#emc | Logs for 2006-02-19

[00:22:33] <CIA-8> 03alex_joni * 10emc2/src/hal/drivers/ (hal_vti.c hal_vti.h): ran through indent, more easy to read, fixed most of the warnings, errors, still a few left which I can't make out (design issues)
[00:22:40] <giacus> g nite
[00:49:36] <CIA-8> 03jmkasunich * 10emc2/scripts/emc.in: allow command line args to be passed from the ini file DISPLAY entry to the selected GUI
[01:06:38] <jmkasunich> jmkasunich is now known as jmk-away
[02:05:51] <gene> Hi all
[02:06:09] <gene> Anybody alive?
[02:07:09] <skunkworks> sort of - hi Gene
[02:07:46] <gene> looking for cgode help
[02:08:55] <gene> Hummm, let me rephrase that, I'm looking for gcode help
[02:09:13] <skunkworks> understood - subroutines?
[02:09:51] <gene> yep, gives this error:
[02:09:56] <gene> emc/task/emctask.cc 270: interp_error: Unknown word starting with a
[02:09:57] <gene> Unknown word starting with a
[02:09:59] <gene> Interpreter stack: - read_operation_unary - read_unary - read_real_value - read_c - read_one_item
[02:10:00] <gene> emcTaskPlanRead() returned 180
[02:10:02] <gene> emcTaskPlanClose() called at emc/task/emctaskmain.cc:1136
[02:10:03] <gene> Outgoing motion id is 33.
[02:10:05] <gene> emcTaskIssueCommand()
[02:10:06] <gene> Issuing EMC_TASK_PLAN_SYNCH -- (+516,+12, +0,)
[02:10:08] <gene> emcTaskPlanSynch() returned 0
[02:10:09] <gene> emcTaskIssueCommand() returning: 0
[02:10:25] <gene> Here is the code:
[02:10:31] <gene> (sub uses vars #11-20, passed in as 1-10 )
[02:10:33] <gene> ( #11 = ystart #12 = zstart #13 = Inum #14 = Jnum )
[02:10:35] <gene> o1000 sub
[02:10:36] <gene> #1 = [#11]
[02:10:38] <gene> #2 = [#12]
[02:10:39] <gene> #3 = [#13]
[02:10:41] <gene> #4 = [#14]
[02:10:42] <gene> G1 F100 Y#1 Z#2
[02:10:44] <gene> G3 Y#1 Z#2 I#3 J#4
[02:10:45] <gene> (return to start in case not full circle)
[02:10:47] <gene> G1 Y#1 Z#2
[02:10:48] <gene> o1000 endsub
[02:10:50] <gene> (Program starts here)
[02:10:51] <gene> ( assign #11 first )
[02:10:53] <gene> n0010 #11 = -0.5
[02:10:54] <gene> ( assign #12 next )
[02:10:56] <gene> n0020 #12 = -0.5
[02:10:57] <gene> ( assign #13 best )
[02:10:59] <gene> n0030 #13 = 100
[02:11:00] <gene> ( and finally assign #14 )
[02:11:02] <gene> n0040 #14 = 0.0
[02:11:03] <gene> ( set x start pos)
[02:11:05] <gene> n0050 #15 = 0.0
[02:11:07] <gene> ( set how far x moves )
[02:11:09] <gene> n0060 #16 = 2.0
[02:11:11] <gene> ( and how big a steps to move )
[02:11:14] <gene> n0070 #17 = 0.025
[02:11:15] <gene> ( and now the while statment )
[02:11:17] <gene> o1001 while [ #15 LE #16 ]
[02:11:19] <gene> ( make sure x is there )
[02:11:21] <gene> n0090 g1F100x#15
[02:11:23] <gene> ( call the subroutine )
[02:11:25] <gene> n0100 call [#11] [#12] [#13] [#14]
[02:11:27] <gene> ( increment position value )
[02:11:30] <gene> n0110 #15 = [ #15 + #17 ]
[02:11:32] <gene> o1001 endwhile
[02:11:34] <gene> n0130 g0z0
[02:11:35] <gene> n0140 g0x0y0
[02:11:37] <gene> n0150 M02
[02:11:41] <gene> lots of useless comments, trying to pin the line number down...
[02:14:06] <gene> Here it is w/o the comments:
[02:14:11] <gene> sub uses vars #11-20, passed in as 1-10 )
[02:14:13] <gene> ( #11 = ystart #12 = zstart #13 = Inum #14 = Jnum )
[02:14:14] <gene> o1000 sub
[02:14:16] <gene> #1 = [#11]
[02:14:18] <gene> #2 = [#12]
[02:14:19] <gene> #3 = [#13]
[02:14:21] <gene> #4 = [#14]
[02:14:22] <gene> G1 F100 Y#1 Z#2
[02:14:24] <gene> G3 Y#1 Z#2 I#3 J#4
[02:14:25] <gene> G1 Y#1 Z#2
[02:14:27] <gene> o1000 endsub
[02:14:28] <gene> (Program starts here)
[02:14:30] <gene> n0010 #11 = -0.5
[02:14:31] <gene> n0020 #12 = -0.5
[02:14:33] <gene> n0030 #13 = 100
[02:14:34] <gene> n0040 #14 = 0.0
[02:14:36] <gene> n0050 #15 = 0.0
[02:14:37] <gene> n0060 #16 = 2.0
[02:14:39] <gene> n0070 #17 = 0.025
[02:14:40] <gene> o1001 while [ #15 LE #16 ]
[02:14:42] <gene> n0090 g1F100x#15
[02:14:43] <gene> n0100 call [#11] [#12] [#13] [#14]
[02:14:45] <gene> n0110 #15 = [ #15 + #17 ]
[02:14:47] <gene> o1001 endwhile
[02:14:49] <gene> n0130 g0z0
[02:14:51] <gene> n0140 g0x0y0
[02:14:53] <gene> n0150 M02
[02:16:30] <fenn> god i hate gcode so much
[02:17:16] <gene> I'm beginning to wonder why this wasn't in C, I can do simple things in that.
[02:17:46] <fenn> "but then you'd have to teach people how to program" they say
[02:19:25] <gene> Chuckle. Funy thing, back in my amiga days, I was very productive in arexx, cause it looked so much like interpreted C
[02:21:20] <gene> Back to this, the one thing I miss in the debug options is outputting the value of the var that caused the ulcer
[02:22:26] <skunkworks> I don't understand (have not played with subs yet). Do you have this example?
[02:22:26] <gene> These messages, at least to me, are less than helpfull :(
[02:22:30] <skunkworks> (subroutine test -- on exit X=2.4 Y=3.5 Z=4.6)
[02:22:30] <skunkworks> o100 sub (begin subroutine)
[02:22:30] <skunkworks> (inside sub O100)
[02:22:30] <skunkworks> o[150+50] call [2+#1][3+#2][4+#3] (call a nested subroutine)
[02:22:30] <skunkworks> o100 endsub
[02:22:30] <skunkworks> o200 sub
[02:22:32] <skunkworks> g0 x#1 y#2 z#3
[02:22:34] <skunkworks> o200 endsub
[02:22:36] <skunkworks> o100 call [.1] [.2] [.3]
[02:22:38] <skunkworks> o100 call [.4] [.5] [.6]
[02:22:40] <skunkworks> (subroutine test -- on exit X=2.4 Y=3.5 Z=4.6)
[02:23:01] <gene> whats the filename?
[02:24:26] <skunkworks> http://cvs.sourceforge.net/viewcvs.py/*checkout*/emc/emc2/nc_files/Attic/interp_test_sub.ngc?rev=
[02:24:52] <skunkworks> http://cvs.sourceforge.net/viewcvs.py/emc/emc2/nc_files/Attic/interp_test_sub.ngc?rev=1.1&hideattic=0&view=log
[02:27:01] <gene> Gotta go mount it as a samba share to save that, brb
[02:32:44] <gene> Ok, this one runs, print and compare I guess. Thanks
[02:39:20] <gene> Got that error, wrong syntax on the call statement. Now the G2-3 command is biting me with a zero radius arc message
[02:46:14] <jepler> gene: I just had an idea. For "debugging", you could use an M1xx command to output some values to the terminal.
