#emc-devel | Logs for 2007-11-03

[00:00:08] <cradek> * cradek tries the obvious thing
[00:00:14] <SWPadnos> I wonder if you need to have an M199 script
[00:00:23] <cradek> an hour ago I didn't need to have one
[00:00:30] <jepler> you don't need to have one
[00:00:44] <cradek> well m199 is now an error for some reason
[00:01:20] <jepler> m2?
[00:01:26] <cradek> m2,m3 works
[00:02:29] <jepler> have to go, good luck with this :-P
[00:02:34] <jepler> I'll read back later
[00:03:14] <jepler> interp_read.cc:1128: "Bad character '\\%03o' used" : "Bad character '%c' used", letter));
[00:03:27] <jepler> is the only place this can come from, but I can't see how it wouldn't affect mdi m3
[00:03:40] <cradek> affects mdi m199, not mdi m3
[00:03:46] <cradek> mdi m3 still works
[00:05:14] <jepler> in mdi you get the same error, or do you get "unknown m code used"?
[00:05:18] <jepler> I'd expect the latter
[00:05:29] <cradek> oh sorry
[00:05:36] <cradek> you're right
[00:06:10] <cradek> I didn't read it
[00:06:47] <cradek> something's stomping on _readers[]
[00:07:01] <cradek> that's not going to be fun to find.
[00:07:06] <jepler> no, not very fun
[00:07:19] <jepler> what happens if you load gcode in the preview that has an mcode (possibly excluding m2)
[00:07:32] <jepler> those would be the "same" instance of the interpreter, unlike the mdi case
[00:07:42] <cradek> I'm going to try to examine _readers[] in gdb before I do anything else
[00:07:48] <jepler> ok
[00:07:52] <jepler> I really must go
[00:07:54] <cradek> bye
[00:08:07] <jepler> look in axis, not task!
[00:08:18] <cradek> thanks
[00:12:30] <cradek> (gdb) p _readers
[00:12:31] <cradek> $3 = {{__pfn = 0, __delta = 0} <repeats 256 times>}
[00:12:31] <cradek> it's empty
[00:12:43] <cradek> but, default_readers is fine
[00:12:57] <cradek> this surely has to do with the new code that disables certain letters
[00:14:18] <cradek> no, that's simpler than I thought
[00:26:13] <cradek> oh there it is. a permanent 'bad number format'
[00:26:13] <cradek> (I loaded a program into the preview and get this now)
[00:26:13] <jmkasunich__> get out the big can of "bug-b-gon"
[00:26:13] <cradek> well I'm happy to finally see it
[00:26:26] <cradek> now if I had any clue how to find it...
[00:36:46] <alex_joni> hi jmkasunich__
[00:36:46] <jmkasunich__> hi alex
[00:36:46] <jmkasunich__> jmkasunich__ is now known as jmkasunich
[00:36:46] <alex_joni> hmm.. is DH still working for you?
[00:36:46] <jmkasunich> doesn't seem to be
[00:36:46] <jmkasunich> nothing posted on their emergency site, but the main site still isn't coming up
[00:36:46] <alex_joni> hmm.. it's back up now
[00:36:46] <jmkasunich> strange
[00:37:06] <jmkasunich> it wasn't a joomla thing, other non-joomla stuff was down as well
[00:37:09] <jmkasunich> and is back now
[00:37:52] <alex_joni> yup.. my ssh died, that's how I noticed
[00:45:37] <SWPadnos> my ssh session was fine (then I closed it)
[00:46:15] <jmkasunich> ah, so you're the one who broke it!
[00:52:11] <alex_joni> SWPadnos: can you group:+w /home/.jared/cncman/www.linuxcnc.org/administrator/components/com_mambelfish/contentelements ?
[01:04:03] <alex_joni> * alex_joni brings out the cattle prod
[01:14:15] <alex_joni> * alex_joni prods SWPadnos
[01:18:18] <cradek> ok I know a lot more
[01:18:28] <cradek> the 'm' error only happens when a file is not yet loaded
[01:18:36] <cradek> this can only happen with a lathe config
[01:18:44] <alex_joni> I thought axis starts up with a file loaded..
