#emc-devel | Logs for 2008-08-24

[01:49:06] <jmkasunich_> jmkasunich_ is now known as jmkasunich
[01:58:44] <jepler> I'm glad seb doesn't hesitate to tell me when I screwed up the code :)
[02:00:37] <cradek> that's a good thing, we all have the same goal - to have it right
[02:01:38] <cradek> jepler: have Z feedback, hooking up the amp now
[02:16:54] <jepler> cradek: oh cool
[02:17:04] <jepler> anything that was less than straightfoward?
[13:11:50] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/common/ (Document_Header.lyx GPLD_Copyright.lyx): changed section to chapter
[13:34:04] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/common/GPLD_Copyright.lyx: changed paragraph layout to match rest of document
[13:35:53] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/Submakefile: added Getting Started to pdfs at least I got that part to work...
[14:04:11] <cradek> jepler: my trunk checkout is getting that trap 14 often. the stop/restart is not required like we first thought.
[14:15:46] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/common/Getting_EMC.lyx: added file
[14:17:36] <alex_joni> trap 14?
[14:17:57] <cradek> 'realtime task killed, trap 14' or something like that
[14:18:18] <cradek> jepler found somewhere it means 'page fault'
[14:18:31] <alex_joni> ah, right
[14:18:41] <alex_joni> on what task is it?
[14:18:47] <alex_joni> or is it RT generally?
[14:18:50] <cradek> I don't even know that
[14:19:00] <alex_joni> I think I had that while I was wirting the STG driver
[14:19:01] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/gui/axis.lyx: add manual too change info
[14:19:14] <alex_joni> if I messed up something it would do that sometimes instead of locking up the PC
[14:19:20] <cradek> yeah I've caused it too, it's just a bad pointer deref or something.
[14:19:34] <alex_joni> might be 5i20 related
[14:21:54] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/gcode/main.lyx: add manual tool change info
[14:23:06] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/Getting_Started.lyx: added file
[14:49:11] <cradek> hi matt
[14:55:01] <mshaver> Hey!
[14:55:13] <mshaver> Did you get my e-mail?
[14:55:28] <cradek> yes, seeing you reminded me to go look at your website
[14:55:40] <mshaver> Good!
[14:56:03] <alex_joni> BigJohnT: check the stuff in, we'll sort it out
[14:56:32] <mshaver> I'm not sure why I was instructed to send this to you, but I have done so, so I guess that's done!
[14:56:48] <cradek> mshaver: I wanted to see your pocketing algorithm, but it's not there :-)
[14:56:58] <mshaver> I hope it's considered useful...
[14:57:10] <cradek> yes, I don't quite understand what's going on either
[14:57:22] <mshaver> I'm probably going to port Jon Elson's
[14:57:55] <mshaver> The boltcircle algo is pretty goo (If I do say so myself)
[14:58:19] <mshaver> not goo, I mean good, although is pretty gooey too
[14:58:42] <cradek> ok so this web page is a demonstration of the scheme, but it will eventually have a different gui that also runs emc?
[14:59:37] <mshaver> Well, actually, this is really just a test interface for Javascript that generates g-code.
[14:59:55] <mshaver> The EMC GUI os written in QT.
[15:00:00] <mshaver> is
[15:00:08] <BigJohnT> ok alex_joni
[15:00:33] <mshaver> In QT, you can run Javascript as a scripting language.
[15:01:08] <cradek> that qt gui is gpl?
[15:01:18] <mshaver> So, it is thought that new "wizards" will be able to be added to the QT gui.
[15:01:26] <mshaver> GPL- yes.
[15:01:50] <mshaver> Wally, the fellow at the workshop has written it.
[15:01:55] <cradek> neat. is it released yet? can I get it?
[15:02:21] <mshaver> Yes, and yes.
[15:02:37] <mshaver> Wally is supposed to be "done" on sep 1
[15:02:38] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/common/Document_Header.lyx: added file
[15:02:59] <mshaver> He's been working on it, trying to get it "presentable"
[15:03:48] <mshaver> You know how it is: You should "release early & often", but you feel self-conscious about releasing something that's less than perfect.
