#emc-devel | Logs for 2010-02-25

[13:15:55] <CIA-2> EMC: 03mshaver 07master * r593ce181cdc0 10/configs/smithy/ (11 files): Changed eztrol plugin library version numbers in anticipation of the 2.3.5 release.
[13:19:02] <CIA-2> EMC: 03mshaver 07v2.4_branch * r4b4a86797513 10/configs/smithy/ (11 files): Changed eztrol plugin library version numbers in anticipation of the 2.3.5 release.
[13:23:17] <fenn> i recommend a complete rewrite of AXIS in likepython http://www.staringispolite.com/likepython/
[13:24:26] <alex_joni> I recommend submitting a patch for review
[14:39:03] <mozmck> fenn: um, like, you know, I don't know about that!
[14:47:47] <skunkworks_> fenn: !
[15:00:24] <skunkworks_> logger_dev: bookmark
[15:00:24] <skunkworks_> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emcdevel/2010-02-25.txt
[16:05:25] <cradek> seb!
[16:11:27] <seb_kuzminsky> hey! :-D
[16:11:44] <seb_kuzminsky> the garage has never been closer to accepting your mill :-)
[16:11:50] <cradek> yay!
[16:11:54] <cradek> have power yet?
[16:12:07] <seb_kuzminsky> not yet... but i'm hoping to get it early next week
[16:12:18] <seb_kuzminsky> then i need to buy a vfd & some other stuff
[16:12:44] <seb_kuzminsky> the steel rods you use to roll the mill around, they're 1/2" by about 4 feet?
[16:13:46] <cradek> yes I think they're 1/2
[16:13:59] <cradek> I bet they're 5-6 feet but for the bridgeport they don't have to be that long
[16:14:21] <cradek> you can have a couple if you remind me
[16:14:53] <cradek> be sure you have a long prybar - the normal 2' nail remover probably doesn't cut it
[16:17:52] <cradek> on another note - from the buildbot failure email, I can't find my way to the failed build's output. is there a better url you could put in there?
[16:18:09] <cradek> brb
[16:41:19] <seb_kuzminsky> yeah the buildmaster email could be better
[16:41:27] <seb_kuzminsky> it's because of how i've set up the various builders
[16:42:50] <seb_kuzminsky> on a checkin, the "master-checkin" build runs, it triggers "master-build"
[16:42:59] <seb_kuzminsky> if master-build fails, it stops there and sends out an email
[16:43:12] <seb_kuzminsky> if master-build worked, it triggers master-package-rt and master-package-sim
[16:57:32] <seb_kuzminsky> this is good because if the build fails we get just one email, instead of one email per platform/config
[16:57:49] <seb_kuzminsky> but it's bad because it's hard to find the failure, as you say
[16:57:51] <cradek> yeah that sure sounds like a good feature - we used to get bombarded
[17:19:36] <seb_kuzminsky> i talked to the guys on #buildbot, there's a way to generate a custom email, and the info we want is accessible, but they admitted their data structures are a huge undocumented mess, so i'm gonna save this for some time when i'm not supposed to be doing my day job ;-)
[17:21:41] <jepler> boo for day jobs
[17:22:01] <seb_kuzminsky> yes, but hooray for paychecks
[17:22:10] <cradek> seb_kuzminsky: thanks for trying
[17:22:32] <cradek> seb_kuzminsky: I'm sure if I try hard, I'll find a way to survive a little bit longer
[17:22:49] <cradek> (in other words, it's not a big deal)
[17:22:53] <seb_kuzminsky> for now i guess the way to do it is to click on the "Buildbot URL" link, then the waterfall, then look for red at the top
[17:23:02] <cradek> ok
[17:23:04] <seb_kuzminsky> yeah agreed, this is not a show stopper ;-)
[17:23:14] <seb_kuzminsky> but still, automation *should* be totally awesome :-)
[17:30:09] <CIA-2> EMC: 03mshaver 07v2.4_branch * r61dff8e1d922 10/configs/smithy/924.ini: Fixed two typos in 924.ini.
