Back
[02:42:45] <seb_kuzminsky> jepler: are you here?
[02:43:49] <jepler> seb_kuzminsky: what's up?
[02:44:26] <seb_kuzminsky> i want to commit something to 2.2 and as you're the release manager i thought i should ask you first
[02:44:37] <seb_kuzminsky> it's a change to bfload that jmk and swp and i have been talking about
[02:45:03] <seb_kuzminsky> it adds a new command-line syntax to bfload, while keeping the old one, so it's backwards compatible
[02:45:18] <seb_kuzminsky> i *think* it won't break anything... passes all my little half-assed tests
[02:45:36] <seb_kuzminsky> also, i think there are no known users of bfload, but there may be lurkers we dont know about
[02:45:41] <jepler> is it necessary to support the new devices?
[02:45:57] <seb_kuzminsky> it will be if/when i commit the hostmot2 drivers to 2.2
[02:46:03] <seb_kuzminsky> in the next month or so
[02:47:05] <jepler> I think it's fine, it sounds low impact and it's for a good reason
[02:47:14] <seb_kuzminsky> cool :-)
[02:48:19] <seb_kuzminsky> what's your thinking on 2.3?
[02:48:31] <jepler> for 2.3 (almost) anything goes
[02:48:44] <jepler> are you asking about retaining the old commandline syntax?
[02:48:48] <seb_kuzminsky> when do you think it will come out?
[02:49:01] <seb_kuzminsky> just curious really
[02:50:19] <jepler> I don't have a specific schedule in mind -- none of the major features that developers are working on (5axis, non-trivial kinematics, and improved cutter compensation come to mind) are really ready for general consumption
[02:50:59] <CIA-30> EMC: 03seb 07v2_2_branch * 10emc2/src/hal/utils/ (Submakefile bfload.c bitfile.c bitfile.h):
[02:50:59] <CIA-30> EMC: This adds support for the new-style command-line syntax, while maintaining
[02:50:59] <CIA-30> EMC: support for the old syntax. The old syntax is marked as "deprecated".
[02:51:58] <CIA-30> EMC: 03seb 07v2_2_branch * 10emc2/src/hal/utils/ (epp.c epp.h): Userspace epp driver, needed for the new bfload.
[02:52:09] <jepler> once one or more "major features" seems like it's ready for users, then I'd start to think about a specific schedule
[02:52:09] <seb_kuzminsky> thx jeff
[02:52:23] <seb_kuzminsky> makes sense
[02:52:56] <jepler> when we were getting close to 2.2 we found ourselves frequently saying "well, you can't really do that in the released version, but..." -- that's a good clue that you should get a new version in the hands of more users
[02:53:09] <jepler> you're in north america, right?
[02:53:50] <seb_kuzminsky> colorado
[02:54:00] <jepler> has anybody tried to talk you into attending cnc workshop yet? It's a good chance to meet some of the other developers and some of our users.
http://www.cnc-workshop.com/
[02:54:06] <jepler> it's in Illinois
[02:54:21] <jepler> near the western border with Iowa
[02:54:39] <seb_kuzminsky> man, i'd love to go
[02:54:50] <seb_kuzminsky> but i dont think my irl commitments will allow it...
[02:55:04] <jepler> too bad -- maybe another year
[02:55:05] <seb_kuzminsky> i have 2 little kids and a busy day job
[02:55:09] <seb_kuzminsky> maybe in another year or two
[02:57:57] <jepler> I'll hold you to that
[02:58:41] <seb_kuzminsky> is that where The Mazak is? that thing looks sweet
[02:58:51] <jepler> yes
[02:59:36] <seb_kuzminsky> how did you and cradek get into emc? you two are friends irl right?
[02:59:40] <jepler> yes
[03:01:08] <jepler> he had a CNC mill and wanted something better than the bundled software. he found emc(1) and while it had its problems, it was better than the other control. (and linux wasn't so much trouble, because we've both been long-time unix users)
[03:01:52] <seb_kuzminsky> what do you guys make? or is it all about the "machine as robot" software aspect?
[03:01:55] <jepler> I used it occasionally to mill circuit boards, and then I got interested in writing a new GUI
[03:02:13] <jepler> and later on, I started working more on all parts of the system
[03:03:21] <jepler> sadly enough, I've probably worked on the emc software 100x as long as I've actually used it to run a physical machine.
[03:03:27] <seb_kuzminsky> hehe
[03:03:58] <seb_kuzminsky> do you have a feel for how many real users of emc there are?
[03:05:23] <jepler> not really; SWPadnos has occasionally looked at the webserver logs to estimate how many ubuntu/emc machines are online (they regularly check for software updates), but one thing we see on irc and the mailing lists is that people don't necessarily have their shop machines online
[03:05:57] <seb_kuzminsky> i saw the blag entry about stuart stevenson's machine shop
[03:09:32] <seb_kuzminsky> brb garage
[03:09:52] <jepler> I'll be heading off for the night .. see you later
[03:14:28] <seb_kuzminsky> bye jeff
[03:19:31] <jepler> seb_kuzminsky: will you also backport the manpage and fix up the packaging to include it?
[03:20:20] <seb_kuzminsky> sure
[04:50:57] <seb_kuzminsky> i can't build the 2.2 deb because the configs/5axis directory is empty and src/Makefile INSTALL_CONFIGS chokes on it
[04:51:07] <seb_kuzminsky> fix the makefile or fix 5axis?
[04:58:33] <seb_kuzminsky> or... fix cvs so you can remove directories! bwahaha
[05:00:39] <CIA-30> EMC: 03seb 07v2_2_branch * 10emc2/debian/emc2.files.in: Added a manpage for bfload, and added it to the emc2 package.
[05:00:47] <CIA-30> EMC: 03seb 07v2_2_branch * 10emc2/docs/man/man1/bfload.1: Added a manpage for bfload, and added it to the emc2 package.
[13:12:00] <jepler> anybody have an opinion about whether the "fix" for debounce should be to make the code match the manpage or vice versa? Using the built in array parsing for kernel parameters removes quite a bit of code from debounce.c:
[13:12:05] <jepler> debounce.c | 100 ++++++++++++-------------------------------------------------
[13:12:08] <jepler> 1 file changed, 20 insertions(+), 80 deletions(-)
[14:43:29] <jmkasunich> jepler: I think when I wrote that code I didn't know about arrays as insmod args
[14:43:58] <jmkasunich> they seem like a better way of doing things, and if the docs already describe that syntax it seems like a no-brainer
[15:18:47] <CIA-30> EMC: 03jepler 07TRUNK * 10emc2/src/hal/components/debounce.c: make the cfg= parameter match the documentation
[15:32:09] <alex_joni> hi
[15:42:43] <CIA-30> EMC: 03fenn 07TRUNK * 10emc2/docs/src/gcode/main.lyx: more explanation of how g64 works
[16:20:25] <skunkworks> jmkasunich:
http://www.cnczone.com/forums/showthread.php?t=25929&page=6
[16:20:36] <skunkworks> even petew put in his 2 cents