#emc-devel | Logs for 2007-11-07

[00:00:47] <eugenio> 555555555555555555555555555555555555555ooooooooooooooooooooooooooooooooooooooo333333333333333333333333333333333333333~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~yyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyyy
[00:01:43] <alex_joni> well.. I can see a pattern
[00:09:08] <tissf> hello, I have a small problem with make docs
[00:09:18] <alex_joni> tissf: ask away
[00:09:34] <alex_joni> (that means go ahead..)
[00:09:56] <tissf> Traceback (most recent call last):
[00:09:58] <tissf> File "../docs/src/lyxtree.py", line 724, in <module>
[00:10:02] <tissf> doc, outfile = parse(sys.argv[1:])
[00:10:01] <tissf> File "../docs/src/lyxtree.py", line 693, in parse
[00:10:03] <tissf> for line in infile: p.feed(line.strip('\n'))
[00:10:04] <tissf> File "/home/francis/emc2.TRUNK/docs/src/lyxparser.py", line 53, in feed
[00:10:06] <tissf> self.do_command(what, tokens[1:])
[00:10:09] <tissf> File "/home/francis/emc2.TRUNK/docs/src/lyxparser.py", line 79, in do_command
[00:10:09] <tissf> if f: f(tokens)
[00:10:11] <tissf> File "../docs/src/lyxtree.py", line 229, in do_layout
[00:10:12] <tissf> self.push('layout', class_=tokens[0])
[00:10:13] <tissf> IndexError: list index out of range
[00:10:15] <tissf> make: *** [objects/hal_tutorial_fr.xml] Erreur 1
[00:10:32] <tissf> too hard for me !
[00:10:57] <alex_joni> seems there's something wrong with a list index..
[00:11:07] <alex_joni> is this with code from CVS?
[00:11:12] <alex_joni> or did you add/change something?
[00:11:21] <alex_joni> (cvs diff -u might be helpfull)
[00:11:42] <tissf> not commited
[00:12:09] <alex_joni> tissf: can you pastebin.ca the output from cvs diff -u ?
[00:12:26] <tissf> yes
[00:12:29] <alex_joni> I don't think I can help with the lyxparser error, and jepler is not here right now
[00:12:56] <tissf> nut urge
[00:12:59] <alex_joni> you could also commit this.. and it'll get fixed sometimes
[00:13:00] <tissf> urgence
[00:14:21] <tissf> This is not urgent
[00:14:35] <tissf> i can wait for jeff
[00:14:46] <alex_joni> it might get late :P
[00:15:07] <alex_joni> but he'll still need the stuff on pastebin.ca when he gets here
[00:17:18] <tissf> cvs diff -u is short
[00:22:56] <CIA-18> 03alex_joni 07v2_2_branch * 10emc2/debian/ (changelog emc2.files.in):
[00:22:56] <CIA-18> package stepconf.glade (seems it was somehow forgotten before),
[00:22:56] <CIA-18> stepconf doesn't run without/
[00:22:57] <CIA-18> 03alex_joni 07v2_2_branch * 10emc2/src/Makefile:
[00:22:57] <CIA-18> package stepconf.glade (seems it was somehow forgotten before),
[00:22:57] <CIA-18> stepconf doesn't run without/
[00:24:20] <tissf> http://borderousse.no-ip.org/www/dobson/fichiers/output_cvs
[00:24:33] <tissf> you can take a look Alex ?
[00:28:00] <alex_joni> tissf: soon
[00:28:12] <tissf> ok
[00:28:26] <CIA-18> 03alex_joni 07TRUNK * 10emc2/debian/emc2.files.in: package stepconf.glade
[00:28:26] <CIA-18> 03alex_joni 07TRUNK * 10emc2/src/Makefile: package stepconf.glade
[00:29:00] <alex_joni> tissf: I also need hal_tutorial_fr.lyx
[00:29:08] <alex_joni> maybe you could check it in
[00:29:27] <alex_joni> (if you don't check in the other stuff, it surely won't bother anything)
[00:29:37] <tissf> with pleasure
[00:30:12] <alex_joni> :-)
[00:30:21] <alex_joni> cvs add hal_tutorial_fr.lyx
[00:30:25] <tissf> translation's finish, i have just verrifiing
[00:30:30] <alex_joni> cvs commit -m "" hal_tutorial_fr.lyx
[00:30:45] <alex_joni> then you're sure it's only that what you commit
[00:32:29] <tissf> Now I have not sent hal_tutorial_fr.lyx
[00:32:47] <tissf> not commited by me
[00:33:02] <tissf> i have just commited HAL intro
[00:33:24] <tissf> intro_fr.lyx
[00:35:28] <SWPadnos> I kind of thought we should have just named all the .var and .nml files the same
[00:38:41] <tissf> here the filename is tutorial_fr.lyx not hal_tutorial_fr.lyx
[00:39:16] <alex_joni> tissf: I took the name out of your cvs diff
[00:39:25] <alex_joni> maybe you called it differently in the Submakefile ?
