#emc-devel | Logs for 2009-05-19

[00:05:06] <jepler> argh, I was bound to forget something
[00:05:13] <jepler> in this case, axis enable pins
[02:54:22] <SWPLinux> hey jepler, did you notice the problem I had with extra_setup() ?
[11:19:53] <CIA-48> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/drivers/hostmot2.lyx: add info about stepgen output
[11:20:48] <CIA-48> EMC: 03bigjohnt 07v2_3_branch * 10emc2/docs/src/drivers/hostmot2.lyx: add info about stepgen output
[12:36:55] <CIA-48> EMC: 03bigjohnt 07v2_3_branch * 10emc2/docs/src/hal/basic_hal.lyx: add info on functions and threads
[12:37:53] <CIA-48> EMC: 03bigjohnt 07TRUNK * 10emc2/docs/src/hal/basic_hal.lyx: add info on functions and threads
[13:10:48] <steves_logging> steves_logging is now known as steve_stallings
[14:58:44] <cradek_> cradek_ is now known as cradek
[15:09:18] <steve_stallings> steve_stallings is now known as steves_logging
[15:41:20] <jmkasunich> found a buglet - dunno if it is axis or the interp itself that is making the assumption that causes it
[15:41:33] <jmkasunich> I'm working on a long part - longer than my machine table
[15:41:41] <jmkasunich> putting slots every 4"
[15:42:11] <jmkasunich> I fixture the part so I can reach one end, touch off there, set to zero, mill first couple slots, no problem
[15:42:23] <jmkasunich> then I re-fixture the part so I can do a couple more slots
[15:42:33] <jmkasunich> I can't reach the end of the part, but I can reach one of the previous slots
[15:42:43] <jmkasunich> I touch off on that slot, setting Y to that coordinate
[15:42:54] <jmkasunich> (say 8")
[15:43:32] <jmkasunich> the new program makes slots at 12 and 16", and does not contain a line that mentions any Y coord less than 8
[15:43:59] <jmkasunich> but the axis backplot assumes that I start at Y = 0, so it draws a line that begins outside the work envelope
[15:44:09] <jmkasunich> and I get the "program exceeds machine limits" warning
[15:44:22] <jepler> 2.2 or 2.3?
[15:44:26] <jmkasunich> 2/2
[15:44:27] <jmkasunich> 2.2
[15:44:33] <jepler> in 2.3, chris changed how initial rapids are treated, for exactly this reason
[15:44:45] <jepler> I suspect, but don't know, that 2.3 wouldn't have the rapid shown and would not give the warning
[15:44:46] <jmkasunich> it doesn't draw them at all, right?
[15:45:00] <jepler> right, I think that the zero or more rapids before any other movements are not show
[15:45:03] <jepler> n
[15:45:14] <cradek> fbofw
[15:45:24] <jmkasunich> ?
[15:45:25] <jepler> fixed bug of the week?
[15:45:33] <cradek> for better or for worse
[15:45:35] <jepler> (no, what's "f"?)
[15:45:36] <jepler> oh
[15:45:58] <cradek> in 2.2 I'm pretty sure you can say continue and it will work fine though
[15:46:07] <jmkasunich> I sure hope so, thats what I'm about to do
[15:46:14] <jmkasunich> nfw do I have time to update
[15:46:32] <jmkasunich> I took today and tommorrow off from work, and I don't know if it will be enough
[15:46:33] <cradek> since gcode is only move destinations, you can't know where the machine starts
[15:49:36] <jmkasunich> it's working
[15:50:50] <skunkworks> If there is anything you guys can think of - we are driving down friday.
[15:51:58] <jepler> we'll let you know
[15:57:32] <skunkworks> I will be bringing version 38 of the h-bridge ;) (and some extra parts) (steves_logging wants to play a little also) Maybe stuart has something big to stick it in ;)
[16:00:19] <jmkasunich> the cinci ;-)
[16:00:40] <jmkasunich> but don't break it - I hope to be doing some machining with it on Sat/Sun
[16:01:38] <skunkworks> heh
[16:07:28] <skunkworks> Cool - for your pcb mill?...
[16:09:14] <jmkasunich> yep
[16:09:32] <jmkasunich> I'm building the main table frame now, I hope to do the table itself there
[16:10:00] <jmkasunich> a couple surfaces on the frame are 24" long and need to be straight and parallel to the ballscrew centerline
[16:10:18] <jmkasunich> then mate table plate to frame, and cut t-slots, etc - 26" long and also need to be parallel to the screw
[16:10:31] <skunkworks> Very neat
[16:10:50] <jmkasunich> I got a great ballscrew from ebay for $80
[16:11:25] <jmkasunich> Hiwin, 25mm diameter, 4mm pitch, preloaded, C3 accuracy grade (8 micron), ~8 years old, but still in factory wrapping
[16:11:40] <alex_joni> hi guys
[16:11:43] <jmkasunich> hi alex
[16:11:53] <alex_joni> jmkasunich: very busy?