[02:47:31] <gene> Are ypu saying I could print a message on the terminal saying that 'at line xxx, variable #12 had a value of [#12] ???
[02:48:15] <jepler> No, it won't be anything quite that .. clear
[02:48:44] <gene> Are ypu saying I could print a message on the terminal saying that 'at line xxx, variable #12 had a value of [#12]' ??? note final quote, thats what I was afraid of
[02:49:11] <gene> So what sort of message can I output to the terminal?
[02:51:10] <jepler> in the emc 2 distribution, there's 'nc_files/M101' (a shell script)
[02:51:17] <jepler> if your program contains a line like
[02:51:20] <jepler> M101 P1 Q100
[02:51:31] <jepler> that script will be invoked, and it will have access to the values '1' and '100
[02:51:31] <jepler> '
[02:51:58] <gene> and these correspond to the $vars?
[02:52:03] <gene> and these correspond to the #vars?
[02:52:07] <jepler> no
[02:52:43] <jepler> It's just like any other g-code that takes a number
[02:52:48] <jepler> so you could write M101 P#1 Q100
[02:53:37] <gene> Could I do M101 P#1 P#2 P#3 P#4 ??
[02:53:48] <jepler> Can you do G1 X1 X2 X3 X4 ?
[02:53:59] <gene> obviously not
[02:54:01] <jepler> forget it, I'm not even sure M1xx works in emc2
[02:54:18] <gene> It is there in the nc_files dir
[02:55:21] <gene> Does this have to be pasted into the program then, with the messages added to it?
[02:56:24] <jepler> /*! \todo FIX-ME Impliment these at a later stage... */
[02:56:24] <jepler> /* user-defined M codes */
[02:56:35] <jepler> M1xx doesn't work at all in emc2, so it's a moot point
[02:58:22] <gene> Oh fudge... Cause this looks expandable to do exactly as I need. So does it not work in emc2?
[02:58:30] <jepler> that's what the source code says
[02:59:43] <gene> So what broke it? Or is that not obvious...
[03:00:05] <jepler> All I know is in the comment I pasted above, next to the disabled
[03:00:06] <jepler> code
[03:00:11] <jepler> someone needs to "Impliment" it
[03:00:20] <gene> Ahh, :(
[03:01:00] <gene> which .c file is it in?
[03:01:14] <jepler> rs274ngc/interp_convert.cc
[03:02:41] <gene> src tree?
[03:03:17] <jepler> there's only one file of that name in the emc2 tree
[03:03:24] <jepler> emc2/src/emc/rs274ngc/interp_convert.cc
[03:04:10] <gene> Yeah I just found it. c++ stuff?
[03:04:23] <jepler> C++ is like a pig with lipstick on it
[03:04:37] <fenn> hahahaha
[03:05:32] <gene> humm, looks like semi normal c to me
[03:05:39] <fenn> it is
[03:05:49] <jepler> umm, and something about, when you make the sausage, you'd be just as well-off if you'd just had "C"
[03:05:49] <fenn> notice the "extern C {"
[03:06:21] <gene> Yeah, but it still smells like a pig, whats that? I'm rusty
[03:06:26] <fenn> i'm not going to get into this conversation.. sorry
[03:06:56] <gene> ahh, Gee, you wouldn't want to miss the fun would you?
[03:07:35] <gene> What line # area is this in?
[03:07:46] <fenn> basically it boils down to "i cant be bothered to fix it cause i'm lazy"
[03:08:04] <gene> Perfect description of this old fart too
[03:08:48] <fenn> jepler: you seen petev's interpreter?
[03:08:52] <jepler> fenn: no
[03:09:02] <fenn> its quite easy to understand
[03:09:07] <gene> found it line 1724 starts
[03:09:53] <fenn> jepler: fenn.dyndns.org/pub/emc/emc3/src/emc/interpreter/
[03:10:48] <fenn> in particular http://fenn.dyndns.org/pub/emc/emc3/src/emc/interpreter/rs274ngc/rs274ngc.g
[03:10:50] <jepler> fenn: I began my own rs274ngc interpreter once but since it was for the project that became axis, exact compatability with emc was a must...
[03:11:01] <jepler> (emc3? we're not even done with emc2 yet)
[03:11:18] <fenn> i know.. i got disgusted with emc when i found out it didnt do what it promised
[03:11:38] <fenn> kinematics is a dirty hack
[03:12:01] <jepler> oh, the parsing is easy
[03:12:09] <jepler> it's the converting to .. whatever is next .. that is hard
[03:12:13] <fenn> yeah
[03:12:50] <jepler> http://unpy.net/cgi-bin/viewcvs.cgi/rs274py/rs274/__init__.g?rev=1.9;content-type=text%2Fplain
[03:13:00] <fenn> i had a lot of ideas about what comes next but i guess i'm too antsy to sit down and write code for hours
[03:13:35] <gene> I don't know as I can say what this does even if the protective comments were removed
[03:14:51] <gene> That link looks simple enough if I truelly understood that snake :)
[03:15:45] <CIA-8> 03jepler * 10emc2/src/emc/rs274ngc/interp_convert.cc: make M1xx work?
[03:16:07] <jepler> Now, I ran this simple program:
[03:16:08] <jepler> #1=3.14159
[03:16:08] <jepler> m101 p1 q#1
[03:16:08] <jepler> m2
[03:16:30] <jepler> and buried in the emc2 diagnostic output is
[03:16:31] <jepler> M101 P1.000000 Q3.141590: put your code here
[03:17:02] <fenn> i take it that's bad
[03:18:04] <gene> No, if it can work, then I'd just do a call for each number I wanted to see, and that would be very helpfull
[03:18:35] <fenn> * fenn was thinking it would output M101 P1 Q#1
[03:18:46] <gene> You got that output with emc
[03:19:52] <gene> And I assume one could also M101 P#1 Q#2?
[03:21:27] <jepler> gene: Just like other places that g-code will take "a number", you can put an expression for P- and Q-.
[03:21:34] <jepler> gene: so, complicated way of saying "yes".
[03:21:52] <gene> Thats IF that code was uncommented?
[03:24:00] <gene> Humm, looks active, but gvim is coloring that block. I'll try this then
[03:25:34] <jepler> http://pastebin.com/562167
[03:25:46] <jepler> this is what I checked in, though I'm sure pastebin mangled the whitespace
[03:26:12] <jepler> and ate @@s for some reason
[03:27:51] <gene> It doesn't fly, doesn't error either:
[03:28:08] <gene> emcTaskPlanCommand(n0100 M101 P#11 Q#12) called. (line_number=24)
[03:28:10] <gene> emcTaskPlanExecute(0) return 0
[03:28:12] <gene> emcTaskPlanRead() returned 0
[03:28:13] <gene> emcTaskPlanLine() returned 25
[03:28:15] <gene> emcTaskPlanCommand(n0101 M101 P#13 Q#14) called. (line_number=25)
[03:28:16] <gene> emcTaskPlanExecute(0) return 0
[03:29:21] <gene> And there are supposedly valid values in those vars (I think) :(
[03:32:16] <gene> Ok, patch applied, refresh me on rebuilding please
[03:34:02] <gene> Will this still work?
[03:34:06] <gene> #!/bin/bash
[03:34:07] <gene> cvs up -rTESTING
[03:34:09] <gene> cd src
[03:34:10] <gene> ./configure --enable-run-in-place
[03:34:12] <gene> make clean
[03:34:13] <gene> make
[03:34:15] <gene> sudo make setuid
[03:34:16] <gene> cd ..
[03:34:18] <gene> scripts/emc &
[03:34:19] <gene> exit 0
[03:34:34] <cradek> gene: stop while you have your sanity, and generate your g-code with a high level language
[03:34:40] <jepler> no, because my change is not in TESTING, it's in HEAD
[03:34:44] <jepler> otherwise the instructions may be right
[03:35:00] <cradek> cvs up -dP -A
[03:35:04] <cradek> hi jeff
[03:35:10] <gene> Humm, you mean the pacthing I just did will be removed?
