#emc-devel | Logs for 2010-10-11

[00:23:12] <CIA-2> EMC: 03cradek 07master * r477a5f098e79 10/src/emc/usr_intf/touchy/mdi.py: remove debug output
[00:23:13] <CIA-2> EMC: 03cradek 07master * ra31f034a5cb3 10/src/emc/usr_intf/touchy/ (mdi.py touchy.glade touchy.py): touchy: add preference for tool setting method
[00:23:14] <CIA-2> EMC: 03cradek 07master * rc1688f186586 10/src/emc/usr_intf/touchy/touchy.py: touchy: save maxvel setting across runs
[01:08:43] <jepler> cradek: until such a time as a brilliant idea comes up, a checkbutton in a menu should suffice
[01:12:02] <cradek> yeah I did the obvious for touchy...
[01:30:20] <cradek> maybe your upcoming inspiration for the new AXIS touch off dialog will include some insight into a good way to do it there.
[02:03:36] <jepler> my impression was that it's a "set once and forget" kind of thing, not something that goes in the touch-off dialog. Am I wrong?
[03:10:40] <cradek> jepler: I'm not sure. I am pretty sure I will want to use both, because it's fairly frequent that none of the table is reachable (when covered with something like a large plate)
[03:11:01] <cradek> I bet on a lot of machines, especially with bigger tables, people will never need to change it
[03:14:29] <cradek> I think it doesn't have to be on the main touch of dialog (which is already plenty complicated) but it seems like it shouldn't require a restart of emc to change (like for an ini entry)
[11:49:41] <jthornton_> jthornton_ is now known as jthornton
[12:47:34] <CIA-2> EMC: 03jepler 07master * r2dfdae1ea3bd 10/ (2 files in 2 dirs): axis: add choice of g10/g11 tool touch off
[12:47:37] <CIA-2> EMC: 03jepler 07master * rfae7d1eb6d1c 10/share/axis/tcl/axis.tcl: debugging tool to show letters duplicated and available in a menu
[13:01:38] <cradek> sweet!
[13:08:20] <SteveStallings> SteveStallings is now known as steves_logging
[13:50:11] <CIA-2> EMC: 03cradek 07master * reef7aed2d670 10/src/emc/usr_intf/touchy/touchy.py: Make embed tabs look like the others
[13:50:55] <cradek> huh, there's a key for clearing mdi history
[13:51:59] <cradek> jepler: should those radiobuttons have a separator above them?
[13:53:13] <cradek> skunkworks: when are you going to have first chips?
[13:54:05] <skunkworks> I don't know - very soon. maybe next weekend? we need to get all the way covers back on also
[13:55:18] <skunkworks> have to alteast make an 'emc2-axis' ... ;)
[13:55:41] <cradek> yeah!
[14:00:33] <skunkworks> that reminds me - I have to call mesa and order a 7I48 - we are offically out of encoder counters and +/-10v
[14:02:41] <cradek> it's sure nice to have lots of those
[14:02:47] <cradek> will you replace the 7i33 or keep it too?
[14:04:17] <skunkworks> planning on one 7i33 and one 7i48
[14:04:42] <skunkworks> that will give us 10 - that should be enough for a while.
[14:04:47] <skunkworks> ;)
[14:04:47] <jepler> the 7i48 is the one that andy added support for?
[14:05:11] <skunkworks> I don't kno who added it - I thought cradek needed/added it.
[14:05:15] <cradek> no - you wrote the patch and I updated and pushed it a while back
[14:05:28] <jepler> oh, yay me then.
[14:05:32] <skunkworks> heh
[14:05:47] <skunkworks> again - I have to run trunk anyways... :)
[14:08:41] <skunkworks> after the tool changer and gearshifting - dad said 'and emc is free?'
[14:10:42] <cradek> weird - emc starts with a file loaded, created somehow in /tmp, that contains just two percents
[14:10:51] <cradek> I don't remember seeing that
[14:11:19] <skunkworks> I hope dad can get the oiling figured out this week.
[14:11:36] <cradek> you make him work all week long?
[14:11:36] <skunkworks> *a solution to the y oiling.
