#emc | Logs for 2007-07-09

[00:08:09] <jmkasunich_> jmkasunich_ is now known as jmkasunich
[02:45:03] <tomp> jepler: you're right, i had a system wide comp and a local comp
[03:05:01] <Skullworks-PGAB> this is bad.
[03:05:31] <Skullworks-PGAB> I'm begining to understand EAGLE, and its fun.
[03:05:39] <Jymmmm> Skullworks-PGAB: You thought she was 18, but....
[03:05:48] <Skullworks-PGAB> this is going to get expensive.
[03:06:49] <Skullworks-PGAB> I never laid a hand on her - I was in another state - those pics are all photoshopped...
[03:07:14] <Jymmmm> And your semen got in her, how?
[03:08:52] <Skullworks-PGAB> only navel activity around here is a weapons testing site - and I'm not playing any Pat Benatar...
[03:09:45] <Skullworks-PGAB> naval - my bad
[03:10:03] <Skullworks-PGAB> err now i"m really confused
[03:12:37] <Jymmmm> lol
[09:48:34] <archivist> * archivist scrolls up reading the weekends chatter.. jymm using CB mic connectors for motor wires WTF!!!
[10:01:40] <sebjames> Just did my first flame cut!
[10:03:21] <sebjames> So, I have drilling, motion and torch control. That means I just need to make user M commands (M1xx) occur _after_ the last traverse completes, and hack emc so that I can send a G commadn for "execute a homing sequence on the X axis" and I 'll be in business
[10:03:41] <archivist> got your drill motor fuse problem fixed then?
[10:12:15] <sebjames> archivist: Yes. Although the schematic looked just like a star/delta starter, in fact the motor has two sets of poles - it has to to have two distinct speeds
[10:13:22] <sebjames> So all I have to do is make sure not to switch both logic lines high from both low. Although they are mechanically interlocked via secondary contactors it still appears to be possible to close them together enough to short out the supply via the fuses.
[10:13:28] <archivist> we have a two speed 3phase motor on one of the lathes here
[10:13:46] <sebjames> Does it direct online start in each speed?
[10:14:17] <archivist> it goes through low to high
[10:14:39] <archivist> but only a small motor
[10:15:27] <archivist> actually we have two of them both on schaublin lathes
[10:16:23] <anonimasu> hm
[10:22:51] <sebjames> Must vary the input frequency to vary the speed?
[10:23:02] <sebjames> Does it have a driver unit?
[10:23:08] <archivist> no
[10:23:31] <sebjames> I'm trying to understand the task execution code at the moment
[10:23:42] <sebjames> emctaskmain.cc
[10:24:27] <sebjames> At the moment, user commands get issued immediately, and all my user commands should be queued such that they execute after the previous command is complete
[10:25:33] <anonimasu> sebjames: I disagree
[10:26:02] <anonimasu> sebjames: I used user M codes to turn on cutting oxygen for my torch.. and they always did execute when supposed to..
[10:26:10] <anonimasu> though that was "external M codes"
[10:32:11] <sebjames> external M codes?
[10:32:52] <sebjames> I'm talking about the script s in nc_files that get executed.
[10:33:16] <sebjames> They definitely get executed before a traverse is complete
[10:35:42] <anonimasu> that's odd..
[10:35:45] <anonimasu> mine didnt do that
[10:36:10] <sebjames> emcSystemCmd() forks and execs the script. emcTaskIssueCommand() is the function which calls emcSystemCmd()
[10:39:40] <sebjames> emcTaskIssueCommand is the function which executes tasks immediately. emcTaskIssueCommand is called either by emcTaskPlan() or emcTaskExecute()
[10:41:11] <sebjames> Something is up here, in emcTaskCheckPreConditions EMC_SYSTEM_CMD would trigger a wait for motion and io, but it seems that the system command tasks are being issued directly by emcTaskPlan()
[10:41:34] <sebjames> ...circumventing emcTaskCheckPreConditions()
[10:56:32] <sebjames> Weird! The order of commands is wrong...
[10:57:24] <sebjames> When I turn on EMC_DEBUG_TASK_ISSUE, then I see my user M command gets acted on before the previous command
[10:59:14] <sebjames> Or strictly, before the previous motion command.
[11:03:30] <sebjames> If the EMC_SYSTEM_CMD task is placed after an EMC_TRAJ_LINEAR_MOVE task in the user program, it is executed _before_ that EMC_TRAJ_LINEAR_MOVE task. If the EMC_SYSTEM_CMD task is placed after, for example, a pause command, then it is executed after the pause command
[11:12:04] <martin_lundstrom> hello people
[11:12:36] <Dallur> hey martin
[11:12:43] <martin_lundstrom> Dallur: I added a small piece to the config
[11:12:58] <Dallur> great, was it a bug or ?
[11:15:23] <martin_lundstrom> a ignore input on Z down, just after input arc ok goes up. My arc ok comes up to fast and sends z down onto the surface and results in no beam. I hope thetimer will do me good. I found this feature in another thc manual.
[11:15:53] <martin_lundstrom> its a timer and a and gate
[11:16:27] <alex_joni> sebjames: sounds like a bug to me
[11:17:13] <Dallur> martin: great, nice to have that bug fixed :)
[11:17:23] <Dallur> martin: have you submitted it already to cvs ?
[11:17:52] <martin_lundstrom> no
[11:17:59] <Dallur> martin: if not you can mail it to me dallur (at) dallur.com and I can submit it :)
[11:18:06] <martin_lundstrom> nice
[11:18:16] <martin_lundstrom> simpler for me
[11:18:48] <Dallur> I will also update the diagram to reflect the change :)
[11:19:08] <martin_lundstrom> i already did that
[11:19:37] <Dallur> martin_lundstrom: ohh great, then if you send me it to I can submit both
[11:19:45] <martin_lundstrom> but it does not look very nice, i messedup the nice formating og you arrows
[11:20:20] <Dallur> martin_lundstrom: no matter, it's all fixable :)
[11:20:28] <martin_lundstrom> :)
[11:20:33] <Dallur> I must be in a good mood today, nothing but ;)
[11:22:30] <martin_lundstrom> The theme to sound of music on the radio here, feels funny
[11:24:14] <Dallur> gtg, bbl (picking up a couple of parts for the table)
[11:24:16] <martin_lundstrom> Dallur: Did you yet get your hand on your new thc?
[11:24:18] <martin_lundstrom> ok
[12:15:42] <sebjames> alex_joni: Agreed
[12:46:12] <alex_joni> sebjames: I suggest you file a bug report at sourceforge
[12:46:17] <alex_joni> you can even assign it to me
[12:46:19] <alex_joni> ;)
[12:46:27] <alex_joni> sebjames: that way it won't get forgotten
[12:50:59] <sebjames> alex_joni: I'm just working through it now. Runnign with DEBUG = EMC_DEBUG_INTERP | EMC_DEBUG_INTERP_LIST. Seeing some strange line_numbers
[12:51:34] <sebjames> In interpl.cc
[12:55:49] <a-l-p-h-a2> a-l-p-h-a2 is now known as a-l-p-h-a
[13:24:15] <rayh> You guys know of any rt issues with a usb mouse?
[13:29:21] <alex_joni> rayh: only SMI which takes care of some USB on intel machines
[13:33:13] <rayh> Hey thanks Alex. I'll try one here and see if I get any errors.
[13:33:29] <rayh> How you doing, Alex?
[13:36:16] <skunkworks> rayh: Hi - keeping busy?
[13:36:47] <skunkworks> I have had usb keychain drives cause latency errors..
[13:36:56] <skunkworks> but that is it
[13:39:07] <rayh> Hi skunkworks. Thanks for that.
[13:39:37] <rayh> The notes will need to read. Do not plug in a usb stick while running your mill!
[13:40:05] <rayh> I'll try the several sticks I have here and see what shakes out.
[13:40:31] <rayh> Do you remember if that was an intel chipset and mobo?