[03:35:30] <jepler> and for the record, I agree with cradek. My remark was intended to be a 1-second bit of advice, not an hour-long development session...
[03:35:53] <gene> How about the tp.c blending problem?
[03:36:00] <jepler> if you want to rebuild with just my change, then you don't need to 'cvs up'
[03:36:15] <jepler> I don't know about any tp.c blending problem.
[03:36:21] <gene> Ok, I'll nuke that line
[03:38:01] <gene> lots of following errors as it approaches a direction change
[03:38:30] <gene> Thats why I'm running TESTING since tp.c in that version doesn't blend
[03:39:30] <cradek> that's not true.
[03:39:46] <cradek> there are no changes in tp between TESTING and HEAD
[03:41:58] <gene> Somethings fixed!
[03:43:09] <jepler> cradek is right; there are no differences between TESTING and HEAD in the src/emc/kinematics directory
[03:44:16] <gene> so I really should do a cvs up -A then?
[03:44:56] <gene> anyway it built and ran, but m101 still outputs nothing.
[03:46:02] <gene> What does your M101 look like?
[03:46:29] <jepler> it's the one CVS gave me
[03:46:38] <gene> Mine too
[03:47:39] <gene> so let me revert from TESTING then & try that
[03:49:39] <gene> Humm, problems I think, makefile stuff
[03:50:19] <gene> Makefile:5: *** missing separator. Stop.
[03:50:20] <gene> Makefile:5: *** missing separator. Stop.
[03:50:22] <gene> Makefile:5: *** missing separator. Stop.
[03:50:37] <gene> This was at the end of the make
[03:51:19] <gene> are we back to that same fix we had to make to the Makefile a week ago, cause pwd wasn't working?
[03:51:53] <jepler> I'm going to bed.
[03:51:55] <jepler> goodnight.
[03:51:57] <jepler> and good luck.
[03:51:58] <Jymmm> jepler!
[03:52:32] <Jymmm> jepler: I got Onsrud to run my gcode and gonna send me back some pics
[04:06:11] <gene> Ok, I just got it to accept that mess at the top of the Makefile, but now its looking for Makefile.inc in the /usr/src/kernel-source- directory & it doesn't exist there :(
[04:06:35] <cradek> what did you change?
[04:08:36] <cradek> I just built HEAD on bdi 4.30 and it works fine
[04:23:57] <gene> The Makefile had to be fixed again, I hard coded the path into BASEPWD and exported i, then it worked
[04:24:26] <gene> You recall we had a problem the other day with that.
[04:24:37] <cradek> yes, but I thought I fixed it
[04:24:40] <cradek> what bdi is yours?
[04:24:44] <gene> 4.30
[04:25:03] <cradek> do you want to troubleshoot now?
[04:25:13] <cradek> If there's still a problem, I want to fix it right.
[04:26:01] <gene> Well, that I got fixed for the time being. There was a whole load of strange stuff in it that I commented out & then fixed the real problem, I'll paste it here
[04:26:23] <cradek> if you had HEAD, I know what's in it, because it's the same as mine
[04:26:44] <cradek> but I want to know what goes wrong with it
[04:26:51] <gene> # Delete the default suffix rules
[04:26:53] <gene> .SUFFIXES:
[04:26:55] <gene> .PHONY: default userspace modules clean modclean depclean install
[04:26:57] <gene> #<<<<<<< Makefile
[04:26:59] <gene> #include /home/gene/emc2/src/Makefile.inc
[04:27:00] <gene> #=======
[04:27:02] <gene> #ifeq ($(BASEPWD),)
[04:27:03] <gene> #BASEPWD := $(shell pwd)
[04:27:05] <gene> BASEPWD = /home/gene/emc2/src
[04:27:06] <gene> export BASEPWD
[04:27:08] <gene> #endif
[04:27:09] <cradek> oh that's because of a merge conflict
[04:27:09] <gene> @echo "BASEPWD ="$BASEPWD
[04:27:11] <gene> include $(BASEPWD)/Makefile.inc
[04:27:12] <gene> #>>>>>>> 1.58
[04:27:14] <gene> Thats my fix
[04:27:16] <gene> the @echo diesn't work tho
[04:27:28] <cradek> that's because you had a modification then updated.
[04:27:32] <cradek> just rm Makefile
[04:27:35] <cradek> then cvs up
[04:27:42] <cradek> you'll get the HEAD version
[04:28:37] <gene> I just did that and a 'cvs up -dP'
[04:28:44] <cradek> ok
[04:28:49] <cradek> now it should build fine.
[04:30:39] <gene> ./configure --enable-run-in-place && make clean && make && sudo make setuid
[04:30:45] <gene> right?
[04:30:50] <cradek> right
[04:30:58] <gene> Ok its building
[04:31:26] <gene> but I figure I'd better stop the old one :)
[04:32:46] <gene> I did play with M101 a bit, it will spit out one var and its contents at a time, and showe me that a "#1 =[#11] isn't working, so the subroutine winds up with all zero vars
[04:33:14] <gene> I did play with M101 a bit, it will spit out one var and its contents at a time, and showe me that a "#1 = [#11]" isn't working, so the subroutine winds up with all zero vars
[04:34:20] <cradek> sorry, I can't help with that, I've never used that stuff.
[04:35:26] <gene> So why isn't that working? Or does the transfer to the subroutine actually make a new set of # vars for the sub, numbered from 1 to last arg furnished?
[04:35:36] <cradek> sorry, I can't help with that, I've never used that stuff.
[04:35:54] <gene> Oh, I thought you had a hand in that
[04:36:17] <gene> I'll see if the code does that automaticly
[04:36:23] <cradek> no, I think it's crazy even though I understand some people want it
[04:37:07] <cradek> I can't imagine writing complex g-code by hand when every high level language can print some strings into a file.
[04:37:25] <gene> Well, if successfull, I will have made a nearly 7000 line program to carve the seat for anew z axis nut assembly into about 50 lines
[04:38:12] <cradek> I can sure understand that would be better if you're forced to write it by hand...
[04:38:29] <gene> I'd say thats a savings from 2 standpoints, the most important being the removal of cut&paste errors
[04:39:36] <gene> Itwould be nice if I was expert enough in a cad program that could output gcode, but I've been bamboozled by those things since back aboiut lightwave for the amiga days.
[04:41:00] <gene> Has any one written a text to gcode convertor one where a text description written in a fixed format style could then output the gcode to carve that part?
[04:41:48] <cradek> I don't know of something like that
[04:43:12] <gene> Gahh, load probs, I'll see if they repeat
[04:43:53] <gene> No, first time startup glitch I guess
[04:46:27] <gene> Well, it worked, sorta, but the table ran about 69 inches. Its a 8 inch table :)
[04:46:40] <cradek> that doesn't seem quite right
[04:50:06] <gene> It wasn't, and M101 has apparently quit working again, I can't find any of its outputs but lemme reset everything and restart, then about before it runs off the shell history
[04:52:46] <cradek> gene: I think maybe tomorrow we'll move TESTING. I'll try Mxxx before I do it.
[04:54:52] <gene> Let me look and see if that edit I put in is gone again
[04:55:12] <cradek> I thought jeff checked it in to HEAD
[04:55:19] <cradek> you have HEAD right?
[04:57:03] <gene> Now I do, and that editing is still in that file
[04:57:37] <gene> Now I do if a cvs up -A takes it back to head
[04:57:58] <gene> Humm, and I've since done a cvs up -dP
[04:58:32] <cradek> yes -A gives you HEAD
[04:58:39] <gene> What time is it at your 20 Chris?
[04:58:47] <cradek> almost 2300
[04:59:05] <gene> cst then, its 23:58 here
[04:59:12] <cradek> yes
[04:59:58] <gene> We should give this up till tomorrow & maybe I can make it work again, it was very handy while it did.
