#emc-devel | Logs for 2007-06-10

[00:22:43] <petev> cradek, u there?
[00:23:05] <petev> I'm seeing some strange behavior with task modes
[00:23:53] <petev> I can step a few blocks in auto mode, then switch to mdi and issue a command, then EMC takes off running the program like it's in auto mode, evne though it's still in MDI mode
[01:03:47] <petev> M6 also doens't seem to be dealing with pre-conditions, as it does not stop the spindle, does not generate an error if the spindle is running, and starts a tool change while the spindle is running
[01:33:25] <cradek> good catch, please fix
[01:33:49] <petev> if I was familiar with this code, I would
[01:33:57] <petev> first I wanted to make sure it wasn't just me seeing this
[01:34:12] <petev> do yuo see both issues?
[01:37:04] <jepler> petev: I am sure you're right about the spindle one even without testing
[01:37:16] <jepler> I have not seen the other one
[01:37:26] <jepler> do you think there's a pattern about the MDI you issue that causes the behavior?
[01:37:38] <petev> I was using tkemc, maybe it's an artifact of the way it does things?
[01:37:59] <petev> no, it's very repeatable, I don't think it has to do witht he actual g-code
[01:38:28] <petev> I can try it with axis and see what happens
[01:39:12] <petev> M6 should probably turn off coolan too, though I don't feel so strongly about this
[01:39:26] <jepler> petev: can you give me an exact sequence of keypresses for tkemc? I'll try to reproduce it now.
[01:39:36] <jepler> or a capture of the NML messages issued, that might be even better
[01:40:00] <petev> I loaded a program, stepped a few blocks which started the spindle, then switched to MDI mode
[01:40:17] <petev> after entering one MDI block, the program continued as in auto mode
[01:40:28] <petev> but tkemc still showed it was in MDI mode
[01:40:44] <jepler> so something like this? F1 F2 F4 o (choose spiral.ngc) step step step F4 "G4P1" RET
[01:40:48] <petev> I don't think it really matters what the g-code is
[01:40:59] <petev> yes
[01:41:40] <petev> there was also other strange behavior when opening a file while paused in auto mode, but I couldn't duplicate it reliably
[01:41:46] <jepler> er, F5 "G4P1"
[01:41:52] <petev> it would start running the g-code from some random line
[01:42:05] <jepler> I suspect that these operations (open file, change to mdi mode) should be rejected when auto mode with the program paused
[01:42:09] <petev> yeah, I don't use the F keys, so you know better than me
[01:42:43] <petev> most commercial controls will allow it, and come back to where you were in auto mode
[01:42:49] <jepler> yes, I duplicated the behavior
[01:43:06] <petev> switching to manual from auto in EMC turned off the spindle and seemed to close the auto mode completely
[01:43:29] <petev> do you see the same thing with axis?
[01:43:36] <jepler> emc doesn't work that way and I suspect a lot of work would be required to let it work that way
[01:43:38] <jepler> no, I was testing with tkemc.
[01:46:33] <jepler> I know I will not try to fix either of these bugs right away. If you don't plan to either, please file these as items on sourceforge
[01:47:01] <petev> ok, I'm still working on NML stuff, so I better file them
[01:48:28] <petev> do you think the auto/mdi mode issue is purely tkemc, or also the interp?
[01:49:11] <jepler> I would call it task
[01:49:21] <jepler> task should prevent an invalid mode switch, such as auto-with-program-paused to mdi
[01:49:42] <petev> so is it task or interp?
[01:49:49] <jepler> task
[01:49:59] <petev> and the m6 is interp?
[01:57:52] <jepler> petev: right
[01:58:51] <jepler> petev: I doubt that rs274ngc manual specifies that *any* of these things are turned off
[01:59:05] <jepler> petev: probably we should take care to add a specification of this to our manual at the same time
[01:59:20] <jepler> e.g., if we decide for sure to not change the state of coolant
[01:59:31] <petev> I logged the M6 against the interp, do you think it should have been something else?
