] <tom3p> no i installed it, not updated/upgrade
[01:34:11] <klick0> ahhh
[01:34:16] <klick0> yea, i see what yer saying
[01:34:22] <klick0> the 2.3 docs don't have it
[01:34:24] <klick0> lemme look
[01:35:18] <klick0> well it's there
[01:35:26] <klick0> 2.3 docs, section 2.2.8
[01:35:43] <klick0> NO_FORCE_HOMING actually has the longest description in that section
[01:36:43] <tom3p> i added the [TRAJ] line, and can touch off w/o errs, but i get no bullseyes. did you?
[01:36:45] <klick0> I searched the 2.2 docs initially, which is why I didn't have much luck eheh
[01:37:05] <klick0> no bullseyes, but i can run, which is like how it used to be
[01:37:12] <klick0> it just allows me to run without homing
[01:37:20] <klick0> or i should say, without homing in that session
[01:38:00] <klick0> you don't need bullseyes, they just mean that you homed at one point during your session
[01:38:42] <klick0> looking at most of the old screenshots, it appears as if the people who did documentation did not home much either, they must not like the new forced homing either ehhe
[01:39:37] <cradek> doing sim for doc screenshots and running a large powerful machine are very different tasks - for the latter you really really need to home. it can be a very big deal if you forget.
[01:40:56] <klick0> yea, i knew you were gonna defend that statement :)
[01:41:32] <cradek> that feature came about when about three times in a day I issued a tool change (MDI) on a big machine without homing - BONK right onto the limit switch at full rapid speed (EMC thought it was moving to the tool change location)
[01:41:53] <klick0> i actually did have my first incident today, i've been running my machine for a year and a half. nothing to do with homing, i simply forgot I had a clamp somewhere, and the machine forced it's way through it, it was a plastic clamp and man it exploded, lfying all through my shop
[01:42:06] <cradek> oops, glad it missed you
[01:42:16] <tom3p> so this just zeros the axis wherever the hell it wakes up? thats useless
[01:42:17] <klick0> yea me too
[01:42:21] <cradek> did you have safety glasses on?
[01:42:24] <klick0> of course
[01:42:27] <cradek> yay
[01:42:40] <cradek> that is a very minor incident then :-)
[01:43:02] <klick0> yea i know, but i've never really had an incident, so it's really my first
[01:43:08] <klick0> it let out a hell of a POW though
[01:43:14] <cradek> I once turned on the spindle (varispeed, so it started at whatever speed it was last running, which was all the way up in this case) with the wiggler pointer in it
[01:43:33] <cradek> got to about 3500 rpm in no time, and man did that pointer fly
[01:43:39] <klick0> hehe, yea
[01:43:42] <cradek> I actually found it! I was amazed.
[01:43:45] <klick0> that's generally not good
[01:43:53] <klick0> eheh, yea
[01:44:10] <cradek> missed me, yay
[01:44:26] <klick0> yea i've never got hit by anything, god i hope that doesn't happen
[01:44:35] <cradek> glad I take up just a small portion of the 360 degrees around the spindle
[01:44:43] <klick0> precisely
[01:45:00] <klick0> hopefully the spindle wasn't at eye/face/throat level
[01:45:01] <cradek> so anyway, now it chides me if I don't home :-)
[01:45:09] <klick0> most anything else could probably take it... to an extent
[01:45:09] <cradek> chest
[01:45:22] <klick0> yea
[01:45:33] <klick0> i agree with homing, it's just my machine is homed
[01:45:42] <klick0> it just doesn't think it is :)
[01:45:49] <klick0> i checked after months, there was no drift
[01:45:57] <klick0> well, there was less then a 20th of an inch
[01:46:06] <klick0> which i'm gonna say is no drift
[01:46:19] <klick0> i don't really run my machine tons though
[01:46:27] <klick0> so months for me is like days for you
[01:46:31] <cradek> all machines move a bit when you turn them off
[01:46:46] <klick0> i swear i checked, i've done it multiple times
[01:46:53] <klick0> after months it's barely off
[01:47:00] <cradek> it could be the amount it moves is fine for you
[01:47:19] <tom3p> http://linuxcnc.org/docs/EMC2_User_Manual.pdf section 2.3 'HOMING' tells the trick about 'NO_FORCE_HOMING = 1'
[01:47:20] <cradek> if you use a tenths indicator I bet you see it relax and move as soon as you turn it off
[01:47:48] <klick0> not sure this makes a difference, i have the universal stepper controller, so when the motors power down, it still watches the encoders, and EMC knows if it moved a bit, records that.. then shuts off.... so really it doesn't move
[01:47:48] <cradek> this mill moves about .0003 on all axes as soon as I turn the servo amps off - you see it on the readouts
[01:47:53] <klick0> you'd think Z axis might, but it doesn't
[01:48:21] <klick0> well yea
[01:48:24] <cradek> ah, if it has real encoders with usc, it may actually keep track
[01:48:47] <klick0> i don't mean that it doesn't move any... it just doesn't move enough to affect me, but i'm not talking inches, definitly less then a tenth of an inch over months
[01:48:56] <tom3p> that was not what was posted and damn hard to find. i appreciate the help but there aint no section 2.2.8, and the normal searches do not lead you to that info.
