#emc | Logs for 2006-04-21

Back
[00:00:00] <cradek> there will be a lot of large machinery there that most of us don't have access to at home
[00:00:07] <paelscrit> sounds fun
[00:00:12] <cradek> yep
[00:01:05] <paelscrit> well im gonna go see if i cant get a live disk running with qcad
[00:01:14] <paelscrit> so i can use it at work ;)
[00:01:24] <cradek> have fun
[00:01:30] <paelscrit> bye
[00:08:19] <CIA-4> 03compile-farm 07BDI-TNG (2.4.18-rtai) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot3_log.txt
[00:18:49] <CIA-4> 03jepler 07HEAD * 10axis/tcl/axis.tcl: fix error 'invalid command name .tabs' when pressing ctrl-tab
[00:19:20] <CIA-4> 03jepler 07maint1_2 * 10axis/tcl/axis.tcl: fix error 'invalid command name .tabs' when pressing ctrl-tab
[00:19:35] <jepler> cradek: thanks for catching that bug
[00:19:58] <jepler> I knew where to look for it
[00:20:23] <CIA-4> 03compile-farm 07BDI-4.20 (2.6.10-adeos) * 10v2_0_branch/: build PASSED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot5_log.txt
[00:20:39] <jepler> cradek: does axis 1.3 look OK in the ubuntu gnome desktop now?
[00:21:19] <cradek> I think it looks fine
[00:21:38] <jepler> there were some color problems for awhile
[00:21:49] <cradek> right, nothing was white iirc
[00:22:14] <cradek> let me look, I think I'm running enough of gnome here
[00:22:43] <cradek> it looks right
[00:23:05] <CIA-4> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot4_log.txt
[00:24:03] <jepler> it's about time to consider a 1.3a1
[00:24:38] <CIA-4> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10v2_0_branch/: build FAILED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot4_log.txt
[00:25:38] <CIA-4> 03jepler 07HEAD * 10axis/rs274/options.py: colors were reversed
[00:26:14] <cradek> hey, you fixed that bug already, I didn't even see it
[00:26:21] <cradek> what colors?
[00:26:56] <jepler> the "OK" vs "out of bounds" color for the dimensions
[00:27:19] <cradek> you're right, they sure are red
[00:27:22] <cradek> I hadn't noticed
[00:33:57] <cradek> I'm tempted to say we should use 1.3 with the next TESTING release
[00:34:14] <cradek> its backplot is
[00:34:22] <cradek> far superior
[00:34:33] <jepler> We should do the menu re-org and call it 1.3a1
[00:34:51] <cradek> think we (you?) can do that without breaking the menus?
[00:35:07] <jepler> all it takes is typing
[00:35:12] <cradek> I always screw up .some.of.the.paths
[00:35:27] <jepler> look how long it took for this .tabs problem to show up
[00:35:31] <Jymmm> famous last words
[00:35:43] <jepler> Jymmm: sometimes it takes more typing later on, after you thought you were done.
[00:35:48] <cradek> true
[00:35:51] <Jymmm> or less
[00:36:04] <Jymmm> lol
[00:36:55] <Jymmm> fuckit... lets just pipe /dev/random/ to files till we get what we want =)
[00:37:41] <Jymmm> That would be interesting though
[00:42:09] <CIA-4> 03compile-farm 07BDI-4.20 (2.6.10-adeos) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot5_log.txt
[00:51:24] <CIA-4> 03cradek 07HEAD * 10emc2/src/Makefile: allow clean to remove bad dep files
[00:55:40] <CIA-4> 03compile-farm 07Ubuntu 5.10 (2.6.12-magma) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot7_log.txt
[01:00:17] <CIA-4> 03cradek 07v2_0_branch * 10emc2/src/Makefile: allow clean to remove bad dep files
[01:02:57] <CIA-4> 03compile-farm 07Ubuntu 5.10 (2.6.12-magma) * 10v2_0_branch/: build PASSED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot7_log.txt
[01:06:56] <CIA-4> 03compile-farm 07BDI-2.18 (2.2.18-rtl3.0) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot2_log.txt
[01:08:42] <CIA-4> 03compile-farm 07BDI-TNG (2.4.18-rtai) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot3_log.txt
[01:14:21] <CIA-4> 03compile-farm 07BDI-4.20 (2.6.10-adeos) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot5_log.txt
[01:17:37] <CIA-4> 03compile-farm 07BDI-2.18 (2.2.18-rtl3.0) * 10v2_0_branch/: build PASSED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot2_log.txt
[01:19:12] <CIA-4> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10emc2head/: build PASSED ; see http://linuxcnc.org/compile_farm/emc2head_slot4_log.txt
[01:23:31] <CIA-4> 03compile-farm 07BDI-TNG (2.4.18-rtai) * 10v2_0_branch/: build PASSED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot3_log.txt
[01:28:35] <CIA-4> 03jepler 07HEAD * 10axis/tcl/axis.tcl: menu reorganization
[01:28:43] <cradek> wooo
[01:28:56] <jepler> cradek: try it out, let me know if you want to tweak it
[01:29:35] <jepler> I discovered one bug while testing it, but I think it's an emc2 bug. While stepping, the menu still looks like in the "paused" state: resume and step are active, pause is not.