[00:39:44] <tissf> ah ah !
[00:39:58] <alex_joni> sorry
[00:40:04] <alex_joni> no.. it's late, and I'm blind :P
[00:40:10] <alex_joni> +hal/tutorial_fr.lyx \
[00:40:31] <tissf> ouf
[00:44:12] <jepler> alex_joni: thanks for fixing the stepconf thing
[00:44:14] <jepler> tissf: here I am
[00:44:28] <alex_joni> yay
[00:44:45] <tissf> ok i can't found !
[00:44:58] <jepler> hal_tutorial_fr.lyx is probably the file I need to see
[00:45:32] <alex_joni> it's hal/tutorial_fr.lyx
[00:45:41] <jepler> oh it was committed? I didn't see that part.
[00:45:57] <tissf> yes but at this time it's full of typos
[00:46:10] <tissf> i commit
[00:46:24] <tissf> i have to add it
[00:46:32] <alex_joni> I didn't see the commit yet
[00:46:46] <alex_joni> tissf: a great thing about cvs is that you can have lots of versions :P
[00:46:48] <tissf> please wait,
[00:46:52] <jepler> I have a file from nov 5, is that the right one?
[00:47:14] <tissf> ?
[00:47:20] <tissf> goot question :)
[00:47:35] <alex_joni> tissf: cvs diff is your friend
[00:48:16] <tissf> done
[00:48:19] <CIA-18> 03tissf 07TRUNK * 10emc2/docs/src/ (Submakefile docs.xml index.tmpl index_fr.tmpl): French translation
[00:50:15] <jepler> did you "cvs add" the file ?
[00:50:17] <CIA-18> 03tissf 07TRUNK * 10emc2/docs/src/hal/tutorial_fr.lyx: French translation (partial)
[00:50:19] <jepler> ah there it is
[00:50:25] <alex_joni> yup
[00:50:30] <jepler> OK I get the same error
[00:50:50] <alex_joni> jepler: when you find 5 minutes please doublecheck the stepconf fix..
[00:51:02] <alex_joni> I did test the 2.2 one, didn't test the TRUNK ..
[00:51:20] <tissf> hello jepler
[00:55:25] <jepler> tissf: the error is occurring where you wrote "mise on oeuvre", with the special single character instead of "oe"
[00:55:47] <jepler> if it's not too offensive to the eyes of the reader, I'd suggest writing "oe" for now
[00:55:57] <tissf> ok i look that
[00:59:55] <jepler> U+00e6 230 LATIN SMALL LETTER AE "æ" Letter, lowercase
[00:59:55] <jepler> U+0153 339 LATIN SMALL LIGATURE OE "½" Letter, lowercase
[01:00:26] <tissf> jepler, fixed, thank you
[01:00:27] <jepler> hm, this is unrelated to the problem, but why is one of these called a "ligature" and the other not? (looks like the special characters themselves didn't survive the trip on IRC)
[01:00:55] <alex_joni> irssi & utf doesn't mix well
[01:00:57] <SWPadnos> strange - they came through correctly here
[01:01:01] <SWPadnos> but half the time, they don't for me
[01:01:06] <jepler> huh, my own irc client redisplayed oe as 1/2
[01:01:16] <SWPadnos> huh
[01:02:16] <alex_joni> I get a ? inside a circle
[01:02:37] <alex_joni> I get a black ? inside a white circle
[01:04:27] <tissf> I am very surprised because I never use that character !