[01:49:20] <tom3p> i gotta unfsck these ini s now
[01:50:01] <klick0> is the development for EMC funded directly by any organization?
[01:50:08] <klick0> or is it purely free open source developers?
[01:50:11] <cradek> no funding
[01:50:15] <klick0> I thought NIST funded it initially
[01:50:23] <cradek> that was 15 years ago
[01:50:26] <klick0> heheh
[01:50:27] <klick0> ok
[01:50:32] <klick0> but nothing now?
[01:50:36] <cradek> nope
[01:50:42] <klick0> i'm sure donations don't do to much
[01:51:16] <cradek> I do it because I use it and believe in home shop cnc and I want to give people that
[01:51:36] <klick0> yea
[01:51:45] <cradek> it's a happy accident that it works in industry too. we need our industry here.
[01:51:46] <klick0> well i use it, i'm glad you guys are doing it
[01:51:49] <cradek> thanks
[01:52:37] <klick0> 5 or 6 years ago I wrote my own program to run my machine, but it was so much work,a nd EMC was light years ahead of mine.. i finally switched after attempting to do it myself for a few years
[01:52:47] <cradek> money donations we never know what to do with. we have had one that I recall (5 years) and we used it to pay for the CDs we gave away at cnc workshop that year
[01:53:29] <klick0> i converted my program to focus on the CAM aspect, and use emc to execute obviously
[01:54:04] <klick0> i just programmed pocketing the other day, I made a puzzle with my daughters name it in :)
[01:54:10] <cradek> one year someone paid for my motel room during the week of cnc workshop. often people buy us food etc. I really appreciate those gestures.
[01:54:24] <cradek> if you have a good pocketing algorithm, do share - it's a hard problem we've all tinkered with one time or another
[01:55:13] <klick0> well it's only a day old, so it's not to advanced, i don't have any "no pocket" areas inside the pocket
[01:55:20] <cradek> ah
[01:55:22] <klick0> just any closed like polygon, it pockets
[01:55:29] <klick0> so perhaps that's not what you call a "pocket"
[01:55:36] <klick0> but to me it is, i used to do it by hand in cad
[01:56:13] <cradek> any closed? even the ones where the path breaks apart?
[01:56:29] <klick0> yea, that's what i'm working on fixing
[01:56:32] <klick0> that is detected
[01:56:40] <cradek> heh that's when it gets nasty
[01:56:41] <klick0> and seperated, but not quite working how i want it
[01:56:51] <cradek> please keep us updated, seriously
[01:57:09] <klick0> what CAM software do you use?
