#emc-devel | Logs for 2007-12-06

[03:23:44] <CIA-19> EMC: 03jepler 07TRUNK * 10emc2/src/ (configure configure.in): improve the end-of-configure text
[13:46:11] <jepler> I thought we'd fixed things so that the unused elements of position were zero'd out
[13:46:14] <jepler> >>> s = emc.stat()
[13:46:15] <jepler> but it looks like we haven't.
[13:46:17] <jepler> >>> s.poll()
[13:46:20] <jepler> >>> s.position
[13:46:23] <jepler> (0.0, 0.0, 0.41200000047683716, -2.3145821281100598e+92, -9.0449115485932757e+157, nan, -8.7293045983278263e+157, -9.0449115485932757e+157, 2.9151278656264617e+215)
[13:56:50] <CIA-19> EMC: 03jepler 07v2_2_branch * 10emc2/debian/changelog: make g38.2 probe log zero for axes that do not exist
[13:56:51] <CIA-19> EMC: 03jepler 07v2_2_branch * 10emc2/src/emc/motion/ (control.c motion.c): make g38.2 probe log zero for axes that do not exist
[13:58:15] <CIA-19> EMC: 03jepler 07TRUNK * 10emc2/src/emc/motion/ (control.c motion.c): merge from branch: make g38.2 probe log zero for axes that do not exist
[14:07:47] <SWPadnos> hmmm. jepler, I just looked over the configure cnahge from last night, and it looks like there's an extra fi in there
[14:08:02] <SWPadnos> at least, I see a +fi lne, but no removed fi or added if
[14:08:06] <SWPadnos> line
[14:08:32] <SWPadnos> unless I look a few lines below. nevermind
[14:12:17] <jepler> thanks for checking up on me
[14:12:27] <SWPadnos> heh - no problem ;)
[14:13:04] <SWPadnos> after marking as read the 147 commit messages I missed while traveling, I actually started looking at the 5 new ones
[14:18:08] <jepler> SWPadnos: do you think would I find a Core 2 Duo T7250 (2.0GHz) to be markedly faster than a Pentium M 1500 (for compiling software, etc)?
[14:18:27] <SWPadnos> hmmm. let me think about that
[14:18:29] <SWPadnos> yes
[14:18:38] <SWPadnos> especially with -J<something>
[14:18:43] <jepler> of course
[14:20:04] <SWPadnos> does the T7250 have the 4M cache as well?
[14:21:05] <jepler> this page says 2M
[14:21:14] <SWPadnos> ok
[14:21:36] <SWPadnos> if you can get a T7200 or T7400, that may be better - I think those are 4M cache versions
[14:22:35] <jepler> yeah the next CPU up on this laptop for $50 more is a 4M cache version
[14:23:04] <jepler> oh bizarre: "the 9-cell battery will protrude from the front of the notebook approximately 1", creating a soft-touch wrist-rest."
[14:23:04] <SWPadnos> that's probably worth it, especially for something like compiling
[14:23:09] <SWPadnos> heh
[14:24:02] <SWPadnos> oh. I should fire up one of the "google OS development" boards I got
[14:30:59] <skunkworks_> anyone seen this? over my head http://www.ee.ust.hk/~eeyoyo/thesis.pdf
[14:32:59] <SWPadnos> it has a lot of pages
[14:33:08] <skunkworks466> that is what I thought
[14:33:10] <skunkworks466> ;)
[14:35:08] <SWPadnos> I could do without the thanks to Jesus
[14:35:30] <skunkworks466> I didn't see that.. I saw him thinking his wife and kids.
[14:35:36] <skunkworks466> and dad
[14:35:46] <SWPadnos> and last, the giver of all talents blah blah blah
[14:40:00] <SWPadnos> maybe I'll go back to shopping for computers or something
[14:51:13] <cradek> jepler: thanks for fixing that already (I am sure I got zeroes when I was testing it)
[14:53:17] <jepler> cradek: you were probably running a 9-axis machine
[14:53:31] <jepler> that reminds me, please contribute the simulated probing configuration to CVS
[14:53:40] <cradek> I don't think I was. I was on max.
[14:53:43] <jepler> oh, hm
[14:53:53] <cradek> oh well
[14:57:15] <SWPadnos> hmm. strange how NewEgg has the AMD 2.3 GHz quad-core CPU selling for $30 less than the 2.2GHz quad-core CPU
[14:57:28] <SWPadnos> $269 and $299 respectively
[14:59:20] <jepler> phenom or opteron?
[14:59:24] <SWPadnos> phenom
[14:59:41] <cradek> phoneme? pheromone?
[14:59:55] <SWPadnos> phantasmagorical
[15:00:26] <jepler> one says AM2 the other says AM2+ (socket).