[02:01:06] <petev> I remember reading some doc about pre-conditions many years ago, but I can't remember what/where it was
[02:13:10] <jmkasunich> petev: those bug reports look good
[02:13:16] <petev> ok
[02:14:57] <jepler> yeah thanks
[02:15:11] <petev> no problem, the least I can do
[02:15:13] <cradek> the spec says tool change stops the spindle. It does explicitly say that coolant doesn't turn off though.
[02:15:30] <petev> yeah, I can see coolant going either way
[02:23:15] <petev> bbl
[02:33:40] <jmkasunich> http://linuxcnc.org/compile_farm/festcam.shtml
[02:35:44] <jepler> petev: I have a 1-liner fix but I have other changes in my working file. want to test and check it in?
[02:36:24] <jepler> (for the spindle problem)
[02:37:19] <jepler> petev: http://pastebin.ca/554282
[03:03:29] <jepler> petev: I have a 1-liner fix but I have other changes in my working file. want to test and check it in?
[03:03:41] <jepler> (for the spindle problem)
[03:03:45] <jepler> petev: http://pastebin.ca/554282
[03:27:37] <jmk-solo> jmk-solo is now known as jmk2
[03:27:46] <petev> let me take a look
[03:30:30] <petev> what was this line doing?
[03:30:31] <petev> settings->spindle_turning = CANON_STOPPED;
[03:34:28] <petev> seems to be working, let me halscope the pins
[03:35:35] <petev> ok, looks good
[04:22:32] <jepler> petev: I'm not sure -- it may be a very old remnant
[04:22:45] <steves_logging> steves_logging is now known as steve_stallings
[04:22:53] <petev> it looks like some of the effect of the macros is not part of them
[04:23:16] <petev> I see that line after every macro call
[04:23:20] <petev> the same with the other macros
[04:26:56] <steve_stallings> Hi, can JMK, cradek, or jelper confirm that Americas Best Value Inn is the hotel where you are staying. I was printing maps and realized that it is on Public Sq in Knoxville about 3 miles Southeast of Galesburg.
[04:55:55] <cradek> yes that's where we are
[04:56:08] <cradek> it's on the little traffic circle (google maps mislocates it)
[04:59:14] <steve_stallings> Thanks, I've had my reservation for weeks and never looked at the map.
[05:02:13] <steve_stallings> Going to be a long, lonely drive. Matt Shaver cannot make it and I get to drive 17 hours myself. Expect to be there Tuesday late.
[05:02:26] <cradek> man that's too bad
[05:03:03] <cradek> take lots of breaks, sing along to the radio, etc - get here safely
[05:03:25] <steve_stallings> Yea, no Ray, no Smithy, no Matt, who shall I try to pick on?
[05:03:56] <steve_stallings> Yep, I do OK for 6 hours or so, then need a nap. I have a minivan with a mattress in the back.
[05:04:01] <cradek> darn, I hadn't heard the no smithy part yet
[05:04:12] <cradek> maybe we'll put you to work coding
[05:04:40] <steve_stallings> not sure you would like the results, does Ubuntu support Z-80 assembly code?
[05:05:04] <cradek> hey I brought the coffin - matt was going to take it back east for me - will you have room to take it?
[05:05:36] <steve_stallings> coffin? how big/heavy is it? for that matter, WHAT is it?
[05:06:10] <cradek> it's the nist lathe, it's heavy/big enough to not move far by yourself
[05:06:40] <steve_stallings> I thought that was a Sherline setup...
[05:06:59] <cradek> we'll see, it's no big deal either way
[05:07:07] <cradek> yes it is a sherline - the longer one
[05:07:13] <steve_stallings> I could switch to driving the truck if necessary.
[05:07:30] <cradek> no don't do that - if it fits, it fits
[05:08:14] <steve_stallings> Sherline, even on a base 3 feet long would fit.