[13:42:50] <cradek> I definitely haven't seen any problem with USB mice
[13:46:49] <alex_joni> hi ray, pretty good thanks
[13:49:08] <hcseb> hcseb is now known as sebjames
[13:52:49] <skunkworks> rayh: I think one of the computers for sure had the intel chipset. The other I am not sure.
[13:56:05] <rayh> Thanks.
[13:56:19] <sebjames> alex_joni: Have you time to look at emcTaskMain.cc?
[13:57:07] <anonimasu> rayh: I hve the same with usb sticks
[14:11:36] <alex_joni> sebjames: in 5 minutes?
[14:14:18] <alex_joni> sebjames: I can now
[14:14:46] <alex_joni> anything in particular?
[14:15:19] <sebjames> alex_joni: COol
[14:15:24] <sebjames> Yes
[14:15:33] <sebjames> Around line 940
[14:16:07] <sebjames> I'm just going to paste some output into pastebin...
[14:17:10] <alex_joni> sebjames: my first impression by looking at the stuff is that it should work..
[14:17:41] <sebjames> http://pastebin.com/944987
[14:17:49] <sebjames> It mostly works...
[14:18:26] <alex_joni> please use pastebin.ca
[14:18:30] <alex_joni> .com works like crap ;)
[14:18:59] <sebjames> I just found that out...
[14:19:02] <sebjames> Very slow
[14:19:08] <sebjames> Sure I'll use .ca
[14:19:12] <sebjames> Do I need to paste there too?
[14:19:20] <alex_joni> no, it opened now
[14:19:23] <sebjames> ok
[14:19:32] <alex_joni> what should I be looking at?
[14:19:42] <sebjames> That's debugging output, some of it is my own debugging output
[14:19:51] <alex_joni> I see that..
[14:20:05] <sebjames> But see the block where the motion command G1Y148F500 is interpreted?
[14:20:51] <sebjames> Well, it has no NML_INTERP_LIST::append() call, unlike the others
[14:20:55] <alex_joni> yeah, I see it..
[14:21:06] <alex_joni> but this is just interpreting
[14:21:25] <alex_joni> the actual execution is further down (Issuing ..)
[14:21:32] <sebjames> Yes, I know....
[14:21:39] <sebjames> But that bit is out of order
[14:21:44] <alex_joni> which one?
[14:21:59] <alex_joni> ah, I see
[14:22:02] <sebjames> lines 66 and 68 should be the toher way round
[14:22:19] <sebjames> I have programmed move, followed by Outfeed clamp up (M125)
[14:22:28] <sebjames> But I get Outfeed clamp up, then move......
[14:23:13] <sebjames> line 57 is the NM_INTERP_LIST::append() call for the move
[14:23:47] <sebjames> And because it happens after line 46 - the append call for the user command, they seem to execute out of order. :(
[14:24:30] <sebjames> The line_number specifies the correct order, but they're getting appended in the wrong order.
[14:25:09] <sebjames> Any idea why the append call for the motion command gets delayed? I can't see anything obvious.
[14:25:57] <alex_joni> * alex_joni is thinking
[14:26:01] <alex_joni> and grepping :D
[14:26:25] <alex_joni> it would be usefull to look at the interp list
[14:26:30] <alex_joni> and check the exact order in there
[14:26:45] <alex_joni> after interpretation is done, before execution
[14:27:01] <alex_joni> then we can see if it's a matter of wrongly appending to the interp list, or executing out of order
[14:28:43] <sebjames> OK
[14:29:05] <sebjames> I also want to look at the code which defines interp.line()
[14:29:14] <sebjames> That is what emcTaskPlan() calls
[14:29:39] <sebjames> Where is Interp::line()?
[14:30:33] <sebjames> rs274ngc/
[14:31:19] <alex_joni> rs274ngc.hh
[14:31:25] <alex_joni> int line() { return sequence_number(); }
[14:32:07] <alex_joni> int Interp::sequence_number()
[14:32:07] <alex_joni> {
[14:32:07] <alex_joni> return _setup.sequence_number;
[14:32:07] <alex_joni> }
[14:32:20] <alex_joni> that one is in rs274ngc_pre.cc
[14:32:39] <alex_joni> sebjames: can you run the same program again with debug set to max?
[14:32:44] <alex_joni> and pastebin the whole thing..
[14:34:29] <sebjames> Ok
[14:34:33] <sebjames> 2 minutes
[14:38:25] <x3rox> Hello, since some hours I am the lucky owner of a broken CNC mill (electronics faulty, but mechanically intact and in good confition). Has steppers with 8°/Step and 2.5A. Is this usual, or are this steppers with very low resolution?
[14:38:53] <sebjames> http://pastebin.ca/610506
[14:39:29] <x3rox> ... the steppers drive spindles with ~4mm/turn
[14:40:10] <bill2or3> most steppers are 1.7°/step
[14:40:20] <cradek> you're right, I think that's .09mm/step, which is quite low resolution
[14:40:50] <bill2or3> but they're probablly geared down with a belt/pulley or something.
[14:41:37] <x3rox> Oh! Doesn't matter, in worst case I take newer steppers. But they told me that the mill has 0.015mm resolution, so did they microstepping?
[14:42:10] <x3rox> This is an old Gravograph VX92 mill
[14:43:29] <cradek> sounds like some of the specs you were given may not be right, you'll have to check it for yourself
[14:44:15] <bill2or3> in any case, lucky you, free is free.
[14:44:17] <x3rox> Can somebody tell me a driver IC for 2.5...3 A? Should have a built in chopper (for power limitation) and be able to drive a bipolar stepper (4 wires)
[14:44:47] <cradek> yes free is a nice price
[14:44:55] <sebjames> alex_joni: The Outgoing motion ids are out of order...
[14:44:56] <cradek> what voltage?
[14:45:08] <bill2or3> drivers are.. complicated. Gecko Drive's seem to be the standard, but they are more than free.
[14:45:27] <alex_joni> sebjames: I notice the NML_INTERP_LIST::append are out of order too
[14:45:39] <alex_joni> sebjames: let's move to #emc-devel
[14:46:17] <jepler> x3rox: L297+L298 (1 L298 per motor) is good for 46V, 2A continuous and there are free board designs available. L297+L298 (2 L298 per motor) should in theory be good for 46V, 4A continuous power but I've never seen a board that does this.
[14:46:42] <jepler> (each L298 contains two H-bridges, and you can "gang" them together for increased current according to the datasheet)
[14:46:54] <sebjames> Ok.
[14:46:56] <jepler> L297 is half-stepping, constant current chopper
[14:48:33] <x3rox> Yes, this is what I think, too. The L297/298 is also very ugly, because it has so many discrete parts. (why doesn't it have internal clamp diodes?) Do you think that 2A are sufficient, too?
[14:49:00] <x3rox> (Dont' want to buy 3 gecko drives)
[14:50:29] <jepler> http://www.pmdx.com/PMDX-150/index.html and http://www.xylotex.com/4axBoard.htm are both based on an "allegro" chip with slightly higher A rating but lower supply voltage
[14:52:32] <x3rox> But the lower voltage results in lower maximum speed, because I will loose torque earlier. Is this what I really want? :-/
[14:58:51] <jepler> x3rox: my understanding is that lower amps decreases torque in the constant-torque (low speed) operating region, while lower volts decreases high-speed torque
[15:01:09] <x3rox> This is what I also found out by reading a lot of docs. So it will probably be the better solution to use higher current at lower voltage. But I wonder that there are no ICs with much higher current. Do the gecko drives use discrete mos-fets? Some support more than 8A !
[15:01:20] <jepler> consider the graph on page 6 of http://geckodrive.com/photos/Step_motor_basics.pdf and imagine that the curves are labeled "1x rated value" and ".5x rated value"...
[15:01:53] <jepler> yes, I think they do
[15:02:21] <x3rox> Therefore....
[15:03:20] <x3rox> So if I understand right, the L297/298 pair is "state of the art" for standard mills. Right?
[15:04:31] <jepler> the main lack of the L297 is microstepping
[15:05:07] <x3rox> Yes. And this, together with this ancient steppers of my mill (8°/step) will not be good...
[15:06:46] <x3rox> Are there steppers available which only need 0.75A, having good torque?