[17:37:19] <seb_kuzminsky> mmm, drive rations
[17:37:55] <cradek> -lMAX_VELOCITY = 1.875
[17:38:00] <cradek> +MAX_VELOCITY = 1.875
[17:38:06] <cradek> someone didn't test before pushing...
[17:38:58] <seb_kuzminsky> at least he caught it before 2.4.0 shipped :-)
[17:39:09] <cradek> true
[17:41:05] <seb_kuzminsky> do we have any way to validate .ini files without actually loading their .hal files? read & parse but dont run
[17:42:13] <cradek> to catch this, we'd have to declare that either a missing MAX_VELOCITY is an error or an unused key lMAX_VELOCITY is an error
[17:42:26] <cradek> I bet currently neither of those is an error
[17:42:44] <seb_kuzminsky> i see
[17:43:16] <cradek> I guess I don't have much idea how to test for a "valid" ini file
[17:44:09] <cradek> some things should be errors, like when you switched around MIN and MAX
[17:44:58] <seb_kuzminsky> i guess there's a long list of rules like "must have MAX_VELOCITY", "[TASK]TASK must be a known executable we install", etc
[17:45:51] <seb_kuzminsky> <shrug>
[17:46:41] <cradek> boooring
[17:48:08] <CIA-2> EMC: 03mshaver 07v2_3_branch * rdfdbbe9aa4b0 10/configs/smithy/924.ini: Fixed two typos in 924.ini.
[17:49:04] <CIA-2> EMC: 03mshaver 07master * r16762f91bd95 10/configs/smithy/924.ini: Fixed two typos in 924.ini.
[18:08:22] <JT-Work> hi seb_kuzminsky
[18:09:47] <seb_kuzminsky> hi John
[18:10:24] <JT-Work> the link on your buildbot for the 2.4 deb was wrong when I tried it the other day
[18:10:50] <seb_kuzminsky> jeff said the same thing a few days ago and i fixed it (i think)
[18:10:57] <seb_kuzminsky> could you see if it works for you now?
[18:11:10] <JT-Work> one second
[18:13:24] <JT-Work> Oh heck I can't try from here but it looks correct.
[18:14:19] <JT-Work> you might add this link to that page too http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UPDATING
[18:15:04] <seb_kuzminsky> i think that link already is on the buildbot page
[18:15:12] <seb_kuzminsky> "Upgrading from 2.3"
[18:15:44] <JT-Work> slap on the forehead moment for me
[18:16:02] <seb_kuzminsky> heh - thanks for the feedback
[18:16:10] <JT-Work> ok
[18:16:16] <seb_kuzminsky> have you run the debs? did they work for you?
[18:16:30] <JT-Work> yes, they are on both my plasma and the hardinge
[18:16:37] <seb_kuzminsky> cool! :-)
[18:20:41] <seb_kuzminsky> btw i wont be going to the cnc workshop this summer :-(
[18:33:51] <micges> seb_kuzminsky: why?
[18:36:34] <seb_kuzminsky> it conflicts with a big group camping trip i'm going on with friends & family
[18:44:19] <micges> ah ok
[18:44:54] <micges> I'm planing camping this summer too :)
[18:44:59] <seb_kuzminsky> nice :-)
[18:45:26] <micges> bought new car for this
[18:45:48] <micges> old one couldn't get further than 100km witohut fixing
[18:57:16] <jepler> that's no good
[18:57:21] <jepler> seb_kuzminsky: that's too bad
[18:58:11] <seb_kuzminsky> jepler: yeah i
[18:58:13] <seb_kuzminsky> i'm bummed
[18:58:17] <alex_joni> emc2 fest in the woods?
[18:58:23] <alex_joni> routers only
[18:59:04] <cradek> seb_kuzminsky: maybe we'll all be able to go to mpm for another week some other time of the year?
[18:59:12] <seb_kuzminsky> ooh, yes!!