[05:00:20] <cradek> I don't understand what changed it
[05:00:45] <cradek> sometimes I get confused about what you are doing/have done
[05:00:58] <gene> Me either, that file still looks ok to me, but let me paste that for your checking
[05:01:31] <cradek> just use cvs diff
[05:01:38] <cradek> to see if it matches HEAD
[05:02:03] <gene> for (int index = 100; index < 200; index++) {
[05:02:04] <gene> if (block->m_modes[index] == index) {
[05:02:06] <gene> if (USER_DEFINED_FUNCTION[index - 100] != 0) {
[05:02:07] <gene> (*(USER_DEFINED_FUNCTION[index - 100])) (index - 100,
[05:02:09] <gene> block->p_number,
[05:02:10] <gene> block->q_number);
[05:02:12] <gene> } else {
[05:02:13] <gene> CHK(1, NCE_UNKNOWN_M_CODE_USED);
[05:02:15] <gene> }
[05:02:16] <gene> }
[05:02:18] <gene> }
[05:02:20] <gene> return INTERP_OK;
[05:02:26] <cradek> do you get the unknown m code error?
[05:03:24] <CIA-8> 03cradek * 10emc2/ (src/Makefile debian/rules): install halcmd manpage
[05:03:47] <gene> Mmm, me not ken this:
[05:03:55] <gene> RCS file: /cvsroot/emc/emc2/src/emc/rs274ngc/interp_convert.cc,v
[05:03:57] <gene> retrieving revision 1.15
[05:03:58] <gene> diff -r1.15 interp_convert.cc
[05:04:00] <gene> 1728d1727
[05:04:01] <gene> < #if 0
[05:04:03] <gene> 1731c1730
[05:04:05] <gene> < for (index = 100; index < 200; index++) {
[05:04:06] <gene> ---
[05:04:08] <gene> > for (int index = 100; index < 200; index++) {
[05:04:09] <gene> 1742d1740
[05:04:11] <gene> < #endif
[05:04:28] <gene> No, no unknown code error
[05:05:27] <cradek> so yours don't match HEAD
[05:06:11] <cradek> oh are you using anonymous cvs? maybe it doesn't have this change yet.
[05:06:20] <cradek> I always forget that it's a little behind.
[05:06:22] <gene> apparently not, yes anon
[05:06:45] <cradek> but now it looks like if you do cvs up -dP -A you'll get this change.
[05:06:49] <gene> but at startup, I get this on the terminal:
[05:06:59] <gene> emcTaskInit: adding user-defined function ../../nc_files/M101
[05:07:17] <cradek> ok, I think it has always said that, even though they didn't work (?)
[05:07:28] <gene> Oh :(
[05:08:03] <cradek> try one more cvs up and I bet it'll work
[05:08:17] <gene> Going now
[05:08:43] <cradek> you should see "P src/emc/rs274ngc/interp_convert.cc"
[05:08:51] <cradek> or something like that
[05:11:05] <gene> No, the diff said cvs update: Updating src/emc/rs274ngc
[05:11:06] <gene> M src/emc/rs274ngc/interp_convert.cc
[05:11:07] <gene> cvs update: Updating src/emc/task
[05:11:33] <cradek> M means you have a modification in that file
[05:11:44] <cradek> so maybe we read the diff backward
[05:11:54] <cradek> ... I can only reliably read cvs diff -u
[05:11:59] <gene> let me rerun the up -dP -A again then
[05:12:47] <gene> and it still says M as before
[05:13:03] <gene> is the -dP messing with it?
[05:13:07] <cradek> no
[05:13:14] <cradek> try cvs diff -u
[05:13:56] <gene> cvs diff: Diffing src/emc/rs274ngc
[05:13:58] <gene> Index: src/emc/rs274ngc/interp_convert.cc
[05:13:59] <gene> ===================================================================
[05:14:00] <gene> RCS file: /cvsroot/emc/emc2/src/emc/rs274ngc/interp_convert.cc,v
[05:14:02] <gene> retrieving revision 1.15
[05:14:04] <gene> diff -u -r1.15 interp_convert.cc
[05:14:05] <gene> --- src/emc/rs274ngc/interp_convert.cc 11 Feb 2006 04:50:26 -0000 1.15
[05:14:07] <gene> +++ src/emc/rs274ngc/interp_convert.cc 19 Feb 2006 05:13:31 -0000
[05:14:08] <gene> @@ -1725,10 +1725,9 @@
[05:14:10] <gene> These are M100-199 with optional P & Q parameters.
[05:14:12] <gene> */
[05:14:13] <gene> /*! \todo Another #if 0 */
[05:14:15] <gene> -#if 0
[05:14:17] <gene> /*! \todo FIX-ME Impliment these at a later stage... */
[05:14:18] <gene> /* user-defined M codes */
[05:14:19] <gene> - for (index = 100; index < 200; index++) {
[05:14:21] <gene> + for (int index = 100; index < 200; index++) {
[05:14:23] <gene> if (block->m_modes[index] == index) {
[05:14:24] <gene> if (USER_DEFINED_FUNCTION[index - 100] != 0) {
[05:14:25] <gene> (*(USER_DEFINED_FUNCTION[index - 100])) (index - 100,
[05:14:27] <gene> @@ -1739,7 +1738,6 @@
[05:14:29] <gene> }
[05:14:31] <gene> }
[05:14:33] <gene> }
[05:14:35] <gene> -#endif
[05:14:37] <gene> return INTERP_OK;
[05:14:39] <gene> }
[05:14:45] <cradek> ok I can read that one
[05:14:49] <cradek> you have patched in jeff's change
[05:14:54] <cradek> BUT it is not in anon cvs yet
[05:14:54] <gene> yes
[05:15:14] <gene> should it not be reverting that
[05:15:17] <cradek> when it is, you will probably get the message "merge: version 1.16 changes are already in your file"
[05:15:25] <cradek> no
[05:15:29] <gene> that figures
[05:15:31] <cradek> cvs will not wipe out your changes.
[05:15:43] <cradek> if you want it to, remove the file and cvs up
[05:15:55] <cradek> the M means "locally modified"
[05:16:17] <gene> then they won't work again, but they were with that same change in teh TESTING code
[05:16:35] <cradek> I bet they are working.
[05:16:45] <cradek> there's little difference in this area between TESTING and HEAD
[05:17:27] <gene> me scrartches head in puzzlement I can't find any output in the trace, only that the call was made without errors
[05:18:26] <gene> Maybe its lost in teh noise of a full debug, this works even with debug at 1?
[05:19:32] <cradek> let me try my M101
[05:20:21] <gene> Turning off all teh debug noise make it visible again, so it is working
[05:20:33] <cradek> oh good
[05:20:37] <gene> damn I wish I could spell the
[05:20:39] <cradek> I don't think mine is...
[05:21:37] <gene> Anyway Chris this will help a lot, so its time I hit the sack, pretty chilly tonight in single digits
[05:21:46] <cradek> ok goodnight gene
[05:21:56] <gene> Thanks a bunch
[05:48:53] <CIA-8> 03cradek * 10emc2/docs/man/man1/halcmd.1:
[05:48:53] <CIA-8> document the new way loadrt/unloadrt work, which allows the removal of
[05:48:53] <CIA-8> a paragraph from the BUGS section
[05:55:29] <CIA-8> 03cradek * 10emc2/src/emc/task/emccanon.cc: remove debug output
[06:38:35] <Jymmm> 6000 Year old tree ---> http://news.yahoo.com/s/ap/20060217/ap_on_sc/ancient_tree
[06:51:19] <bill__> bill__ is now known as bill20r3
[09:10:41] <Imperator_> logger_aj: bookmark
[09:10:41] <Imperator_> See http://solaris.cs.utt.ro/irc/irc.freenode.net:6667/emc/2006-02-19#T09-10-41
[10:08:48] <robin_z> meep?