[15:07:12] <x3rox> Maybe I replace this old steppers by better ones...
[15:33:53] <jepler> .75A is pretty low current
[15:33:57] <jepler> why .75A?
[15:34:48] <skunkworks> jepler: he left..
[15:34:49] <skunkworks> ;)
[15:35:12] <jepler> I know
[15:40:12] <skunkworks> would there be better position control if you could do a velocity pid loop inside the emc's position loop? Using the encoder velocity pin?
[15:40:24] <skunkworks> all within emc
[15:48:27] <awallin> are we updating www.linuxcnc.org or has it crashed ?
[15:48:45] <alex_joni> hrmm.. not that I knew about it
[15:49:08] <Dallur> hmmf wasn't like that an hour or two ago
[15:49:47] <alex_joni> Error 2 = FATAL ERROR: Connection to database server failed.
[15:49:59] <alex_joni> probably a problem with the mysql server
[15:50:38] <awallin> anyone know what an electrospindle is??
[15:53:30] <CIA-8> 03alex_joni 07TRUNK * 10emc2/src/emc/task/emctask.cc: call flush_segments() when we get to a custom M code, so order doesn't get mangled
[16:02:08] <alex_joni> awallin: it's back up
[16:05:19] <CIA-8> 03alex_joni 07v2_1_branch * 10emc2/src/emc/task/emctask.cc: backport fix for out-of order planning of motion and custom M-codes.
[16:05:58] <awallin> alex_joni: ah, good. someone on cnczone was asking about HSM and control software etc. so I pointed him to the usual emc resources
[16:08:52] <CIA-8> 03alex_joni 07v2_1_branch * 10emc2/debian/changelog: mention backport fix for out-of order planning of motion and custom M-codes.
[16:24:59] <skunkworks> awallin: nice posts.
[16:48:44] <acemi_> acemi_ is now known as acemi
[16:51:20] <jepler> Radius to end of arc differs from radius to start: start=(X0.2508,Y0.2880) center=(X-1231.2414,Y1232.8410) end=(X1.3786,Y1.5556) r1=1742.3432 r2=1742.2444
[16:51:36] <jepler> hm yes, something seems fishy about these numbers
[16:52:13] <cradek> uh-oh
[16:56:35] <Dallur> jepler: variable radius curves = splines,, bug,, or feature ? :)
[16:57:25] <jepler> Dallur: bug
[16:57:29] <jepler> ask me a harder question
[16:58:38] <alex_joni> * alex_joni hopes those are inches
[16:59:31] <cradek> arcs are hard
[16:59:34] <Dallur> jepler: a train leaves New York headed to London, assuming it accelerates at 10m/s and that it's not affected by air/track resistance how fast would it be going when it would need to start slowing down ?
[17:00:14] <archivist> * archivist thinks the water resistance should slow it nicely
[17:00:15] <jepler> Dallur: that depends on the water resistance which is not given in the problem
[17:00:18] <skunkworks> umm - train won't work over water..
[17:00:56] <archivist> heh even the questions in here have bugs
[17:01:01] <Dallur> :)
[17:02:30] <archivist> and straight line involves some rock and a hard place
[17:05:01] <Dallur> archivist: might also include some magma
[17:05:13] <archivist> hehe
[17:42:30] <skunkworks> awallin: les watts that used to be on here.. (hasn't been on for a while) Used emc1 in a high speed router system.. I am thinking with all the work done on emc2 that it is better.
[17:43:13] <skunkworks> he had issues with que starvation IIRC - which I believe has been greatly improved with emc2
[18:06:34] <skunkworks> anyone have any thoughts on the pid velocity loop in side the position pid loop? Say the pluto velocity as the feedback and the commanded velocity (output of the position pid) as the commanded?
[18:06:51] <skunkworks> instead of it being in the servo amplifier?
[18:08:10] <skunkworks> * skunkworks is just thinking out loud
[18:10:43] <awallin> skunkworks: queue starvation: yep, might be better in emc2, wonder if performance benchmarking can be done in sim mode. not many have a hsm mill controlled by emc2...
[18:14:49] <awallin> hi lerman
[18:15:08] <lerman> hello.
[18:15:30] <awallin> what's up? G-wiz work?
[18:15:48] <lerman> Haven't had a chance to work on it recently.
[18:16:42] <awallin> ok... I'm messing around with this drop-cutter idea for CAM
[18:17:24] <lerman> My most recent work using python was somewhat dissapointing. Stuff that worked on windows failed on linux.
[18:17:38] <lerman> Drop cutter?
[18:17:43] <skunkworks> awallin: I remember jepler/cradek showing a halscope of queue starvation..
[18:18:16] <awallin> lerman: the idea that you drop a cutter down along the z-axis until it touches the part model (STL format triangulation)
[18:18:39] <lerman> Then move to a new spot?
[18:19:35] <awallin> you first compute a dense grid in the x,y plane somewhere above the part. then at each point you drop down the cutter to the model
[18:20:22] <lerman> Since any point in XY will have at least two faces (front and back) and maybe more (undercuts), you will need to determine the frontmost face at any point.
[18:21:07] <awallin> yes, you compute a z-value for the vertices, the edges and the face of each triangle. then you take the highest value from all of these as the cutter position
[18:21:45] <anonimasu> that sucks..hello
[18:22:10] <anonimasu> err wrong place :)
[18:22:11] <anonimasu> hi.
[18:23:06] <anonimasu> awallin: you mran Z plunge roughting?
[18:23:43] <awallin> it can be used for plunge roughing, but for raster finish/semi-finish machining too
[18:23:50] <anonimasu> hm ok?
[18:23:53] <anonimasu> never seen that one
[18:24:09] <anonimasu> awallin: did you get anywhere with the pocketing algorithm?
[18:24:21] <awallin> moving the tool in zigzag/whatever patterns around the model
[18:24:48] <awallin> pocketing: it's just an idea for now. I want to do 2D offset calculations and traditional pocketing toolpaths first
[18:24:55] <anonimasu> awallin: agreed
[18:25:06] <anonimasu> awallin: I'm really interested in cam stuff but my math is a bit weak..
[18:25:09] <anonimasu> :/
[18:25:47] <awallin> are you good at c++/GUI/OpenGL stuff? I'm weak in that department!
[18:26:36] <anonimasu> depends on what you mean by good
[18:26:57] <anonimasu> most of what I code nowdays for UI is mono.. as that's crossplatform enough for me
[18:27:12] <anonimasu> Win CE
[18:27:15] <anonimasu> (crap edition)
[18:28:31] <awallin> I wonder if the gui etc. could be done in python, or if it really needs to be C++ for speed
[18:28:45] <awallin> OpenGL with a lot of surfaces/lines etc. might require some speed
[18:28:49] <anonimasu> yeah..
[18:28:54] <anonimasu> that needs c++ for the renderer..
[18:29:00] <anonimasu> though the rest of the UI dosent need to be fast..
[18:29:11] <anonimasu> it needs to be easy extendable.. that's it I guess..
[18:31:28] <awallin> yes. are there any cool technologies for plugin/extendability stuff? I'm clueless about this but I've heard buzzwords like corba and object databases and such..