[01:57:13] <cradek> none
[01:57:40] <klick0> heeh, alright
[01:57:41] <cradek> the only nonfree software I use is autocad R12 I bought around 1994
[01:57:55] <klick0> i used QCad with my custom CAM software
[01:57:57] <klick0> then i use EMC
[01:58:03] <klick0> works for me
[01:58:07] <cradek> neat
[01:58:18] <klick0> i only do 2.5D stuff
[01:58:29] <cradek> me too, almost always
[01:58:43] <klick0> to get some dimensional stuff, I name layers "inner 0.25" and my cam softwrae only cuts down to 0.25 inches, or something like that
[01:59:05] <klick0> or "pocket 0.10" to pocket to a certain depth
[01:59:29] <klick0> works for me
[01:59:33] <cradek> I might do fancier projects if I had good 3d software, but I don't and won't
[01:59:51] <klick0> i have used blender to create some depth maps... then run the depth map though emc
[02:00:19] <cradek> yeah image-to-gcode is pretty neat. I might have to try that some, now that I have a mill running emc
[02:00:45] <cradek> bbl
[02:00:48] <klick0> later
[02:19:27] <tom3p> from the hitchikers guide "for though it has many omissions and contains much that is apocryphal..."
[03:41:20] <roh> klick0 you should take a look into heekscad. sounds like generating gcode emc likes from the dxf you do should be easy with it
[10:44:17] <archivist> * archivist just read the thread oscillation thread, very interesting
[11:25:57] <tarzan> archivist, link?
[11:29:11] <archivist> http://news.gmane.org/gmane.linux.distributions.emc.user
[11:29:57] <archivist> title is threading Z oscillation depends on encoder PPR
[11:36:13] <MarkusBec_away> MarkusBec_away is now known as MarkusBec
[12:55:32] <tomp> tomp is now known as tom2p
[12:55:49] <tom2p> tom2p is now known as tom3p
[14:29:33] <tomp> tomp is now known as tom3p
[14:41:12] <EbiDK|AWAY> http://www.geekologie.com/2009/06/its_about_time_haynes_manual_f.php
[15:14:13] <skunkworks_> it is started...
http://www.cnczone.com/forums/showthread.php?t=91965
[15:16:33] <cradek> ouch
[15:17:29] <archivist> hehe
[15:17:59] <cradek> wonder what OS most people are running it on currently
[15:18:45] <skunkworks_> I have seen some posts on vista - so 2000, xp and vista
[15:18:56] <skunkworks_> not bad
[15:43:19] <Valen> I wonder if mach works under wine ;->
[16:27:07] <pjm__> evening all!
[16:27:16] <pjm__> pjm__ is now known as pjm
[16:29:09] <skunkworks_> pjm: how goes the mesa conversion?
[16:29:54] <pjm> well so far i'm dead impressed with the 7i43
[16:30:01] <pjm> it is absolutly wicked!
[16:30:18] <skunkworks_> :) good
[16:30:20] <pjm> I've currently got all the steppers / VFD migrated over to it, along with limit switches
[16:30:35] <pjm> so now have max and min limits for X/Y and max for Z
[16:31:20] <pjm> and for the VFD I'm using the stepgen in velocity mode to generate a pulse train for the speed control
[16:31:58] <pjm> when i was using the parport, at high speeds when it was making a 5 to 6KHz pulse train, u could hear the motor wobbling in speed a little, but with the 7i43 it is solid as a rock
[16:34:18] <skunkworks_> heh
[16:35:15] <pjm> so anyway the next step is to fix the encoders and move the pendant over to the 7i43
[17:44:15] <cradek> UKers: are there metric versions of loc-line or do you find the inch versions on metric machines?
[17:44:45] <cradek> I ask because I have measurements (not taken by me) that don't make sense, trying to figure out the right loc-line to buy
[17:48:24] <jymm> you guys make servo so complicated... I/O cards, encoders, blah, blah, blah... here's the easy solution...
http://allmotion.com/
[17:50:15] <archivist_attic> * archivist_attic pokes jymm in the get a clue about the complete loop cell
[17:50:34] <jymm> archivist_attic: totally overrated
[17:50:49] <pjm> archivist btw did u have any photos of the UK engineering show on line, and how was the show?
[17:50:49] <cradek> we have fading! clip art!
[17:51:19] <jymm> cradek: is that like eco friendly cipart?