[15:03:48] <cradek> is this EZ-Trol?
[15:03:52] <mshaver> yes
[15:04:02] <cradek> http://www.cnczone.com/forums/showpost.php?p=481749&postcount=2
[15:04:11] <mshaver> Version 1 is being used now.
[15:04:20] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/Submakefile: added getting started to the pdf...
[15:04:44] <mshaver> Wally has refactored the code as he said that V1 was a complete mess
[15:04:54] <mshaver> let me look at your link...
[15:05:29] <cradek> I'm confused by how it seems to be GPL but also not freely available
[15:05:48] <alex_joni> maybe it's more than just the GUI
[15:06:03] <mshaver> I don't know who "nanmol" is...
[15:06:04] <cradek> oops, here's the thread: http://www.cnczone.com/forums/showthread.php?t=59870
[15:06:11] <mshaver> hold on...
[15:06:19] <cradek> someone at weber
[15:06:23] <mshaver> I go look @ thread!
[15:07:42] <mshaver> OK...
[15:07:46] <mshaver> Uh...
[15:08:02] <mshaver> Here's the whole story as I know it:
[15:08:16] <CIA-36> EMC: 03jepler 07TRUNK * 10emc2/docs/src/Submakefile: make building html from documents in docs/src but not in a subdir work
[15:08:49] <mshaver> Smithy wanted to do a new GUI.
[15:09:05] <mshaver> They wanted it to have CAM functionality.
[15:09:18] <mshaver> Weber makes Synergy.
[15:09:43] <mshaver> Smithy collaborated with Weber on the GUI.
[15:10:09] <mshaver> It's basically 4 tabs, Auto, Manual, Settings, and Mill.
[15:10:31] <mshaver> Wally wrote all the tabs except Mill.
[15:10:43] <mshaver> Mill is Synergy "Lite".
[15:10:48] <skunkworks> mill is the xynergy?
[15:10:50] <skunkworks> ah
[15:10:59] <mshaver> yes
[15:11:35] <mshaver> The folks at Weber, especially Bob, are _very_ smart.
[15:11:57] <mshaver> And have a long time experience with CNC programming.
[15:12:19] <mshaver> The system started out on HP/UX workstations.
[15:13:02] <mshaver> When that line of computers wasn't made anymore, it was ported to Unix, then Windows NT, then to Linux.
[15:13:40] <mshaver> So, the UI is sort of "unique"...
[15:13:50] <cradek> haha
[15:14:06] <mshaver> It has some good ideas in it.
[15:14:19] <mshaver> But they are'nt really "obvious".
[15:14:37] <mshaver> There's also no documentation.
[15:14:48] <mshaver> None, that I know of.
[15:14:54] <mshaver> I've asked.
[15:15:12] <mshaver> What they said makes a lot of sense.
[15:15:49] <mshaver> "Since people never read the docs, we don't supply them. They'll call us no matter what."
[15:16:09] <mshaver> I can actually symathize with this line of thinking.
[15:16:20] <mshaver> sympathize I mean.
[15:16:56] <mshaver> Anyway, the programs architecture is rather well thought out.
[15:17:37] <mshaver> There's a UI. The Linux one is written in QT. The QT GUI for Synergy is GPL.
[15:18:27] <mshaver> The GUI gathers together the user input and calls a program called "webmon", which is sort of a dispatcher.
[15:18:51] <mshaver> The webmon calls script files to do the actual work.
[15:18:57] <cradek> darnit, I've got to run - I'll read back in a bit
[15:19:10] <mshaver> The script files are written in Rocky Mountain Basic.
[15:19:16] <mshaver> OK!
[15:21:08] <mshaver> Rocky Mountain Basic, as I understand it, was included on the HP workstations mainly for use controlling GPIB equipped instruments, og which HP made many.
[15:21:20] <mshaver> not og, of!
[15:21:35] <alex_joni> seems it's still available
[15:22:59] <mshaver> Yes, sort of, but the folks at Weber re-implemented and extended it with commands that they needed.