[01:18:50] <alex_joni> oh, not for lathe
[01:18:54] <cradek> the error happens because Interp::init fails, because there is not a var file for axis's interp yet
[01:19:25] <cradek> the 'invalid number format' error was my but that I introduced today, and probably not what has been reported a few times
[01:19:36] <cradek> so, I'm still lost about that one
[01:20:12] <cradek> I wonder what a good lathe splash screen would be
[01:20:40] <jmkasunich> the classic chess pawn?
[01:20:59] <cradek> we sure could use that
[01:21:35] <alex_joni> otoh, the splash screen is not intended for actual cutting
[01:21:43] <cradek> haha
[01:21:43] <alex_joni> (even if a lot of people do that..)
[01:21:53] <cradek> it's the first thing everyone cuts
[01:22:08] <alex_joni> I know :)
[01:22:19] <alex_joni> but read the comment on line 1
[01:22:22] <cradek> the pawn would cut nicely if we added css,fpr
[01:22:31] <cradek> it would be a good demo
[01:23:03] <CIA-40> 03cradek 07TRUNK * 10emc2/share/axis/tcl/axis.tcl: having this menu in two places is confusing.
[01:23:10] <cradek> (for some reason I've never cut the pawn)
[01:23:22] <alex_joni> where is SWPadnos, when I need him
[01:23:59] <CIA-40> 03cradek 07TRUNK * 10emc2/src/emc/rs274ngc/gcodemodule.cc: not used for anything, but called now
[01:31:38] <tissf> good night all
[01:33:01] <alex_joni> good night tissf
[01:34:47] <cradek> ok I'm back to not knowing what that error is that others are getting
[01:51:02] <SWPadnos> oh - her eI am
[01:51:05] <SWPadnos> here I
[01:51:49] <alex_joni> finally
[01:51:57] <alex_joni> can you please +w on that folder ?
[01:52:11] <SWPadnos> wthe folder, or all the files in the folder?
[01:52:16] <SWPadnos> -w
[01:52:43] <alex_joni> only the folder
[01:52:51] <SWPadnos> ok
[01:52:54] <SWPadnos> done
[01:52:57] <alex_joni> thx
[01:53:02] <SWPadnos> sure
[01:53:29] <alex_joni> hmm.. can't write to folder
[01:53:48] <alex_joni> drwxr-xr-x 2 cncman pg167418 4096 Feb 28 2006 contentelements
[01:53:48] <SWPadnos> it's g+w
[01:54:06] <SWPadnos> oh - contentelements - missed that
[01:54:35] <SWPadnos> ok, done
[01:58:17] <alex_joni> darn, doesn't work as I'd hoped it would
[01:58:52] <SWPadnos> how can you tell?
[02:29:33] <jmkasunich> eww, somebody joomlified the compile farm page
[02:29:44] <SWPadnos> heh
[02:30:37] <alex_joni> jmkasunich: sorry.. that was me :)
[02:30:42] <jmkasunich> heh
[02:30:51] <alex_joni> you can still reach the original one at c_f
[02:30:52] <jmkasunich> as long as it all works I guess I don't care
[02:40:32] <jepler> cradek: too bad
[02:50:49] <alex_joni> http://www.linuxcnc.org/component/option,com_fireboard/Itemid,16/lang,en/
[03:09:51] <cradek> jepler: I found other things, so it's ok
[03:10:09] <cradek> jepler: if we care about touch-off working before loading a file, there's some work to do that I'm not sure about
[03:11:32] <SWPadnos> shouldn't the vars be loaded before a g-code file is loaded?
[03:11:37] <SWPadnos> arbitrarily long before, even
[03:11:44] <cradek> in axis's interpreter, it happens together
[03:12:01] <cradek> fixable but invasive enough that I want jepler to do it :-)
[03:12:05] <SWPadnos> heh
[03:12:06] <SWPadnos> me too! :)
[03:12:17] <cradek> seems I don't like to fix anything that can't be fixed with one line of code
[03:12:38] <SWPadnos> that's not true. I think you used several lines in that last patch
[03:12:54] <cradek> yeah but I didn't like it.