[17:51:27] <jymm> clipart
[17:51:29] <archivist_attic> pjm
[17:51:29] <archivist_attic> http://www.collection.archivist.info/archive/djcpd/PD/2009/2009_10_16_model_engineer/
[17:51:34] <pjm> thanks
[17:52:04] <archivist_attic> were not that many really good models this year
[17:52:19] <archivist_attic> Harrogate show was better
[18:12:36] <cradek> is 5052 fairly tarnish/fingerprint resisting?
[18:15:31] <archivist_attic> series are alloyed with magnesium, derive most of their strength from work hardening. It is suitable for cryogenic applications and low temperature work. However is susceptible to corrosion above 60°C
[18:18:12] <cradek> 60c is no problem - this is for things that will be handled at room temp
[18:19:32] <archivist_attic> seems its used in boats and planes, but they may treat in those applications
[18:20:29] <archivist_attic> most ally is ok for handling as it self protects
[18:22:50] <jymm> 5052 being what? alum? ss? something else?
[18:23:21] <cradek> yes Al
[18:23:27] <jymm> k
[18:31:59] <frallzor> hello boys, girls or a little bit of both
[18:41:12] <archivist_attic> machines!
[18:41:58] <frallzor> hello mr machine
[18:42:12] <frallzor> or shall I say, meep meep
[18:53:15] <frallzor> anyone here into Artcam?
[18:57:29] <jymm> frallzor: Sure,
http://tinyurl.com/yzd8sf5
[18:59:13] <archivist_attic> * archivist_attic wants real urls not ubscured ones, specially at a public archive
[19:00:30] <jymm> archivist_attic: tough
[19:05:54] <robh_> cradek, loc-line here in UK varys make to make on fit. some is tight & stiff some is ok. and yes its all in imperial we have u may get metric now im not sure never looked. most is 1/4", 3/8", 1/8", 1/16", 1/2" (this is bore size), threads are 1/4" NPT, 1/8" NPT, 1/2" NPT (i prob missed some)
[19:20:11] <cradek> robh_: thanks
[20:38:21] <jymm> cradek: Hey, does your ttf to gcode thingy take into consideration the tooling being used? v-tip, ballnose, etc?
[20:38:52] <cradek> no, it only traces the font as-is
[20:39:22] <jymm> cradek: traces, asin just outline the font itself?
[20:39:39] <cradek> yes
[20:40:43] <jymm> does it use in/mm dimensiions for the final output, as in give me 3" tall letters?
[20:41:08] <cradek> neither, it uses the units in the font (whatever they are) and you can scale the result by editing a scaling factor in the output file
[20:41:42] <jymm> Oh, so *I* apply the 72 or 96 dpi to get a final dimension.
[20:41:44] <andypugh> Yeah! Why doesn't stepconf let me set up a machine in Points?
[20:42:18] <cradek> andypugh: you can use whatever units you want, but only mm and inch make sense in gcode (g21, g20)
[20:43:18] <andypugh> Actually, I guess it is totally arbitrary, but as you say, if you set up in points then switching to inches would give you something wierd
[20:43:33] <andypugh> However, you could switch between mils/microns...
[20:43:38] <jymm> http://www.google.com/search?q=convert+1in+to+points
[20:43:40] <cradek> right, that's the only reason not to use parsecs or furlongs
[20:43:56] <jymm> lightyears?
[20:44:24] <andypugh> I think parsecs are bigger than mightyears
[20:44:45] <andypugh> But a megaparsec.barn is 2/3rd of a teaspoonful.
[20:44:46] <cradek> % units parsec lightyear
[20:44:46] <cradek> * 3.2615638
[20:44:56] <cradek> not by much...
[20:45:40] <jepler> $ units parsec
[20:45:40] <jepler> Definition: au / tan(arcsec) = 3.425107e-40 kg
[20:45:45] <jepler> hm, there's something wrong with my units
[20:45:55] <cradek> Definition: au / tan(arcsec) = 3.0856776e+16 m
[20:46:03] <cradek> buh?
[20:46:42] <jepler> $ units au
[20:46:42] <jepler> Definition: atto u = 1.6605387e-45 kg
[20:46:53] <cradek> uh no
[20:47:14] <cradek> what is a u?