[01:04:47] <tissf> I cant found it on my keyb
[01:04:59] <alex_joni> maybe lyx tried to be smart?
[01:05:04] <tissf> yes
[01:07:06] <tissf> jepler>huh, my own irc client redisplayed oe as 1/2, my irc client too
[01:08:15] <tissf> good night all, thank you for help :)
[01:08:24] <alex_joni> good night tissf
[01:08:37] <jepler> good night tissf
[01:08:51] <alex_joni> jepler: latency-test is nice :D
[01:09:08] <alex_joni> I get max-jitter 9157 for servo thread
[01:09:19] <alex_joni> and 11186 for base-thread
[01:09:29] <SWPadnos> if only there were a text mode version :)
[01:09:55] <jepler> ttyvcp
[01:09:59] <SWPadnos> heh
[01:09:59] <alex_joni> haha
[01:10:06] <alex_joni> libascii-vcp
[01:10:11] <SWPadnos> halttyvcpomg
[01:10:20] <SWPadnos> oops, I forgot the ponies
[01:10:24] <SWPadnos> !
[01:37:41] <jepler> haven't hooked it up yet, but your wish is my commmand: http://emergent.unpy.net/index.cgi-files/sandbox/latcurses.png http://emergent.unpy.net/index.cgi-files/sandbox/latency_tty.py
[01:37:47] <jepler> bbl
[01:38:34] <SWPadnos> cool
[07:08:33] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[07:53:17] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[08:03:10] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[08:11:19] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[08:23:54] <alex_joni> jepler: call it latency-text :D
[11:00:31] <tissf_> tissf_ is now known as tissf
[13:17:10] <CIA-18> 03tissf 07TRUNK * 10emc2/docs/src/hal/tutorial_fr.lyx: French translation
[13:17:08] <CIA-18> 03tissf 07TRUNK * 10emc2/docs/src/ (Master_HAL_fr.lyx Submakefile): French translation
[13:21:03] <tissf> Who wrote the intro of HAL?
[13:21:32] <cradek_> john k
[13:21:54] <tissf> ok thanks
[13:22:11] <cradek_> cradek_ is now known as cradek
[13:22:42] <jepler> hi chris, tissf
[13:23:02] <tissf> hi jepler
[13:23:00] <cradek> hi
[13:26:48] <jepler> tissf: it looks like you forgot to run lyx2lyx before your last checkin of Master_HAL_fr.lyx
[13:26:58] <skunkworks> The step/dir mesa config isn't done yet - right?
[13:27:08] <jepler> skunkworks: no, it's not
[13:27:12] <skunkworks> Ok - thanks
[13:27:23] <tissf> oops ! exact sorry
[13:27:41] <tissf> a moment please I correct
[13:29:14] <alex_joni> hi guys
[13:29:29] <alex_joni> cradek: I didn't manage to get it to error.. tried at least 20 times
[13:29:48] <alex_joni> with RT, without, with sim, with stepper, with low accels, with high accels, etc..
[13:32:21] <jepler> cradek: what's the size you expect the .tar.gz produced by dpkg-buildpackage to be?
[13:35:20] <CIA-18> 03tissf 07TRUNK * 10emc2/docs/src/Master_HAL_fr.lyx: French translation update
[13:36:24] <tissf> now i move to dapper !
[13:40:48] <alex_joni> jepler: I think he mentioned that the sim tar.gz contains RT objects inside
[13:41:39] <jepler> alex_joni: yeah that's why I'm asking
[13:45:24] <jepler> I guess I also need to rename the directory before packaging
[13:56:28] <jepler> * jepler wrestles with autoconf
[13:58:11] <skunkworks> * skunkworks roots for jepler.
[13:58:23] <jepler> no, I'm pretty sure I lose :(
[13:59:19] <jepler> I think I have to make VERSION be generated by src/configure..
[14:08:37] <jepler> or not have it at all
[14:08:50] <jepler> anybody know if we have documentation telling the user to look at the file VERSION in the source tree?
[14:09:26] <jepler> I didn't find any in src/docs ..
[14:09:59] <jepler> it's so stupid that configure requires the version to be hardcoded in this way
[14:56:57] <cradek> alex_joni: well crap. I wonder what's different. I get it regularly.
