#emc-devel | Logs for 2008-11-21

[01:21:17] <CIA-38> EMC: 03jepler 07TRUNK * 10emc2/src/emc/rs274ngc/interp_read.cc: fix reporting of error with illegal character values above 127
[03:54:12] <cradek> sadly, all my matrox cards that gave me direct rendering without latency problems on dapper fail to do that same thing on hardy
[03:56:41] <LawrenceG> that sucks.... would be interesting to see the driver diffs to see what they broke
[04:00:36] <cradek> it's no big deal - direct rendering isn't needed for AXIS
[04:04:37] <CIA-38> EMC: 03seb 07TRUNK * 10emc2/src/libnml/nml/nml.cc: fix a segfault (robotic code review ftw)
[04:46:32] <CIA-38> EMC: 03seb 07TRUNK * 10emc2/src/hal/utils/halrmt.c: fix an array overrun
[04:51:17] <cradek> think I wrote that one...
[04:59:49] <CIA-38> EMC: 03seb 07TRUNK * 10emc2/src/libnml/cms/tcp_srv.cc:
[04:59:49] <CIA-38> EMC: Fix a segfault.
[04:59:49] <CIA-38> EMC: unregister_port() dereferences client_ports willy-nilly, so let's only
[04:59:49] <CIA-38> EMC: call it if client_ports is not NULL
[05:25:16] <CIA-38> EMC: 03seb 07TRUNK * 10emc2/src/libnml/buffer/memsem.cc: fix a segfault
[05:32:05] <CIA-38> EMC: 03seb 07TRUNK * 10emc2/src/hal/utils/halrmt.c: fix a memory leak
[05:40:54] <CIA-38> EMC: 03cmorley 07TRUNK * 10emc2/src/hal/classicladder/ (5 files): changes to put filename into shared memory so ladder remembers filename when GUI is reloaded-fixes SAVE button not working when GUI is loaded with axis
[05:42:21] <CIA-38> EMC: 03cmorley 07TRUNK * 10emc2/src/hal/classicladder/config_gtk.c: fix update of MODBUS request array
[05:44:48] <CIA-38> EMC: 03seb 07TRUNK * 10emc2/src/emc/ini/initraj.cc: reverse the order of the checking to avoid an access-past-end-of-array error
[05:51:31] <CIA-38> EMC: 03cmorley 07TRUNK * 10emc2/src/hal/classicladder/zEMC_README.txt: update
[07:23:39] <alex_joni> logger_dev: bookmark
[07:23:39] <alex_joni> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2008-11-21.txt
[11:51:17] <CIA-38> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/ladder/classic_ladder.lyx: remove note about save not working thanks for fixing this Chris
[16:42:52] <CIA-38> EMC: 03cradek 07TRUNK * 10emc2/debian/changelog: another new feature
[17:20:16] <BigJohnT> what does halui.max-velocity.count do?
[17:24:06] <SWPadnos> it lets you hook up an encoder to control the new velocity limit, I imagine
[17:24:48] <cradek> yes
[17:25:04] <cradek> halui stuff has 'counts' wherever a normal gui would have sliders
[17:25:16] <BigJohnT> ok
[17:25:20] <SWPadnos> does that have a corresponding enable input (like the various jog count inputs)?
[17:25:37] <SWPadnos> so you can share a single encoder among several .count inputs
[17:27:38] <CIA-38> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/gui/axis.lyx:
[17:27:38] <CIA-38> EMC: add
[17:27:38] <CIA-38> EMC: another new feature to the manual
[17:32:28] <cradek> SWPadnos: no, I think that is missing from all the halui inputs
[17:32:41] <SWPadnos> hmmm, ok
[17:32:51] <SWPadnos> right, encoder jogging was moved into motion
[17:32:55] <SWPadnos> which has them
[17:34:39] <BigJohnT> I assume without testing that increase and decrease are for a faster and slower button of some sort either real or pyVCP ...