[15:00:47] <SWPadnos> hmmm
[15:01:51] <jepler> I haven't quite figured out the AM2 / AM2+ thing, or whether a 1-year-old AM2 motherboard will be able to use one of these CPUs (though it's not listed on the official CPU compatability list for my motherboard)
[15:02:26] <SWPadnos> yeah. I lost track of AMD sockets when the 1207/F/1207+ ... stuff happened
[15:03:59] <jepler> hm, among other interesting information this page lists a "Phenom X3". http://en.wikipedia.org/wiki/Socket_AM2+
[15:06:48] <SWPadnos> I wonder what that would mean
[15:07:18] <jepler> apparently it means triple core. "Phenom triple-core (Toliman)" (model number would start with 7). http://en.wikipedia.org/wiki/AMD_Phenom
[15:07:31] <SWPadnos> oh - DDR2 and DDR3 support, for socket AM3
[15:07:44] <SWPadnos> oh right, I remember reading about those now
[15:07:55] <SWPadnos> if all 4 cores don't work, just sell it as a 3-core chip
[15:12:43] <jepler> I wonder how many dies they can salvage that way
[15:12:57] <SWPadnos> dunno
[15:13:09] <jepler> intel's had quad cores on the market first and they don't seem to have decided to do this
[15:13:25] <SWPadnos> I'd probably be a little concerned that the process wasn't right for the chip, and that other cores might fail sooner than expected
[15:13:54] <SWPadnos> Intel's quads are actually two duals slapped together, so they wouldn't put a defective dual into a quad package
[15:14:03] <jepler> oh, I see
[16:09:49] <jepler> man -- I want to reward Dell for offering systems without Windows preinstalled and at a slight savings, but they are missing some options (e.g., you can't choose bluetooth in D830N (freedos) but you can in D830 (XP/Vista)) and at least with their promotions today the $40 savings you get for FreeDOS vs XP Home is probably eaten up by the Free Shipping on the XP system.
[16:18:21] <cradek> but you don't need a new laptop...
[16:19:07] <jepler> I keep forgetting that
[16:20:55] <cradek> aluminum sure is expensive.
[16:26:10] <skunkworks_> only if you buy it. How is the max coming? I suppose the shop is taking away from that :)
[16:40:05] <skunkworks_> and work, eating and sleeping..
[16:47:01] <cradek> other than buying it I can only think of one way to get aluminum...
[16:47:39] <cradek> max is coming along fine. I got the second rotary (C) assembled last night
[16:48:05] <cradek> without a surface grinder, which is what I actually needed
[16:53:27] <skunkworks_> next purchase :)
[16:53:41] <skunkworks_> I would think stuart would have an extra ;)
[16:57:19] <cradek> he'd let me use his. too bad about the 10 hours of driving though.
[17:00:23] <skunkworks_> yeck
[17:01:46] <SWPadnos> with the price of gasoline, it doesn't take many trips to equal the cost of your own grinder
[17:01:53] <SWPadnos> plus the aggravation :)
[17:54:20] <skunkworks_> I am sure you could get a inexpensive one from enco or the lik
[17:54:24] <skunkworks_> like
[18:45:08] <skunkworks_> This is how mach does rigid tapping.. :p http://www.hossmachine.com/Videos9.html#videos53&54
[18:47:32] <cradek> ouch
[18:47:42] <cradek> but, he's doing "peck tapping"
[18:47:53] <cradek> (he should get some better taps instead)
[18:49:15] <skunkworks_> he said he stalled it once so far - so he added more pecking cycles
[18:49:48] <cradek> a thru tap in 1/2 aluminum should take about 3 seconds :-)
[18:50:06] <SWPadnos> that's about the same as how EMC2 does rigid tapping
[18:50:29] <SWPadnos> except for the fact that he's keeping the spindle load low enough so it doesn't slow down :)
[18:50:30] <cradek> this is actually a pretty clever hack he's got going
[18:50:53] <cradek> SWPadnos: no, he's got his spindle hooked to a stepper configured as the A axis
[18:51:02] <SWPadnos> oh, I didn't notice that part
[18:51:17] <cradek> he must pull the belt off to do regular machining
[18:51:38] <SWPadnos> ok. I noticed the belt, but didn't think further than "that's a strange arangement"
[18:52:34] <SWPadnos> it does look like hand tapping might be faster
[18:52:54] <cradek> yeah but this is going to be much more upright
[18:53:37] <cradek> this is what someone would have had to do with emc a few years ago...
[18:53:49] <SWPadnos> yep
[18:54:05] <skunkworks_> but we can make fun of it now.. ;)
[18:54:11] <SWPadnos> nyah nyah
[18:54:40] <cradek> I still think it's pretty clever considering what he's working with
[18:54:56] <skunkworks_> yes - we had discussed this at the cnc workshop.
[18:55:31] <cradek> neat little tool holders - I wonder if they are tormach?
[18:55:51] <skunkworks_> did you see his tool changeer?
[18:56:00] <cradek> loading now
[18:56:47] <skunkworks_> he has turned that little mill into a monster.
[18:56:48] <skunkworks_> http://www.cnczone.com/forums/showthread.php?t=30385
[18:57:47] <cradek> must be air powered somehow?
[18:58:19] <skunkworks_> yes
[18:58:53] <skunkworks_> I think he used belvel springs to clamp it - air cylinder to unclamp.