[18:31:44] <anonimasu> hm, I mean it has to be easy to add new dialogs for settings and stuff :)
[18:32:29] <awallin> ah, yes that too
[18:32:46] <awallin> cross platform would not be wrong either. win/linux/macX
[18:47:45] <skunkworks> logger_emc: bookmark
[18:47:45] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-07-09.txt
[19:03:26] <jepler> .q/query cradek
[19:03:38] <jepler> oops
[19:17:37] <maddash> --- emc2/src/emc/ini/initraj.cc.new2007-07-09 15:12:57.000000000 -0400
[19:17:37] <maddash> +++ emc2/src/emc/ini/initraj.cc2006-10-05 21:23:54.000000000 -0400
[19:17:37] <maddash> @@ -84,7 +84,6 @@
[19:17:37] <maddash> char home[LINELEN];
[19:17:37] <maddash> EmcPose homePose = { {0.0, 0.0, 0.0}, 0.0, 0.0, 0.0 };
[19:17:38] <maddash> double d;
[19:17:40] <maddash> - int iniretval;
[19:17:43] <maddash>
[19:17:45] <maddash> if (NULL != (inistring = trajInifile->find("AXES", "TRAJ"))) {
[19:17:47] <maddash> if (1 == sscanf(inistring, "%d", &axes)) {
[19:17:48] <maddash> @@ -104,9 +103,9 @@
[19:17:51] <maddash> }
[19:17:53] <maddash> axes = 0;// default
[19:17:55] <maddash> }
[19:17:57] <maddash> - if (0 != (iniretval=emcTrajSetAxes(axes))) {
[19:17:59] <maddash> + if (0 != emcTrajSetAxes(axes)) {
[19:18:02] <maddash> if (EMC_DEBUG & EMC_DEBUG_CONFIG) {
[19:18:03] <maddash> - rcs_print("bad return value from emcTrajSetAxes: %d\n",iniretval);
[19:18:05] <maddash> + rcs_print("bad return value from emcTrajSetAxes\n");
[19:18:07] <maddash> }
[19:18:09] <maddash> return -1;
[19:18:11] <maddash> }
[19:18:15] <maddash> er, whoops
[19:18:17] <maddash> http://pastebin.ca/611046
[19:18:19] <maddash> stupid X11 paste
[19:18:20] <cradek> please use pastebin for long pastes
[19:18:23] <maddash> jepler: ^^
[19:18:50] <maddash> cradek: sorry about that, I got confused with the contents of the X11 clipboard
[19:18:51] <cradek> also, your diff is backward
[19:19:00] <maddash> eh?
[19:19:00] <cradek> ah, I've done that
[19:20:42] <maddash> silly me. hold on.
[19:20:43] <maddash> http://pastebin.ca/611055
[19:21:55] <jepler> that one won't build
[19:22:18] <jepler> + int iniretval;
[19:22:19] <jepler> + if (0 != (emctsa=emcTrajSetAxes(axes))) {
[19:22:45] <maddash> * maddash apologizes profusely. brb.
[19:23:14] <jepler> you're also editing a very old version of that file -- the last change in CVS was 2007/03/22
[19:23:35] <jepler> if it's useful to print the return value for one failure, why isn't it for the others? I notice it's not printed anywhere else
[19:24:53] <jepler> a glance at emcTrajSetAxes doesn't reveal that it has very interesting error values (just -1 or whatever usrmotWriteEmcmotCommand returns)
[19:28:25] <maddash_> seems like the guy across the isle has been messing with my sources. jepler: http://pastebin.ca/611079
[19:28:38] <maddash_> maddash_ is now known as maddash
[19:29:11] <cradek> haha, that one's backward again
[19:29:38] <maddash> jepler: the patch is near trivial, but I thought that it would be nice to have something besides "emcTrajSetAxes gave a bad return"
[19:29:56] <cradek> it would be nice if it would tell you what that means
[19:30:30] <cradek> something that would help the user answer "ok so what do I have to do about it?"
[19:30:56] <maddash> good idea. should I tack on something to decode iniretval?
[19:31:18] <cradek> maybe - I haven't studied what it might mean when this happens
[19:31:41] <cradek> but I think adding "-1" (or any number) doesn't help make an error message clearer
[19:32:08] <alex_joni> cradek: you think?
[19:32:15] <cradek> :-P
[19:32:23] <alex_joni> * alex_joni loves error messages like : internal error -75237
[19:32:43] <maddash> btw, this is done on 2.1.5
[19:33:06] <cradek> you should be working with the current cvs trunk if you want to submit patches for incorporation
[19:33:36] <jepler> cvs will help you get your patches the right way around, every time
[19:33:47] <cradek> also true
[19:35:03] <maddash> didn't know that. cvs only supplies me with a regular dose of cold medicine.
[19:35:15] <alex_joni> cvs diff -u
[19:35:49] <alex_joni> that will get a diff from your working copy to the repository
[19:36:31] <maddash> why is there a .vimrc in the cvs?
[19:40:30] <maddash> I'm filtering out everything with "trunk-premerge7" and now everything in ini/ and motion/ is 2 years old -- do i have the right tag?
[19:42:43] <jepler> no
[19:42:49] <jepler> you want to check out with no "-r" tag
[19:44:44] <alex_joni> cvs -d:...: co emc2
[19:45:39] <maddash> ah.
[20:04:38] <maddash> hal_parport only works with my pci parport when I load and unload parport_pc -- what's going on?
[20:05:05] <jepler> maddash: that is intentional
[20:05:14] <robin_sz> all this talc is beginning to get up my nose ... is there noting else to use with latex to make it slip on?
[20:06:05] <jepler> maddash: if the I/O ports are reserved by some other driver (typically parport_pc) then the hal_parport driver will not load
[20:08:43] <alex_joni> maddash: you can also try 'loadrt probe_parport' instead of loading&unloading parport_pc
[20:08:45] <maddash> jepler: maybe I wasn't clear. not only does halparport not work with parport_pc loaded, it won't work after i've blacklisted parport-pc and reboot.
[20:09:25] <jepler> maddash: maybe your system is one where probe_parport is needed. http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting#Parallel_port_no_longer_works_in_EMC_2_0_1_or_later_emc_starts_but_motors_don_t_turn
[20:09:25] <alex_joni> maddash: hal_parport won't load if there is any other kernel module already registered at the IO address you are after
[20:09:27] <maddash> alex_joni: ok. i've been wondering for a while what probe_parport is used for.
[20:10:07] <maddash> alex_joni: but i blacklisted parport_pc so that it doesn't get loaded at all. so doesnt that mean that the io addr is free?
[20:10:19] <jepler> maddash: sorry, I should have read what you said more closely
[20:11:06] <jepler> "In EMC 2.0.1, the Linux parallel port driver must be completely disabled for EMC's hal_parport driver to load. However, as we learned after the release of 2.0.1, this disabled certain parallel ports that are "PNP" (plug and play) devices. The new probe_parport realtime module performs the probing for one type of PNP port. If probe_parport doesn't allow your card to work, please contact the EMC developers.
[20:11:12] <jepler> "
[20:12:27] <maddash> hmm, i should've grepped the doc srcs. let me try this out. brb.
[20:12:37] <alex_joni> maddash: that's from the wiki..
[20:13:12] <Rugludallur> robin_sz: I have used teflon and silicone spray on latex/rubber with good results, as long as there's no kerosene (which destroys the latex/rubber)
[20:17:07] <anonimasu> robin_sz: there's always pills..
[20:17:33] <maddash> where's jmkasunich? haven't seen him ever since we discussed switching the homing to level-detection
[20:25:56] <ohiopctechDOTcom> yo! i wasn't sure where else i would find so many genius people to ask about this at one time... so here it goes: i was in the dumpster @ one of the E-Check places that they just shut down around town ehre and found this 19" rack enclusure that says seimens on the front... i think it may be some kind of pressure tester/regulator. it has a lcd on the front and a button matrix. it won't power on bu it hink it has some
[20:26:41] <ohiopctechDOTcom> i'm going to take some pics of the guts here in a few and upload them to my website if anyone may be interested in the parts inside it...
[20:27:16] <alex_joni> front and a button matrix. it won't power on bu it
[20:27:16] <alex_joni> hink it has some
[20:27:23] <alex_joni> it got cut off there
[20:27:40] <ohiopctechDOTcom> it won't power on bu it hink it has some good parts in it. would anyone know what it may be?
[20:27:49] <ohiopctechDOTcom> *that was the rest of that line
[20:27:56] <alex_joni> not by that description :P
[20:27:59] <ohiopctechDOTcom> lol
[20:28:05] <ohiopctechDOTcom> i'm taking some pics right now
[20:32:34] <anonimasu> :)
[20:32:36] <anonimasu> *curious*
[20:33:42] <ohiopctechDOTcom> ok i got some shots of the front/back & inside... getting them off the camera now
[20:34:28] <ohiopctechDOTcom> there is a cpu board inside that says "pc/chip (chips)" on the biggest ic chip on the board
[20:34:38] <ohiopctechDOTcom> is it a single board computer maybe?