[20:47:18] <andypugh> 2.5"
[20:47:24] <cradek> atomicmassunit
[20:47:24] <jepler> Definition: atomicmassunit = 1.66053873e-27 kg = 1.6605387e-27 kg
[20:47:27] <cradek> huh
[20:47:35] <andypugh> Sorry, 1U is 1.75 in
[20:47:47] <jepler> $ units U
[20:47:48] <jepler> Definition: retmaunit = 1.75 in = 0.04445 m
[20:47:55] <jepler> got that one
[20:48:21] <andypugh> the amu is normally called an amu, not a u
[20:48:48] <jepler> hm, looks like I have a copy of a rather old units.dat with some local units added..
[20:48:58] <jepler> it overrides the system one rather than just overlaying it
[20:49:19] <jepler> perd- 1|12 # one dozenth
[20:49:19] <jepler> perg- 1|12 perd # one grossth
[20:49:19] <jepler> pong pergfurlong
[20:49:28] <jepler> and it looks like (some of) my local units are just jokes
[20:49:30] <archivist> barn 10^-24 sq meters was named because to an electron the area is a barn door
[21:30:04] <jymm> It's been awefully quiet in here the last few months, kinda creepy. I though it was just the summer months, vacation thing, but it's Autumn now and still kinda quiet.
[21:30:33] <andypugh> <tumbleweed>
[21:31:02] <cradek> you mean on this channel?
[21:31:09] <jymm> yeah
[21:31:28] <andypugh> Mailing list is lively-ish. Dunno about the forum
[21:31:44] <cradek> well I hate to say it, but I think most of the emc-related talk has moved to #emc-devel because of all the noise in here - sometimes it's pretty bad.
[21:31:52] <archivist> oscillating threading is.....
[21:32:06] <MattyMatt> heavy dev on emc? in my experience any correlation between IRC & coding is negative
[21:32:27] <MattyMatt> aka what cradek said
[21:33:03] <cradek> lately a lot has been going on actually - I committed my new gui, currently jepler and I are working on sub-servoperiod probe position snapshots
[21:33:12] <jymm> It was just an observation. I don't think I've seen it this quiet in here for a few years.
[21:33:19] <cradek> gearing up for a new (maybe the last) 2.3 release
[21:33:37] <MattyMatt> cradek what is used for the GUI?
[21:33:58] <MattyMatt> what lib, and does it use x-win?
[21:34:01] <cradek> MattyMatt: gtk, it's a gui specifically for touchscreens
[21:34:29] <jymm> cradek: a rewrite of AXIS, or more based on TK
[21:34:30] <cradek> it's on git master if you want to play with it
[21:34:38] <cradek> nothing in common with AXIS
[21:35:15] <cradek> no tcl or tk code
[21:35:30] <jymm> ah
[21:36:13] <jymm> if you ever get a chance to, toss up some screenshots of what you have thus far
[21:36:21] <cradek> it's on git master if you want to play with it
[21:36:22] <MattyMatt> I was thinking yesterday any X based one is going to be dodgy with RTOS, but it was idle speculation really
[21:36:35] <cradek> MattyMatt: yeah that's not our (long) experience
[21:43:07] <cradek> some video cards screw up realtime, but most don't, and yay for that
[21:43:15] <MattyMatt> I have no experience of git. I've just got the hang of svn mostly
[21:43:44] <cradek> if you need help building a devel tree, there's good stuff on the wiki, or ask for help if you are stuck
[21:43:50] <MattyMatt> is it the cards or the drivers or X implementation?
[21:44:25] <cradek> cards/drivers, not sure, the worst offender is nvidia + nvidia's proprietary driver
[21:44:40] <cradek> but easy enough to avoid...
[21:45:04] <MattyMatt> not on this machine :) I'm playing with cuda
[21:45:22] <cradek> a fish?
[21:45:27] <MattyMatt> but I'll put the Matrox in the cnc machine
[21:46:05] <MattyMatt> cuda = NV gpu lib, which will be obsolete by next year when we'll all use OpenCL
[21:46:15] <cradek> The 'Cuda, based on the Formula S option, was available with either the 340, 383 and new for 1969 the 440 Super Commando V8.
