#emc-devel | Logs for 2007-11-20

[05:09:09] <Unit41> hahaha niiiice
[05:09:20] <Unit41> emc works flawlessly
[13:45:01] <CIA-19> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: improve legibility of running line in text area
[13:45:47] <CIA-19> 03jepler 07v2_2_branch * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: backport: improve legibility of running line in text area
[13:46:29] <CIA-19> 03jepler 07v2_2_branch * 10emc2/debian/changelog: note improvement
[14:10:25] <jepler> tissf: will you backport gcode_fr.html to version 2.2? It looks like it's done ..
[14:10:34] <jepler> tissf: it would be nice to have it in the next release, which I may make this weekend
[16:58:50] <tissf> jepler: yes I commit now
[16:59:31] <CIA-19> 03tissf 07v2_2_branch * 10emc2/docs/html/gcode_fr.html: French translation
[16:59:31] <CIA-19> 03tissf 07v2_2_branch * 10emc2/docs/src/ (index.tmpl index_fr.tmpl): French translation
[17:11:38] <CIA-19> 03tissf 07TRUNK * 10emc2/docs/html/gcode_fr.html: French translation update
[17:11:46] <CIA-19> 03tissf 07v2_2_branch * 10emc2/docs/html/gcode_fr.html: French translation update
[19:32:31] <jepler> tissf: thanks you
[19:32:33] <jepler> er, thank you
[19:34:42] <cradek> http://www.linuxcnc.org/docview/html//gcode_main.html#sub:G43,-G49:-Tool
[19:35:05] <cradek> I think we have a bug with TLO, but I want someone to say they agree with me before I change it
[19:35:34] <cradek> to see what I mean:
[19:35:37] <cradek> g49
[19:35:40] <cradek> g0x0y0z0
[19:35:49] <cradek> g43.1k1 (add one inch tool length)
[19:35:53] <cradek> g0x1
[19:36:11] <jepler> this issue aside, I think I see a doc bug -- "motion is commanded on the same line as G43.1" should probably apply to G43 too
[19:36:23] <cradek> this leaves the tool uncompensated. it only comps if you move Z.
[19:36:39] <cradek> but the spec says 'at the end of a subsequent movement'
[19:37:01] <cradek> jepler: actually with G43 it's allowed I think
[19:37:03] <cradek> let me try it
[19:37:27] <cradek> yes it's allowed
[19:37:30] <jepler> oh is the additional requirement because the presence of I- K- give an error when there's a non-arc movement, and do something wrong if it is an arc?
[19:37:33] <jepler> nevermind
[19:37:38] <cradek> yes exactly
[19:37:49] <cradek> (I don't like it but I don't see any way around it)
[19:38:48] <jepler> so after the g0x1 you want the tool tip to rise to z0 instead of remaining at z-1
[19:39:07] <cradek> jepler: yes that's what I want and I'm > 50% sure that's what the docs say should happen
[19:39:27] <jepler> I am wondering if there's a weasel pronoun there: "a subsequent motion" vs "the subsequent motion"
[19:39:44] <cradek> sure is
[19:40:21] <jepler> hm, that description is *not* a cut&paste from the nist document. I figured it was. http://www.isd.mel.nist.gov/personnel/kramer/pubs/RS274NGC_3.web/RS274NGC_33a.html#1013768
[19:40:43] <cradek> I was just reading there too
[19:40:51] <cradek> ngc says nothing about how it works
[19:42:50] <jepler> if a tool length is negative, the change you propose could cause a plunge where there was none before
[19:43:22] <cradek> yes
[19:48:03] <jepler> I wonder where I got that phrase from, "a subsequent motion"
[19:48:14] <cradek> oh you wrote it?
[19:48:19] <jepler> apparently
[19:48:38] <jepler> and it seems to have been part of the initial text for g43, not an addition later on
[19:49:10] <cradek> well it doesn't matter, it only means I can't ask the docs how it's supposed to work
[19:50:57] <CIA-19> 03jepler 07TRUNK * 10emc2/debian/emc2.files.in: packaging fixes: new files
[19:51:21] <CIA-19> 03jepler 07TRUNK * 10emc2/debian/changelog: debian/rules binary requires that the changelog be properly formatted
[19:52:03] <jepler> so you envision a potential problem if the tool length is positive but the movement G0X1 doesn't retract Z?