[20:34:50] <alex_joni> let's see them pics
[20:35:00] <ohiopctechDOTcom> gimme another sec...
[20:35:25] <alex_joni> it passed..
[20:35:38] <LawrenceG> new goodies....
[20:35:43] <ohiopctechDOTcom> sonofabi...
[20:35:55] <robin_sz> Rugludallur, silicone spray?
[20:36:00] <anonimasu> ohiopctechDOTcom: www.imagebin.org
[20:36:19] <ohiopctechDOTcom> anonimasu: ohiopctech.com ;)
[20:36:38] <maddash> :( probe-parport didn't do it. I even patched it to output a simple "Hello world" message at init to make sure it was getting loaded ( and it was).
[20:36:41] <anonimasu> ah ok
[20:37:21] <robin_sz> been playing with that Rabbit processor board today, interesting pseudo realtime stuff on it
[20:38:12] <jepler> robin_sz: can you develop for rabbit with linux/gcc yet, or do you have to use their own dev environment?
[20:38:24] <alex_joni> own dev afaik
[20:38:34] <robin_sz> jepler, there own dev thing ... seems quite good
[20:38:44] <alex_joni> not free enough though
[20:38:50] <ohiopctechDOTcom> is pseudo realtime like uhh.. not realtime?
[20:38:56] <anonimasu> ohiopctechDOTcom: where is that images?
[20:38:59] <alex_joni> mostly realtime :P
[20:39:05] <anonimasu> thoose..
[20:39:06] <Rugludallur> robin_sz: yup, I got some from a company called Fabory, use it for rubber/latex parts which are sensitive to oil
[20:39:26] <robin_sz> Rugludallur, interesting ... I'll have to try that.
[20:39:37] <robin_sz> Rugludallur, and its ok for skin?
[20:40:06] <Rugludallur> robin_sz: sure, no issues
[20:40:15] <robin_sz> * robin_sz notes that down
[20:40:39] <ohiopctechDOTcom> so these pics are kinda huge and some are semi-crappy shots.. they are uploading by ftp now...
[20:40:44] <Rugludallur> robin_sz: also great for cleaning windshield wipers (wd40 destroys them)
[20:40:46] <anonimasu> oh ok
[20:40:55] <anonimasu> wd40 is like kerosene..
[20:40:56] <anonimasu> :D
[20:40:57] <robin_sz> coo
[20:41:01] <Rugludallur> anonimasu: yup
[20:41:02] <anonimasu> err contains lots of it if I remember it right
[20:41:07] <anonimasu> great for cleaning stuff..
[20:41:07] <ohiopctechDOTcom> who in thier right mind would use wd40 to clean windshield wipers?
[20:41:09] <robin_sz> * robin_sz didnt reallise people cleaned windscreen wipoers
[20:41:24] <anonimasu> heh..
[20:41:26] <robin_sz> probably the same sort of people that clean their cars I guess
[20:41:37] <anonimasu> I clean my mill table with it sometimes..
[20:41:48] <Rugludallur> ohiopctechDOTcom: a bunch of ppl seem to, it makes them work great for a couple of days untill the rubber cracks and hardens
[20:41:49] <ohiopctechDOTcom> i use it to clean my stones
[20:41:51] <robin_sz> it must be .. .oh, 5, 10 years since we leaned a car
[20:41:57] <robin_sz> cleaned
[20:41:58] <ohiopctechDOTcom> wd-40 and rub two stones together
[20:42:04] <robin_sz> stones?
[20:42:15] <ohiopctechDOTcom> stones for stoning things
[20:42:20] <robin_sz> ahh.
[20:42:21] <ohiopctechDOTcom> sharpening stones
[20:42:22] <robin_sz> stones
[20:42:26] <Rugludallur> ahh, I have not stoned people in a while
[20:42:32] <robin_sz> I remember those from the old days
[20:42:33] <anonimasu> lol
[20:42:34] <ohiopctechDOTcom> deburring stones
[20:42:38] <skunkworks> everybody must get stoned
[20:42:40] <ohiopctechDOTcom> polishing stones
[20:42:50] <robin_sz> I use diamond plated these days
[20:42:59] <anonimasu> Rugludallur: yeah it was a while since we burnt any witches too..
[20:43:30] <Rugludallur> anonimasu: you can tell they are witches if they are heavier than ducks
[20:43:41] <robin_sz> rubbish
[20:43:49] <robin_sz> its if they weight the same as a duck
[20:43:56] <skunkworks> * skunkworks was quoting bob dylan
[20:43:57] <anonimasu> :D
[20:45:42] <alex_joni> lets start with skunkworks
[20:46:00] <alex_joni> skunkworks: or is that too late?
[20:46:21] <skunkworks> ?
[20:46:34] <skunkworks> you want to stone me?
[20:46:35] <anonimasu> http://ohiopctech.com/cnc/cubes/DSC00812.JPG
[20:46:38] <anonimasu> I need to make a few..
[20:46:46] <ohiopctechDOTcom> http://ohiopctech.com/wtfisit/ <-- pics of umknown seimens equipment
[20:47:00] <anonimasu> ohiopctechDOTcom: did you consider detaching the small cubes?
[20:48:03] <ohiopctechDOTcom> i haven't done anything with it but plug it in yet.. i know i may be able to use the torroid in it for my cnc pwoer supply ;D
[20:48:43] <skunkworks> one of these days I will finish this http://www.electronicsam.com/images/woodcube.jpg
[20:48:52] <anonimasu> try finding a model number
[20:49:05] <ohiopctechDOTcom> hey that's badass skunkworks
[20:49:40] <maddash> well, boys and girls, I overcame my problems with a simple boot-time script that unloads parport/_pc
[20:49:51] <maddash> jepler: , alex_joni^^
[20:49:56] <anonimasu> maddash: well done
[20:50:16] <ohiopctechDOTcom> anonimasu: is says "divmat" on the back
[20:50:20] <skunkworks> Thanks - http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Oword#Sample_1_One_side_of_a_ball_in_cage
[20:50:28] <maddash> anonimasu: it's a hack. most likely an unreliable one.
[20:50:34] <maddash> skunkworks: is that a furball?
[20:50:46] <anonimasu> ohiopctechDOTcom: it looks custom made :P
[20:50:58] <Rugludallur> ohipctech: it's a NOX generator
[20:51:20] <anonimasu> heh
[20:51:46] <Rugludallur> ohiopctech: at least that's what is says on the back :)
[20:52:03] <ohiopctechDOTcom> so what is a nox generator good for?
[20:52:05] <anonimasu> cool
[20:52:06] <ohiopctechDOTcom> lol
[20:52:10] <ohiopctechDOTcom> a nonworking one at that
[20:52:40] <ohiopctechDOTcom> i figure if nothing else it will be my new cnc comtrol rack mount cabbinet
[20:52:43] <skunkworks> getting high?
[20:52:54] <skunkworks> nox generator
[20:52:57] <ohiopctechDOTcom> i think the actual NOX gen part has been removed
[20:52:58] <alex_joni> http://www.environsys.com/pdf/CAI_600SGD.pdf
[20:53:01] <alex_joni> that looks a lot like it
[20:53:19] <ohiopctechDOTcom> ok...
[20:53:27] <ohiopctechDOTcom> so this one only has half of what that pdf has
[20:53:33] <ohiopctechDOTcom> it only has on bubble on the front
[20:53:41] <anonimasu> hehe
[20:54:09] <ohiopctechDOTcom> ty alex_joni
[20:54:20] <alex_joni> ohiopctechDOTcom: warm-up time: 1 hour
[20:54:25] <anonimasu> there's 2 models..
[20:54:32] <alex_joni> maybe you need to leave it plugged in longer ROFL
[20:55:03] <alex_joni> anonimasu: different supplier & all.. but it looks very similar
[20:55:17] <ohiopctechDOTcom> alex_joni: it has been rained on and all kind of stuff... the lcd only has bars across it when i turn it on
[20:55:38] <JymmmmEMC> ohiopctechDOTcom: that could just be the contrast setting
[20:55:40] <anonimasu> yep
[20:55:46] <alex_joni> ohiopctechDOTcom: that means the lcd might still be useable
[20:55:58] <ohiopctechDOTcom> yes.. i was thinking of that..