[15:24:01] <mshaver> Their library of scripts represented a considerable investment of time and development effort, and contain a lot of "know how" that they didn't want to lose.
[15:26:43] <mshaver> Anyway, I almost don't remember where I was going with all this.
[15:27:04] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/Submakefile: added file
[15:27:05] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/common/Linux_FAQ.lyx: added file
[15:28:05] <mshaver> The long and short of it is that EZ-Trol was developed by Smithy with Weber's help (as fas as I know anyway). Also, as far as I know, all the QT code is GPL.
[15:28:19] <mshaver> I guess tha's where I was going.
[15:29:05] <mshaver> cradek: I'll be back in a little while - got to pick up one of my sons from a friend's house!
[15:29:10] <alex_joni> does it run with only the GPL stuff?
[15:29:25] <alex_joni> or do you need for it to be linked to the other parts too?
[15:29:28] <mshaver> Well, it will run and control EMC.
[15:29:39] <alex_joni> right, that's what I was asking
[15:29:43] <mshaver> But the Mill tab won't generate any code.
[15:29:49] <alex_joni> obviously
[15:30:53] <mshaver> Back in a little bit!
[15:40:21] <alex_joni> BigJohnT: getting started looks good
[15:42:36] <BigJohnT> thanks alex_joni
[15:43:28] <BigJohnT> all I'm lacking is getting the html part done but I'm kinda stuck...
[15:43:33] <alex_joni> 2 things I notice..
[15:43:38] <alex_joni> in the legal section
[15:43:52] <alex_joni> it's the GNU Free Documentation Licence
[15:43:58] <alex_joni> but the chapter says GPL
[15:44:12] <alex_joni> (probably been like this for a long time.. but I never studied it carefully :)
[15:44:14] <BigJohnT> ok
[15:44:33] <BigJohnT> yes it was
[15:44:38] <alex_joni> and 2). the index is kinda empty
[15:44:53] <alex_joni> not sure it makes a lot of sense :)
[15:45:00] <BigJohnT> that's on the list :)
[15:45:27] <BigJohnT> what part
[15:45:37] <alex_joni> having the index empty
[15:45:49] <alex_joni> if you plan to add stuff to it, then it'll be fine
[15:46:16] <BigJohnT> I have not done any index tags yet...
[15:46:25] <alex_joni> right
[15:49:22] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/common/GPLD_Copyright.lyx: fix
[15:50:48] <alex_joni> BigJohnT: that's not what I meant :(
[15:52:01] <BigJohnT> :( what was it?
[15:52:38] <alex_joni> section 1.2 was titled "GNU General Public Licence"
[15:52:53] <alex_joni> it should be "GNU Free Documentation Licence"
[15:53:20] <BigJohnT> ok
[15:54:05] <alex_joni> the other reference in the text is correct as "GNU General Public Licence" also known as GPL
[15:55:12] <alex_joni> too bad I'm on 8.04.. I would fix it myself
[15:56:41] <BigJohnT> I can put it back
[15:57:13] <alex_joni> please do..
[15:57:25] <alex_joni> so section 1.2 is "GNU Free Documentation Licence"
[15:59:35] <BigJohnT> in the original for some reason the first paragraph is duplicated in the second, any reason for that?
[16:01:22] <alex_joni> what do you mean?
[16:02:46] <BigJohnT> Copyright (C) 2000 Free Software Foundation, Inc. 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
[16:02:48] <BigJohnT> Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not
[16:02:50] <BigJohnT> allowed.
[16:03:09] <BigJohnT> this paragraph is in there twice in a row
[16:05:14] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/common/GPLD_Copyright.lyx: fix the fix
[16:05:24] <alex_joni> I don't see it
[16:05:45] <BigJohnT> it's not there now :) but it was...
[16:06:06] <BigJohnT> it looked to me to be a copy and paste error
[16:12:45] <alex_joni> BigJohnT: still missed one
[16:12:47] <alex_joni> -GPL Free Documentation License Version 1.1, March 2000
[16:12:48] <alex_joni> \layout Standard
[16:12:48] <alex_joni> +GNU Free Documentation License Version 1.1, March 2000
[16:12:59] <BigJohnT> ok
[16:17:34] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/common/GPLD_Copyright.lyx: GPL 2 GNU
[16:33:47] <alex_joni> BigJohnT: can I ask you something?