[05:08:36] <cradek> it's 2x2x5 feet maybe
[05:08:44] <cradek> (I'm bad at estimating)
[05:08:53] <cradek> no it can't be 5
[05:09:00] <cradek> 4?
[05:09:11] <steve_stallings> either that is a big empty box, or you are really bad at guessing
[05:09:17] <cradek> or both
[05:09:32] <cradek> it's surprisingly big for a sherline, but not crazy big
[05:09:48] <steve_stallings> fully enclosed, AKA shielded from careless hands?
[05:10:11] <cradek> yes
[05:11:17] <cradek> after finding an 11" piece of paper, I'm guessing 20x12x4'
[05:12:02] <cradek> no sweat - like I said, if it fits, it fits
[05:12:04] <steve_stallings> anyway, if you can get it to Galesburg in a Subaru, I can get it to DC area
[05:12:20] <cradek> it fits easily in the full size station wagon
[05:12:28] <steve_stallings> OK
[05:12:29] <cradek> I really appreciate it
[05:13:24] <cradek> I hate to ship it again - the box was damaged a bit when he sent it to me
[05:14:11] <steve_stallings> shipping big things with some heavy stuff in the middle of lots of air is asking for trouble
[05:15:03] <cradek> the box is pretty good - I think the problem is when one person tries to move it - you have to drag one end on the ground
[05:15:34] <steve_stallings> Enco seems to specialize in this, only they think "peanuts" are rigid packing
[05:15:34] <cradek> that broke some of the box a bit, but it could be glued - the "parts" came with it
[05:16:12] <cradek> those are uselesspeanuts are useless
[05:16:30] <cradek> oops
[05:16:57] <steve_stallings> I have received double wall cardboard boxes with a Morse 4 shank poking out the side.
[05:17:23] <cradek> heh
[05:17:30] <cradek> hmm I better get to sleep!
[05:17:37] <steve_stallings> Peanuts do nothing to hold a 30 pound chuck.
[05:17:38] <cradek> safe trip, and see you tuesday
[05:17:46] <steve_stallings> Later, see you there.
[05:18:51] <steve_stallings> steve_stallings is now known as steves_logging
[05:19:26] <cradek> goodnight
[12:19:35] <cradek> good morning!
[12:21:21] <cradek> wow was I wrong to guess 9:00 as the time I'd be ready
[12:44:58] <alex_joni> hi guys
[12:45:29] <cradek> hi alex!
[12:46:36] <cradek> wireless is dodgy here...
[12:50:47] <alex_joni> that
[12:50:52] <alex_joni> that's not nice ;)
[12:51:14] <alex_joni> how was the trip? settled in ok?
[12:51:58] <cradek> yes just fine
[12:54:13] <cradek> the trip was nice enough - only about 450 miles
[13:05:41] <alex_joni> ah, that's quite close ;)
[13:05:50] <alex_joni> did that 4 times in the last 2 weeks
[13:06:21] <cradek> you travel a lot lately!
[13:12:04] <jepler> cradek: good morning
[13:12:24] <jepler> cradek: can you hear the trains on that side of the hotel?
[13:13:37] <cradek> yes it's pretty trainy
[13:13:55] <cradek> I'm probably more used to that than you are - I slept through the whole night easily
[13:14:14] <jepler> I didn't hear one after I went to sleep until 7:30 AM or so
[13:14:18] <cradek> is your wireless dodgy? mine's pretty bad
[13:14:22] <jepler> that's when I decided I wouldn't sleep anymore
[13:15:11] <jepler> it's OK right now, the latency is a bit long though.
[13:16:28] <cradek> I woke up at 6:30 like normal - 12:30 is a normal bedtime I guess
[13:17:26] <jepler> hm just had a hiccough
[13:17:46] <cradek> mine routinely stops for 30secs and then comes back
[13:20:38] <jepler> I notice mine was switching between access points, but setting it to just one hasn't helped
[13:20:53] <cradek> I also tried that
[13:21:35] <jepler> did you just have a cut-out?