[14:25:23] <Imperator_> Oh my, now you can buy this toys with EMC2 !!! http://www.thecooltool.com/produktgruppe.php?language=e&pg_id=6
[15:06:53] <jepler> morning
[15:11:08] <Imperator_> morning
[15:28:00] <dmessier> good monin' all
[15:46:24] <dmessier> im hacking my way thru some self inflicted CATIA training
[16:00:58] <robin_z> catia? intersting .. but is it as good as Solidorks?
[16:01:08] <robin_z> solidworks even
[16:01:44] <robin_z> they are both built off the same core libraries
[16:01:51] <robin_z> from Dassault
[16:02:11] <robin_z> but, Solidworks is supposed to be easier to use for 99% of things
[16:02:58] <robin_z> and about 1/3rd of the price :)
[16:23:24] <staggerlytom> good morning to all
[16:24:56] <fenn> and to all a good night
[16:25:06] <CIA-8> 03cradek * 10emc2/src/Makefile: install M101 executable
[16:26:45] <staggerlytom> has interp_convert.cc been changed to allow user mcodes?
[16:29:30] <jepler> staggerlytom: Yeah, I basically just removed an #if 0
[16:30:27] <staggerlytom> right, been doing that on my local, but it was done by Fred P, and I didnt know why
[16:30:51] <staggerlytom> did he have somethign in mind?
[16:32:08] <jepler> I don't know
[17:22:04] <A-L-P-H-A> hehe... solidorks. :D
[17:42:29] <Jymmm> Mornin Folks!
[17:43:28] <staggerlytom> good morning jymmm
[18:00:12] <staggerlytom> cradek: saw you suggest carbide to cut FRP. is 10% cobalt good? ( cant find 3/16 sqr for trammel )
[18:01:15] <cradek> staggerlytom: don't know, I know HSS is destroyed instantly, never tried cobalt
[18:02:19] <Jymmm> frp?
[18:04:16] <staggerlytom> fiberglass reinforced plastic.. the stuff you make pcbs on (the stuff below the copper)
[18:05:05] <Jymmm> ah
[18:05:12] <staggerlytom> and yes hss looks like I ran it backwards
[18:22:22] <staggerlytom> bye
[19:08:41] <gene> Humm, Nickserv says my nick is owned by someone else. WTH?
[19:09:11] <gene> hello everybody
[19:10:53] <gene> Am I late?
[19:12:22] <gene> A live body?
[19:12:52] <skunkworks> did you get your sub working?
[19:13:15] <gene> I gave up on that, and made it all into a while routine
[19:13:29] <gene> But its still not working
[19:14:03] <gene> Let me get my teat out of the nuker, and I'll paste some things in for your perusal
[19:14:16] <gene> Let me get my tea out of the nuker, and I'll paste some things in for your perusal
[19:15:58] <gene> Basicly I had it working cutting a full circle with the G19 G3 command, but then I tried to clip part of the circle and it got arc errros now
[19:16:43] <gene> Here isa run output
[19:16:47] <gene> COMMENT: do arc cut
[19:16:49] <gene> emc/task/emctask.cc 270: interp_error: Radius to end of arc differs from radius to start
[19:16:50] <gene> Radius to end of arc differs from radius to start
[19:16:52] <gene> M101 P1.000000 Q0.450000
[19:16:53] <gene> M101 P4.000000 Q0.050000
[19:16:55] <gene> M101 P2.000000 Q-0.100000
[19:16:56] <gene> M101 P5.000000 Q-0.450000
[19:16:58] <gene> M101 P3.000000 Q-0.450000
[19:16:59] <gene> M101 P6.000000 Q-0.050000
[19:17:47] <A-L-P-H-A> Anyone know of a good cleaning solution?... specifically it's carbon steel and plastic. It's a pocket knife.
[19:18:13] <A-L-P-H-A> it's just got a lot of grime and grease inside the the folding mech.
[19:18:23] <gene> P1 is start pos of Y, P4 is end pos of Y, P2 is start Z, P5 is end Z, P3 is Y ancher (J value) and P5 is Z anchor(K value)
[19:18:33] <gene> The line of code is:
[19:19:00] <gene> n0290 G19 G3 Y#2 Z#5 J#3 K#6
[19:19:25] <gene> And it moves to the start posiitons fine.
[19:20:08] <gene> WD40 should be ok
[19:20:43] <A-L-P-H-A> gene, next time www.pastebin.com
[19:21:05] <gene> And that does?
[19:21:29] <A-L-P-H-A> instead of you pasting into the channel, it goes into a website container. Where you can just post the link.
[19:22:03] <A-L-P-H-A> gene, wd40? it won't leave it slippery/greasy afterwards?
[19:22:41] <gene> Yeah but its a wax, and should also lube the pins etc
[19:23:14] <gene> Alpha, its asking for a login of some kind
[19:23:34] <A-L-P-H-A> gene, it shouldn't. just type in gene. it's just to differentiate you from others.
[19:23:58] <Jymmm> http://pastebin.com/
[19:24:32] <Jymmm> A-L-P-H-A: lighter fluid
[19:24:46] <gene> Ok, whole proggy pasted & sent
[19:25:06] <Jymmm> gene you need to paste the link that was created
[19:25:38] <gene> Ths one?
[19:25:41] <gene> http://pastebin.com/563097
[19:27:09] <alpha1125> alpha1125 is now known as A-L-P-H-A
[19:27:11] <Jymmm> Jymmm is now known as Jymmmm
[19:27:37] <A-L-P-H-A> gene, example. http://pastebin.com/563105
[19:27:45] <A-L-P-H-A> gene, it's so it doesn't clutter up the chat channel.
[19:30:00] <gene> new link with out puts
[19:30:03] <gene> http://pastebin.com/563110
[19:34:22] <skunkworks> gene - I don't know how you are making a circle with that data. are you running absolut i and k?
[19:35:59] <gene> from 3.5.2 in the manual I is x j is y k is x and all are offsets from the starting position of the cut AIUI
[19:36:20] <gene> from 3.5.2 in the manual I is x j is y k is X and all are offsets from the starting position of the cut AIUI
[19:36:30] <gene> from 3.5.2 in the manual I is x j is y k is Z and all are offsets from the starting position of the cut AIUI
[19:36:48] <gene> got it right I think, damnd typus
[19:37:26] <skunkworks> unless I put the cordiates in cad - unless I entered them wronge - you can't make a circle from them - hense the error you are getting
[19:38:06] <skunkworks> or arc
[19:38:47] <gene> In other words Y starts at +.45 and is anchored to a point -0.45 from there with the J var, right?
[19:39:29] <gene> And Z starts at +0.050, anchored to a K of -0.050?
[19:42:56] <gene> And both of those anchors then set the point of the radii at Y=0.0 and Z=0.0 according to my sometimes twisted thinking
[19:44:01] <skunkworks> arc starts at .45,-.1 ends at .05,-.45 with an i,k of -.45,-.45
[19:44:16] <skunkworks> is what I was getting
[19:44:24] <gene> Thats the correct figure then?
[19:45:22] <gene> mmm, I = X not Y in G19 mode
[19:46:04] <skunkworks> I ment j,k - sorry
[19:46:23] <skunkworks> is that the numbers you posted (do they look correct?
[19:47:06] <gene> Not exactly, see the P# outputs in the pastebin posting
[19:48:48] <gene> n0290 G19 G3 Y#2 Z#5 J#1 K#1 I just tried, also dies
[19:49:10] <gene> Radius to end of arc differs from radius to start
[19:50:23] <gene> Can I do without one of j,k?
[19:51:25] <skunkworks> no then they would default to zero
[19:51:38] <dave-e> hi all....have some ubuntu questions
[19:51:59] <skunkworks> still trying to wrap my head around your program - I have to run but will be back in a bit.'
[19:52:48] <gene> Ok
[19:53:21] <alex_joni> hi dave
[19:53:22] <alex_joni> shoot
[19:53:23] <gene> Is that Pauls territory?