[01:29:46] <jepler> but that's not new
[01:30:53] <cradek> this is a thing of beauty
[01:30:59] <jepler> it ... is?
[01:31:11] <cradek> all things being relative, yes
[01:31:24] <cradek> I mean it's tremendously better
[01:31:58] <SWPadnos> what's the axis CVS server address?
[01:32:01] <SWPadnos> (or name)
[01:33:01] <jepler> um jas
[01:33:21] <jepler> export CVSROOT=:pserver:anonymous@axis.unpythonic.net:2401/cvsroot
[01:33:41] <SWPadnos> thanks
[01:35:53] <CIA-4> 03compile-farm 07BDI-4.20 (2.6.10-adeos) * 10v2_0_branch/: build PASSED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot5_log.txt
[01:38:05] <CIA-4> 03jepler 07HEAD * 10axis/TODO: Update TODO with 1.3 in mind
[01:38:07] <CIA-4> 03cradek 07HEAD * 10axis/rs274/options.py: remove debug output
[01:43:22] <CIA-4> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10v2_0_branch/: build PASSED ; see http://linuxcnc.org/compile_farm/v2_0_branch_slot4_log.txt
[02:00:29] <CIA-4> 03jmkasunich 07HEAD * 10infrastructure/farm-scripts/ (index.shtml results_cia): shorter IRC message - omit log URL on PASS
[03:34:46] <wholepair> hello -
[03:35:20] <SWPadnos> ah - just in time for bed ;)
[03:36:06] <wholepair> Oh well - how come I don't see a user list - im new to IRC chat and ubuntu - ussually there is a user list
[03:36:39] <SWPadnos> dunno - I have one
[03:36:46] <SWPadnos> what chat program are you using
[03:36:48] <SWPadnos> ?
[03:37:09] <wholepair> X-Chat(2.4.4) came with ubuntu
[03:37:22] <SWPadnos> ah. I'm not familiar with it
[03:37:30] <SWPadnos> (with using it at any rate)
[03:37:37] <wholepair> \?
[03:37:45] <wholepair> thats not a command
[03:37:57] <SWPadnos> try /help
[03:38:29] <jmkasunich> I'm using xchat from ubuntu, I have a user list
[03:38:29] <wholepair> \names
[03:38:39] <jmkasunich> but I don't remember if I had to do anything to get it
[03:38:43] <SWPadnos> use the forward slash
[03:39:35] <wholepair> "/names" prints the users in the chat area - but doesn't show the window - I will look through the drop downs again
[03:40:12] <SWPadnos> it may be behind the main window
[03:41:28] <fenn> i heard new versions of xchat had a crappy UI
[03:42:14] <fenn> hm i guess 2.4.4 isnt that new
[03:42:34] <wholepair> it had a user list yesterday
[03:42:59] <fenn> try dragging on the sides on the window
[03:45:12] <wholepair> aha - I got it!
[03:45:43] <wholepair> it was in settings - preferences
[03:46:22] <wholepair> is anyone going to be around awhile to maybe help me finish my emc2 config
[03:47:32] <SWPadnos> unfortunately, it looks like a few of us are headed off to sleep
[03:48:23] <jmkasunich> wholepair: if you have a quick question I can try to answer
[03:48:29] <jmkasunich> but I need to get to bed soon
[03:48:38] <jmkasunich> (almost midnight here)
[03:49:17] <wholepair> I just finished pasteing a copy of the sim folder in my home directory like the wiki said
[03:49:44] <jmkasunich> ok
[03:49:54] <wholepair> I want to know where to put the ini and the hal files I edited (in same folder I copied)
[03:50:16] <wholepair> which files should I over-write
[03:50:56] <wholepair> these are the files i edited: http://www.bikegeeks.com/EMC2_Configuration/
[03:51:24] <jmkasunich> you copied from /etc/emc2/sample-configs/sim to where?