[19:52:50] <cradek> it bothers me that you can have arbitrarily many moves after G43 without the compensation you asked for
[19:53:21] <cradek> I bet since tool changes often happen "up", people aren't going to see any difference.
[19:56:55] <cradek> I'm running into trouble when canceling tool offset (G49), tool plunges into the part.
[19:57:04] <cradek> (fanuc 16i)
[19:58:40] <cradek> I'm not finding much.
[20:14:27] <CIA-19> 03tissf 07v2_2_branch * 10emc2/docs/html/gcode_fr.html: French translation update
[20:15:01] <CIA-19> 03tissf 07TRUNK * 10emc2/docs/html/gcode_fr.html: French translation update
[20:57:09] <CIA-19> 03jepler 07v2_2_branch * 10emc2/debian/ (changelog emc2.files.in): include new documentation in package
[21:15:05] <jepler> I looked in the irc archives and didn't find any discussion of g43 at around the time I committed the change
[21:15:22] <cradek> ok
[21:15:33] <cradek> we should change it to document the current behavior (which is looking more and more right to me)
[21:16:12] <jepler> you wanna help me find the words?
[21:18:35] <cradek> do not cause any motion. However, next time a compensated axis is moved, the offset will be applied so that move's endpoint is in the expected location.
[21:19:29] <jepler> "expected"?
[21:19:38] <jepler> well *I* expected it to move ...
[21:20:04] <cradek> in the offset location?
[21:21:40] <jepler> G43 and G43.1 change subsequent motions by offsetting the Z and/or X coordinates by the length of the tool. G43 and G43.1 do not cause any motion. The next time a compensated axis is moved, the move's endpoint is the compensated location.
[21:22:07] <cradek> looks good
[21:22:23] <jepler> OK
[21:22:53] <cradek> it's still not precisely right
[21:23:11] <cradek> g43.1 i1 k1 / g0 x1
[21:23:15] <jepler> "the move" vs "the axis"?
[21:23:19] <cradek> a "compensated axis [was] moved"
[21:23:25] <cradek> but the endpoint is half-compensated
[21:23:50] <jepler> ... The next time a compensated axis is moved, that axis's endpoint is the compensated location.
[21:24:00] <cradek> yes
[21:24:16] <cradek> (I have not tested on lathe)
[21:24:29] <jepler> but we agree that if it doesn't do that, it's a bug?
[21:24:39] <cradek> grnnnrrg
[21:24:41] <cradek> yes
[21:41:16] <jepler> argaraggargargr
[21:41:20] <jepler> date: 2007/11/16 20:32:24; author: fenn; state: Exp; lines: +2802 -2728
[21:41:20] <jepler> added named parameters section, cleaned up wording of numbered parameters.
[21:41:36] <cradek> offs
[21:41:49] <jepler> lyx2lyx seems to introduce spurious whitespace changes :(
[21:42:31] <cradek> someone trustworthy told me recently that lyx "isn't all that bad" but I still don't like it.
[21:42:39] <jepler> and then when you re-edit it with lyx that natively writes file version 221 (as I just did) those spurious whitespace changes get changed back
[21:43:37] <jepler> it does suck that not editing with the one annointed version of lyx causes these problems
[21:43:46] <jepler> maybe the files should just be marked -kb like gifs
[21:46:33] <jepler> argh the loggers have died again
[21:46:36] <jepler> er, cia I mean
[21:46:38] <jepler> * jepler kicks CIA-19
[21:46:38] <CIA-19> ow
[21:46:46] <jepler> CIA-19: don't you have something to say to me?
[21:47:04] <CIA-19> 03jepler 07TRUNK * 10emc2/docs/src/gcode/main.lyx: resave with lyx 1.3. no substantive difference
[21:47:43] <CIA-19> 03jepler 07TRUNK * 10emc2/docs/src/gcode/main.lyx: fix documentation of g43/g43.1 to match behavior
[22:35:29] <tissf> jepler: do not cause any motion. The effect of the compensation will occur only with the next movement of the compensated axis.