[20:56:30] <ohiopctechDOTcom> i know the torroid is good (to use for my cnc power supply maybe) since the thing at least turns on
[20:56:47] <anonimasu> :)
[20:56:58] <JymmmmEMC> ohiopctechDOTcom: But the toroid is rated 1VA less than what you need!
[20:57:10] <ohiopctechDOTcom> and if i clean the enclosure and paint it it will be a great enclosure for the stepper drivers/power supply on the cnc
[20:57:49] <alex_joni> ohiopctechDOTcom: now you only need a 23U enclosure
[20:58:22] <robin_sz> ohiopctechDOTcom, looks complete to me
[20:58:30] <JymmmmEMC> alex_joni: There's one at work, I'm tempted to ask my boss about it, but I got no place for it!
[20:58:31] <robin_sz> the thing with the heatsink is probably the NOX gen
[20:58:33] <ohiopctechDOTcom> this thing weighs about 30# imagine trying to drag it out of a huge huge construction dumpster by your self while dodging bees and wasps!
[20:58:44] <ohiopctechDOTcom> *at 3 in the morning
[20:59:11] <ohiopctechDOTcom> 23u? a rackmont cabinet?
[20:59:20] <JymmmmEMC> ohiopctechDOTcom: a half cabinet
[20:59:24] <ohiopctechDOTcom> ahhh
[20:59:32] <JymmmmEMC> full cabinet is 42U
[20:59:32] <robin_sz> ahh, no
[20:59:39] <ohiopctechDOTcom> i have been working on building my own rack for audio equipment :D
[20:59:39] <robin_sz> the NOX gen is optional
[20:59:50] <robin_sz> so i guess it never had one
[20:59:53] <ohiopctechDOTcom> good observation robin
[21:00:00] <JymmmmEMC> ~ 1,75" per U
[21:00:04] <ohiopctechDOTcom> yes, that half of the case is empty
[21:00:07] <robin_sz> the heatsink thing must be the analyzer section then
[21:00:36] <robin_sz> * robin_sz might have made "a few" 19" cases in his time ...
[21:00:41] <ohiopctechDOTcom> it's got some awesomely built manifolds in it
[21:00:46] <ohiopctechDOTcom> (the silver tubes)
[21:00:54] <robin_sz> ebay them :)
[21:01:06] <skunkworks> * skunkworks seconds that
[21:01:19] <ohiopctechDOTcom> so i gues sthere is a few solenoids and valves in there too from a quick glance
[21:01:43] <robin_sz> is that a PC104 bus control n the back?
[21:02:02] <ohiopctechDOTcom> is that what it is?
[21:02:37] <ohiopctechDOTcom> from the back i know there is a torroid, maybe some sort of a power supply board.. then a board that says "pc/chip" "chips"
[21:03:02] <skunkworks> pcchips?
[21:03:34] <ohiopctechDOTcom> the date on one of the chips on that board is: 31-10-94
[21:03:52] <robin_sz> looks PC104 to me ...
[21:03:52] <ohiopctechDOTcom> it's hooked to the lcd and the valves and everything else in the cab
[21:04:07] <ohiopctechDOTcom> i figured it was some sort of off the shelp 1 board pc
[21:04:13] <ohiopctechDOTcom> off the shelf*
[21:04:15] <robin_sz> yep
[21:04:32] <ohiopctechDOTcom> hmm
[21:04:43] <ohiopctechDOTcom> i might have to pull that out and try to get it running
[21:04:57] <ohiopctechDOTcom> will it run emc? lol
[21:05:45] <ohiopctechDOTcom> there was an entire copier machine in there too
[21:05:52] <ohiopctechDOTcom> there is no way i could drag that out alone
[21:05:53] <robin_sz> best left in there
[21:06:03] <JymmmmEMC> ohiopctechDOTcom: if it has at least 16MB ram, you might be able to run TurboCNC on it
[21:06:12] <robin_sz> hint : if dumpster diving, stick to useful stuff
[21:06:13] <ohiopctechDOTcom> let alone get it home without a truck, it won't fit in my lumina europsport ;)
[21:06:16] <anonimasu> dont count on it..
[21:06:29] <maddash> heh turbocnc
[21:06:43] <ohiopctechDOTcom> tubocnc=headaches
[21:06:49] <maddash> anyonr actually have the source code to that thing? I'm dying to take a look at its TP
[21:07:00] <JymmmmEMC> ohiopctechDOTcom: Copier machines are worthless, toss an add on craigslist and they'll give them to you.
[21:07:08] <anonimasu> maddash: buy it
[21:07:09] <ohiopctechDOTcom> if you reg the software you can get the source...
[21:07:24] <robin_sz> this is the thing with the PIC chips and the encoders on steppers?
[21:08:01] <ohiopctechDOTcom> robin_sz: ?
[21:08:04] <cradek> maddash: no point in looking at source when you can't reuse it or improve it
[21:08:21] <robin_sz> why can;t you improve it?
[21:08:23] <cradek> besides isn't turbocnc the one that stops at the end of every segment?
[21:08:26] <bill2or3> unless you need a bad example, I suppose.
[21:08:31] <anonimasu> cradek: yep
[21:08:32] <cradek> robin_sz: you can't give it away when you're done
[21:08:32] <bill2or3> (assuming bad code..)
[21:08:34] <anonimasu> cradek: and drops steps.
[21:08:42] <ohiopctechDOTcom> did turbo cnc come from stepster?
[21:08:43] <robin_sz> cradek, thats not the same as cant improve it
[21:08:45] <cradek> heck that's easy to write :-)
[21:08:53] <ohiopctechDOTcom> i think stepster source is available i saw it somewhere
[21:09:02] <alex_joni> good night all
[21:09:05] <anonimasu> List of moves :D go through it.. and stop at the end of each one
[21:09:22] <ohiopctechDOTcom> later alex_joni
[21:11:11] <JymmmmEMC> maddash: why dont you like tcnc?
[21:11:37] <anonimasu> JymmmmEMC: personally I hate it because it never worked well.
[21:11:48] <anonimasu> JymmmmEMC: it did omit outputting pulses.
[21:11:58] <robin_sz> it would be ideal for JymmmmEMC's machine I suspect
[21:11:59] <skunkworks> worked fine for me.. Never had issues with lost steps
[21:12:12] <skunkworks> just limites
[21:12:13] <ohiopctechDOTcom> ROFL @ robin_sz
[21:12:14] <skunkworks> limites
[21:12:18] <skunkworks> limited
[21:13:14] <JymmmmEMC> robin_sz: Sorry to disapoint you, I'm using QBASIC to drive my machine.
[21:13:19] <ohiopctechDOTcom> i don't see any reason whatsoever to use something like turbocnc when you have the emc2/ubuntu distribution... or on old hardware the puppy linux/emc dist.
[21:13:28] <skunkworks> and the guy that owns it seems to have fallen off the face of the earth. No developement for a while
[21:13:31] <maddash> JymmmmEMC: i don't don't like tcnc
[21:13:32] <anonimasu> yeah
[21:13:37] <anonimasu> a good tp would rock with tcnc..
[21:13:38] <robin_sz> JymmmmEMC, a great choice, should be fine
[21:13:52] <anonimasu> it's easy to get working quickly and runs on really old hardware..
[21:13:53] <maddash> JymmmmEMC: except that it abuses the "open-source" moniker
[21:13:53] <JymmmmEMC> ohiopctechDOTcom: EMC = 90ipm, TCNC = 120ipm
[21:14:13] <cradek> it's sad when good programs die when their developers lose interest, because they aren't fully free
[21:14:16] <JymmmmEMC> maddash: It's not open-source, though you can get the source.
[21:14:54] <ohiopctechDOTcom> is there multiple way to do everything in program source liek there is in machine shops?
[21:15:11] <anonimasu> ohiopctechDOTcom: what?