[03:51:42] <jmkasunich> /home/<you>/emc2/myconfig or something like that?
[03:52:19] <wholepair> yes - what you said - my name shows up in the list to choose friom when I start EMC2
[03:52:59] <jmkasunich> do you have the URL of the wiki page you followed?
[03:53:18] <jmkasunich> I really need to understand _exactly_ what you have so I don't give you the wrong advice
[03:53:26] <wholepair> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?CustomizingConfigsOnUbuntu
[03:55:20] <wholepair> I think I have an idea - the sim is for a simulator - do I need to make a new folder for my machine configuration?
[03:55:45] <jmkasunich> hang on
[03:55:50] <jmkasunich> I'm following the wiki directions
[03:56:05] <jmkasunich> (I usually do this kind of stuff from the keyboard, I want to make sure I have what you have)
[03:56:22] <wholepair> OK - Im sorry - how does emc2 populate the list of machines to choose from?
[03:56:47] <SWPadnos> it searches the install dir (etc/emc2) and your home dir for configs
[03:56:58] <Jymmm> hoe dir?!
[03:57:01] <Jymmm> home dir?!
[03:57:11] <jmkasunich> it looks at every directory in /etc/emc2/sample-configs, and in your home directory <you>/emc2/configs
[03:57:13] <SWPadnos> ho dir
[03:57:35] <wholepair> gothidy hidie ho directory
[03:57:58] <jmkasunich> ignore jymmm
[03:58:21] <Jymmm> jmkasunich Did you not see the correction?
[03:58:25] <jmkasunich> ok, I did what the wiki page says, and when I start emc, I get a tree showing all the configs
[03:59:08] <Jymmm> Now, why the hell would an application be looking in someone's home direcory for configuration files, is beyond me.
[03:59:09] <jmkasunich> then /etc/emc2/sample-configs has a bunch of configs under it
[03:59:31] <jmkasunich> because thats where they put the config for they're specific machine
[03:59:41] <fenn> Jymmm: its a pretty normal thing to do, but its usually hidden by starting the directory name with a "."
[03:59:42] <SWPadnos> Jymmm, that's perfectly normal behavior for many UNIX/Linux program
[03:59:45] <SWPadnos> s
[03:59:47] <wholepair> im not allowed to edit the other files - when I tried it said I did'nt have permission
[03:59:57] <SWPadnos> and that's why ;)
[04:00:03] <wholepair> but I can edit the files in my home directory
[04:00:09] <jmkasunich> right, you don't want to mess with the samples
[04:00:19] <wholepair> right!
[04:00:37] <jmkasunich> so, if Jymmm is done confusing the issue, can we focus on wholepair's situation please?
[04:00:57] <Jymmm> prefernce files is dofferent from config files
[04:01:10] <jmkasunich> wholepair: you have a config directory called /home/<you>/emc2/configs/sim
[04:01:18] <jmkasunich> and you edited some files in there, right?
[04:01:42] <wholepair> yes I have that directory - in it has a sim folder with ini hal var and other foles
[04:01:54] <jmkasunich> ok, now what is the question?
[04:01:54] <wholepair> no I did'nt edit files in there
[04:02:15] <jmkasunich> then what files did you edit?
[04:02:43] <wholepair> I have two files on my desktop - jgstepper.ini and jgpinout.hal - same as ones here: http://www.bikegeeks.com/EMC2_Configuration/
[04:02:57] <jmkasunich> ok, no idea how they got on your desktop
[04:03:11] <wholepair> alex-joni helped me edit them from work - I was on an xp machine at work
[04:03:14] <jmkasunich> did you make them from scratch or something?
[04:03:16] <jmkasunich> oh, ok
[04:03:37] <jmkasunich> so you mailed them to yourself, or posted them on the web and then downloaded to your desktop or some such thing
[04:03:45] <jmkasunich> now I begin to understand whats happening
[04:03:48] <jmkasunich> ;-)
[04:03:52] <wholepair> I took them from online and edited them at work and then copied them from my webage to my ubunto desktop
[04:03:58] <jmkasunich> got it
[04:04:06] <jmkasunich> (I was pretty confused for a while)
[04:04:07] <wholepair> now where do I put them?
