#emc-devel | Logs for 2009-11-01

[00:40:45] <jepler> hi dgarr
[00:48:53] <dgarr> hi -- i noticed today that short (print, messages) were not see immediately but never saw this on simulator. a patch for consideration:
[00:48:57] <dgarr> http://www.panix.com/~dgarrett/stuff/0001-make-print-messages-available-immediately.patch
[00:50:40] <jepler> looks pretty obvious
[00:51:00] <jepler> though I'm a bit surprised stdout isn't defaulting to line mode
[00:51:26] <jepler> you are running emc on the terminal, or in some other way?
[00:51:53] <dgarr> i start it from a terminal
[00:53:16] <jepler> redirected with | or > ?
[00:55:04] <dgarr> i use aliases like "emc nameofinifile" but there may be a pipe on that one, moment to turn it on and check
[00:58:21] <dgarr> yes i have a pipe like: alias mill="rip;emc -v nameofinifile 2>&1 | tee -ia logfilename"
[00:58:44] <cradek> ahh
[00:58:45] <jepler> ok, that makes sense
[00:58:46] <dgarr> so that probably caused the behavior i saw
[00:59:15] <dgarr> but adding the flush made it work with that
[01:00:42] <dgarr> are there downsides to including the fflush(stdout)?
[01:00:51] <dgarr> for (print, messages)?
[01:01:25] <jepler> probably not
[01:05:39] <CIA-65> EMC: 03jepler 07master * r664fecd25c79 10/src/emc/rs274ngc/interp_convert.cc: make (print, messages) available immediately
[01:07:39] <dgarr> thanks -- i'd forgotten about the pipe (but the log files have been useful on occasion)
[01:39:01] <pcw_home> I wondered about wireless probe delays, i guess if its constant it can be factored out (knowing velocity)
[03:01:51] <jepler> yeah but it'll end up being gross
[03:02:06] <jepler> 'night all
[03:04:27] <pcw_home> nite
[03:04:38] <pcw_home> bye
[15:08:29] <skunkworks> http://groups.yahoo.com/group/mach1mach2cnc/message/114275
[15:10:30] <skunkworks> http://groups.yahoo.com/group/mach1mach2cnc/message/114286
[15:19:12] <alex_joni> heh
[15:19:16] <skunkworks> I don't care what other people say - I like emc ;)
[15:19:21] <skunkworks> Hi alex
[15:20:58] <alex_joni> hey sam
[15:21:44] <skunkworks> dad just got 6 amc drives - they are 30a peak 80v that will work good for playing with the puma. I
[15:21:59] <skunkworks> I might pick up some cheaper ones also.
[15:22:03] <alex_joni> cool
[15:22:24] <alex_joni> 30A probably will work for all 6 axes
[15:22:26] <alex_joni> :P
[15:22:36] <skunkworks> the brushless ones go pretty cheap - and they can be used brushed also
[15:23:29] <cradek> heh, "it only gets done if it
[15:23:30] <cradek> suits one of them, or you do it yourself."
[15:23:44] <alex_joni> wonder how that works when applied to mach
[15:24:04] <alex_joni> "it only gets done if ??, or you forget about it"
[15:24:05] <cradek> surely "it only gets done if it suits them, otherwise you're totally screwed"
[15:25:04] <mozmck> cradek: you're correct there, I've run into that.
[15:26:32] <cradek> "but at least you can go to your grave feeling entitled to a fix since you're out $x00"
[15:26:35] <cradek> ok I'll stop now
[15:26:44] <cradek> good morning everyone :-)
[15:26:52] <mozmck> :) morning!
[15:27:00] <jepler> hi cradek
[15:27:08] <jepler> did you find the latency/rise time number for your probe?
[15:27:17] <cradek> no, I slept instead
[15:27:28] <jepler> that's often a good idea
[15:28:06] <mozmck> what do you think about this: http://groups.yahoo.com/group/mach1mach2cnc/message/114255
[15:28:41] <mozmck> he's not sure SCurve is really that valuable...