[17:35:36] <cradek> yes
[17:35:47] <BigJohnT> ok
[17:36:16] <cradek> works just like the feed-override inputs
[17:36:27] <BigJohnT> thanks
[17:36:29] <cradek> increase/decrease by 'scale' value, and/or use counts
[17:36:48] <BigJohnT> I was just fixing to ask that one :)
[17:42:55] <jepler> I agree that someone should add enables to the halui encoder-type inputs
[17:43:00] <jepler> if that was the question
[17:45:36] <cradek> I also agree that someone should do that
[17:47:04] <CIA-38> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/gui/halui.lyx: add max vel info
[17:47:04] <BigJohnT> I third the motion
[17:47:29] <cradek> I nominate SWPadnos to do this, since he brought it up
[17:47:41] <SWPadnos> thanks, I accept
[17:47:41] <BigJohnT> I second the nomination
[17:47:48] <SWPadnos> there's no time limit though, right?
[17:47:59] <BigJohnT> before 2.3 comes out
[17:48:03] <cradek> it'll only take you a few minutes
[17:48:10] <cradek> (really, I bet so)
[17:48:10] <SWPadnos> ok, as long as we delay 2.3 untli I'm done ;)
[17:48:12] <cradek> really
[17:48:15] <cradek> trust me
[17:48:43] <BigJohnT> * BigJohnT goes to throw more wood on the fire and find his hip boots
[17:49:03] <SWPadnos> well, it would take someone else just a few minutes, but I'd have to make a macro-like thing for all of the "standard" kinds of things that halui uses, like toggles and encoders and stuff
[17:49:09] <SWPadnos> otherwise it would be hack-ish
[17:50:32] <SWPadnos> huh, now what's missing from these photos? http://cgi.ebay.com/HIGH-TORQUE-1125-OZ-IN-STEPPER-MOTOR-42-FRAME-STEP_W0QQitemZ260294804197QQihZ016QQcategoryZ41942QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[17:51:01] <cradek> hmm
[17:51:22] <SWPadnos> well it may be hidden, but it's hard to tell
[17:52:20] <SWPadnos> (here's a hint, how does the electricity get into the motor?)
[17:54:23] <jepler> it's like all laptop auctions on ebay: "think it works but I can't check because I don't have <this>"
[17:54:31] <SWPadnos> heh
[17:54:45] <SWPadnos> I'm thinking it may be a new motor for the, um, wireless age :)
[18:50:41] <skunkworks> we have a ton of steppers that have screw terminals under the back cover
[18:51:14] <SWPadnos> can you send me a few?
[18:51:35] <SWPadnos> cheap :)
[18:54:14] <skunkworks> heh - I think all we have are accounted for projects.
[18:56:00] <SWPadnos> dang
[18:56:11] <SWPadnos> I've got a G540 on the way, and I have no motors to run with it
[18:56:27] <SWPadnos> I don't want to spend a lot on motors just so I can make an EMC2 config for it :)
[18:56:49] <BigJohnT> $19 from automation direct
[18:56:56] <SWPadnos> yeah, tiny though
[18:57:11] <BigJohnT> yea, how big do you need?
[18:57:26] <SWPadnos> well, I'd like them to be potentially useful
[18:57:44] <SWPadnos> whatever that means
[18:58:07] <BigJohnT> LOL, did you purchase the 540 with a project in mind?
[18:58:22] <SWPadnos> no, Mariss is sending it to me on a 1E6 year loan :)
[18:58:51] <BigJohnT> what is 1E6?
[18:59:01] <SWPadnos> 1 * 10^6 (1 million)
[18:59:22] <BigJohnT> that is a long time
[18:59:40] <SWPadnos> yeah. I hope he doesn't mind if I forget by then :)
[19:00:11] <BigJohnT> he seems like a nice guy to deal with the time I spoke with him
[19:00:21] <SWPadnos> yep
[19:00:23] <BigJohnT> so nema 23 size?
[19:00:47] <SWPadnos> actually, NEMA34 could be directly attached to my Bridgeport for testing
[19:02:23] <SWPadnos> I guess the G251 specs limit the motor size a bit anyway, so maybe I will get some from AD
[19:03:06] <BigJohnT> the 23's are $29 and $39
[19:03:11] <SWPadnos> yep
[19:04:55] <BigJohnT> I use the $39 ones on my plasma table
[19:05:10] <SWPadnos> how do they perform?
[19:05:47] <BigJohnT> they do real will with the G203v
[19:06:12] <SWPadnos> are you around 50V?