[03:13:28] <alex_joni> you can always rewrite it into a single line
[03:13:28] <SWPadnos> oh, I see
[03:13:43] <SWPadnos> removingwhitespacedoescompactthecodequiteabit
[03:14:09] <SWPadnos> especially on windows, where each line has a cr *and* an lf at the end!
[03:18:27] <jepler> hah
[03:22:13] <cradek> goodnight
[03:22:25] <SWPadnos> night
[03:23:48] <jepler> goodnight
[03:25:05] <alex_joni> good night
[05:39:23] <fenn> oh boy another web forum
[13:20:47] <tissf_> tissf_ is now known as tissf
[14:21:22] <CIA-18> 03jepler 07TRUNK * 10emc2/debian/control.in: I'll be the release manager for the 2.2 cycle
[15:24:02] <cradek> jepler: what do you think about fixing that axis problem - as I see it, either we have to rearrange the interp init so there's a var file copy available right away, or we need a lathe splash screen of some kind
[15:25:38] <cradek> the splash screen would be the change we're most likely to get right
[15:26:34] <alex_joni> or left
[15:26:52] <alex_joni> I never was sure about lathe directions :D
[15:29:49] <alex_joni> * alex_joni hates having connection issues
[15:34:04] <jepler> cradek: hm
[15:40:14] <jepler> I think having a lathe splash screen would be fine, but I'd also like to fix the bug
[15:40:59] <cradek> I can't figure out what a lathe splash screen would be
[15:41:10] <cradek> if we have a logo it's completely unmachinable
[15:41:16] <jepler> yeah
[15:41:23] <cradek> if we have a program that you can cut, it's not a splash screen
[15:41:32] <cradek> so maybe just fixing the bug is better
[15:41:39] <alex_joni> how about having both?
[15:42:04] <cradek> both?
[15:42:09] <alex_joni> say the first half of the program is some lathe work, turning & threading
[15:42:11] <CIA-18> 03jepler 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/axis.py: fix "touch off" when no gcode has been loaded yet (e.g., lathe with no splash gcode)
[15:42:42] <alex_joni> and below that you have some text, whatever,m which is separated in the program by an optional stop
[15:43:13] <cradek> true, the text could be in air outside the part
[15:43:16] <alex_joni> so when the user runs it, it only does the first part
[15:44:57] <cradek> I think that's the best idea so far, if we want to do it
[15:49:50] <jepler> anybody get the behavior described on the mailing list in 2.1.7: pressing "-" in MDI turns off the spindle
[15:51:37] <cradek> jepler: I thought 2.1.7 would increase the spindle by 100rpm with F12. what does it do?
[15:51:49] <cradek> maybe he's not running 2.1.7...?
[15:52:51] <cradek> I can't run 2.1.7 on anything right now without some gymnastics
[15:53:09] <jepler> no, 2.1.7 seems to change by +-1
[15:53:13] <alex_joni> I can run it on a vm
[15:53:24] <cradek> yuck.
[15:53:44] <cradek> I haven't used 2.1 for so long I assume things are fixed in it
[15:54:11] <alex_joni> lets just release 2.2.0 this weekend
[15:54:23] <cradek> I'd like that too
[15:54:38] <jepler> oh dear god
[15:54:41] <jepler> is it really ready?
[15:54:50] <cradek> but I wish I had more time to spend on it - I should be doing garage work
[15:54:49] <alex_joni> jepler: it won't ever be
[15:55:08] <alex_joni> it's always like this.. (remember 2.0.0 ?)
[15:55:12] <cradek> jepler: do you know of anything glaringly wrong?
[15:55:16] <jepler> cradek: well, no
[15:55:23] <jepler> but like the rest of the developers, I don't actually use the software
[15:55:27] <alex_joni> are we satisfied with the arc issues?