[13:22:02] <cradek> I have them constantly
[13:22:17] <jepler> I am just wondering if we get the same ones, since we're probably on the same AP
[13:25:20] <cradek> I'll watch the time on the next one
[13:27:37] <jepler> 08:27:10
[13:27:51] <cradek> 8:27:08
[13:28:05] <cradek> hmm
[13:28:13] <cradek> it came back about when yours did too
[13:29:57] <jepler> that tells me it's not our individual machines
[13:29:57] <jepler> 8:28:50 or so
[13:30:40] <cradek> 8:30:15
[13:30:53] <cradek> to 8:30:45
[13:33:19] <cradek> also I'm not getting the usual "AP out of range/AP in range" so I don't think it's reception
[13:42:08] <cradek> jepler: are you ready?
[13:42:13] <cradek> jmk is ready
[13:42:32] <jepler> cradek: yes
[17:57:53] <alex_joni> how's it going?
[18:00:10] <jepler> alex_joni: not bad
[18:00:23] <jepler> john k is working on getting his stepper demo set up
[18:00:28] <jepler> chris has been cutting threads
[18:00:57] <jepler> I'm working on CSS and feed/rev for lathes, just started the interpreter so far
[18:01:24] <jmkasunich> got the stepper demo working
[18:06:39] <alex_joni> nice
[18:07:05] <alex_joni> let me know if there's anything I can help with
[18:57:19] <jepler> alex_joni: I'll see if I can't figure out how to partition the work ...
[19:02:46] <alex_joni> jepler: great
[19:02:56] <alex_joni> btw.. hi.. you just appeared on cam :P
[19:02:57] <cradek> hi alex
[19:03:15] <alex_joni> hi chris ;)
[19:08:09] <LawrenceG> Hi Guys... is that jepler and cradek on the fest cam???
[19:08:38] <cradek> yep
[19:09:05] <LawrenceG> cool.... web page is showing "2006" workshop...
[19:09:37] <LawrenceG> I wish you guys werent so far away.... would be great to join you
[19:10:21] <alex_joni> LawrenceG: you're talking about far away?
[19:10:58] <LawrenceG> google shows about 41 hours drive and about a 60km swim
[19:11:58] <LawrenceG> alex... your swim is a little more challenging
[19:12:24] <alex_joni> probably :P
[19:12:52] <alex_joni> Drive:9,447 km (about 30 days 14 hours)
[19:13:14] <alex_joni> Swim across the Atlantic Ocean
[19:13:17] <alex_joni> Entering United States (Massachusetts)5,572 km
[19:13:29] <alex_joni> that's probably a hard swim
[19:13:42] <LawrenceG> :}
[19:14:43] <LawrenceG> fest cam lighting is much better than I remember from last year
[19:20:46] <alex_joni> who else is already there?
[19:21:08] <cradek> for emc folks, just us three so far
[19:21:18] <alex_joni> shouldn't JonE be already around?
[19:21:41] <cradek> no, I think he's coming later, but I don't remember exactly what he said
[19:21:46] <cradek> I think swp should be here soon
[19:22:12] <alex_joni> not quite that big of a crowd
[19:22:29] <jmkasunich> it officially doesn't start till tomorrow
[19:22:51] <alex_joni> hey jmk, don't see you in that pic
[19:23:11] <cradek> he's underneath the camera
[19:23:16] <jmkasunich> hiding
[19:23:16] <alex_joni> ahh, ok
[19:23:18] <petev> what's the URL for the web cam?
[19:23:23] <alex_joni> http://linuxcnc.org/compile_farm/festcam.shtml\
[19:23:26] <jmkasunich> every oncei in a while you see the top of my head
[19:23:31] <jmkasunich> and my bald spot
[19:23:32] <alex_joni> without the trailing \
[19:23:39] <alex_joni> jmkasunich: see it now
[19:24:26] <jepler> petev: in the nist gcode document I see that for G84 and G86 it is explicitly mentioned that "The spindle must be turning clockwise before this cycle is used"; I believe the document allows other canned cycles such as G81 with the spindle not commanded to turn.