[16:12:12] <alex_joni> was wondering if you read back last night, about the hm2-stepgen
[16:12:48] <jepler> in TRUNK I want to change some things about the locations of executables.
[16:13:14] <alex_joni> * alex_joni goes back to his robot simulation
[16:13:19] <jepler> basically, I want to do two things: add a "libexec" dir for executables that usually don't make sense to invoke directly, just from inside emc
[16:13:29] <alex_joni> it's quite heavy on the head when 15 joints are moving
[16:13:37] <jepler> and put executables from scripts and tcl into either bin or libexec as appropriate
[16:13:56] <alex_joni> jepler: sounds good
[16:13:56] <jepler> (an example of a program that isn't usually invoked directly is module-helper)
[16:15:08] <jepler> (of course there will be a fair number of judgement calls, and that's the part I'm not looking forward to)
[19:48:27] <jepler> sf is going to force us to a different issue tracker?
[19:48:38] <jepler> sounds like the perfect moment to decide to set up a bugzilla that we host
[19:49:06] <cradek> ack
[19:49:14] <jepler> http://apps.sourceforge.net/wordpress/sourceforge/2009/05/12/upcoming-feature-deprecation-built-in-forums-docmanager-task-manager/
[19:49:17] <cradek> oh was that in the mail that I didn't read?
[19:49:21] <jepler> yeah probably
[19:49:34] <jepler> "TaskManager: replaced by TaskFreak!, dotProject, Trac ticketing"
[19:50:24] <cradek> !!
[19:51:00] <cradek> heh, http://apps.sourceforge.net/limesurvey/sourceforge/verification.php
[19:51:02] <jepler> trac is a piece of software that poorly integrates bugs, wiki, and source code history viewing
[19:51:17] <jepler> I've never heard of Task!Freak!!
[19:51:26] <cradek> ?!!
[19:51:33] <jepler> I don't have experience with the others
[19:51:53] <cradek> replacing one thing with three things can't be a win, can it?
[19:52:15] <jepler> btfoom
[19:52:24] <jepler> taskfreak is php+sql
[19:52:52] <jepler> and so is dotproject
[19:53:16] <jepler> oh, is dotproject the one michael tried at work? I see that their front page has a gantt chart, which we absolutely don't need in emc
[19:56:05] <jepler> oh and dotproject also has a "calendaring" app
[20:06:43] <jmkasunich> I really wish people would stop fixing things that aren't broken
[20:11:16] <jepler> none of us knows how crappy the sf bug tracker is "inside"
[20:11:38] <jepler> but it's silly that they did two visual redesigns of it in the last year, and now they're ditching it
[20:12:09] <cradek> maybe they discovered that smearing some gradients on top doesn't make it stop sucking
[20:17:18] <jmkasunich> it doesn't? darn
[20:20:08] <jmkasunich> holy shit, they don't even eat their own dog food - it must be bad stuff
[20:23:20] <jepler> yes, as far as I can tell neither of them uses it, at least for public consumption
[20:35:14] <alex_joni> jepler: are you sure they are closing bug tracking?
[20:35:30] <jepler> alex_joni: that's how I interpret the message
[20:35:35] <alex_joni> I don't ..
[20:35:44] <alex_joni> they are closing the task manager
[20:35:53] <alex_joni> which we don't really use
[20:36:10] <alex_joni> I tried to set up some things a long time ago, but nobody followed them ;)
[20:36:32] <alex_joni> we don't use discussion forums
[20:36:36] <alex_joni> we don't use docmanager
[20:36:41] <alex_joni> we don't use diary and notes
[20:36:49] <jepler> well they get a big "fail" in communicating, then
[20:37:38] <alex_joni> I don't see any mention of "tracker"
[20:37:48] <alex_joni> which is the stuff that keeps bugs & feature requests
[20:37:54] <jepler> I didn't remeber "tasks", and so I just assumed they changed the name of "tracker" in there somewhere
[20:38:11] <alex_joni> hopefully not ;)
[20:38:15] <jepler> oh look at that out of date and useless information in "tasks"
[20:38:50] <alex_joni> it was so usefull, no matter they're closing it down
[20:41:55] <jepler> darn, I was looking for a chance to take one more service off of sf
[20:45:55] <alex_joni> heh
[20:46:17] <jepler> bbl
[20:51:13] <alex_joni> good night all
[21:17:00] <jepler> cvs downtime coming up in the next few hours
[21:17:15] <jepler> I'll mention it again just before it occurs
[21:17:17] <jepler> should be short
[21:51:12] <jepler> ok, now's the time. shuold be back in 10-15 minutes.
[22:17:10] <jepler> cvs is now booting .. it should be available in a minute or so
[22:18:12] <jepler> .. should now be up