#emc-devel | Logs for 2010-08-17

[07:14:12] <alex_joni> Jymmm: small deer ;)
[07:33:36] <alex_joni> Jymmm: http://failblog.files.wordpress.com/2010/08/e91fae7e-1b92-4801-a8fd-0ee6c32d1a66.jpg
[07:41:05] <Jymmm> alex_joni: That's like telling someone "DONT LOOK", so what do they do? They look!
[12:42:07] <mozmck> alex_joni: I wonder if this could be related to the problem we are seeing with > 2 cpus (at least AMD): https://mail.gna.org/public/adeos-main/2010-08/msg00000.html
[12:57:17] <alex_joni> mozmck: sounds like it
[12:57:28] <alex_joni> although from your stories it's more frequent than every 100 boots
[12:57:47] <alex_joni> another thought.. what if we limit CPUs to 2 ?
[14:19:59] <cradek> that was close - I almost got sucked into the latest g33 debate
[14:20:53] <alex_joni> heh
[14:40:33] <JT-Work> come on you know you want to :)
[14:50:21] <cradek> meh
[14:50:36] <cradek> has all been said before
[15:00:25] <jepler-> The g33.1 documentation says the final move is 'An unsynchronized move back to the original coordinate.' Is it a rapid or a feed?
[15:00:37] <cradek> rapid
[15:04:32] <skunkworks> now when someone says 'fudge' the threading cycle to make tapered threads - is that because they don't know how to use math?
[15:05:24] <cradek> a spec for a tapered thread specifies the pitch a certain way - to actually cut it you need to program it a different way
[15:06:24] <cradek> you need to know the tiniest bit of trig to program the tapered thread (although you also need the same amount of trig to calculate the endpoints to get the taper before you thread it)
[15:06:41] <skunkworks> that is what I would think
[15:08:05] <cradek> for a thread that measures 12tpi when you're measuring 20 degrees from it, you get 12/cos20 along it
[15:08:44] <cradek> as I'm sure you know... :-)
[15:09:00] <skunkworks> heh
[15:28:44] <jepler-> skunkworks: besides the unintended change to axis.py, did you find that the "commanded spindle speed" change worked for your needs?
[15:29:39] <skunkworks> Yes - I don't see any issue with it. (I have no problem multiplying it by 60 for my needs)
[15:29:51] <skunkworks> solves my issue
[15:30:04] <skunkworks> (of not knowing the commanded spindle speed)
[15:30:34] <skunkworks> Thanks again for that. You're a life saver!
[15:32:40] <jepler-> in that case I'll stick it on master
[15:35:01] <jepler-> jepler- is now known as jepler
[15:35:27] <CIA-5> EMC: 03jepler 07master * ra1c66d4949f1 10/src/emc/motion/ (control.c mot_priv.h motion.c): Put commanded spindle speed on a hal pin