[14:11:46] <skunkworks> he is retired. yes
[14:12:01] <cradek> skunkworks:
[14:12:01] <cradek> http://timeguy.com/cradek-files/emc/2010-10-11_07-43-51_23.jpg
[14:12:02] <cradek> http://timeguy.com/cradek-files/emc/2010-10-11_07-44-03_831.jpg
[14:12:02] <cradek> http://timeguy.com/cradek-files/emc/2010-10-11_07-44-19_652.jpg
[14:15:18] <skunkworks> cradek: very nice!
[14:16:17] <cradek> skunkworks: I was lucky to have the nice belt box with bearings that I could adapt
[14:17:33] <cradek> I had to elongate the worm's shaft to fit through the pulley all the way
[14:18:00] <cradek> I made it a little long so I can have an exposed setscrew for the resolver there (there is a hole in it, behind the overly long key in the photo)
[14:21:52] <skunkworks> isn't it fun?
[14:26:57] <cradek> rewarding yes - lots of hours in it though, and I'm not near done yet
[14:31:35] <skunkworks> look at us - it has been a few years :)
[14:39:03] <alex_joni> skunkworks: heh @ 'and emc is free?'
[14:43:48] <skunkworks> yes - He is very impress with the flexibillity of it.
[14:44:00] <skunkworks> logger_dev: bookmark
[14:44:00] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2010-10-11.txt
[14:56:20] <skunkworks> alex_joni: did you see the gearshift video?
[14:59:31] <jepler> skunkworks: I didn't
[15:05:32] <cradek> active_codes.append("G%d.%d" % (i/10, 1%10))
[15:06:04] <cradek> anyone spot the bug? took me embarrassingly long to find it.
[15:09:14] <CIA-2> EMC: 03cradek 07v2.4_branch * ra427ac719b4e 10/src/emc/usr_intf/touchy/emc_interface.py: touchy: fix three-digit active gcodes being displayed wrong
[15:11:13] <skunkworks> jepler: http://www.youtube.com/watch?v=22dWg3GbywE
[15:11:50] <skunkworks> you can see the led's on the black solid state relays count up in binary. (bottom 4 of the top 5 relays)
[15:19:26] <alex_joni> skunkworks: yeah
[15:28:48] <jepler> skunkworks: worst techno music ever
[15:29:23] <jepler> skunkworks: is every other gear reversed, or is that just part of the way you're testing it?
[15:29:50] <skunkworks> I was testing it that way (m3/4 every other gear)
[15:30:45] <skunkworks> the background noise initally was wisconsin public radfio.
[15:30:48] <skunkworks> radio
[15:33:53] <skunkworks> jepler: I could have worked around the bug by doing a if(gear & 2 == 1) sol = true; ?
[15:34:06] <alex_joni> skunkworks: yeah
[15:34:21] <jepler> skunkworks: not quite
[15:34:26] <cradek> sol = !!(gear & 2)
[15:34:54] <jepler> x & 2 is either 0 or 2, so x & 2 == 1 will never be true
[15:34:55] <cradek> sol = (gear & 2)? 1: 0
[15:36:01] <jepler> > C++ has a separate Boolean data type ('bool'), but with automatic conversions from scalar and pointer values that are very similar to those of C. This approach was adopted also by many later languages, especially by some scripting ones such as AWK and Perl
[15:36:10] <jepler> wtf wikipedia -- awk is later than C++?
[15:36:24] <cradek> hah
[15:36:29] <jepler> > As of the 1999 standard, SQL specified a Boolean data type with four possible values: true, false, unknown or null
[15:36:39] <cradek> hahaha
[15:36:42] <cradek> keep 'em coming
[15:37:10] <jepler> > In Eastern Outer Mongolia, age is traditionally determined based on the number of full moons since conception for girls, and the number of new moons since birth for boys.
[15:37:17] <alex_joni> true, false, maybe, not sure
[15:37:35] <micges> jepler: irc channel emc version is outdated
[15:37:43] <alex_joni> micges: and webpage too
[15:38:01] <jepler> jepler has changed the topic to: EMC2 development -- http://linuxcnc.org/ | Latest release: EMC 2.4.5
[15:38:01] <alex_joni> http://juve.ro/blog/projects/01286810746
[15:38:14] <micges> jepler: cool
[15:38:34] <micges> wiki is updated
[15:38:48] <cradek> thanks, release guys!
[15:39:19] <cradek> alex_joni: cute
[15:41:15] <skunkworks> alex_joni: neat :)
[15:43:35] <skunkworks> oh - so my understanding of what returned was off. duh. I had assumed that 1 was returned. duh.