[22:35:31] <tissf> The reader know Z is offseted.
[22:35:33] <tissf> ??
[22:40:54] <cradek> tissf: on a lathe, X and Z may both be offset
[22:41:26] <tissf> yes X/Z
[22:42:35] <cradek> I don't understand: what is your question about the g43 docs?
[22:44:49] <tissf> I think the sentence added to earlier is unnecessary or poorly worded.
[22:45:20] <cradek> sorry I disagree
[22:46:05] <cradek> it says exactly when the offset is applied. The docs did not have that information before.
[22:46:15] <alex_joni> cradek: what's the command touch off uses ?
[22:46:20] <cradek> g10
[22:47:12] <alex_joni> do you remember the incantaion (not to look it up now..) say I want to set X0
[22:47:35] <cradek> g10 l2 p1 x0 removes any x offset in g54
[22:48:03] <alex_joni> so the current position will read as 0 ?
[22:48:08] <cradek> no
[22:48:21] <cradek> for that you put g10 l2 p1 x[current position]
[22:48:27] <cradek> or negative, I can never remember which
[22:48:36] <alex_joni> darn
[22:49:04] <alex_joni> I want to do it via halui pin
[22:49:23] <alex_joni> but [current position] is not an option :/
[22:49:33] <cradek> interesting
[22:49:49] <cradek> you should examine how axis gets what it needs (from the stat buffer somehow I'm sure)
[22:49:57] <cradek> seems like you should be able to do it
[22:49:58] <alex_joni> if it would have been g10 L2 p1 x0, it would have worked easily via MDI_COMMAND
[22:50:05] <cradek> oh right
[22:50:07] <alex_joni> without extending halui
[22:50:13] <cradek> yeah I wish it was that easy
[22:50:30] <alex_joni> I know how to do it myself.. but this is a config question from a 2.2.x user
[22:51:47] <alex_joni> maybe G92 would work
[22:51:53] <alex_joni> * alex_joni tries to understand the docs
[22:53:50] <alex_joni> hmm.. G92X0 seems to do the trick
[22:54:49] <tissf> 2. If G49 is programmed, USE_TOOL_LENGTH_OFFSET(0.0) is called, tool_length_offset is reset to 0.0 in the machine settings model, and the value of current_z in the model is adjusted. The effect of tool length compensation is illustrated in the screen shot below. Notice that the length of the tool is subtracted from the z setting so that the tool tip appears at the programmed setting. You...
[22:54:51] <tissf> cradek: sorry,
[22:54:53] <tissf> 1. If G43 Hn is programmed, A USE_TOOL_LENGTH_OFFSET(length) function call is made (where length is the value of the tool length offset entry in the tool table whose index is n), tool_length_offset is reset in the machine settings model, and the value of current_z in the model is adjusted. Note that n does not have to be the same as the slot number of the tool currently in the spindle.
[22:54:54] <tissf> ...should note that the effect of tool length compensation is immediate when you view the z position as a relative coordinate but it does affect actual machine position until you program a z move.
[22:55:31] <alex_joni> it doesn
[22:55:33] <alex_joni> 't
[22:55:35] <alex_joni> ??
[22:55:55] <alex_joni> crap, can't type.. shouldn't it be "it doesn't affect.." ?
[23:03:06] <tissf> good night all
[23:03:14] <alex_joni> good night tissf
[23:08:33] <CIA-19> 03alex_joni 07TRUNK * 10emc2/docs/src/gui/halui.lyx: document MDI_COMMAND usage
[23:13:55] <CIA-19> 03alex_joni 07v2_2_branch * 10emc2/docs/src/gui/halui.lyx: backport MDI_COMMAND documentation
[23:26:55] <alex_joni> good night all
[23:41:03] <cradek> alex_joni: if it still works right, I think halui will switch to MDI mode, send the command, wait for it to complete, and then switch back to AUTO mode