[14:57:18] <cradek> (it's stopping my probing work)
[14:57:44] <cradek> I guess I could assume it's not really broken, and remove the check (here)
[14:58:51] <cradek> jepler: looks like it's 6-7MB
[14:59:24] <jepler> cradek: OK
[14:59:35] <cradek> (I just removed the bogus stuff and retarred it)
[14:59:43] <skunkworks> logger_dev: bookmark
[14:59:43] <skunkworks> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2007-11-07.txt
[15:03:28] <cradek> alex_joni: http://pastebin.ca/764745
[15:05:43] <cradek> err no, I thought that made it fail every time, but it only failed the first time I tried it, now it doesn't
[15:07:48] <jepler> I'm thinking of making 2.2.1 to fix the stepconf issue.
[15:07:54] <jepler> should I just sign the whole repository?
[15:18:09] <cradek> I don't know
[15:30:50] <alex_joni> I don't think so..
[15:32:55] <cradek> jepler: if you put the files in place and rsync up, I can easily rsync down, sign, and rsync up
[15:33:37] <cradek> most people are going to use 2.1 for a while. I'd hate for them to all get errors from the update manager etc
[15:34:38] <alex_joni> right, my thoughts too
[15:35:27] <jepler> ok
[15:37:01] <jepler> this look like an OK commandline to rsync download? jepler@zenbot:~/repo$ rsync -a emcboard@linuxcnc.org:emc2 .
[15:38:38] <jepler> well if the path was right..
[15:41:58] <cradek> alex_joni: did you increase the probe feed rate?
[15:42:19] <alex_joni> cradek: using F-word?
[15:42:23] <alex_joni> nope, didn't try that
[15:43:39] <alex_joni> * alex_joni runs home
[15:47:10] <cradek> I bet I know what it is
[15:47:25] <cradek> task commands a stop because the probe hit
[15:47:38] <cradek> but, it takes a while to actually stop, especially since I'm using F100
[15:48:24] <cradek> after task commands the stop, it tells the interp to go on and read the next thing, and the interp knows it just probed, so it expects the queue to be empty, but it hasn't emptied yet since the tp is actually still stopping
[15:48:57] <cradek> I was able to stop at the error and I see the last issued move (tpAddLine) IS the probe move. there's nothing else wrongly on the queue.
[15:51:15] <cradek> I have no evidence to support this yet :-)
[15:54:19] <skunkworks> * skunkworks agrees
[15:54:21] <skunkworks> ;)
[15:55:18] <cradek> no, I'm wrong
[15:55:23] <skunkworks> crap
[15:55:34] <cradek> I get the error when it's still moving down in the probing move
[15:55:40] <cradek> I think way before the probe even touches
[17:15:27] <cradek> yes this is it. sometimes at the beginning of the probe move, emcTaskPlanClearWait() is called at emctaskmain.cc:876.
[17:16:08] <cradek> (not that I know why)
[20:45:17] <CIA-18> 03jepler 07v2_2_branch * 10emc2/VERSION: prepare for 2.2.1 release
[20:45:17] <CIA-18> 03jepler 07v2_2_branch * 10emc2/debian/changelog: prepare for 2.2.1 release
[20:45:18] <CIA-18> 03jepler 07v2_2_branch * 10emc2/src/configure: prepare for 2.2.1 release
[20:46:01] <cradek> jepler: want to wait until we check out pause/step?
[20:46:26] <cradek> unfortunately I think alex is hiding today
[20:46:33] <jepler> cradek: I would really like to get a version out with working stepconf
[20:50:36] <CIA-18> 03jepler 07v2_2_branch * 10emc2/debian/changelog: prepare for 2.2.1 release (fix version number)
[20:52:35] <SWPadnos> can the syntax hilighting stuff be added for 2.2.1?
[20:53:00] <jepler> ??
[20:53:05] <SWPadnos> one sec
[20:53:41] <SWPadnos> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Highlighting_In_Gedit
[20:54:21] <cradek> there is not an easy/good way to edit that system file
[20:54:30] <SWPadnos> ok - I was just going to ask about that
[20:55:14] <SWPadnos> just putting the XML file in the correct dir will still add the option of hilighting, it just won't be done automatically
[20:55:38] <SWPadnos> that file should probably be submitted to the gedit folks though
[20:56:26] <jepler> especially the last step ("3. Adding a text/ngc mime-type") looks tough to get right
[20:56:46] <SWPadnos> yep. that's the part that makes it automatic.