[16:34:27] <BigJohnT> sure
[16:35:12] <alex_joni> what version of lyx-qt do you have installed?
[16:35:15] <alex_joni> is it 1.3.7 ?
[16:35:37] <BigJohnT> one moment
[16:35:59] <alex_joni> maybe paste dpkg -l | grep lyx
[16:36:16] <BigJohnT> yes, 1.3.7 of Tue, Jan 17, 2006
[16:37:12] <BigJohnT> dpkg -l | grep lyx
[16:37:21] <BigJohnT> ii lyx-common 1.3.7-0ubuntu4 High Level Word Processor - common files
[16:37:21] <BigJohnT> ii lyx-qt 1.3.7-0ubuntu4 High Level Word Processor - Qt frontend
[16:37:38] <alex_joni> ok, thx
[16:39:19] <BigJohnT> alex_joni: when you add the getting started manual to the document page please update the user manual description
[16:39:38] <alex_joni> I'm working on getting a working lyx first
[16:39:40] <alex_joni> on hardy
[16:40:12] <BigJohnT> ok cool
[16:40:27] <BigJohnT> let me know how you did it when you get it done...
[16:48:39] <alex_joni> yeah.. I'm trying to move over the lyx2lyx from 1.3.7 to the 1.4.x
[16:50:46] <BigJohnT> so when you get that done it won't work on 1.3.7 anymore?
[16:51:26] <alex_joni> no.. lyx2lyx is a program
[16:51:34] <alex_joni> which is included in the lyx package
[16:51:58] <alex_joni> I am trying to use the same converter program on the newer lyx, so the 221 generated files are identical to the ones generated on dapper
[16:52:10] <BigJohnT> the one that converts from one version to another?
[16:52:56] <alex_joni> yes
[17:07:24] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/common/ (Getting_EMC.lyx Linux_FAQ.lyx): added some index entries
[17:09:26] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/config/stepconf.lyx: added some index entries
[17:14:46] <CIA-36> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/Submakefile: missed one file
[17:23:57] <alex_joni> hmm.. doesn't quite work
[17:29:23] <jepler> alex_joni: no, it still introduces spurious changes
[17:36:29] <fenn> why not just edit in a text editor and verify with lyx
[17:39:46] <BigJohnT> the tags would drive you nuts :)
[17:39:54] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/common/GPLD_Copyright.lyx: change section to chapter
[17:39:56] <BigJohnT> cd ..
[17:40:24] <BigJohnT> I see my focus jumped to here :)
[17:40:36] <jmkasunich> all by itself
[17:41:13] <BigJohnT> yep when the commit happened I was in a terminal window about to cd up one level
[17:42:41] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/Getting_Started.lyx: changed 2.3 to 2.2 in title
[18:08:30] <alex_joni> jepler: I think the change is in /usr/share/lyx/lyx2lyx/parser_tools.py
[18:08:45] <alex_joni> but the dapper version is compiled for python 2.4 I think
[18:08:50] <alex_joni> and the hardy is for 2.5
[18:08:56] <alex_joni> so I can't just copy the pyc there
[18:21:31] <alex_joni> how does one apply a foo.diff.gz ?
[18:22:52] <alex_joni> nm
[18:46:25] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/common/Getting_EMC.lyx: added some index entries
[18:48:21] <alex_joni> * alex_joni cusses lyx
[18:48:34] <SWPadnos> darned lyx!
[18:49:14] <alex_joni> SWPadnos: yeah
[18:49:50] <alex_joni> trying to build a package for hardy out of the 1.3.7 sources
[18:49:52] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/common/Linux_FAQ.lyx: added some index entries
[18:50:09] <alex_joni> ../../../boost/boost/config/compiler/gcc.hpp:57:7: warning: #warning "Unknown compiler version - please run the configure tests and report the results"
[18:50:25] <SWPadnos> compiling lyx?
