#emc-devel | Logs for 2009-08-20

[02:02:19] <dgarr> jepler: haltcl sounds neat. If i understand, one can include both current hal files and .tcl files in the ini file as long as the appropriate syntax is used in each?
[02:03:19] <dgarr> for interactivity, tkcon might be useful -- you can source a tcl file and then open the powerful tkcon console
[02:12:31] <jepler> dgarr: yes
[02:13:05] <jepler> there's already a tcl interface to hal, but this makes tcl files usable as [HAL]HALFILEs and adds some sugar like automatic [expr] for the argument of setp..
[06:37:30] <micges_work> good morning all
[06:41:23] <CIA-8> EMC: 03micges 07joints_axes3 * r9d8a404ed4ea 10/src/emc/usr_intf/axis/scripts/axis.py: Make override limits working in Axis
[07:36:52] <micges_work> hi alex_joni
[07:40:23] <alex_joni> hi
[07:40:49] <CIA-8> EMC: 03micges 07joints_axes3 * rd644a65c0866 10/src/emc/usr_intf/axis/scripts/axis.py: Fix bug that override limits is grayed out when machine is on limit
[08:32:41] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[08:52:13] <micges_work> I think feedoverride shouldn't be applied to jog moves, what do you think?
[10:38:31] <alex_joni> micges_work: why not?
[10:40:49] <micges_work> because it doesn't work in ja3
[10:40:54] <micges_work> ;)
[10:48:13] <alex_joni> that's not a reason ;)
[10:50:37] <micges_work> when you operate machine by FO, when you stop machine you don't have to mess up with FO to fast jog machine to show material cutted
[10:51:04] <micges_work> but maybe thats also not a reason ;)
[10:52:20] <micges_work> our solution was added "M50 P0" to escape key
[10:53:00] <micges_work> but thats is less important
[11:05:45] <alex_joni> right ;)
[11:12:57] <CIA-8> EMC: 03micges 07joints_axes3 * r31e2431ecd5c 10/src/emc/motion/ (control.c mot_priv.h motion.c): Add joint free planner current velocity HAL pin
[11:15:04] <skunkworks928> alex_joni: Played with the puma560 - very neat. cool seeing it virtually cut tux ;)
[11:18:44] <alex_joni> skunkworks928: run it in sim, not RT, if you can
[11:18:53] <alex_joni> I saw that the kins have some issues once in a while
[11:21:49] <alex_joni> sometimes the joint values passed to the kins are 0,0,0,0,0,0
[11:21:59] <alex_joni> and that's a singularity where the inverse doesn't work
[11:24:54] <alex_joni> skunkworks928: you can switch to joint mode, jog the joints to another configuration, then switch back to world mode, and play with it some more
[11:30:20] <alex_joni> and the ABC mappings aren't quite right
[11:35:45] <skunkworks928> ah
[11:36:01] <alex_joni> actually .. they are right, but hard to define ;)
[11:36:20] <alex_joni> I think you can have more than one set of ABC angles for the same orientation
[11:36:45] <skunkworks928> I did have some issues where it seemed like I couldn't issue a mdi command. It would seem like it would start to move then stop. could jog though
[11:37:04] <skunkworks928> then the mdi commands seemed to work.
[11:37:10] <alex_joni> did you try to MDI outside the working range?
[11:37:22] <skunkworks928> I don't think so because I could jog it there.
[11:37:41] <skunkworks928> and then it went there after I jogged - but I will play with it some more :)
[11:38:17] <skunkworks928> the abc mapings are a bit above my head (hard to wrap my head around the movements. seems to keep tool tip in the correct location though :)
[11:39:23] <skunkworks928> bbl - have to go to work
[11:39:25] <alex_joni> yeah, the kins seem to be working ok
[11:39:42] <alex_joni> once it's working ok, we'll have to add a tool ;)
[11:39:47] <skunkworks928> heh
[11:40:37] <alex_joni> I am surprised how well the kins are working
[11:41:16] <skunkworks928> :) I am not
[11:54:10] <CIA-8> EMC: 03micges 07joints_axes3 * r352c55515bf3 10/src/emc/motion/control.c: Execute free mode planner calculation only when joint is active, some small cleanup
[12:06:47] <alex_joni> skunkworks928: at about 5800 lines of code, I am
[12:10:22] <CIA-8> EMC: 03alex_joni 07master * rc3c910214b0b 10/lib/python/vismach.py: remove debug prints
[12:16:43] <christel> [Global Notice] Hi all, I'm about to do some re-hubbing and take some servers down in preparation for todays maintenance, it will be somewhat noisy as I swap things around but shouldn't take particularly long. Apologies for the inconvenience, thank you for flying freenode and have a great day!
[12:24:01] <alex_joni> http://juve.ro/blog-files/puma/puma-tux2.jpg
[16:08:39] <jepler> hm -- the hostmot2 5i20.ucf file doesn't specify the IOSTANDARD of the IOBITS<> pins. Are they LVTTL?
[16:12:08] <CIA-8> EMC: 03jepler 07master * r776f70787178 10/src/autogen.sh: make sure timestamps are updated
[16:46:56] <pcw_home> Oops, probably relied on the GUI for setting the IOSTANDARD and missed something
[16:46:58] <pcw_home> I'll fix that and update the 5I20 bit files.
