Back
[11:31:41] <jthornton_> jthornton_ is now known as jthornton
[13:29:59] <jepler> -rw-r--r-- 0/0 2737686 2010-10-09 08:11 emc2-2.4.5/src/emc/usr_intf/pncconf/pncconf.glade
[13:30:02] <jepler> that's a HUGE file!
[13:30:45] <jepler> happily, it compresses REALLY WELL
[13:30:46] <jepler> $ gzip -9 < pncconf.glade | wc -c
[13:30:46] <jepler> 68681
[13:31:01] <jepler> (that's a 97.5% reduction in size)
[15:57:40] <skunkworks> is this normal?
http://pastebin.ca/1958195
[15:57:48] <skunkworks> It loads in hal.\
[15:58:29] <skunkworks> I was developing it in a sim install on my laptop - this is the first time I compiled it on the actual machine
[15:58:52] <skunkworks> (just didn't see those errors on my laptop.
[16:40:07] <jepler> skunkworks: yes, these warnings are normal: WARNING: "rtapi_snprintf" [/tmp/tmpV6koCg/gearshift16.ko] undefined!
[16:40:11] <jepler> the real test is whether the module loads
[16:40:46] <jepler> somebody who is not too tired should figure out how to fix it; it probably involves changes to the rtai and emc2 packaging to privde a list of symbols from each of them when building components
[16:43:31] <skunkworks> seems to load. But I have issues when I am setp'ing some values. but it is probably something I am doing. (componant loads and the function gets added)
[16:43:32] <skunkworks> Thanks
[16:44:05] <CIA-5> EMC: 03jepler 07v2.4_branch * rb5bb7beec1b3 10/ (VERSION debian/changelog): release v2.4.5
[16:47:02] <jepler> well that was only several weeks overdue :-/
[16:48:58] <jepler> researchers have discovered a startling new idea in computer design: make computers out of relays!
http://hardware.slashdot.org/story/10/10/09/155202/Electromechanical-Switches-Could-Reduce-Future-Computers-Cooling-Needs
[16:51:45] <morficmobile> jump to the side?
[17:21:09] <awallin> uhh, switching at 500 kHz... ??
[18:17:29] <cradek> cool, a release!
[18:17:32] <cradek> thanks jepler
[18:38:45] <skunkworks> ok - why would a pin say it is true - but not set the output on the mesa card? The gear comp sets the out put true but the output doesn't turn on. if I unlink the pin and set it true manually it works.
[18:39:29] <skunkworks> it is the 4 shift rails - the first one seems to work - but the last 3 doesn't work.
[18:40:21] <skunkworks> it shows true in hal show. I am setting it by sasol = gear & 1
[18:41:14] <skunkworks> each rail then is = gear & 1 , = gear & 2, = gear & 4, = gear & 8
[18:41:28] <skunkworks> where gear is a number 1-16 that corisponds to the gear picked
[18:41:50] <skunkworks> the first rail works (sasol = gear & 1)
[18:42:11] <skunkworks> but like I say - in hal show - it shows it as true.
[18:42:21] <skunkworks> but the actual relay isn't turning on
[18:42:42] <skunkworks> unlink pin and I can set the output true manually with a setp ....
[18:44:49] <micges> wiki updated
[18:44:59] <micges> to 2.4.5
[18:49:01] <skunkworks> it seems to be an issue with with how I am 'anding' to get the bits. I can set them manually within the comp by manually setting the bits to true
[18:53:11] <archivist> AND do you want to OR
[18:54:54] <skunkworks> I want to take a 1-16 number and convert it to 4 bits
[18:55:08] <jepler> skunkworks: it looks like you've just discovered a bug in hostmot2 :-/
[18:55:13] <jepler> preparing a patch ..
[18:56:24] <jepler> skunkworks: try with this patch:
http://emergent.unpy.net/files/sandbox/0001-hostmot2-handle-a-bit-pin-with-a-value-other-than-0-.patch
[19:17:48] <psha> cradek: may you take a look at
http://psha.org.ru/git?p=psha/emc2.git;a=commitdiff;h=520b6ad2aaa6f256544c5b2ca5338a3320ba9c1a
[19:18:04] <psha> it adds killing of spawned child processes
[19:19:46] <skunkworks> jepler: thank you! that worked
[19:21:00] <skunkworks> If I am not good for much - atleast I am good at finding bugs :)
[19:22:18] <micges> most bug are discovered few hours after bugfix release ;)
[19:22:24] <micges> bugs*
[19:26:09] <cradek> jepler: weird lyx stuff in that patch
[19:27:17] <cradek> yuck, that nasty C feature has sure bitten us a lot of times
[19:28:03] <cradek> psha: I am not at the right computer(s) to apply that - but it looks good to me
[19:28:11] <cradek> psha: I will get it later if nobody else grabs it
[19:28:17] <psha> so it's in touchy-embed branch
[19:28:36] <psha> i'll add this to .axisrc too and update wiki page
[19:28:55] <micges> cradek: I agree, I just imagined my hours spended looking what's wrong
[19:32:59] <skunkworks> it applied no proglem
[19:33:23] <skunkworks> proglem? almost make sense
[19:34:59] <skunkworks> logger_dev: bookmark
[19:34:59] <skunkworks> Just this once .. here's the log:
http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2010-10-09.txt
[19:40:32] <skunkworks> spindle shifts!
[19:41:22] <cradek> yay!
[19:43:48] <skunkworks> we dumped a bit of oil out before we figured out that we left the old spindle speed magnetic pickup out.
[20:04:35] <skunkworks> does spindle at speed work with g4px also?
[20:05:13] <skunkworks> so if I have a spindle speed and then a delay - emc won't start the delay until the spindle is at speed?
[20:05:17] <skunkworks> heh
[20:05:23] <skunkworks> I guess I could test that
[20:07:39] <jepler> cradek: oops, ugh
[20:09:22] <jepler> patch updated at same url:
http://emergent.unpy.net/files/sandbox/0001-hostmot2-handle-a-bit-pin-with-a-value-other-than-0-.patch
[20:13:38] <skunkworks> jepler: should I apply that patch?
[20:13:43] <skunkworks> Or not worry about it?
[20:13:57] <skunkworks> at what point will That get into trunk?
[20:28:50] <jepler> skunkworks: probably not long
[20:28:59] <jepler> I wish the answer was "in v2.4_branch" and "about 24 hours ago" :-(
[20:41:34] <skunkworks> heh - I have to use head anyway. not an issue for me. :)
[20:41:45] <skunkworks> trunk, what ever it is
[20:43:14] <skunkworks> I had no clue what was happening. I could see that the pin was true - but the little light on the ssr wasn't on. ;)
[20:45:48] <psha> bb all
[20:49:21] <CIA-2> EMC: 03jepler 07v2.4_branch * r290bf1737b22 10/src/hal/drivers/mesa-hostmot2/ioport.c: hostmot2: handle a 'bit' pin with a value other than 0 or 1
[20:55:11] <jepler> skunkworks: it's a bug I've personally written several times, so I knew what to look for
[20:57:10] <skunkworks> heh - neat
[21:57:56] <tom3p> is psha pavel?
[22:14:21] <jepler> tom3p: yes
[22:16:15] <tom3p> thx