[19:06:18] <BigJohnT> will/well
[19:06:26] <SWPadnos> oh, right :)
[19:06:45] <BigJohnT> no I'm at 60+ I forget exactly
[19:07:24] <SWPadnos> ah, ok. that's the 276 oz-in one, at 3.82 mH
[19:07:37] <SWPadnos> 62.54V is optimal
[19:07:56] <BigJohnT> I think it is the Antek ps-6n63rxx at 63 nominal
[19:07:57] <SWPadnos> (according to the 32*sqrt(L) formula)
[19:08:03] <SWPadnos> yep, should be about perfect
[19:08:19] <SWPadnos> a bit high for the G251 though, which tops out at 50V
[19:08:28] <BigJohnT> I had to look all that up when I bought it :)
[19:08:32] <SWPadnos> yep
[19:08:59] <SWPadnos> the 166 oz-in motor from AD is pretty ideal for the G251, but they're out of stock (and seem so small ;) )
[19:09:07] <BigJohnT> I ran them for a while on a 24v power supply here in the beer cave and got the results I expected from the lower voltage
[19:13:07] <skunkworks> I thought the 540 was only good for 3a or there abouts
[19:13:14] <SWPadnos> 3.5A, 50V
[19:13:18] <skunkworks> ah
[19:13:24] <skunkworks> It is a nice package.
[19:13:59] <BigJohnT> * BigJohnT heads for town
[19:14:12] <BigJohnT> talk to you guys later
[19:14:26] <SWPadnos> see you
[21:41:52] <CIA-38> EMC: 03cmorley 07v2_2_branch * 10emc2/src/hal/classicladder/ (5 files): change to put filename into shared memory so CL remembers filename when GUI closes-makes SAVE button works in AXIS. Also fix so CL will load a new ladder progra-if REopenned with a filename- instead of ignoring it
[21:59:54] <chester88> Hi guys ! small question. How does one set the level so that rtapi_print_msg are printed?
[22:00:50] <SWPadnos> scho level > /proc/rtapi/debug_level I think
[22:00:56] <SWPadnos> uh, echo
[22:01:07] <SWPadnos> check the proc dir for the correct name :)
[22:03:47] <chester88> no rtai in my proc directory
[22:04:54] <SWPadnos> it's only there when the RT system is loaded
[22:05:10] <SWPadnos> and it's RTAPI, not RTAI
[22:05:23] <SWPadnos> (lower case though)
[22:05:27] <chester88> sorry ya thats what I meant
[22:05:56] <chester88> Then how do I get debug messages to print when running simulator code?
[22:06:15] <SWPadnos> uh. good question
[22:06:21] <chester88> maybe I wasn't clearr about what I am doing..
[22:06:41] <SWPadnos> you can set debug level in an ini file, but I don't know for sure if that affects the "RT" debug level
[22:07:34] <chester88> in classicladder someone wrapped my lines like this: rtapi_print_msg(RTAPI_MSG_INFO, "Loading ladder GUI\n");
[22:07:43] <chester88> Now I want it to print
[22:08:15] <chester88> I read some where that I have to set a 'level' not exactly sure what was meant..
[22:08:38] <SWPadnos> well, for an RT system you'd use the /proc entry. I'm not sure how to do it for a sim system
[22:08:54] <chester88> Is there a level for EMC that is set when run?
[22:09:18] <SWPadnos> yes, that's DEBUG_LEVEL in the ini file, but I'm not sure if that affects the "RT" system
[22:09:26] <SWPadnos> EMC also has its own message levels
[22:10:21] <chester88> K I'm sure that is what is meant- how do you set them?