[18:59:27] <cradek> or your place!
[18:59:37] <cradek> we'll work on your new lathe
[18:59:53] <seb_kuzminsky> heh
[19:00:20] <jepler> seb_kuzminsky: you got a lathe?
[19:00:21] <alex_joni> * alex_joni thinks cradek needs a robot for part swapping
[19:01:08] <seb_kuzminsky> i have a little manual 7x14 that needs a cnc conversion
[19:01:09] <cradek> jepler: he hasn't even decided he needs one yet...
[19:01:25] <cradek> oh right!
[19:01:32] <seb_kuzminsky> i got it at the same time i got my x2
[19:01:37] <alex_joni> seb_kuzminsky: the one you _need_ is a multi turret with synchronized milling
[19:01:41] <alex_joni> that can cut squares
[19:02:05] <seb_kuzminsky> first i need to build another story onto my garage to house all this stuff!
[19:03:10] <alex_joni> 2 stories up, 3 down
[19:03:56] <cradek> maybe you will be set up to make some good quality retrofit kits for those little machines...
[19:04:09] <seb_kuzminsky> yeah i'm pretty excited about that :-)
[20:17:52] <seb_kuzminsky> my new fave coffee-snob drink: dirty chai (aka rasta chai)
[20:18:25] <cradek> does that mean chai and coffee?
[20:19:07] <seb_kuzminsky> chai + espresso :-)
[20:19:38] <cradek> american chai or indian chai?
[20:19:46] <seb_kuzminsky> hmm.. i'm not sure
[20:23:42] <skunkworks_> * skunkworks_ just got a mocca out of the vending machine....
[20:24:26] <seb_kuzminsky> brb coffee!!
[20:28:52] <clytle374> trying to build git hostmot2. I haven't patched anything yet, just testing the setup.
[20:29:18] <clytle374> make: *** [fw/3x20-1/SV24.BIT] Error 127
[20:44:11] <clytle374> ./build.py i20 SVST2_8_72
[20:44:24] <clytle374> workdir /home/cory/emc/hostmot2-firmware/5i20_SVST2_8_72_work
[20:44:35] <clytle374> xst -ifn scr
[20:44:49] <clytle374> exited with 127
[20:44:55] <clytle374> That's all I get
[20:48:11] <jepler> did you download and install Xilinx 9.2? Did you ". /path/to/Xilinx92i/settings.sh"? Use a pastebin site like pastebin.ca or pastebin.com to show all the output starting with the line where you typed "make".
[20:50:57] <clytle374> I ran the setting.sh . Other wise you get other errors. do you want the make -d? If not the above is the whole thing.
[20:53:18] <clytle374> Funny, it will run on this machine. 32 vs 64 bits?
[20:54:24] <clytle374> I'm going to try Xilinx 11.1, unless someone knows it won't work.
[20:54:47] <jepler> beats me
[20:55:10] <jepler> if you run "xst -help" does it show a help message? (wow, on my system it takes 9 seconds for that to complete! not fast)
[20:55:49] <jepler> Over here: $ xst -help
[20:55:49] <jepler> Release 9.2.04i - xst J.40
[20:55:49] <jepler> Copyright (c) 1995-2007 Xilinx, Inc. All rights reserved.
[20:55:51] <jepler> ...
[20:56:07] <clytle374> http://pastebin.com/umQATr8g from before
[20:56:33] <clytle374> no xst
[20:57:06] <clytle374> in hindsight the message was quite explanatory
[20:57:23] <jepler> har
[20:58:28] <clytle374> where does xst come from? it's new to me
[20:59:05] <clytle374> libxstr?
[20:59:10] <jepler> no, it's a part of Xilinx
[20:59:22] <jepler> $ . ~/Xilinx92i/settings.sh ; which xst
[20:59:22] <jepler> /home/jepler/Xilinx92i/bin/lin/xst
[21:02:48] <clytle374> still no good
[21:04:07] <clytle374> I was missing the 32 bit compatibility libs and had problems earlier with the setup of Xilinx
[21:05:17] <clytle374> running xst gives: error while loading shared libraries: libXilinxPkg.so: cannot open shared object file: No such file or directory
[21:05:25] <clytle374> so on to find that.