[16:46:59] <pcw_home> Unfortunately the source has diverged significantly because I changed the setup to 1 file per
[16:47:01] <pcw_home> pinout (so there's no longer a giant ball of all pinouts file) and got rid of the per card type top level files
[16:50:12] <jepler> hi pcw_home
[16:50:24] <jepler> thanks for the clarification
[16:51:51] <jepler> I don't think we have any local changes in the firmware files; they'd be useless since we don't build the firmwares ourselves anyway
[16:52:39] <jepler> but if explicitl IOSTANDARD settings are in the source next time we update the files, that'd be dandy
[16:53:38] <pcw_home> I think JMK made soe scripts for command line building that my new source will break :-(
[16:53:44] <pcw_home> (some)
[16:53:57] <jepler> he and I have both worked on that at times, but none of it is actually being used
[16:54:42] <jepler> I still would like to be able to do a commandline build from our source tree but I'm short on round tuits, and the practical benefit is small
[16:56:39] <pcw_home> I think the new source will actually make it easier FWIW
[16:58:09] <jepler> sounds good
[16:59:23] <pcw_home> So I'll fix the .ucf file and make new 5I20 bitfiles.
[16:59:25] <pcw_home> What to do about the source? (many files deleted and added)
[17:00:40] <jepler> unless you want to learn git, I guess just send a zip or tar file and we'll replace what's there now with the new stuff
[17:00:54] <jepler> the new source control management software is much better about adding/removing files and directories than cvs was
[17:00:59] <jepler> bbl, it's lunchtime here
[17:01:56] <pcw_home> OK I'll send a snapshot
[17:01:58] <pcw_home> Bye!
[22:22:32] <cradek> jepler:
[22:22:38] <jepler> hi cradek
[22:22:46] <cradek> hi!
[22:22:48] <cradek> File "/home/chris/emc2.trunk/lib/python/rs274/glcanon.py", line 78, in draw_dwells
[22:22:52] <cradek> if lathe: axis = 1
[22:22:54] <cradek> NameError: global name 'lathe' is not defined
[22:23:00] <cradek> (in master)
[22:23:05] <jepler> thanks, will look
[22:23:30] <jepler> I did most of the gl refactor in a private branch, but I did push the first commit :-/
[22:26:50] <cradek> wheeeee! my "new" servo amps work
[22:27:13] <cradek> I have to turn all the little knobs a bit more, but things look good.
[22:27:44] <jepler> woo
[22:28:21] <cradek> what a relief
[22:28:23] <jepler> if you just want to get running, delete the 'if lathe' line
[22:28:29] <cradek> ok, thanks
[22:28:44] <jepler> I hope to push a fix in just a minute
[22:30:08] <skunkworks928> cradek: from moses?
[22:30:25] <cradek> nope, from ebay
[22:30:49] <cradek> the stuff from moses is not exactly the same - good for parts
[22:31:18] <skunkworks928> wow - cool
[22:31:53] <CIA-8> EMC: 03jepler 07master * rb6eee9ec1bd6 10/ (2 files in 2 dirs): fix failure to load lathe programs with dwells
[22:32:13] <jepler> git rebase is cool
[22:32:19] <cradek> jepler: this wasn't a lathe
[22:32:26] <jepler> cradek: it wasn't?
[22:32:34] <jepler> oh I guess it wasn't ..
[22:32:48] <jepler> so my commit message was pretty bad, wasn't it
[22:32:51] <jepler> I think the fix is right
[22:33:17] <jepler> anyway, with rebase I've been able to work on this branch for days, make multiple commits, and just now incorporate the fix I made on master
[22:33:27] <jepler> all with a nice linear history, so I can post it as a changeset if I want
[22:33:38] <cradek> cool
[22:33:49] <cradek> uh-oh, help me with git here
[22:34:00] <jepler> oh, you had your local edit and then pulled?
[22:34:09] <cradek> no, I haven't pulled yet
[22:34:11] <jepler> oh
[22:34:23] <jepler> then you want to 'git checkout glcanon.py' before you do
[22:34:29] <skunkworks928> cradek: still have the spindle problem?
[22:34:29] <cradek> locally, I have a (now out of date) master with random_toolchange merged into it
[22:34:41] <jepler> oh, hm
[22:34:51] <cradek> should I reset --hard to origin/master, then pull, then remerge?
[22:35:12] <cradek> seems like a plain pull won't just work
[22:35:29] <cradek> skunkworks928: yep but I have a vfd on the way (should be here tomorrow)
[22:35:42] <jepler> as long as you don't have any changes not on origin/master or origin/random_toolchange
[22:35:47] <cradek> throw money at it - problems go away
[22:35:57] <cradek> oh just try a pull?
[22:36:00] <jepler> .. that you want to retain
[22:36:16] <cradek> oh you're saying that yes I should do what I said earlier?
[22:36:49] <jepler> yes, I think that's a sensible thing to do as long as you don't have any local changes (committed or uncommitted) that you care about
[22:36:57] <jepler> in the branch you're currently on
[22:37:10] <cradek> right, I don't
[22:37:46] <jepler> though just fetching and merging origin/master might be fine too
[22:38:04] <jepler> after discarding the change to glcanon.py that would generate a merge conflict, anyway
[22:38:04] <cradek> looks like it worked...
[22:40:18] <robh_> cradek, what VFD u replacing with? im still looking around whats out there for our spindle motor
[22:42:01] <cradek> a reliance electric 7.5hp 230v drive
[22:43:35] <cradek> jepler: that fixed it - thank you
[22:44:55] <robh_> with brake resistors? have to let us know how it works out be very intrested to see how responsive it is
[22:46:08] <robh_> aah baldor drives i did take a look at theres
[22:46:17] <jepler> cradek: thanks for noticing it
[22:48:28] <jepler> it's a reminder that maybe I should do more than load the splash screen
[22:49:29] <alex_joni> g'night all