[19:53:31] <dave-e> hi alex
[19:53:35] <alex_joni> gene: not at all
[19:53:43] <gene> :)
[19:54:44] <dave-e> install does not bring up X but I kinda know what to do there....boot stalls at checking battery in bluetooth
[19:55:09] <alex_joni> did you try Ctrl-C ?
[19:55:19] <dave-e> I think so
[19:55:32] <dave-e> can go do that
[19:55:51] <dave-e> brb
[19:55:56] <cradek> dave-e: did you try looking on google or ubuntu.com?
[19:55:59] <alex_joni> I know sometimes over here it seems to hang at ntpdate, Ctrl-C makes it skip that very fast
[19:59:37] <dave-e> Cntr-C didn't help...
[20:00:06] <alex_joni> dave-e: ubuntu has some very good support from what I have heard
[20:00:08] <dave-e> it actually says the battery state is [ok] but doesn't go beyond that
[20:00:45] <dave-e> oh...like try their site?
[20:00:53] <alex_joni> yeah, but also phone iirc
[20:01:21] <cradek> yeah, you aren't at the emc part yet, so you can use all the things available from the community-supported ubuntu distribution
[20:01:28] <cradek> I'll look too
[20:01:47] <dave-e> for some reason grub also failed to find my 2-18 install so I'm in real trouble
[20:01:47] <alex_joni> dave-e: http://www.ubuntu.com/support
[20:02:21] <dave-e> not ubuntulinux ???
[20:03:13] <alex_joni> dave-e: I suggest you try /join #ubuntu
[20:03:21] <alex_joni> they have support on IRC too
[20:03:25] <cradek> good idea
[20:03:30] <cradek> I've had help there before
[20:03:38] <dave-e> ok....indeed I'll give it a shot
[20:03:40] <dave-e> tnx
[20:03:46] <alex_joni> dave-e: always ;)
[20:03:50] <cradek> wow, 705 people in that channel
[20:03:57] <dave-e> ouch...
[20:04:03] <alex_joni> too bad I have no idea what kind of problem you have..
[20:04:15] <dave-e> may have to reinstall...
[20:04:16] <alex_joni> dave-e: 705 means you'll get a reply ;)
[20:04:22] <dave-e> hopefully
[20:04:32] <dave-e> bye
[20:04:58] <cradek> dave-e: you can be in more than one channel at a time
[20:05:21] <dave-e> but not sure my brain can handle it
[20:05:27] <cradek> haha
[20:05:37] <cradek> ok just making sure you knew that
[20:05:45] <alex_joni> dave-e: I'm in about 4 channels usually + 2-3 private conversations
[20:06:03] <alex_joni> that's only IRC, there is also IM (usually 4-5 conversations) ;-)
[20:06:14] <alex_joni> oh, and work stuff adds to that..
[20:06:48] <alex_joni> eventually you get used to that
[20:08:35] <Jymmmm> Jymmmm is now known as Jymmm
[20:18:06] <cradek> looks like there are a lot of simultaneous topics in that channel...
[20:20:04] <alex_joni> that might need some filtering (mental filtering)
[20:23:08] <K4ts> hello
[20:31:26] <Jymmm> hi
[20:35:30] <K4ts> hi Jymmm
[20:35:35] <Jymmm> ANyone know of url to read up more on roughing/finishing passes?
[20:41:30] <giacus> Jymmm: maybe some forum..
[20:42:22] <Jymmm> giacus url?
[20:42:38] <giacus> try this http://www.woodweb.com/cgi-bin/forums/cnc.pl
[20:42:49] <giacus> artcam should have a forum too
[20:43:14] <giacus> I think its a toolpath topic
[20:43:33] <Jymmm> No, no, I mena the basics, a tutorial, or intriduction
[20:43:41] <Jymmm> less the typos
[20:43:44] <giacus> uhm.. hard
[20:43:56] <alex_joni> you can't find basics without the typos
[20:43:57] <alex_joni> :)
[20:45:11] <Jymmm> Jymmm is now known as Red70sShow
[20:45:11] <Red70sShow> Red70sShow is now known as Jymmm
[20:46:46] <Jymmm> I'm trying to figure out when, why, and how to use roughing/finishing passes.
[20:48:23] <giacus> I only found the help of the software for that, nothing more..
[20:48:46] <giacus> never seen a good doc around
[20:50:00] <giacus> if it exist, should be a book
[20:51:37] <Jymmm> yeah... like a intro to machining
[20:52:11] <giacus> never found it :(
[20:52:46] <Jymmm> http://www.efunda.com/processes/machining/machin_intro.cfm
[20:55:00] <giacus> too much vague
[20:55:22] <giacus> we could talk weeks about materials, drilling, milling ...
[20:57:17] <giacus> and anyone customize the topic about his point of view
[20:57:36] <giacus> atcam has some nice tutorial about toolpath, but isnt generic
[20:58:31] <giacus> is specific for theyr software
[20:59:00] <giacus> what wikipedia says ?
[20:59:00] <Jymmm> yeah.
[21:05:20] <giacus> Jymmm: http://www.proshoppublishing.com/
[21:05:27] <giacus> that could be nice
[21:14:26] <giacus> alex_joni: do you know any robotics wiki ?
[21:14:42] <giacus> or similar
[21:15:56] <giacus> I just found this: http://en.wikibooks.org/wiki/Robotics
[21:16:04] <giacus> any other good source ?
[21:17:46] <alex_joni> what kind of robotics?
[21:17:59] <giacus> uh ?
[21:18:00] <alex_joni> the kind I know probably aren't common on wikis
[21:18:07] <alex_joni> industrial robots
[21:18:22] <alex_joni> no toys, no autonomous, etc
[21:18:45] <giacus> mm, ok
[21:20:01] <giacus> im looking for GFDL sources around, the only I found is the Open-Pino http://www.symbio.jst.go.jp/PINO/
[21:25:29] <CIA-8> 03alex_joni * 10emc2/src/hal/drivers/hal_vti.c: latest additions from Eric that made it compilable, I also removed some stuff to make it warning-free
[21:27:28] <CIA-8> 03alex_joni * 10emc2/src/Makefile.inc.in: compile the hal_vti module by default, now that it compiles cleanly it shouldn't bother anyone
[21:28:22] <jmk-away> jmk-away is now known as jmkasunich
[21:35:53] <A-L-P-H-A> giacus, that webpage is so poorly designed. :(
[21:36:16] <giacus> A-L-P-H-A: yeah..
[21:36:26] <giacus> havent found better
[21:37:07] <alex_joni> and VERY outdated
[21:37:16] <alex_joni> [Whats New]
[21:37:16] <alex_joni> 11. Dec. 2002 : Add 2.5, updated 3.1, 5
[21:37:24] <giacus> I know ..
[21:38:10] <giacus> my community (roboitalia) could update it, maybe ..
[21:40:34] <A-L-P-H-A> roboitalia? .it?
[21:40:40] <giacus> com
[21:40:48] <A-L-P-H-A> by! it should be .it
[21:40:59] <giacus> dunno why the choosed .com
[21:41:06] <giacus> .org was appropriate
[21:41:11] <A-L-P-H-A> cause they aren't proud enough to use .it. ;)
[21:41:22] <giacus> they started in 2002 with .com
[21:41:28] <giacus> hehe
[21:42:19] <giacus> actually we're going to update the photo gallery
[21:42:34] <CIA-8> 03cradek * 10emc2/src/module_helper/module_helper.c: new hal driver
[21:42:34] <giacus> latest photos are the mine
[21:42:57] <giacus> and tryng to start a GDFL wiki too ..
[21:43:04] <A-L-P-H-A> someone send me a copy of the book for review.
[21:43:49] <giacus> wich book ?
[21:45:45] <A-L-P-H-A> http://www.amazon.com/gp/product/0974389765/ref=sr_11_1/002-5042909-8012855?%5Fencoding=UTF8
[21:46:54] <giacus> oh.. nice. how it is ?
[21:47:48] <giacus> seems a good book
[21:50:22] <giacus> dont know your real name to find it in the reviews
[21:50:35] <giacus> if you already did it
[21:51:22] <skunkworks> Cradek?