[21:09:22] <jepler> I remember hacking the settings.sh on a 64-bit system
[21:09:23] <jepler> elif [ $PLATFORM = "x86_64" ]
[21:09:23] <jepler> then
[21:09:23] <jepler> PLATFORM=lin
[21:09:36] <jepler> try doing this in a copy of settings.sh
[21:09:52] <jepler> (lines 8-10 in my copy)
[21:12:20] <clytle374> okay, changing PLATFORM=lin64 to lin
[21:13:58] <clytle374> That works. Thanks a bunch. I'll write that down.
[21:14:58] <jepler> now you wait .. if you know the eventual pathname of the firmware you're interested in, you can give it on the commandline of make
[21:18:28] <clytle374> I saw that using build.py, I skipped all that trying to reduce the possibility of me introducing an error
[21:18:33] <clytle374> while I'm here, there is a PIN_SV12IM_2X7I48_72.vhd in git
[21:19:48] <clytle374> I haven't looked at the patches yet. one step at a time. Little bitty baby steps;)
[21:27:29] <jepler> yes, but I think it's not yet the updated one from peter (it was missing PWM enable pins)
[21:27:46] <jepler> (I have some local changes I haven't pushed yet)
[21:30:09] <jepler> I think the 7i48 bitfile I put on my website is with the pinfile that is in git, which doesn't have the enable pins
[21:31:05] <clytle374> so are the files you linked me to yesterday what I should be using?
[21:31:39] <clytle374> I just got the computer up and running and tested this morning. I'm happy with the results http://imgur.com/ue2ay
[21:32:35] <jepler> excellent
[21:33:00] <clytle374> Almost sad it's going to be a servo machine.
[21:33:32] <skunkworks_> those atom boards also seem to do decent opengl
[21:34:40] <jepler> anyway, I'm not sure of the impact of the missing enable pins in the pinout. it seems like you can hook enables to the corresponding GPIO pins using hal and the board will work right
[21:35:09] <jepler> other than that the bitfile you would get and the ones on my website should be functionally the same
[21:35:30] <jepler> until you have another change that I haven't pushed, the .PIN files will not correctly describe the muxed qcounter pins
[21:36:58] <clytle374> where is you website?
[21:37:48] <jepler> http://emergent.unpy.net/files/sandbox/0001-untested-support-for-muxed-encoders.patch
[21:38:07] <jepler> http://emergent.unpy.net/files/sandbox/sv12im_2x7i48_72.pin http://emergent.unpy.net/files/sandbox/sv12im_2x7i48_72.bit
[21:38:15] <clytle374> oh, sorry I had that. didn't know it was your website.
[21:38:20] <jepler> yep
[21:38:50] <clytle374> glxgears was still pretty smooth. That flashplayer based moving radar map was the biggest hit of everything running.
[21:41:21] <jepler> anyway, I just pushed some accumulated hostmot2-firmware changes to git.linuxcnc.org including the fix to the .pin file generator and the corrected PIN_*7I48*
[21:43:52] <clytle374> Thanks. Again thanks for the help. I'm going to get my head around this yet. This is all to cool to not know how the guts work.
[21:44:40] <jepler> you seem to be picking stuff up pretty quickly
[21:46:27] <clytle374> Jack of all, master of none.
[21:48:26] <jepler> bbl
[22:52:34] <clytle374> um, well. I guess the saying is a little different now then when grandpa always said it. Wiki says it means competent in all trades, grandpa's version didn't include the competent part.
[23:58:52] <CIA-2> EMC: 03cradek 07v2.4_branch * r495f5af080f5 10/src/emc/rs274ngc/interp_find.cc: make this error message less mysterious