[18:50:31] <alex_joni> yeah :)
[18:50:32] <SWPadnos> err - duh. nevermind
[18:50:40] <alex_joni> it doesn't know about g++-4.2
[18:50:47] <SWPadnos> (I'm just having my first coffee of the day, if that tells you anything :) )
[18:50:52] <alex_joni> but that's only a warning.. so nothing exciting
[18:51:03] <alex_joni> had to hack the heck out of the debian scripts to make it build
[18:54:50] <SWPadnos> I think I'll stay safely on the low level stuff, thanks :)
[18:55:07] <alex_joni> this reminds me of my first days with emc
[18:55:12] <alex_joni> I had no idea what I was doing :)
[18:55:19] <SWPadnos> me too!
[18:55:37] <SWPadnos> I still have no idea when we start talking about task/interp stuff ;)
[18:56:22] <alex_joni> ../../../../boost/boost/signals/slot.hpp:104: instantiated from 'boost::slot<SlotFunction>::slot(const F&) [with F = boost::_bi::bind_t<void, boost::_mfi::mf0<void, QContentPane>, boost::_bi::list1<boost::_bi::value<QContentPane*> > >, SlotFunction = boost::function0<void, std::allocator<boost::function_base> >]'
[18:56:33] <jmkasunich> ewww
[18:56:34] <alex_joni> looks nice, doesn't it?
[18:56:40] <jmkasunich> don't get any of that on you
[18:56:44] <jmkasunich> it might not wash off
[18:56:47] <alex_joni> ../../../../boost/boost/bind.hpp:1142: error: no matching function for call to 'visit_each(boost::signals::detail::bound_objects_visitor&, QContentPane* const&, int)'
[18:57:36] <SWPadnos> you need oil eater and gojo combined to get const& ints off
[19:15:33] <jepler> http://emergent.unpy.net/index.cgi-files/sandbox/lyx2lyx_1.5.3-0jepler1_all.deb http://emergent.unpy.net/index.cgi-files/sandbox/lyx2lyx_1.5.3-0jepler1.dsc http://emergent.unpy.net/index.cgi-files/sandbox/lyx2lyx_1.5.3-0jepler1_i386.changes http://emergent.unpy.net/index.cgi-files/sandbox/lyx2lyx_1.5.3-0jepler1.tar.gz
[19:15:55] <alex_joni> whee.. so I can stop trying?
[19:15:57] <jepler> ^^ lyx2lyx extracted from the lyx on hardy and compiled for dapper.
[19:16:06] <alex_joni> ah, the other way around
[19:17:00] <SWPadnos> just out of curiosity, when you use "new lyx" to edit, it leaves droppings. does the old lyx choke on those, remove them, change them, or what?
[19:17:15] <alex_joni> SWPadnos: you can export as 1.3 from the new lyx
[19:17:21] <alex_joni> and both of them read the files just fine
[19:17:26] <jepler> SWPadnos: if you don't use lyx2lyx, it doesn't work at all. if you use lyx2lyx it introduces undesired whitespace changes that obscure what's really going on
[19:17:27] <alex_joni> the only problem is CVS
[19:17:47] <alex_joni> you get tons of changes even if you only have 1-2 characters changed in a file
[19:17:48] <SWPadnos> oh, I thought there were whitespace problems or something stupid
[19:17:53] <alex_joni> which makes cvs diff useless
[19:17:56] <jepler> yes it is stupid whitespace stuff
[19:18:34] <alex_joni> jepler: so that runs on dapper?
[19:18:36] <SWPadnos> ok, and each version keeps saving in what it thinks is the "right way", so you get 2x as long diffs every change
[19:18:42] <alex_joni> right
[19:18:48] <jepler> (so if we ever switch to the new lyx as the one true version to use, we won't be able to edit on dapper anymore..)
[19:18:52] <jepler> alex_joni: yes
[19:19:09] <jepler> also, my lyx to html converter only knows the format written by dapper-lyx
[19:19:11] <alex_joni> and it also imports from 1.5.x ?