[21:51:52] <cradek> yes
[21:52:02] <alex_joni> santa?
[21:52:14] <giacus> no
[21:52:21] <cradek> skunkworks: your machine is fixed
[21:52:27] <skunkworks> I see that
[21:52:33] <alex_joni> ok, must be too late for santa..
[21:52:38] <cradek> skunkworks: I built new kernel, rtai modules and tested on yours first :-)
[21:52:55] <skunkworks> that is what I was going to ask - I loaded emc2 to play with it and forgot the 1.5gb was still in it.
[21:53:03] <skunkworks> it loaded
[21:53:13] <cradek> I ran it too
[21:53:17] <skunkworks> right ;)
[21:53:55] <jmkasunich> I hope the machine wasn't connected to the computer
[21:54:06] <jmkasunich> the idea of remotly running a mill is scary
[21:54:09] <cradek> I ran sim...
[21:54:15] <skunkworks> what did you end up doing? did you limit the amount of memory?
[21:54:15] <jmkasunich> ;-)
[21:54:24] <skunkworks> no it isn't connected to anything. ;)
[21:54:35] <cradek> yes I took out the kernel option that does a trick to allow up to 4G of ram
[21:54:39] <cradek> rtai seems incompatible with it
[21:56:11] <alpha1125> giacus, nono. I was asking for the book.
[21:56:21] <cradek> wow, what a hostname
[21:56:23] <giacus> Ok
[21:56:24] <CIA-8> 03alex_joni * 10emc2/src/hal/drivers/hal_vti.c: changed the return code to -ENODEV, thanks chris
[21:56:28] <skunkworks> nice. - one more question now - how do I up the screen resolution? now it is at 640X480 with no options that I can see.
[21:56:30] <alpha1125> I know.
[21:56:34] <alpha1125> I got signed off.
[21:56:48] <alpha1125> alpha1125 is now known as A-L-P-H-A
[21:57:28] <skunkworks> no other options that I can see in the display settings
[21:57:59] <alex_joni> skunkworks: you probably don't have the correct monitor definitions in /etc/X11/Xorg.conf
[21:58:11] <cradek> I have a system/preferences/screen resolution
[21:58:21] <cradek> but yeah, it sounds like it did not probe your monitor properly
[21:58:26] <cradek> is it an old monitor?
[21:58:27] <alex_joni> try dpkg-reconfigure xserver?
[21:58:53] <skunkworks> it is a flat screen ;) - It saw it when if first installed ubuntu
[21:59:09] <alex_joni> that was with fb I think
[21:59:14] <skunkworks> that is the other thing I noticed when I started emc2 ;)
[22:00:05] <CIA-8> 03alex_joni * 10emc2/src/hal/drivers/hal_stg.c: ENODEV return code, thanks chris
[22:12:47] <skunkworks> gene: are you still here?
[22:13:50] <skunkworks> Gene: unless you are doing full circles or circle quadrants it is pretty hard to do without trig or autocad.
[22:14:24] <cradek> skunkworks: sometimes the R format is easier to use
[22:14:58] <skunkworks> gene: http://pastebin.com/563402 did
[22:14:59] <skunkworks> http://pastebin.com/563402
[22:15:11] <skunkworks> http://www.electronicsam.com/images/KandT/AXIS.JPG
[22:15:51] <skunkworks> cradek - have never used radius - right it might be easier (do you use a negative radius to tell wich whay the ark goes?)
[22:16:14] <cradek> umm I think so
[22:16:24] <skunkworks> would make sense to me ;)
[22:16:32] <skunkworks> I should read up on it.
[22:19:46] <CIA-8> 03cradek * 10emc2/VERSION: moving TESTING
[22:20:39] <A-L-P-H-A> cradek, alex_joni, is there a need to defrag a linux style partition?
[22:20:47] <gene> Yikes, was out cooking and eating dinner
[22:20:56] <cradek> A-L-P-H-A: no
[22:21:02] <alex_joni> A-L-P-H-A: no
[22:21:10] <skunkworks> gene - nice - where are your priorities?
[22:21:12] <A-L-P-H-A> why's that? [do you happen to know?]
[22:21:33] <alex_joni> skunkworks: no, there is no such thing as negative radius iirc
[22:21:41] <A-L-P-H-A> my priorities... make money. find a girl. pay the girl. leave.
[22:21:41] <alex_joni> but I might be wrong
[22:21:48] <cradek> alex_joni: I think you're wrong
[22:21:50] <gene> Well, that was lunch guys, breakfast was too long ago if you get my drift.
[22:21:53] <skunkworks> how would it know which way to go?
[22:22:08] <alex_joni> cradek: that's a fscked up ontology then
[22:22:11] <cradek> alex_joni: but I might be wrong :-)
[22:22:14] <jmkasunich> are you talking about g-code arcs?
[22:22:16] <skunkworks> you could make 2 arc at a given radius though 2 points
[22:22:17] <cradek> too busy to look
[22:22:18] <alex_joni> jmkasunich: yes
[22:22:22] <alex_joni> cradek: ditto
[22:22:23] <jmkasunich> G2 goes one way, G3 goes the other
[22:22:29] <jmkasunich> no negative radii
[22:22:29] <skunkworks> there we go
[22:22:34] <skunkworks> thanks jmk
[22:22:36] <cradek> ok I'm wrong
[22:22:46] <skunkworks> forgot about that ;)
[22:22:48] <alex_joni> no, we were talking about G2XxxYyyRrr vs. R-rr
[22:22:56] <jmkasunich> heh, G-code is old, and made for machinists
[22:23:08] <jmkasunich> rulers and micrometers don't measure negative
[22:23:10] <alex_joni> but I am sure G2 vs G3 give directions
[22:23:15] <alex_joni> jmkasunich: I AGREE
[22:23:16] <alex_joni> ;)
[22:23:37] <A-L-P-H-A> jmkasunich, yes they do!!! it's called "i" imaginary numbers. :)
[22:23:44] <cradek> The R number is the radius. A positive radius indicates that the arc turns through 180 degrees or less, while a negative radius indicates a turn of 180 degrees to 359.999 degrees.
[22:23:52] <cradek> ok I'm right, and you're all wrong.
[22:23:56] <skunkworks> ;)
[22:24:04] <gene> I've tried both scenarios, essentially the same results, I don't know what the heck I'm doing :(
[22:24:13] <jmkasunich> well ah'll be darned
[22:24:14] <skunkworks> like I have said - I have never used radis
[22:24:20] <cradek> there are two arcs in a certain direction through two endpoints with a certain radius
[22:24:38] <skunkworks> hey that is what I said
[22:24:54] <cradek> a "short" one and a "long" one
[22:24:55] <jmkasunich> cradek: but one curves left and one curves right
[22:25:04] <cradek> jmkasunich: no they don't
[22:25:06] <jmkasunich> assuming you always start at the same endpoint
[22:25:16] <gene> Then R is then wrong designator and thats condfusing me
[22:25:22] <jmkasunich> oh, my bad
[22:25:24] <cradek> jmkasunich: one's not the arc you want, that's why you can't picture it in your head
[22:25:34] <jmkasunich> I can picture just fine
[22:25:34] <gene> WHen I thin R, I think radii, not radians
[22:25:47] <jmkasunich> but I was thinking of two arcs, same endpoints AND same center
[22:25:51] <cradek> jmkasunich: ok, no offense meant
[22:25:54] <jmkasunich> the case you are talking about is two centers
[22:25:58] <cradek> jmkasunich: yeah, different centers
[22:26:00] <K4ts> night
[22:26:04] <cradek> yeah only endpoints and radius are specified
[22:26:07] <gene> and that doesn't fly
[22:26:17] <jmkasunich> one center on each side of the line from endpoint 1 to endpoint 2
[22:26:25] <cradek> jmkasunich: exactly
[22:26:31] <cradek> ok back to work
[22:26:32] <jmkasunich> see, I can picture it ;-)
[22:26:40] <jmkasunich> I just had the wrong picture before
[22:26:58] <cradek> you rarely want the one that bulges out, so it's easy to forget about it
[22:27:59] <skunkworks> gene: did you see what I had wrote to you?