[20:57:13] <SWPadnos> but just having the description file in the right place allows one to manually select this type of hilighting
[20:57:37] <SWPadnos> anyway - if it's a PITA, then it can be left on the wiki
[21:05:13] <cradek> too bad they did not set it up 'smarter' with individual files for new filetypes.
[21:05:28] <jepler> it looks like they have that .. but the person who wrote those instructions wasn't able to discover it
[21:05:34] <cradek> oh...
[21:05:41] <jepler> The following is a list of directories and files that are found inside the MIME database along with brief descriptions:
[21:05:43] <jepler> <MIME>/packages/
[21:05:52] <jepler> Applications which provide information about new MIME types are to install a single new XML file here.
[21:05:57] <SWPadnos> cool!
[21:06:09] <jepler> but there's also /usr/lib/mime which is similar information but not in XML format
[21:06:12] <jepler> * jepler pulls his hair out
[21:06:28] <SWPadnos> I hate mimes
[21:06:36] <jepler> somebody with a stake in this should figure it out and start by putting it in TRUNK
[21:06:50] <cradek> I think there's an echo in here
[21:06:53] <jepler> I believe we should improve the experience of users, and waiting 6 years for a .ngc highlighting to make it into the gedit distribution will not accomplish that
[21:07:17] <jepler> echo?
[21:07:22] <SWPadnos> echo?
[21:07:51] <cradek> I said that last time swp brought it up - put it in trunk, and figure out (and test) the packaging
[21:08:03] <SWPadnos> actually, it should only be 6 months. I'm pretty sure that Hardy Heron would include it, since the Ubuntu releases are tied to Gnome releases, and I don't see any (technical) reason why this couldn't be included
[21:08:12] <SWPadnos> oh, I didn't remember that
[21:08:13] <cradek> or, if I didn't say it, I at least thought it
[21:08:15] <SWPadnos> heh
[21:08:17] <cradek> hmm
[21:08:19] <cradek> sorry then
[21:08:45] <skunkworks>
[21:08:59] <skunkworks> * skunkworks was miming.
[21:09:18] <SWPadnos> I hate you
[21:09:22] <SWPadnos> :)
[21:09:51] <jepler> to whom are you speaking?
[21:10:01] <SWPadnos> err - him
[21:10:02] <SWPadnos> the mime
[21:10:21] <cradek> the visualization stuff is busted too (see the list)
[21:12:36] <fenn> pumagui etc are binaries that get copied into bin/
[21:13:02] <fenn> well, python scripts
[21:13:11] <fenn> it's binary ascii text
[21:13:48] <jepler> Makefile's install target and debian/emc2.files.in also have to be modified when new items are created in bin that should be in the package..
[21:14:44] <jepler> or something
[21:17:09] <fenn> is there a reason each file has to be in the makefile?
[21:17:41] <cradek> because you have to say how/where to install it
[21:17:39] <fenn> instead of ../bin/*
[21:18:34] <jepler> looks like all the xxxgui are not listed
[21:19:26] <jepler> it is hard to make automatic -- for instance, bin/comp goes in emc2-dev, bin/CVS (if it exists) goes nowhere.
[21:20:18] <jepler> fenn: is tracking-test something useful?
[21:21:00] <fenn> it's a neat (i think) demo of what you can do with hal_input
[21:25:41] <SWPadnos> does anyone know the reason why so many people seem to be having trouble with emc.nml (or thinking they do due to bogus error messages)?
[21:27:00] <cradek> has anyone said what the message is?
[21:27:34] <SWPadnos> it seems the IRC channel last night was mostly tech support for "some problem with the NML file
[21:27:41] <SWPadnos> or var file ...