[15:30:01] <cradek> on my mill I turned down the accel because I could really feel it in the floor... it is still very fast
[15:30:12] <cradek> it's true that making the path smoother also makes it slower
[15:31:10] <cradek> I think most people who spend a lot of time worrying about this would not see a significant difference on their machine
[15:31:21] <skunkworks> I don't see how it wouldn't
[15:31:38] <cradek> bbl, making potato panckaes
[15:31:43] <skunkworks> (smoother path + slower)
[15:32:10] <jepler> it strikes me that you could "easily" write a one-axis jerk-limited planner (hal component, not emc) to see if it makes you feel better
[15:32:15] <skunkworks> oh - that sounds good - although we went to the local greasy spoon this morning - yum
[15:33:19] <jepler> we had (regular) pancakes
[15:33:22] <jepler> with real maple syrup, yum
[15:34:04] <mozmck> bbl
[15:34:59] <skunkworks> mmmm - I should not be hungry...
[15:43:40] <alex_joni> should I build dapper packages?
[15:43:45] <alex_joni> I have a VM handy
[15:59:35] <alex_joni> * alex_joni counts bits dribbling in.. really slowly
[16:05:04] <alex_joni> ntpdate: step time server offset 6370401.240536 sec
[16:05:15] <alex_joni> (talk about living in the past :D)
[16:08:10] <jepler> alex_joni: no, I've got it under control
[16:08:44] <alex_joni> jepler: oh, ok.. it's pretty easy for me if you want
[16:09:44] <jepler> alex_joni: I just finished uploading them, in fact -- now it needs cradek to update and sign the repository (I msg'd him)
[16:10:07] <alex_joni> right.. cool
[16:33:43] <cradek> bah, I hate reading pissing contests
[16:34:08] <jepler> I hate it more than you.
[16:34:25] <cradek> "when people say EMC can do rigid tapping, now we'll be able to say we have S curves (but we don't - but art says he has code that would do it if he had time to integrate it!!) yay!!"
[16:34:36] <cradek> "so we win again!! yay us!!"
[16:35:51] <cradek> am I the only one who isn't bothered one bit if someone chooses to use different software than I do?
[16:36:29] <alex_joni> not the only one
[16:36:38] <alex_joni> but onr of the very few
[16:37:18] <cradek> jepler: maybe inside the probe there is a "delay time" switch that configures false trigger immunity
[16:37:33] <alex_joni> "how can people decide to use that free emc, when _I_ payed for my mach licence?"
[16:38:18] <alex_joni> "surely _they_ are wrong..."
[16:39:42] <jepler> that's OK, it's just the mirror of what I think: how can that software possibly be any good? It costs money!
[16:40:30] <alex_joni> jepler: it's not, I tried it
[16:40:31] <alex_joni> :P
[16:43:04] <cradek> jepler: nothing tells me what the options are - guess I'm going to have to open the probe to find it
[16:44:09] <alex_joni> http://blogs.osuosl.org/kreneskyp/2009/06/02/irssi-notifications-via-xmpp/
[16:45:11] <cradek> "Myself and otherses can't make an English senteince to save themselveses's lives"
[16:46:05] <cradek> sorry
[16:46:38] <alex_joni> cradek: lol
[16:56:42] <jepler> cradek: is your probe "MP3"? I found a PDF which says "there is a nominal delay of 140us with a repeatability of 2us from the time the probe actually operates to the MI 12 interface indicating a probe change of state. Enabling the enhanced trigger circuit will add a further nominal 7 milliseconds"
[16:57:25] <jepler> SW2-1 ON enables the enhanced trigger circuit
[16:58:31] <cradek> it is MP700 but I think that might be made up of other parts - not sure
[17:00:52] <jepler> yeah, this isn't quite yours -- it uses 2 9V batteris
[17:00:55] <jepler> batteries
[17:27:21] <cradek> hm, it has 3 unmarked dip switches
[17:32:18] <Dave911> cradek: >>am I the only one who isn't bothered one bit if someone chooses to use different software than I do?
[17:32:20] <Dave911> I'm with you on that one. Use what you want and pay what you want... However, the thing that does bother me is when people (no one here) spread misinformation in order to distort the truth for their own gain.
[17:32:21] <Dave911> FYI - The guy who wrote that last quoted Mach3 message has "download EMC2 live CD" on his todo list today! :-) We had a long chat yesterday..
[17:34:33] <cradek> yeah the spread of misinformation sucks
[17:42:12] <Dave911> I wasn't trying to start a CNC war when I posted those message links yesterday, I just want you guys to know what was being said..
[17:43:31] <Dave911> On a different subject - do you guys have any idea where EMC2 is being used.. Do you log IP's during downloads etc. For instance, is EMC2 very popular in China? I know that India is big into Linux.
[17:58:12] <alex_joni> Dave911: there are some statistics available, but we don't do a proper screening
[19:44:39] <syn.> *** Disconnecting you (syn (Banned))