[19:25:01] <petev> hmm, that seems strange, they are drill cycles
[19:44:59] <alex_joni> chris is having a duh moment :P
[19:45:21] <cradek> who me?
[19:45:32] <alex_joni> it looks like that in the picture
[19:45:40] <alex_joni> (last picture.. just updated)
[19:45:48] <jmkasunich> he's pondering important things
[19:46:05] <alex_joni> oh, really?
[19:46:29] <cradek> I think I need to move next to jmk
[19:46:29] <jmkasunich> probably what to have for dinner
[19:46:39] <alex_joni> jmkasunich: lol, I was guessing the same too
[19:46:53] <alex_joni> cradek: that, or drag jmk into the picture :P
[19:47:20] <jmkasunich> it the camera was lower on the wall, I'd be all you could see
[19:47:24] <jmkasunich> consider yourself lucky
[19:47:47] <alex_joni> heh, ok :P
[19:49:11] <alex_joni> so.. any plan for this week yet?
[19:57:25] <steves_logging> steves_logging is now known as steve_stallings
[19:58:45] <steve_stallings> Are you guys still hoping to work on probing? I found my probe, but am coming up empty handed in my search for the interface box.
[19:59:19] <jepler> chris says we don't need the interface
[19:59:26] <jepler> so bring whatever you found
[19:59:33] <cradek> is it just a contact?
[19:59:45] <cradek> we can hook it to anything
[20:00:28] <steve_stallings> Probe does not work well without the "special" Renishaw box, and excess current might pit the contacts.
[20:01:05] <cradek> hmmm
[20:01:35] <cradek> I was thinking of 5v to parport - not sure if that's what you mean by excess or not
[20:02:29] <steve_stallings> I'll keep looking for a bit. Was the plan to work on surface scans, or workpiece locating.
[20:03:21] <cradek> yeah I think improving surface scans would be great
[20:03:44] <cradek> locating/straight probe currently works correctly - I fixed the "last" few bugs
[20:03:57] <steve_stallings> Renishaw setup senses with less than 1 mA on the probe using controlled current source. I do have another DEA probe that is less valuable but I don't have a connector that mates. Maybe some steady hands on a soldering iron.
[20:05:09] <cradek> I think you shouldn't bring anything really special/valuable
[20:05:18] <cradek> I wouldn't want to test new code with it
[20:12:31] <steve_stallings> Sorry, phone delay... I understand the risk in debugging code and would try to rig precautions such as using a soda straw to the tip extension.
[20:15:07] <cradek> that's a good idea
[20:15:45] <cradek> for the initial work, we could also use simulation - make a virtual object for the virtual probe to sense
[21:15:58] <jmkasunich> damn webcam keeps stopping every so often...
[21:17:13] <jmkasunich> but you can see by the sudden influx of "stuff" that SWP has arrived
[21:19:41] <alex_joni> oh, say hi to SWP
[21:21:15] <jmkasunich> I think he brought about 40 lbs of food
[21:21:26] <jmkasunich> if "food" is spelled "snacks"
[21:25:43] <alex_joni> heh
[21:50:56] <alex_joni> good night guys
[21:51:07] <alex_joni> have a fun and productive week
[22:08:32] <steve_stallings> OK, I found the probe interface and by gosh, it all seems to work!
[22:11:13] <jepler> steve_stallings: great
[22:13:07] <steve_stallings> Now all I have to do is find 200 other things, pack, and drive 17 hours...
[22:26:16] <jepler> see you in 17 hours!
[22:35:30] <steve_stallings> steve_stallings is now known as steves_logging
[23:04:55] <DanielFalck> jepler: is that you in front of the camera?