[19:19:25] <alex_joni> ah, right
[19:19:34] <SWPadnos> err - ok
[19:19:47] <alex_joni> jepler: does your lyx2html choke on the whitespace changes?
[19:20:03] <jepler> alex_joni: no, as far as I know it is fine
[19:20:33] <alex_joni> so the long term plan is to switch to hardy and 221 as the file format
[19:20:41] <jepler> I don't have a long term plan
[19:21:00] <alex_joni> longer than today
[19:21:09] <jepler> (221 is the number of the dapper format and apparently 276 is the hardy format)
[19:21:39] <alex_joni> * alex_joni knows
[19:22:01] <jepler> well, the "user story" we want to satisfy is this: Bob wants to contribute to the docs. He uses Ubuntu Hardy, since that's the version we recommended to him for a new install.
[19:22:11] <alex_joni> oh, we do?
[19:22:15] <jepler> don't we?
[19:22:15] <alex_joni> * alex_joni hides in that case
[19:22:35] <alex_joni> sorry.. it's been a long day without doing much.. so I'm kinda unfocused at times
[19:22:46] <jepler> anyway, if that's the case, then I think the solution to pursue is to make the 276 format the one in CVS, and fix up the build process to convert to 221 when required
[19:23:02] <alex_joni> s/when/if/
[19:23:20] <jepler> and then tell people 'edit docs on hardy, not on old dapper', which will enable more people to edit docs than it will exclude
[19:23:23] <alex_joni> so you could in theory have both in cvs
[19:23:32] <jepler> that makes no sense
[19:23:56] <alex_joni> when someone on hardy edits a file that gets converted saved as 276
[19:24:35] <alex_joni> * alex_joni gives up on 1.3.7 for hardy
[19:25:42] <jepler> I think there's something that keeps 1.5.3 from building on dapper too
[19:27:51] <alex_joni> well. .having lyx2lyx 1.5.3 on dapper should be enough
[19:27:57] <alex_joni> if it then loads the new files
[19:28:35] <jepler> it needs makefile work, too
[19:28:45] <jepler> and a new external package, so it has to wait for 2.3 to get to users
[19:29:03] <alex_joni> external package?
[19:29:07] <jepler> yes, this lyx2lyx package
[19:29:13] <jepler> it's a new (build) requirement
[19:29:22] <jepler> I guess I can be a bit more relaxed about build requirements than runtime requirements..
[19:29:23] <alex_joni> is lyx a build requirement?
[19:29:29] <alex_joni> umm.. yeah, for pdf files
[19:29:35] <alex_joni> ok, 2.3 it is
[19:33:38] <jepler> hm it can take a surprising length of time to run
[19:34:15] <jepler> (lyx2lyx --to 221 Master_HAL.lyx has been running for over a minute now)
[19:34:22] <alex_joni> huh?
[19:34:31] <alex_joni> it only converts a couple bytes?
[19:34:40] <jepler> beats me
[19:34:48] <alex_joni> 3243 bytes
[19:35:02] <jepler> maybe something's wrong
[19:36:05] <jepler> well poo-burgers
[19:36:14] <jepler> this is the best idea I could think of, and it doesn't work either
[19:37:02] <alex_joni> real0m0.048s
[19:37:02] <alex_joni> user0m0.020s
[19:37:02] <alex_joni> sys0m0.028s
[19:37:10] <alex_joni> that's on hardy
[19:39:53] <jepler> /usr/share/lyx/lyx2lyx/lyx2lyx --to 276 Master_HAL.lyx > x
[19:39:57] <jepler> ^^ this works quickly
[19:40:15] <jepler> /usr/share/lyx/lyx2lyx/lyx2lyx --to 221 x
[19:40:17] <jepler> ^^ this hangs
[19:41:09] <jepler> same for me on hardy with lyx2lyx from lyx-common or on dapper with my own lyx2lyx package
[19:42:52] <alex_joni> hmm.. does that do it on any file?
[19:43:02] <alex_joni> I worked on src/common/GPLD*
[19:43:08] <jepler> I don't know; master_hal.lyx was the first one the build system tried
[19:43:31] <alex_joni> and I could open it with lyx on hardy (it gets converted automatically on load), then export as lyx-13 (which saves 221)
[19:43:42] <alex_joni> might be something in Master_HAL.lyx..