[21:15:21] <anonimasu> ohiopctechDOTcom: there are several thousand of ways to do one thing
[21:15:20] <ohiopctechDOTcom> could someone look at tcnc source and write something that does the same thing with other functions? then proceede to make it better?
[21:15:39] <anonimasu> ohiopctechDOTcom: there hardly is a point..
[21:15:50] <ohiopctechDOTcom> jsut curious
[21:15:52] <anonimasu> ohiopctechDOTcom: emc's there and it's really opensource..
[21:15:53] <anonimasu> :p
[21:16:17] <JymmmmEMC> ohiopctechDOTcom: Other than converting it from Pascal to C
[21:16:44] <skunkworks> * skunkworks has the source also somewhere..
[21:16:59] <JymmmmEMC> I'd think it be very cool if it used Frame Buffer =)
[21:17:17] <anonimasu> JymmmmEMC: framebuffer-what?
[21:17:18] <anonimasu> :p
[21:17:26] <ohiopctechDOTcom> soo does anyone want the lcd outta this seimens thing?
[21:17:29] <JymmmmEMC> anonimasu: shush you =)
[21:17:33] <anonimasu> JymmmmEMC: though that stuff is easy on pascal..
[21:17:36] <maddash> JymmmmEMC: of course it isn't open sourced. but it claims to be one.
[21:17:42] <skunkworks> one of the turbocnc programmers is now using emc ;)
[21:17:51] <JymmmmEMC> anonimasu: VGA=794 =)
[21:17:53] <anonimasu> I'd still hate to write that
[21:18:35] <cradek> according to the FSF, "open source" has no meaning except that you can (somehow) get the source. Selling it, or making it available under various restrictions is common.
[21:19:07] <anonimasu> GPL Licenced on the other hand means that it has to be free..
[21:19:13] <JymmmmEMC> I dont have an issue with what he's doing, it's only $40, not like he's asking $1400 to ge tthe source code.
[21:19:54] <skunkworks> although - feeds never worked right for me with g02/3
[21:20:35] <skunkworks> they where always way slower than they should be
[21:21:18] <JymmmmEMC> Hey, when your missing earth ground from mains (like pc + crt), WHY do you get shocked (barefoot)? I know I'm completing the path, but where is the AC coming from that it needs to goto earth ground?
[21:21:26] <ohiopctechDOTcom> i really reall tried to use turbocnc but it drove me nuts using dos
[21:21:38] <kanzure> Hello.
[21:21:50] <skunkworks> your building static electricity in your body
[21:21:57] <ohiopctechDOTcom> i had it runnin my mill, but after previously using emc i knew i had other choices and trashed that install
[21:22:04] <kanzure> I was just directed here from http://www.cnczone.com/forums/showpost.php?p=317867&postcount=7 and yesterday from another thread to linuxcnc.org :)
[21:22:04] <skunkworks> kanzure: hello
[21:22:26] <JymmmmEMC> skunkworks: Hardly, I get the shit zapped out of me, it aint ESD =)
[21:22:28] <ohiopctechDOTcom> hi kanzure
[21:22:34] <kanzure> Hi skunkworks and ohiopctechDOTcom :)
[21:22:50] <kanzure> I am pretty new to this whole machinist area of information.
[21:22:59] <anonimasu> kanzure: Welcome
[21:23:03] <kanzure> Big DIY sort of guy.
[21:23:03] <skunkworks> your the one wanting info on hsm?
[21:23:16] <kanzure> Nope
[21:23:17] <JymmmmEMC> I'm just trying to understand it all
[21:23:29] <kanzure> I mean, I did not post that thread
[21:23:34] <skunkworks> ah
[21:23:38] <skunkworks> welcome
[21:23:50] <kanzure> I am trying to navigate around these threads to find some good general FAQs to sink my teeth into
[21:24:00] <kanzure> but thought that IRC would probably be better ;)
[21:24:32] <skunkworks> This is mainly the EMC2 machine control software..
[21:24:43] <skunkworks> although we get sidetracked ;)
[21:25:00] <ohiopctechDOTcom> a lot of the people here are machinists hobby or pro also though
[21:25:16] <kanzure> Software's a good way to get introduced to a subject
[21:25:23] <kanzure> Maybe I'll go tred through the code soon
[21:25:24] <anonimasu> kanzure: just ask about anything here we try to answer to the best of out abilities :)
[21:25:30] <skunkworks> have you downloaded the livecd and tried it?
[21:25:29] <kanzure> anonimasu: Hey there.
[21:25:44] <kanzure> No, I haven't checked out the livecd- also, I do not have a shop
[21:25:52] <skunkworks> So? ;)
[21:26:05] <ohiopctechDOTcom> kanzure: now's a good time to start building you own DIY cnc!
[21:26:07] <ohiopctechDOTcom> ;)
[21:26:12] <kanzure> yeah, that's what I am thinking
[21:26:16] <kanzure> though I have no particular need quite yet
[21:26:20] <kanzure> Have you guys heard of Dave Gingery?
[21:26:38] <ohiopctechDOTcom> to diy machine casting guy?
[21:26:39] <kanzure> http://www.kk.org/thetechnium/archives/2007/03/bootstrapping_t.php "Bootstrapping the industrial age"
[21:26:40] <kanzure> yeah
[21:26:48] <ohiopctechDOTcom> i have read some of his stuff
[21:26:58] <kanzure> The guy who did his own industrial-shop starting with a bucket of sand and some fires with sticks and stones
[21:27:02] <kanzure> sounded quite novel
[21:27:11] <ohiopctechDOTcom> lol, yes...
[21:27:16] <kanzure> and manly.
[21:27:18] <ohiopctechDOTcom> he is a smart guy
[21:28:19] <skunkworks> kanzure: check out the wiki also http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl
[21:28:19] <ohiopctechDOTcom> i wouldn't mind having one of his mini-lathes
[21:28:39] <kanzure> So what's the cross over between mechanical engineering with their plastic molding and structural modeling and the CNC shops? Metal cutting?
[21:29:08] <ohiopctechDOTcom> cncs makes molds...
[21:29:11] <kanzure> anonimasu: Where have I seen you before? #perl? #physics? Something like that.
[21:29:28] <kanzure> ah, that's right- metal pressing, which I've glanced at before at one time or another
[21:29:32] <kanzure> you'll have to forgive me, ;)
[21:29:47] <skunkworks> bbl
[21:30:02] <anonimasu> kanzure: I dont know
[21:30:23] <anonimasu> #microcontrollers perhaps
[21:30:37] <kanzure> Neat.
[21:31:26] <anonimasu> brb
[21:31:32] <anonimasu> time to have a hot bath then go to bed
[21:31:36] <anonimasu> err time for.
[21:31:57] <ds2> heh gingery heh
[21:32:01] <ohiopctechDOTcom> * ohiopctechDOTcom start ripping the guts out of the seimens 19" rack enclosure
[21:33:45] <kanzure> What's that, ohiopctechDOTcom ?
[21:34:11] <ds2> gingery is great if you don't have a job
[21:34:18] <ohiopctechDOTcom> oh.. i found some old used/non-working equipment
[21:34:22] <ds2> and live out away from the cities
[21:34:44] <ohiopctechDOTcom> i'm going to rip the guts out of f the rack mount enclosure and use it for my cnc controller enclosure
[21:35:45] <kanzure> ds2, haha- hopefully his stuff is useful in other situations as well
[21:36:00] <ohiopctechDOTcom> kanzure: you can ignore about 90% of what i say ;)
[21:36:05] <ohiopctechDOTcom> lol
[21:36:34] <kanzure> pfft, I'll ignore you when I feel like it
[21:36:48] <ohiopctechDOTcom> ok, good deal
[21:36:50] <ds2> kanzure: they are fine to get ideas from but not very time efficient
[21:37:15] <kanzure> Hmm
[21:37:28] <kanzure> Has anybody tried a run-through to see how quickly they could get up to something that would be even close to useful ?
[21:37:34] <kanzure> I mean, there's a certain amount of time that you have to account for
[21:37:35] <ds2> I read his lathe book
[21:37:36] <kanzure> cooling etc.