[21:27:48] <cradek> maybe it's the bogus var file problem
[21:27:55] <cradek> that's expected (and well documented)
[21:28:19] <SWPadnos> yep
[21:28:32] <cradek> maybe a followup message to the list that says "really, really read this" and the wiki link again would be helpful
[21:28:44] <jepler> we really should have fixed the interpreter to be more graceful in the face of this change :(
[21:29:21] <cradek> * cradek tries not to get defensive
[21:29:25] <fenn> or a script to add the necessary whatsits maybe
[21:30:10] <jepler> cradek: the variable load and save machinery is hardly a mess of your making
[21:30:40] <jepler> *why* does it write out the new variables required, but write an unusable value for one of them? geez!
[21:30:58] <cradek> oh is that what it does? nice.
[21:31:13] <SWPadnos> I couldn't even find where that was (using lxr/cvsweb ...)
[21:31:42] <cradek> there's an array of required values
[21:32:25] <cradek> if it works except what jepler says, we could accept 5220=0 and then change it to 1
[21:32:48] <cradek> (that's a one line change that we could get right)
[21:33:46] <jepler> http://pastebin.ca/765193
[21:34:01] <cradek> yep
[21:34:19] <cradek> but also write it to pars[5220]
[21:34:29] <cradek> _setup.origin_index = (int) (pars[5220] + 0.0001);
[21:34:32] <cradek> haha
[21:34:43] <cradek> yay arrays of doubles used as ints
[21:34:45] <cradek> welcome to 1982
[21:56:54] <jepler> that's not a full fix for the problem .. the first time you run with an old var file, emc prints
[21:56:57] <jepler> emc/task/emctask.cc 312: interp_error: Required parameter missing
[21:57:04] <jepler> and hangs at startup. the second time, the change above fixes the value and emc starts
[21:57:19] <cradek> could be worse.
[21:59:53] <jepler> hm, I wonder if this change was important, found it in a local tree not checked in:
[21:59:54] <jepler> - CHKS((settings->speed == 0.0), "Cannot feed with zero spindle speed in feed per rev mode");
[21:59:59] <jepler> + CHKS(((settings->spindle_turning != CANON_CLOCKWISE) &&
[22:00:00] <jepler> + (settings->spindle_turning != CANON_COUNTERCLOCKWISE)),
[22:00:02] <jepler> + "Spindle not turning in G95 G1");
[22:00:41] <cradek> I don't see what difference that makes...
[22:00:59] <jepler> the s-number can be nonzero when the spindle is stopped (M5 or whatever)
[22:01:21] <cradek> oh.....
[22:03:01] <jepler> http://pastebin.ca/765246
[22:03:50] <jepler> this also does away with the "required variable" error -- the variable is set to 0 and if it's in the required array it's added at the next write-out
[22:04:13] <cradek> error messages that do nothing but complain and abort when the error is easily handled are stupid.
[22:04:50] <jepler> hm I bet my change isn't quite right
[22:04:55] <cradek> so is a missing varfile generated?
[22:05:15] <jepler> all the "required" are written out the next time
[22:05:29] <jepler> if I check out sim.var from v2_1_7 and run, the file at exit compares identical to the TRUNK one
[22:05:36] <cradek> nice
[22:05:46] <cradek> so I got all the requireds right? :-)
[22:05:55] <jepler> beats me
[22:06:41] <cradek> try adding #1 to your varfile and make sure it's preserved etc
[22:08:35] <jepler> yes, it preserves a variable #1 and a variable #5386
[22:08:36] <SWPadnos> hmmm
[22:08:49] <jepler> (after one additional fix not shown in that pastebin)
[22:08:50] <SWPadnos> what happens if the var file is missing?
[22:09:00] <SWPadnos> ie, all required parameters aren't there ...
[22:09:07] <jepler> you mean empty? or missing?