[22:28:17] <gene> At that last lnk? Yes
[22:28:28] <CIA-8> 03cradek * 10emc2/VERSION: Done tagging
[22:28:36] <skunkworks> what are you trying to do with your program?
[22:29:22] <gene> carve a hollow in the face of a piece of cast iron to set a tubes side into, the tube holding new acme nuts
[22:30:15] <gene> so the contour of the hollow face exactly matches the tube for a good sweat joint with SS.
[22:31:02] <skunkworks> is it 1/2 a circle?
[22:31:17] <skunkworks> going in the y/z plane?
[22:31:21] <jmkasunich> one pass if you have the right ball end mill ;-)
[22:31:28] <skunkworks> ;)
[22:38:34] <gene> No, its about 100 degrees starting at the rear with z up, descending into the workpiece while y moves forward
[22:38:50] <alex_joni> yay, emc2 working great
[22:39:02] <alex_joni> stepper_mm with 7.5uS BASE_PERIOD
[22:39:02] <gene> But 1/4 circle describes it fairly well
[22:40:22] <gene> The older code I lost, in emc, used a falt nosed mill and assumed the rear corner was the only cutting surface, so the gap opened up a thou or so at the front
[22:40:35] <gene> flat nosed...
[22:41:58] <gene> And adjusting software is a heck of a lot cheaper than adjusting the size of the ball mill :)
[22:42:26] <alex_joni> and easier
[22:42:26] <skunkworks> couple of things (not visualizing it very well) how far do you go ahead each time - n0280 #8 = [#8 + #9] (set new x pos in var) this shows 2 inches - so I think it only goes through your loop once. #8 starts out 0 and #9 is 2
[22:43:26] <skunkworks> should it be n0280 #8 = [#8 + .001] or something like that?
[22:43:40] <cradek> new emc2 & emc2-dev packages in the ubuntu repository
[22:43:41] <gene> It should advance .025" x per pass
[22:44:06] <gene> and when I had it doing full circles, that did work.
[22:44:14] <skunkworks> n0280 #8 = [#8 + #7] (set new x pos in var) then?
[22:45:32] <gene> yur right of course, I'd changed the var order and missed that one.
[22:49:31] <gene> And now its running, but the direction is wrong, lets see if a G2 fixes that, yup, bvut the down stroke is short, bot reaching the bottom of the cut
[22:49:50] <gene> And now its running, but the direction is wrong, lets see if a G2 fixes that, yup, but the down stroke is short, not reaching the bottom of the cut
[22:50:17] <gene> Looks like its only swinging about 70 degrees?
[22:51:52] <cradek> new emc2-axis package in the ubuntu repository
[22:53:02] <gene> and when I try to adjust the y end point, I'm back to
[22:53:16] <gene> Radius to end of arc differs from radius to start
[22:53:18] <gene> M101 P1.000000 Q0.450000
[22:53:19] <gene> M101 P4.000000 Q0.050000
[22:53:21] <gene> M101 P2.000000 Q-0.150000
[22:53:22] <gene> M101 P5.000000 Q-0.450000
[22:53:24] <gene> M101 P3.000000 Q-0.450000
[22:53:25] <gene> M101 P6.000000 Q-0.050000
[22:54:07] <skunkworks> if you change the y end point you have to chang the j and k to match
[22:54:23] <gene> So I need to get out the calculator & redo the Z endpoint too I guess. Right?
[22:55:17] <gene> From the book:J and K are the offsets from the current location (in the Y and Z directions, respectively) of the center of the circle.
[22:55:33] <skunkworks> right
[22:55:39] <gene> And I define the current location as the posotion at the start of the move, right?
[22:55:45] <skunkworks> right
[22:56:24] <gene> So if I remove the z figure, it will calulate in internally and be right?
[22:56:45] <gene> That would be the K value I think?
[22:58:10] <skunkworks> now you lost me again. if you change the ending point of the ark - you need to tell it with j,k where the center of the arc is now. from the arc starting point.
[23:01:33] <gene> Ok, I'm doing that, but when I try to change its ending point, the error is back
[23:02:24] <gene> It seems to me that if I set the anchors with J&K, then only one ending point is needed. Right?
[23:05:18] <gene> Humm backplot optical illusion I guess, set it for yz, and it can be seen to be going exactly where its told
[23:05:51] <gene> I'm 100 feet from the machine and the motors aren't powered.
[23:06:03] <skunkworks> ;)
[23:06:10] <skunkworks> you need to try axis ;)
[23:07:08] <gene> Ok, its working, and now that I know how, I may see if I can make just one figure control the size/depth of the cut. I assume I can do math inside a pair of []'s?
[23:07:44] <gene> I have axis, but not really currrent as I know you are working on it non-stop
[23:07:46] <skunkworks> alex - I tried to run dpkg-reconfigure xserver and it says it needs to be run as root.
[23:08:17] <giacus> http://www.kennislink.nl/upload/129573_962_1112181501982-emc2.jpg
[23:08:19] <skunkworks> yes you should be able to do the math
[23:08:21] <giacus> hehe
[23:08:32] <alex_joni> skunkworks: try sudo in front
[23:08:40] <skunkworks> crap - thanks
[23:08:40] <alex_joni> but that was from memory, it might be wrong
[23:09:06] <gene> Ok, I should be home if not free :) thanks guys
[23:11:00] <skunkworks> says there isn't an xserver package
[23:11:47] <skunkworks> good to hear gene
[23:12:23] <gene> And this beats the heck out 0of 7000 lines of code I used the first time
[23:13:03] <skunkworks> ;) loops are our friend
[23:13:27] <gene> Now I'll go away, and concentrate on the math stuff so I can change one var to change the sixe of the partial cylinder, yeah baby!
[23:19:10] <giacus> g night
[23:23:23] <skunkworks> alex - looked in /etc/X11/Xorg.conf and the monitor was there. rebooted and it was fine ;)
[23:23:37] <alex_joni> skunkworks: ok
[23:23:51] <skunkworks> go figure - cradek must have screwed it up. Last time I let him in my machine. ;)
[23:24:27] <skunkworks> ^definatly joking
[23:26:14] <skunkworks> cradek - shouldn't I be getting a software update notifacation?
[23:27:18] <alex_joni> skunkworks: probably, but not right away
[23:27:33] <skunkworks> ok
[23:27:55] <alex_joni> if you want the updates right away you can do 'sudo apt-get update', or use the System->Update Manager? to get a new list
[23:28:12] <alex_joni> and afterwards you'll see the new packages, or just wait a bit and it'll roll in by itself
[23:28:54] <skunkworks> update manager says I am up-to-date. have to wait I guess. ;)
[23:29:03] <alex_joni> no, there is a button there
[23:29:08] <alex_joni> top-right
[23:29:14] <alex_joni> don't recall how it's called
[23:29:39] <skunkworks> there it is "reload"
[23:29:42] <skunkworks> that worked
[23:30:15] <alex_joni> ok, after that you will get the list below, and the notification shortly after ;)
[23:30:27] <skunkworks> uh - that just cause a hard lockup
[23:30:35] <skunkworks> no mouse or keyboard
[23:30:57] <skunkworks> it shows the emc update and axis but now I am frozen
[23:31:09] <alex_joni> that is ODD
[23:31:22] <skunkworks> you know me ;)
[23:34:26] <skunkworks> reboot and it is installing now.
[23:34:45] <skunkworks> does it require a reboot?
[23:38:02] <alex_joni> if the kernel changed, yes
[23:40:14] <skunkworks> is the fix that cradek did for the 1+gb problem something that will have to be done manually on another computer - or is it something that is fixed in emc2?
[23:42:37] <skunkworks> got to go. Thanks for all your help.
[23:42:43] <skunkworks> bbl
[23:50:05] <alex_joni> night all