#emc-devel | Logs for 2010-10-14

[06:11:14] <CIA-2> EMC: 03cmorley 07v2.4_branch * rafdcdd2e1172 10/src/emc/usr_intf/pncconf/pncconf.glade: raise max values on spinboxes for metric machines
[14:53:16] <Als> Als is now known as AlSMT
[15:34:01] <psha> jepler: i've tested touchy with embeded urxvt
[15:34:12] <psha> keyboard focus is always on main window when using xfce4
[15:34:27] <psha> focus stealing occurs only when i'm using ion3
[15:59:18] <jepler> I haven't tested with touchy. with axis, the ubuntu 10.04 default window manager (metacity) steals focus. this will reduce the usefulness of this feature for most users--I suspect very few of them change the window manager from the default
[16:06:02] <psha> i'm slowly looking trhough tk/tkinter docs to implement solution from xembed spec
[16:06:16] <psha> at least gtk behaves fine
[16:06:32] <psha> so there is hope that tk may be fixed too...
[16:19:44] <qq-> psha, thanks very much for previews helps ;)
[16:30:10] <psha> qq-: i've looked at debian rtai package
[16:30:14] <psha> it's horrible...
[16:30:34] <psha> it's spamming in /usr/ with lot of subdirs
[16:30:43] <psha> testsuite, rtai-py, etc...
[16:30:56] <psha> so ubuntu one is clean one compared to debian :)
[16:33:28] <qq-> psha, yeah, but how to "implement" the right things in debian ?
[16:34:29] <qq-> bugger the upstream about it ?
[16:35:00] <psha> bug report (non-fhs compilant) + cc emc2-dev
[16:35:06] <qq-> beside* bugger
[16:35:20] <psha> upstream = debian here, not rtai
[16:35:41] <qq-> ah, thought rtai ..
[16:36:20] <psha> some parts of rtai-modules may be killed without any harm
[16:36:31] <psha> for example python bindings
[16:36:39] <psha> they are packaged in python-rtai
[16:36:50] <psha> userland is packaged in librtai
[16:36:50] <psha> etc
[16:37:39] <qq-> hmm, all that it's up to my head
[16:41:05] <psha> may you look at our rtai-modules package?
[16:41:13] <psha> and post list of files from it
[16:41:18] <psha> dpkg -L rtai-modules-...
[16:48:38] <qq-> psha, you means remove modules from rtai-modules-* which are provided by python-rtai ?
[16:49:50] <psha> yes
[16:50:01] <psha> leave only modules
[16:52:46] <psha> also config-rtai may be useful
[16:53:32] <psha> other stuff like testsuite, calibration and bin has to go into rtai-utils
[16:59:10] <qq-> psha, so changes have to be made in rtai-source , as i see
[16:59:24] <psha> i've not found where exactly still
[17:00:07] <qq-> rtai-modules-* is implemented by rtai-source
[17:01:28] <psha> have you looked in it? :)
[17:02:02] <qq-> when we build a rtai kernel it use rtai-source to build rtai-modules-*
[17:02:13] <qq-> yes
[17:03:07] <psha> so it needs another patch over rtai-source.tar.g
[17:03:08] <psha> z
[17:03:55] <psha> it's unpacked in /usr/src/modules/rtai and then processed by m-a(?) or somebody similar
[17:04:38] <psha> # don't include files already present in other packages!
[17:04:39] <psha> rm -f debian/$(pmodules)/usr/bin/rtai-config
[17:04:39] <psha> rm -rf debian/$(pmodules)/usr/lib
[17:04:39] <psha> rm -rf debian/$(pmodules)/usr/include
[17:04:48] <psha> /usr/src/modules/rtai/debian/rules
[17:05:12] <psha> at least python modules may be safely killed
[17:05:58] <qq-> now it rest to 'sync' rtai-source , python-rtai, and other rtai package (if it is involved in this process)
[17:06:30] <psha> all rtai packages are built from one source - 'rtai' source package
[17:08:17] <qq-> psha, usually yes, but not work yet , for now rtai-modules-* is made by make-kpkg , which 'looks' in /usr/src/ for some modules , and build them
[17:09:13] <psha> nonono
[17:09:26] <psha> rtai-source is binary package, not source package
[17:10:08] <psha> and rtai-source binary package is built from source package 'rtai'
[17:10:28] <psha> make-kpkg uses tar file from binary package
[17:10:39] <psha> and this archive is created in rtai source package
[17:11:13] <qq-> sure , said same thing , you better explained ;)
[17:11:54] <psha> so /usr/src/modules/rtai/debian/rules is equal to one from rtai source
[17:12:18] <psha> so it's needed only to patch that file to get correct results
[17:12:19] <qq-> yes
[17:15:45] <qq-> hm, not sure now ..
[17:16:36] <psha> what's wrong?
[17:18:02] <qq-> see, it seems to be the same rules used by all Debian , since '97 , so not good place to do our changes
[17:19:10] <qq-> * qq- dinner time
[17:19:17] <psha> old is not equal to correct :)
[17:20:20] <qq-> old is equal standard , can't break standards for 'lite' changes
[17:20:55] <psha> so i don't understand what you mean under 'same rules used by all Debian'
[17:21:04] <psha> debian/rules is like Makefile
[17:21:21] <psha> you place there commands needed to build your package
[17:21:53] <psha> there are common parts of debian rules split into cdbs (common debian build system) package
[17:22:11] <qq-> psha, maybe , we may try and see
[17:25:32] <psha> if you have enought patience i may prepare some patches for test
[17:25:46] <psha> it's evening here so i'm not able to test them today
[18:00:51] <CIA-2> EMC: 03cradek 07master * r4292d75f198c 10/src/emc/rs274ngc/interp_convert.cc: G92 offset should not move the tool touchoff fixture
[18:04:35] <jepler> cradek: yuck, I hope you caught that before the tool was ruined!
[18:05:25] <cradek> my meat-based simulator found the problem by itself
[18:08:01] <cradek> it works better on some days than others
[18:53:58] <qq-> psha, sure, no harry here , thanks
[20:10:02] <DJSprocket> is this the emc channel?
[20:10:17] <cpresser> DJSprocket: the emc-developer channel
[20:10:29] <DJSprocket> so this is where i would ask questions?
[20:10:31] <cpresser> main channel is #emc
[20:10:58] <DJSprocket> or would i have to go to #emc to ask questions
[20:12:02] <micges> goto #emc and ask question, there is many people in there
[20:12:16] <micges> better chance to get answer
[20:13:12] <DJSprocket> how do i join the other channel ?
[20:13:26] <DJSprocket> nevermind got it thans