[15:44:31] <skunkworks> so if I would have done if(gear & 2 == 1) sol = true;
[15:44:37] <skunkworks> so if I would have done if(gear & 2 == 2) sol = true;
[15:44:54] <skunkworks> it would have probably worked correctly.
[15:45:34] <skunkworks> It is suprising I get anything to work :)
[15:48:24] <alex_joni> given there are soo many hardware and software components in a PC, I am most of the times surprised it does work
[15:48:30] <alex_joni> probably the same chance factor ;)
[16:03:26] <jepler> else sol = false;
[16:08:48] <skunkworks> * right :)
[16:21:57] <jepler> personally I'd rather see you leave your code as it was when you discovered the problem, and fix hm2
[16:22:17] <jepler> because it's perfectly legitimate code you wrote
[16:26:15] <skunkworks> Cool
[17:47:50] <roberth_> on pncconf seems in metric mode you can not have more than 1000mm/min and 1000mm/sec^2 on axis setup screens, also when u input encoder pp/rev seems limit of 5000
[17:51:23] <jepler> roberth_: the primary pncconf developer doesn't frequently use irc, so you may have better luck using the mailing list or bug tracker to report this.
[17:57:58] <roberth_> ok no problem thx
[18:18:27] <psha> cradek: here?
[18:18:32] <cradek> hey
[18:18:39] <psha> good evening
[18:18:46] <cradek> good afternoon
[18:18:57] <psha> please merge touchy-embed branch of git://psha.org.ru/psha/emc2.git
[18:19:05] <psha> there is patch for killing child processes
[18:19:12] <psha> to ensure clean exit state
[18:19:34] <cradek> ok, thanks for reminder
[18:26:14] <cradek> thanks, pushed, but I guess CIA is slow today
[18:28:24] <CIA-2> EMC: 03cradek 07master * r0c843c14c043 10/share/axis/tcl/axis.tcl: Add separator to set off the radiobuttons
[18:28:24] <CIA-2> EMC: 03cradek 07master * r376a676d6eb0 10/src/emc/usr_intf/touchy/touchy.py: touchy: Kill spawned child processes
[18:37:38] <jepler> one hassle with push.default=tracking is that tags aren't pushed by deafult
[18:37:43] <jepler> I think I remembered to push the v2.4.5 tag, though
[18:38:01] <cradek> yes I have it
[18:40:04] <jepler> oh good
[19:01:30] <cradek> /usr/local/src/emc2.trunk/bin/touchy:80: GtkWarning: _gdk_drawable_get_source_drawable: assertion `GDK_IS_DRAWABLE (drawable)' failed
[19:01:36] <cradek> pix = gtk.gdk.pixmap_create_from_data(None, pix_data, 1, 1, 1, color, color)
[19:01:50] <cradek> on lucid, I get this, and some icons are missing from buttons. does anyone have any clues about it?
[19:03:01] <cradek> http://www.developpez.net/forums/d982366/general-developpement/programmation-linux/gtk/gtkp/gdk_pixmap_create_from_data-ubuntu-10-0-4-a/
[19:03:23] <cradek> wonder if the answer is in here
[19:04:23] <jepler> https://bugzilla.gnome.org/show_bug.cgi?id=631332
[19:04:30] <skunkworks> http://translate.google.com/translate?u=http%3A%2F%2Fwww.developpez.net%2Fforums%2Fd982366%2Fgeneral-developpement%2Fprogrammation-linux%2Fgtk%2Fgtkp%2Fgdk_pixmap_create_from_data-ubuntu-10-0-4-a%2F&sl=fr&tl=en&hl=&ie=UTF-8
[19:05:04] <jepler> I don't see that he says he doesn't get images, though
[19:05:19] <cradek> I am only guessing the two are related
[19:05:46] <cradek> oh -- not related
[19:05:49] <cradek> I didn't look at the code, duh
[19:06:00] <cradek> this is your invisible cursor hack
[19:17:51] <cradek> answer: set /desktop/gnome/interface/buttons_have_icons = true
[19:21:10] <skunkworks> well that was obvious... ;)
[19:30:50] <jepler> 14:05:20 <cradek> I am only guessing the two are related
[19:30:53] <jepler> ops
[19:33:26] <psha> aaaaa, op on channel
[19:33:31] <psha> * psha hiding under table
[19:41:26] <alex_joni> psha: better?
[19:41:37] <psha> :)