[22:10:34] <SWPadnos> edit the ini file
[22:10:50] <SWPadnos> but again, I don't know if that sets the RTAPI message level
[22:11:30] <chester88> right ok. There must be another way -I hope not using EMC just HAL
[22:11:57] <SWPadnos> well, lemme look around (and hope somebody who knows the answer shows up in the meantime :) )
[22:13:21] <chester88> lol ok Thanks. I checked man pages they don't even mention rtapi_print_msg
[22:17:48] <SWPadnos> huh. it looks like all messages are printed in sim, they just go to stdout vs. stderr
[22:17:53] <SWPadnos> depending on the level
[22:18:15] <SWPadnos> and anything other than RTAPI_MSG_ALL goes to stderr
[22:20:16] <CIA-38> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/gui/halui.lyx: a bit more for max vel
[22:20:26] <chester88> next dumb question how do I see stderr ? :)
[22:21:05] <SWPadnos> run in a terminal and you'll see both stderr and stdout
[22:21:18] <BigJohnT> hi chester88
[22:21:57] <chester88> hey John
[22:22:17] <Lerman_____> Lerman_____ is now known as Lerman
[22:22:41] <chester88> SWPadnos ok I didn't know they both were displayed . Thanks
[22:22:46] <SWPadnos> sure
[22:22:54] <SWPadnos> you can pipe them to different files if you like
[22:23:24] <SWPadnos> something like emc 2>stderr.txt
[22:23:45] <SWPadnos> that would send stderr to the text file, but still show stdout
[22:23:50] <SWPadnos> on the terminal
[22:23:51] <chester88> It doesn't display the messages though there must be more to it. I see I can run halrun with -v for more messages but that seems to fail...
[22:24:00] <SWPadnos> hmmm
[22:24:34] <SWPadnos> I may be wrong about it, but looking at the source leads me to believe that's how it's done (msg_all prints to stdout, all else to stderr)
[22:27:12] <chester88> I see in the man pages : int rtapi_set_msg_level(int level) so something has to set it to make it print...
[22:31:31] <SWPadnos> oh hmm. that's just one mesasge handler I looked at
[22:31:49] <SWPadnos> the level is set to some default, I'm just not sure what that default is
[22:38:13] <SWPadnos> of course, if you're already compiling for yourself, you could just change the default in src/rtapi/sim_common.h
[22:38:25] <SWPadnos> otherwise it
[22:38:41] <SWPadnos> it's probably something simple that I'm missing or it's impossible :)
[22:40:08] <chester88> ya I agree.. I will leave a message in the maillist ! thanks for your help though!
[22:40:51] <SWPadnos> sure
[22:41:02] <SWPadnos> are you compiling for yourself?
[22:41:27] <chester88> Yes I maintain Classicladder for the EMC group.
[22:41:36] <SWPadnos> oh, Chris?
[22:41:44] <chester88> yes !
[22:41:51] <SWPadnos> heh, didn't know that
[22:41:53] <chester88> i hide!
[22:42:10] <SWPadnos> smart
[22:42:18] <chester88> And try to hide my ignorance ...lol
[22:43:57] <SWPadnos> so change the default and recompile
[22:44:04] <SWPadnos> it should only take a couple of minutes
[22:44:38] <SWPadnos> FWIW, I don't see the message you pasted here, but I do see a couple of others from CL (specifically one whenever I close the GUI)
[22:44:50] <SWPadnos> though I should update and check again
[22:45:47] <chester88> Some messages are not wrapped so they will print no matter what.
[22:47:01] <SWPadnos> are you talking about the '\n's?
[22:48:26] <chester88> no
[22:48:52] <SWPadnos> ah, wrapped in rtapi_print_msg instead of stock printf?
[22:48:52] <chester88> some messages just use rtapi_print
[22:48:56] <SWPadnos> ok
[22:49:03] <chester88> yep
[22:57:16] <BigJohnT> I think Chris broke CVS with his last commit :)
[22:58:06] <chester88> uumm no it was John!
[23:00:43] <BigJohnT> I guess I was last
[23:35:27] <BigJohnT> hmm, cvs.linuxcnc.org must be broken
[23:43:30] <SWPadnos> indeed. and yet jepler is still attached (or so it seems)
[23:43:35] <jepler> BigJohnT: yes
[23:43:40] <jepler> vmware crashed
[23:43:45] <SWPadnos> bummer
[23:43:53] <jepler> I'll reboot the hosting machine in a few minutes
[23:44:01] <jepler> that should bring it back
[23:44:02] <BigJohnT> bummer
[23:44:08] <SWPadnos> I wonder if it noticed that there's a new version out
[23:44:47] <jepler> oh it hasn't been updated in years
[23:44:57] <SWPadnos> that's probably not it then :)
[23:45:08] <SWPadnos> (jus a coincidence that I got an update email a few minutes ago)
[23:45:41] <jepler> heh
[23:56:33] <jepler_> cvs should be back now
[23:56:46] <jepler_> jepler_: shoo
[23:58:57] <SWPadnos> huh. interesting dm_crypt question
[23:59:17] <SWPadnos> I wonder if that affects RT latency