[21:38:06] <ohiopctechDOTcom> kanzure: you mean as far as building one of the gingery machines?
[21:38:08] <ds2> my guess, a year or two
[21:38:18] <ds2> since you got to build a small foundry
[21:38:19] <kanzure> ohiopctechDOTcom: yeah
[21:38:23] <kanzure> a year or two, eh?
[21:38:24] <ds2> then go off do woodworking
[21:38:26] <ds2> pour it
[21:38:28] <ds2> hand scrape it
[21:38:29] <anonimasu> kanzure: if we are talking about any cnc machine/lathe/mill it comes down to how much money you want to invest..
[21:38:34] <ds2> etc
[21:38:58] <anonimasu> and precision and amount of work you are prepared to put down.
[21:39:10] <ohiopctechDOTcom> kanzure: you can buy off the shelf part sand have a machine built/running in less than a day, if you want to spend$
[21:39:14] <kanzure> Ah, well, money's not something I have for all of my many hobbies
[21:39:21] <kanzure> yeah, but I'm in it more from the 'from scratch' aspects
[21:39:43] <ohiopctechDOTcom> you would def. learn something from building it the gingery way
[21:40:21] <anonimasu> kanzure: yeah
[21:40:24] <anonimasu> err sorry..
[21:40:25] <kanzure> http://heybryan.org/instrumentation/instru.html
[21:40:46] <kanzure> That's my analytical chemistry instrumentation page that I've been working on a few days ago
[21:41:00] <kanzure> Most of the machines are very costly, but why buy what you can do yourself?
[21:41:04] <ds2> you can go half way into it and buy cheap import machines
[21:41:17] <ds2> just assume they are the raw castings in gingery books and rework it
[21:41:34] <anonimasu> kanzure: because you cant produce anything as good I guess..
[21:41:47] <ohiopctechDOTcom> calibrated/tested
[21:41:50] <anonimasu> yep
[21:41:58] <ohiopctechDOTcom> *to add to anonimasu's statement
[21:42:16] <anonimasu> kanzure: also the people scraping machines to 0.0001 paralellism are people that's been working with it for 30 years.
[21:42:15] <ds2> kanzure: what is costly? > $0.01?
[21:42:42] <kanzure> anonimasu: What do you mean 0.0001 parallelism? I've never seen this notation
[21:42:56] <ohiopctechDOTcom> squareness
[21:43:00] <ohiopctechDOTcom> flatness
[21:43:03] <anonimasu> box ways/squareness/flatness
[21:43:15] <anonimasu> straight angles..
[21:43:16] <kanzure> ds2: Costly means "stuff I don't know" really- and remember, I have no particular projects that I am working on at the moment, so of course there's going to be no budget when I'm talking about these machines and development systems
[21:43:22] <kanzure> hmm
[21:43:33] <kanzure> * kanzure was thinking of parallelism as in the computational principles
[21:44:57] <anonimasu> kanzure: I guess that's the things you pay for..
[21:46:13] <anonimasu> it's by no means impossible to make great machines yourself. :)
[21:50:23] <JymmmmEMC> In response to "ohiopctechDOTcom: you would def. learn something from building it the gingery way"... Yeah a whole lot of profanity is the FIRST thing you'll learn =)
[21:51:15] <JymmmmEMC> But, if you have a "profanity jar" , you'll be able to save up $50,000 in no time =)
[21:51:27] <ohiopctechDOTcom> i already know 45 different way to use "F@#K"
[22:04:41] <ohiopctechDOTcom> ohiopctechDOTcom is now known as chr0n1c
[22:08:11] <chr0n1c> http://sourceforge.net/projects/dn-cam/ <--- i just found this a few minutes ago, i have not tried it yet. "PNG images & DXF files to CNC G code files"
[22:08:26] <chr0n1c> *for linux?
[22:08:47] <kanzure> What is the G code file format?
[22:09:07] <chr0n1c> kanzure: have not read that far
[22:09:13] <kanzure> hehe
[22:09:14] <chr0n1c> i'm stil investigating
[22:09:26] <kanzure> Opera has stalled for me - otherwise I'd be answering that for myself
[22:09:47] <chr0n1c> sourceforge has a crazy script that hangs sometimes
[22:10:02] <kanzure> nah, I think it's the 180 tabs and 2.2k unread RSS items
[22:10:38] <chr0n1c> only 180? and only one browser running? what a lightweight! lol
[22:10:42] <jepler> kanzure: "g code" is a general name for ASCII files used to control CNC machines. For instance, it might say "now move the tool to X=1 Y=2 Z=3 at a rate of F=10 inches per minute": "G0X1Y2Z3F10"
[22:10:59] <kanzure> Ah, so it's just some scripting file format
[22:11:09] <jepler> er, "G1X1..."
[22:11:14] <chr0n1c> it's a very awesome scripting format ;)
[22:11:28] <chr0n1c> n001 f5
[22:11:44] <chr0n1c> n002 s57,000
[22:11:51] <chr0n1c> :|
[22:15:36] <kanzure> I like how Windows XP closes programs by initializing a two megabyte "kill" program, so if it doesn't respond the first time, you get to launch lots and lots of two megabyte kill signals. Woo.
[22:26:05] <skunkworks> I love how if you try to physically speed up the servo - it reverses the the polarity to it - slowing it back down.. :)
[22:41:30] <ds2> what is this obession that people have with wanting to convert images to G code? that topic seems to come up at least once every two month on some forum/list/etc? it seems like a trivial enough exercise
[22:43:48] <kanzure> Later I'll be dropping by again to post a link to my bookmarks from cnczone.com- I have some interesting threads collected into a list that some here may find interesting
[22:59:53] <chr0n1c> http://ohiopctech.com/dp/?q=node/41 <- i jsut updated the links page on my wesite to reflect all my cnc (and other) research links if anyone wants some cool links to browse through for metalworking and such go check it out
[23:00:21] <chr0n1c> (it's a output straight from my firefox browser bookmarks)
[23:00:29] <chr0n1c> an output even*
[23:32:32] <skunkworks> hmm day old Chinese isn't as good as day old pizza
[23:32:40] <chr0n1c> it's better!
[23:32:56] <chr0n1c> i'm actually eating some now
[23:33:06] <chr0n1c> day old chinese that is.. :
[23:33:36] <skunkworks> reboot - back in a bit.
[23:33:38] <chr0n1c> beef with mixed vegtables...
[23:33:46] <jepler> ds2: the simplest such program is very simple, but it's easy to imagine a lot of features to add. (simple, e.g., the 74-line http://timeguy.com/cradek-files/image-to-gcode/gcode.py)
[23:35:04] <ds2> jepler: yes, I know it is simple; I have my own versions... I am just preplexed by the interest it generates
[23:38:03] <jepler> some topics seem to elicit an opinion from everyone under the sun
[23:38:39] <chr0n1c> * chr0n1c missed a bunch of convo during a netsplit or something...
[23:39:25] <chr0n1c> ehh what's the logger commands?
[23:39:33] <chr0n1c> to get the log?
[23:39:33] <jepler> logger_emc: bookmark
[23:39:33] <jepler> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2007-07-09.txt
[23:39:40] <chr0n1c> ty
[23:40:19] <chr0n1c> ahh i only missed a few lines
[23:40:38] <chr0n1c> i love engraving things.. that's my thing with image to g-code
[23:41:36] <chr0n1c> image to .dxf is way more useful though
[23:42:05] <chr0n1c> say for profiling an automotive trim pc. that you flattened out with a hammer and want to make a reproduction with a laser
[23:42:09] <chr0n1c> scan the part and convert it ;)
[23:42:31] <chr0n1c> then pull it into autocad as a .dxf, scale it and pull of dimensions
[23:56:32] <JymmmmEMC> The most common issue with img2gcode is the issue of shadows and light.
[23:57:56] <JymmmmEMC> Someone turned me on to a possible solution, but I have to try it out. Takes a lot of setup and post-processing just to get the photograph.
[23:59:51] <JymmmmEMC> s/have/haven't/