[04:04:27] <jmkasunich> in that /home/<you>/emc2/configs/sim directory
[04:05:29] <wholepair> are you sure I shouldnt make a /home/<me>/emc2/configs/realmachine/ directory and leave sim alone incase I want to use its config
[04:05:40] <jmkasunich> you can do that if you want
[04:05:51] <wholepair> ok I will do that - hold on
[04:05:55] <jmkasunich> probably better that way
[04:06:02] <jmkasunich> cd ~
[04:06:11] <jmkasunich> cd emc2/configs
[04:06:15] <jmkasunich> mkdir realmachine
[04:06:23] <jmkasunich> cd realmachine
[04:06:34] <wholepair> im not ding it cammand line - hold on
[04:06:39] <jmkasunich> ok
[04:06:43] <wholepair> doing - command
[04:06:57] <jmkasunich> I understood ;-)
[04:07:42] <jmkasunich> once you do that, when you start emc you should see your realmachine directory, with your ini file inside it
[04:08:17] <jmkasunich> actually, with only one ini file, you'll just see the directory
[04:08:35] <wholepair> yes - I do only see the directory
[04:08:55] <jmkasunich> the problem with creating a directory instead of copying it is that there are files other than your ini and hal file
[04:09:02] <jmkasunich> if you copy, you get everything
[04:09:35] <jmkasunich> do you still have copies of those two files on your desktop?
[04:09:56] <wholepair> yes I still have the copies
[04:10:08] <jmkasunich> ok, go to the home/<you>/emc2/configs dir
[04:10:15] <jmkasunich> and make a copy of the sim directory
[04:10:34] <jmkasunich> you can call it whatever you want, sim2, or mymachine, or whatever
[04:10:49] <jmkasunich> if you already have a realmachine (that you just made) don't call it that
[04:11:18] <jmkasunich> once you have a copy of sim with another name, copy the two files from your desktop into it
[04:11:29] <wholepair> ok - I have a copy with all the files in it - i called it JGreal
[04:11:52] <wholepair> my files are copied in
[04:12:04] <wholepair> is the a confiliction in the folder now
[04:12:15] <jmkasunich> ok, so when you start emc, in /home/<you>/emc2/config you should see sim, with three configs (I think, mini, axis, tkemc)
[04:12:27] <jmkasunich> and you should see realmachine (with nothing under it)
[04:12:37] <jmkasunich> and you should see JGreal with four configs in it
[04:12:45] <jmkasunich> axis mini tkemc, and yours
[04:12:52] <jmkasunich> is that what you see?
[04:13:18] <wholepair> I deleted realmachine - but I see JGreal with axis, mini, tkemc, and, JGstepper
[04:13:24] <jmkasunich> ok
[04:13:38] <jmkasunich> you can click on JGstepper and then click OK, and it should run
[04:14:09] <jmkasunich> if you want to get rid of the other choices, go to the JGreal directory and delete axis.ini, mini.ini, and tkemc.ini
[04:15:45] <wholepair> it's not working - I click on JGstepper and the pretty picture comes up - but the program interface never starts :-(
[04:16:08] <fenn> do you have the .nml files?
[04:16:09] <jmkasunich> are you starting emc from a terminal or an icon
[04:16:50] <fenn> start it from a terminal and tell us what error you get :)
[04:16:52] <jmkasunich> if its a terminal, there will be some messages there telling you what is happening
[04:17:04] <jmkasunich> if its from an icon, you are running blind and you should try a terminal instead
[04:17:07] <wholepair> there is a nml file in the directory - and I am starting from icon
[04:17:19] <jmkasunich> open a terminal and start it from there instead
[04:17:27] <jmkasunich> just type "emc" and hit return
[04:18:11] <jmkasunich> do you get some messages in the terminal?
[04:19:00] <wholepair> the terminal said it can't find core_stepper.hal - then it shuts down - im not blind anymore!
[04:19:20] <jmkasunich> yay
[04:19:34] <wholepair> there is a core_sim.hal in the folder
[04:19:48] <jmkasunich> the reason it can't find core_stepper is because you made the directory by copying sim
[04:19:54] <jmkasunich> and sim doesn't use the stepper
[04:19:58] <jmkasunich> two choices:
[04:20:18] <jmkasunich> 1) you can copy the complete stepper directory from the samples, and then copy your files into it
[04:20:19] <wholepair> I got it - I will edit the ini file
[04:20:25] <jmkasunich> no
[04:20:29] <jmkasunich> you need core-stepper
[04:20:32] <wholepair> no - ok
[04:21:01] <jmkasunich> 1) you can copy the complete stepper sample dir, or
[04:21:16] <jmkasunich> 2) you can copy core_stepper.hal from the stepper sample to your JGreal directory
[04:21:36] <wholepair> I will do 2) - is that ok?