[15:55:48] <cradek> alex_joni: I am
[15:55:52] <alex_joni> ok, me too
[15:56:07] <alex_joni> that would be imo the biggest issue I know of
[15:56:18] <alex_joni> rest is cosmetics )
[15:56:30] <jepler> I guess I have to figure out how to sign the package, then.
[15:56:42] <alex_joni> debuild does it for you
[15:56:50] <cradek> if you set DEBEMAIL in the environment to match your key, it just works
[15:56:59] <cradek> (I think)
[15:57:14] <alex_joni> * alex_joni always builds packages using "debuild"
[15:57:20] <cradek> I forget that nobody has done this but me...
[15:57:39] <alex_joni> cradek: I think I tried it once on a local package..
[15:58:30] <cradek> I have to go - I can help more after dark
[15:58:52] <jepler> my key isn't on the machine where I build packages, so that's the first problem...
[15:59:00] <alex_joni> well.. it's dark here already :)
[15:59:00] <cradek> jepler: I think if you use it for a couple hours, you'll either find little things, or be a lot more confident. That's what I did and I'm a lot more confident now.
[15:59:20] <cradek> right now we're building for dapper, right?
[15:59:37] <alex_joni> right
[15:59:40] <jepler> yes, dapper only
[15:59:41] <cradek> I think we should ignore the CD issues for the moment
[15:59:54] <jepler> cradek: go work on your garage!
[15:59:56] <alex_joni> * alex_joni volunteers for a new LTS CD
[15:59:58] <cradek> we need a wiki page for updating configs
[15:59:58] <jepler> this will be here when you get back
[16:00:05] <alex_joni> when horny hardon is out
[16:00:14] <cradek> ok, bye
[16:00:30] <alex_joni> I'll do the wiki now
[16:00:40] <jepler> don't do too much before the packages are ready!
[16:00:50] <jepler> but I guess everything but the announcement is pretty under-the-radar...
[16:01:04] <alex_joni> indeed :)
[16:17:18] <CIA-18> 03alex_joni 07TRUNK * 10emc2/docs/UPDATING: information on updating configurations for 2.2.x
[16:26:49] <alex_joni> jepler: there is an issue which needs addressing before you make the package
[16:26:59] <alex_joni> all configurations have a NC_FILE= setting
[16:27:08] <alex_joni> NC_FILES
[16:29:04] <jepler> you mean PROGRAM_PREFIX?
[16:30:42] <jepler> that is patched up by pickconfig when copying a read-only configuration
[16:32:04] <alex_joni> yeah, PROGRAM_PREFIX
[16:32:12] <alex_joni> I meant for the sample configurations
[16:32:36] <alex_joni> so far chris always replaced PROGRAM_PREFIX=/usr/.../ before building the pachage
[16:33:57] <jepler> why bother, when it'll just be overwritten when pickconfig copies the configuration to ~/emc2/configs ?
[16:35:25] <alex_joni> well.. just for people who only run sample configs, without copying to ~
[17:26:58] <jmkasunich_> what is cradek doing in (to?) his garage?
[17:29:14] <jepler> jmkasunich_: the "old garage" is now the shop; he's building the "new garage" for the cars.
[17:29:24] <jmkasunich_> oh - big project
[17:29:35] <jmkasunich_> seems late in the year to be starting something like that
[17:29:37] <jepler> I haven't seen it in a few weeks
[17:29:39] <jmkasunich_> or is it almost done?
[17:29:47] <jepler> I think the shingles may be going on this weekend, not sure though
[20:36:37] <tissf> Halgui is implemented now?
[20:39:15] <tissf> In the manual HAL intro it is indicated (not implemented yet)
[20:59:39] <alex_joni> tissf: no
[21:27:33] <tissf> alex_joni: thk
[21:58:35] <alex_joni> * alex_joni heads to bed
[21:58:37] <alex_joni> good night all
[22:36:03] <tissf> godd night alex
[22:36:35] <tissf> good night