[18:59:44] <cradek> I can think of all sorts of designs for power drawbars right up until I realize the spindle still has to turn
[19:00:00] <skunkworks_> http://www.cnczone.com/forums/showthread.php?t=43651
[19:00:11] <skunkworks_> that is his toolchanger thread
[19:00:19] <skunkworks_> heh
[19:04:33] <cradek> jeez his Z is slow
[19:08:59] <skunkworks_> yes
[19:10:18] <cradek> I would tell him he could use my drill press and tapping head...
[19:13:47] <skunkworks_> yes. (I was thinking the same thing)
[19:14:58] <skunkworks_> (having him use your drill press and tapping head)
[19:15:24] <cradek> I did a bunch of 1/2-13 holes that way. it's **fast**
[19:29:55] <cradek> I just found why it's so slow: http://www.cnczone.com/forums/attachment.php?attachmentid=29635&d=1168883992
[19:30:33] <cradek> "FET-3/Dynamo kit($329) from stepperworld.com"
[19:35:02] <jepler> it's a dumb unipolar L/R drive with expensive transistors?
[19:35:20] <cradek> looks like
[19:35:31] <cradek> it might have step/dir circuitry (probably PICs)
[19:36:09] <cradek> with his 10? pitch ballscrews and those big steppers, I couldn't figure out why it was so slow...
[19:36:29] <jepler> those black package ICs look like they must be power, because they have the funny double pins at the center. http://www.stepperworld.com/images/fet3Pic.jpg
[19:36:49] <jepler> the yellow/orange ICs -- dunno what they are
[19:37:07] <cradek> http://www.cnczone.com/forums/attachment.php?attachmentid=31317&d=1171070322
[19:37:11] <cradek> wow look at his pendant!
[19:37:15] <jepler> but from the silkscreen text it does look like it takes step/direction
[19:37:30] <jepler> wow
[19:37:39] <SWPadnos> the yellow/orange ones look like resistor packs to me
[19:37:55] <cradek> yes the yellow are definitely resistors
[19:38:42] <cradek> for $329 you could almost buy three geckos. what a joke.
[19:38:59] <jepler> I think the $329 kit includes motors
[19:38:59] <SWPadnos> but this is so simple - it just plugs right into the parallel port
[19:39:39] <SWPadnos> yep, 190 oz-in
[19:39:59] <SWPadnos> heh, and a power supply
[19:40:09] <jepler> on the sp3 you can see that the funny power chips are ucn8048 (not 100% sure of the digits) http://www.stepperworld.com/images/SP3pic.jpg
[19:40:31] <SWPadnos> that looks like 5804B
[19:40:32] <jepler> er, ucn58048?
[19:40:51] <jepler> ah yes
[19:41:20] <cradek> and some nice fast 1n4001 diodes I'm sure
[19:42:58] <jepler> I played a bit with zenbot last night. I found that 1/4 stepping works a little better than 1/2 stepping, but 1/8 stepping still works for crap.
[19:43:15] <jepler> I dunno why that would be
[19:43:58] <cradek> in this thread he does say it's R8 so those are probably tormach or similar
[19:44:24] <cradek> (R8 seems so small for use in a full sized mill)
[21:26:13] <tissf> hi
[21:27:46] <jepler> hi tissf
[21:29:30] <tissf> jepler: problem with accent in checklinks
[21:29:40] <tissf> python ../docs/src/checklinks.py ../docs/html/gcode_fr.html ../docs/html/gcode_main_fr.html
[21:29:41] <tissf> Anchors used in gcode_fr.html but not defined in gcode_main_fr.html:
[21:29:43] <tissf> sub:-Messages-d�boguage
[21:29:44] <tissf> sub:G54-�-G59.3:
[21:29:46] <tissf> sec:G33,-G33.1:-Broche-synchronis
[21:29:48] <tissf> sub:G81:-Cycle-de-per�age
[21:29:50] <tissf> sec:G0:-Interpolation-lin�aire
[21:29:52] <tissf> sub:S:-R�glage-broche
[21:29:53] <tissf> sec:M100-�-M199:
[21:29:55] <tissf> sub:G81-�-G89:
[21:29:56] <tissf> sec:G1:-Interpolation-lin�aire
[21:29:58] <tissf> sub:F:-R�glage-vitesse
[21:29:59] <tissf> sub:G61,-G61.1,-G64:
[21:30:01] <tissf> sub:G10:-Coordonn�es-pi�ce
[21:30:02] <tissf> sec:G76:-Filetage-pr�programm
[21:30:04] <tissf> make: [checkref_fr] Erreur 1 (ignorée)
[21:30:05] <tissf> all these links are correct but with accents
[21:30:43] <tissf> I have to remove these accents ? or you can update the script ?
[21:30:49] <jepler> do the links work when you use a browser?
[21:30:58] <tissf> Yes
[21:31:50] <tissf> It's just messages, but they are ignored
[21:32:41] <tissf> jepler: I don't like all these warning :)
[21:33:25] <jepler> I'll look into it
[21:34:17] <tissf> jepler: tnak you
[21:34:23] <tissf> thank
[22:02:28] <jepler> yes it's a problem of encoding
[22:02:36] <jepler> I'm not sure how to untangle it but I'll keep working
[22:36:00] <tissf> jepler: This is not very important!