[04:21:40] <jmkasunich> copying the one file (option 2) is probably easier and better
[04:21:48] <wholepair> doing
[04:22:50] <SWPadnos> shouldn't the files refer to stepper.tbl and stepper.nml, etc.?
[04:23:10] <wholepair> are you asking me? i dunno
[04:23:14] <jmkasunich> oh fsck
[04:23:19] <fenn> all the sample configs just point to emc.nml dont they?
[04:23:19] <SWPadnos> nope - jmk
[04:23:23] <jmkasunich> yeah, SWP is probably right
[04:23:27] <SWPadnos> not any more, unfortunately
[04:23:34] <SWPadnos> I'd say option 1 is the best bet
[04:23:37] <jmkasunich> that was a bad bad bad decision on our part
[04:23:51] <jmkasunich> those files should be called emc.nml, emc.var, etc
[04:23:51] <SWPadnos> I agree. what was the reasoning, if I may ask?
[04:24:06] <jmkasunich> not entirely sure
[04:24:15] <wholepair> Ok I will start over - del jgreal - and copy stepper folder to my home directory
[04:24:22] <jmkasunich> carryover from the old arrangement where everything was in one directory maybe
[04:24:26] <SWPadnos> could be
[04:24:37] <jmkasunich> wholepair: that is probably best
[04:24:49] <wholepair> doing
[04:24:55] <jmkasunich> the idea is to start with the sample that is closest to what you need, and then tweak it
[04:25:12] <SWPadnos> I thikn it helps developers the most - we're the only people that are actually likely to use more than one configuration on a single PC
[04:25:30] <jmkasunich> it doesn't even matter for us
[04:25:33] <SWPadnos> I suspect Alex started with stepper when originally doing the files
[04:25:52] <jmkasunich> you can have an emc.nml in the sim dir, and another one in the stepper dir, and another one ......
[04:25:58] <SWPadnos> right
[04:26:35] <jmkasunich> at this point I dunno if we want to screw around with it for 2.0, but by 2.1 I think we should fix that
[04:26:49] <jmkasunich> functionally identical files should have identical names
[04:26:57] <wholepair> oh shit I lost my original files - I will get them off the web again - where is the trash?
[04:27:11] <jmkasunich> I dunno
[04:27:12] <SWPadnos> good question. it seems to me that any layout changes are better done before the release, even if there are already people using the "development" layout
[04:27:19] <jmkasunich> thank goodness for the web
[04:27:39] <SWPadnos> (file layout)
[04:28:14] <jmkasunich> I tend to agree, but there is also the "stop changing shit and release the damn thing" factor
[04:28:32] <jmkasunich> we should discuss it tomorrow on emc-devel
[04:28:35] <jmkasunich> (or on list)
[04:28:43] <SWPadnos> true, but then there's the "why did you change shit after release?" factor ;)
[04:28:44] <wholepair> "stop changing shit and release the damn thing!"
[04:29:18] <jmkasunich> wholepair: the problem with that is once we release it, even more people start using it
[04:29:34] <SWPadnos> heh - I just requested a load of ubuntu CDs - the warning is just a front (we'll see if they actually ship)
[04:29:37] <jmkasunich> then when version 2.1 comes out (with the changes) even more people have to mess with the changes
[04:29:56] <jmkasunich> it might not be a front
[04:30:13] <jmkasunich> it might be easier for them to rout the request to /dev/null than to turn off the page on the website
[04:30:14] <SWPadnos> right - so making decisions about the arrangement of things is best done before release
[04:30:17] <SWPadnos> heh
[04:30:29] <wholepair> ok - I started emc from terminal and pickid my ini file - now I got TKEMC running - it is very blue
[04:30:42] <jmkasunich> blue is good
[04:30:49] <jmkasunich> that means your ini file is OK
[04:30:57] <SWPadnos> heh - what speed is your PC?