[19:43:50] <jepler> GPLD works for me too
[19:43:56] <jepler> but still .. it has to be dependable
[19:44:22] <jepler> it has to work on 100% of the files in the docs
[19:45:12] <alex_joni> right.. so it's probably a problem in lyx2lyx
[19:49:32] <jepler> it fails on all the "master" lyx files (or at least takes a very long time to run), and crashes with a python exception on CodeNotes
[19:50:23] <alex_joni> wonder if it tries to convert the included files aswell
[19:50:25] <jepler> and it's also crashed with a python exception on two more files
[19:50:35] <jepler> alex_joni: no, strace shows no sign of opening other files
[19:51:24] <jepler> it also crashed on hal/rtcomps_fr hal/tutorial_fr hal/tutorial gcode/tool_compensation config/ini_homing
[19:51:30] <jepler> so that is a terrible terrible batting average
[19:51:33] <jepler> sigh
[19:51:43] <alex_joni> * alex_joni sighs too
[19:52:13] <jepler> all the ones I bothered to test are the same on hardy, so it's not mispackaging of lyx2lyx.
[19:53:06] <alex_joni> ok.. so we submit a bugreport to lyx
[19:53:10] <alex_joni> and hope they fix it by 2.3 :)
[19:53:20] <jepler> http://www.mail-archive.com/lyx-devel@lists.lyx.org/msg43633.html
[19:54:56] <jepler> no, that must be something else
[19:55:33] <alex_joni> that is related to 220 vs. 221
[19:55:41] <jepler> sigh, I wish I believed there was a better system to use instead of lyx
[19:55:46] <jepler> I'd force everyone to change to it
[19:58:38] <alex_joni> tex? :)
[19:58:55] <jepler> don't tempt me
[19:58:56] <alex_joni> ah, I know.. OO :)
[19:59:37] <alex_joni> bbl
[20:01:18] <jepler> see you later
[20:04:16] <CIA-36> EMC: 03bigjohnt 07v2_2_branch * 10emc2/docs/src/config/stepconf.lyx: added some index entries
[21:48:30] <alex_joni> good night all
[22:01:35] <jepler> fwiw I just looked at MPG behavior compared to feed override. I think that the jogwheel will go at (machine max speed)*(feed override %)/100. I didn't see anything I didn't expect.
[22:01:58] <jepler> in both 95% and 100% FO I could get ahead of the MPG
[22:04:48] <jepler> notable exception: setting FO to 0%, giving the MPG a good whirl, then setting it up from 0%. In this case, the jog appears to be lost!
[22:05:17] <jepler> but setting FO to 1%, whirl, 0%, 100%, the move isn't lost
[22:10:04] <jmkasunich> if (emcmotStatus->net_feed_scale < 0.0001 ) {
[22:10:04] <jmkasunich> /* don't jog if feedhold is on or if feed override is zero */
[22:10:04] <jmkasunich> break;
[22:17:10] <jmkasunich> some people in that thread are talking of "jog override" as if it is distinct from feed override
[22:17:23] <jmkasunich> we've had that arguement, or a related one, before too
[22:17:38] <jmkasunich> some folks want feed override and rapid override as two independent things
[22:17:55] <jmkasunich> today, the override is a linear scale on everything - feeds, rapids, jogs, homing moves, etc
[22:23:14] <SWPadnos> isn't Jon's description of how the MPG works incorrect?
[22:23:40] <SWPadnos> he says that the MPG provides both velocity and position information, so why does override change anything
[22:23:55] <jmkasunich> he's babbling
[22:24:09] <SWPadnos> AFAIK, the only difference between vel mode and pos mode is that after some timeout, the machine stops when in vel mode
[22:24:11] <SWPadnos> that's what I thought
[22:24:19] <jmkasunich> wheel jogs are affected by feed override just like every single other move in EMC
[22:24:37] <SWPadnos> but in both modes, the wheel updates the new "target position", and that's all
[22:24:41] <jmkasunich> SWPadnos: there's no timeout
[22:25:25] <SWPadnos> does it set target_pos to "current_pos + increment"?