[21:46:52] <cradek> ah, I don't keep up with that kind of stuff...
[21:46:59] <MattyMatt> ah, the hard drive. seagate
[21:47:12] <MattyMatt> :)
[21:47:25] <cradek> yep that evil fish on the hard drive
[21:47:25] <archivist> that was a car
[21:47:38] <cradek> all those 2G drives failed - we had a ton at customer sites.
[21:48:09] <cradek> I hated the sight of that fish after a while
[21:48:11] <cradek> bad omen
[21:48:32] <MattyMatt> a GUI can sidestep all drivers by using raw VGA
[21:49:05] <MattyMatt> the only drawback is that VGA limit
[21:49:59] <cradek> and no acceleration like bitblt - so important for a gui
[21:50:01] <MattyMatt> you lose hardware acceleration, but gain predictability
[21:50:16] <cradek> * cradek hugs his stack of old matrox cards
[21:51:18] <MattyMatt> I want an 850 to go with my 400, to try and get that 30 bit mode working
[21:53:03] <MattyMatt> Atari ST managed a usable GUI with no bitblt
[21:53:41] <MattyMatt> at least my memories are of usability :)
[21:53:49] <cradek> heh, funny how that goes :-)
[21:54:53] <MattyMatt> ST was a good system for specific real time apps, like Steinburg Pro 24
[22:12:47] <skunkworks> wee - garage doors openers now plugged into outlets instead of long extension cords.. :)
[22:15:46] <andypugh> My garage door openers are plugged into my shoulders....
[22:27:33] <MattyMatt> mine are in the hollow volcano of my dreams
[22:29:37] <andypugh> When you build your dream hollow volcano, save some money and don't install the self-destruct device
[22:30:18] <MattyMatt> the escape rocket will trigger an eruption
[22:31:07] <MattyMatt> it confuses the radar >:)
[22:31:50] <andypugh> You see, that sonds good, but one hal-pin wired wrongly and you will trigger the explosion flushing the toilet. And how are you ever going to test it?
[22:33:02] <mhaberler_> anybody figured out how to pull emc2-dev into eclipse with jig?
[22:33:27] <mhaberler_> s/jig/jgit/
[22:34:03] <andypugh> Sorry, speaking only for myself, I understand less than half the words.
[22:35:46] <mhaberler_> eclipse is an IDE which is comfortable like an old shoe once you get the hang of it
[22:35:48] <mhaberler_> just want to use it as a browsing aid since it's got an excellent indexer
[22:36:43] <mhaberler_> jgit is an eclipse plugin for git access which doesnt like me just yet
[22:37:05] <MattyMatt> this linear drive I got only has 14.5" travel. pointless having a 22" wide table now
[22:37:32] <MarkusBec> the only ide I use is emacs :D :P
[22:38:23] <MattyMatt> mhaberler_, does it autocommit when you save? is it supposed to?
[22:39:56] <mhaberler_> not on saving a file, commit is a seperate operation handled by plugins (cvs, svn, git etc)
[23:03:42] <archivist> I gave up on eclipse after loading an XML file killed it
[23:07:03] <jymm> archivist: how big was the xml file ?
[23:10:23] <frallzor> http://www.weebls-stuff.com/toons/Amazing+Horse/ enjoy this video =)
[23:12:43] <jymm> That's... just dumb.
[23:13:27] <archivist> 16 meg:)
[23:14:01] <archivist> it was a mysql manual section
[23:23:38] <andypugh> MattyMatt : You could try selling it back to eBay better advertised. But words like "CNC" and "Mach3!" in the title
[23:32:44] <MattyMatt> I'll just make a narrower table, but my new design looks like overkill. 4 shelf sliders each on their own thick wood spar for alignment
[23:34:42] <MattyMatt> putting 2 sliders on each of 2 spars would be neater and solid but less adjustable
[23:41:38] <MarkusBec> MarkusBec is now known as MarkusBec_away