[22:09:10] <SWPadnos> either
[22:09:17] <jepler> emc/task/emctask.cc 312: interp_error: Unable to open file
[22:09:25] <jepler> (missing)
[22:09:40] <jepler> runs when empty
[22:09:45] <SWPadnos> though empty is a very nice thing - the fix is `touch /path/pt/my/config/blah.var`
[22:12:33] <CIA-18> 03jepler 07TRUNK * 10emc2/src/emc/rs274ngc/rs274ngc_pre.cc: these errors are easily fixed, and were otherwise provoked by people upgrading from version 2.1
[22:14:32] <CIA-18> 03jepler 07v2_2_branch * 10emc2/src/emc/rs274ngc/rs274ngc_pre.cc: merge rev 1.50: make graceful upgrade from v2.1.x var files possible
[22:16:26] <CIA-18> 03jepler 07v2_2_branch * 10emc2/debian/emc2.files.in: add xxxgui, tracking-test, teach-in to packages
[22:16:31] <CIA-18> 03jepler 07v2_2_branch * 10emc2/src/Makefile: add xxxgui, tracking-test, teach-in to packages
[22:16:36] <CIA-18> 03jepler 07v2_2_branch * 10emc2/debian/changelog: note changes
[22:21:05] <jepler> OK, now I'm really making the 2.2.1 release...
[22:22:46] <SWPadnos> I missed discussion of the impetus for this quick release, can you sum it up quickly (or should I read the logs?)
[22:24:53] <jepler> SWPadnos: stepconf was completely broken
[22:25:04] <SWPadnos> oh, ok. that seems like enough
[22:25:10] <jepler> the var file thing seems to be tripping up a lot of users
[22:25:14] <SWPadnos> plus the var file fix
[22:25:14] <SWPadnos> yep
[22:25:49] <SWPadnos> are people reasonably sure that the nml file errors are bogus results of var file problems?
[22:26:04] <jepler> I didn't see any "nml file errors" personally
[22:26:34] <skunkworks> I think cradek would also like to fix the step/resume and probing issue. maybe 2.2.2?
[22:26:35] <SWPadnos> hmmm. there were certainly some issues with that last night. Alex and I helped jlmjvm fix that
[22:27:13] <SWPadnos> and maybe it was ADam_1 or someone else with the var file problem, which may also have generated a bogus emc.nml-related error message
[22:28:47] <jepler> I didn't do anything that had to do with nml files
[22:30:23] <SWPadnos> I know there were some errors related to the emc.nml file, but they were mentioned while we were debugging a var file problem as well. It's possible that the emc.nml error was actually a warning, and the var file caused the error
[22:32:12] <SWPadnos> ok - here's one of the many problems jlmjvm had last night: http://pastebin.ca/764200
[22:33:03] <SWPadnos> ok. he just hadn't copied the new file over. here's a more informative log: http://pastebin.ca/764153
[22:33:27] <SWPadnos> an installed EMC2 has configs in the /usr/share/emc2 dir?
[22:34:00] <SWPadnos> I think I'll change the UPDATING page to explicitly tell people what file to copy into their config dir.
[22:38:11] <jepler> beats me what that error is. I'd be uncharitable, but I'd guess he removed his emc.nml file somehow
[22:38:18] <jepler> or replaced it with something else
[22:38:23] <jepler> like a var file
[22:38:40] <SWPadnos> one never knows ...
[22:38:53] <SWPadnos> many PEBCAKs there
[22:39:43] <jepler> anyway, with 2.1.1 no manual change to the var file will be needed anymore
[22:39:53] <SWPadnos> right
[22:40:03] <jepler> so working on the perfect wording and placement of the warning may be effort misspent
[22:40:15] <SWPadnos> looking at the UPDATING page, it says "you'll need a fresh emc.nml", but doesn't really say what that means
[22:40:43] <SWPadnos> no, I'll make the instructions explicitly tell the users what file to copy
[22:47:07] <jepler> hm, is it clear that that's only for users upgrading from 2.0.x?
[22:47:47] <SWPadnos> I guess it should be, since it's in the 2.0->2.1 section
[22:48:22] <SWPadnos> it could be that the person/people who have had a problem had tried to change that file, and caused their own issues
[23:20:50] <jepler> jepler@zenbot:~/repo$ ./put
[23:20:50] <jepler> emcboard@linuxcnc.org's password:
[23:20:50] <jepler> building file list ...
[23:20:50] <jepler> 453 files to consider
[23:20:50] <jepler> ./
[23:20:53] <jepler> sent 16192 bytes received 26 bytes 1908.00 bytes/sec
[23:20:53] <jepler> total size is 695271803 speedup is 42870.38
[23:20:56] <jepler> OK, now I'm pretty confident that I can update the repository...
[23:21:38] <SWPadnos> that's a reasonable speedup