[22:25:33] <SWPadnos> rather than "old_target + increment"
[22:25:51] <jmkasunich> it just says "how fast is the machine moving right now", and "how long (in distance) would it take me to stop", then if the user has wound in more than that stopping distance, the target is set to "here + stopping distance"
[22:25:53] <SWPadnos> (I could look it up, but I'm being lazy :) )
[22:26:01] <SWPadnos> ok
[22:26:34] <SWPadnos> which might cause some acceleration, making the stopping distance longer ...
[22:27:14] <SWPadnos> taget = here+min(stopping_distance, jog_target)
[22:27:18] <SWPadnos> target
[22:36:50] <jmkasunich> target = position target from user;
[22:36:55] <jmkasunich> if (vel mode) then
[22:37:11] <jmkasunich> if ( target > here + stop) then target = here + stop
[22:37:32] <jmkasunich> repeat 2nd if for target < here - stop, if going the other way
[22:38:04] <jmkasunich> it won't cause accel
[22:38:27] <SWPadnos> there must be accel somewhere, or there will never be any motion ...
[22:43:45] <jmkasunich> I must have misunderstood you
[22:44:21] <SWPadnos> I'm assuming that the free planner looks at the target position at some point, and decides that it can go faster and still be able to stop at the end
[22:44:39] <jmkasunich> the output of control.c:handle_jogwheels is a target position
[22:44:44] <SWPadnos> yes
[22:44:56] <jmkasunich> that target is the same target used for keyboard jogs, or whatever
[22:45:11] <jmkasunich> that target is INPUT to the free mode planner, which handles accel, decel, etc
[22:45:30] <jmkasunich> for example, if you do a 1" incremental jog, the target jumps by 1" instantly
[22:45:40] <jmkasunich> the free mode planner is responsible for ramping that
[22:45:53] <SWPadnos> right
[22:46:37] <jmkasunich> current position and velocity are available as part of the joint data, as is accel
[22:46:42] <SWPadnos> in vel mode, if you move the wheel one click with the increment sert to 1", does the machine move 1" before stopping, or does it stop before the inch is done since you stopped turning the wheel?
[22:46:50] <jmkasunich> so calculating stop dist is just simple math
[22:47:10] <jmkasunich> it will stop before 1"
[22:47:52] <jmkasunich> oh, I see the issue
[22:47:55] <jmkasunich> and I see how I avoided it
[22:48:08] <jmkasunich> the stop dist is not "stop dist at current speed", it is "stop dist at max speed"
[22:48:38] <jmkasunich> make that "stop dist at (max speed * feed override)"
[22:48:57] <SWPadnos> right
[22:49:14] <jmkasunich> IOW, the fastest it could possibly be moving right now, even if it isn't already moving that fast, or at all
[22:50:18] <SWPadnos> I guess I'm unclear as to how the target gets reset to something closer than the jog target, but it's not terribly important for me to understand it I guess (not this instant anyway)
[22:50:43] <SWPadnos> oh - now I see it, I Think
[22:50:55] <jmkasunich> control.c grep for handle_jogwheels
[22:51:04] <jmkasunich> its really not very complex
[22:51:37] <SWPadnos> no, I'm sure it's easier to understand when your wife isn't bugging you about going for a walk RIGHT NOW!
[22:58:57] <jmkasunich> so go already
[23:26:56] <cradek> I'm pretty sure I fixed velocity mode / feed override interaction once already
[23:31:18] <cradek> yep finally found it: http://cvs.linuxcnc.org/cvs/emc2/src/emc/motion/control.c.diff?r1=1.129;r2=1.130
[23:33:53] <cradek> of course I might not understand the current problem
[23:34:14] <jmkasunich> the current "problem" isn't what anybody thought it was
[23:34:36] <jmkasunich> people wanted velocity mode and didn't know it existed ("problem" 1)
[23:34:54] <jmkasunich> people think there should be separate FO sliders for everything under the sun (problem 2)