[04:31:02] <jmkasunich> there are three different GUIs (at least)
[04:31:10] <jmkasunich> you can pick one that isn't so blue if you want
[04:31:21] <SWPadnos> mini is quite gray
[04:31:35] <jmkasunich> axis is also nice
[04:31:38] <wholepair> maybe I should bring it out to the shop in the backyard now and hook it up to my machine - there is axis too
[04:31:49] <jmkasunich> its past my bedtime
[04:31:54] <SWPadnos> oh - duh. I misread the notice on the ubuntu site
[04:32:07] <jmkasunich> I'll be around tomorrow evening, and others will be around during the day
[04:32:13] <wholepair> good night -
[04:32:17] <SWPadnos> it says they're not taking any requests for the new release (dapper), not "no new requests
[04:32:30] <SWPadnos> night, jkm
[04:32:32] <SWPadnos> grrr
[04:32:34] <SWPadnos> jmk
[04:32:42] <jmkasunich> if you want axis, go in the ini file and uncomment the line that says "# DISPLAY = axis" and comment the one that says DISPLAY = tkemc
[04:32:52] <jmkasunich> "comment" = put a # in front of the line
[04:33:03] <jmkasunich> goodnight everybody
[04:33:12] <jmkasunich> good luck wholepair
[04:33:32] <wholepair> got-it - still need help with one last pinout - my E-stop - but I can wait till tomorrow
[04:34:24] <SWPadnos> that'll be a pretty easy one
[04:34:32] <wholepair> SWPadnose - its a P3 running at 850mhrz
[04:34:41] <fenn> ahah padnose
[04:34:54] <wholepair> drop the e
[04:35:06] <SWPadnos> ok - that should be fast enough for axis to run smoothly
[04:35:17] <SWPadnos> padded nose
[04:35:33] <fenn> wholepair: i really like your wave propagation stuff
[04:35:45] <wholepair> thanks fenn
[04:35:46] <fenn> i've been wanting to do something like that, but with artificial holograms
[04:36:30] <wholepair> fenn: what do you mean - artificial
[04:36:45] <fenn> computer generated
[04:36:51] <wholepair> fenn: do you have a machine
[04:36:56] <fenn> yeah but it sucks
[04:37:11] <wholepair> why does it suck?
[04:37:13] <fenn> i could carve it in foam and cast aluminum into it
[04:37:20] <fenn> it cant handle more than styrofoam
[04:37:35] <fenn> and even then its not too reliable
[04:38:07] <wholepair> oh - you need more rigidity - and precision too - maybe you will upgrade someday
[04:38:10] <fenn> its made out of 2x4's and gas pipes
[04:38:15] <fenn> i'm building a hexapod right now
[04:38:29] <fenn> i need to put motors on the lathe too
[04:38:37] <fenn> too many projects
[04:38:54] <wholepair> ????????????? 2*4's ??????????????? I hope youre not using that for the hexapod
[04:39:09] <fenn> no i got some 0.1" wall stainless tubing today
[04:40:10] <fenn> hoping its not the kind that breaks bits
[04:40:25] <fenn> 316/316L
[04:40:50] <wholepair> we have a tig welder in our shop - I got a fancu robot at the scrap yard last summer - http://www.flickr.com/photos/jeffreygarman/sets/72057594054963039/
[04:41:10] <fenn> * fenn sulks
[04:42:01] <wholepair> 316L - is the more machinable variety - your good with emc2 - im just bi\eginning - how are you gonna control your hexapod?
[04:42:55] <fenn> well supposedly emc can do the math, but i'm skeptical that it will come close to the machine's potential after looking at the source
[04:43:22] <fenn> because of the way accel/velocity limits are handled
[04:43:31] <wholepair> souce - you mean source code.
[04:43:37] <fenn> yeah i always make that typo
[04:44:10] <wholepair> have you tried any other pc controlers - mach2 - turbocnc - yeager
[04:44:29] <fenn> i have used turbocnc
[04:45:09] <wholepair> have you looked at its source - I paid the reg fee and got it - but I don't know pascal
[04:45:29] <SWPadnos> emc is the only one with (a) different kinematics and (b) true servo control
[04:45:35] <fenn> i dont know pascal either.. but at least there are (or were) working examples of hexapods controlled with emc
[04:45:58] <wholepair> cool - your name sounds familiar - are you on cnczone
[04:45:59] <SWPadnos> at least, the only affordable one
[04:46:06] <fenn> it would be easier to rewrite everything from scratch than to try to hack code that wasnt designed for it
[04:46:24] <fenn> i hate cnczone with a passion
[04:47:19] <wholepair> hate is ussually a passionate emotion - why don't you like(imeanhate) cnczone
[04:48:20] <fenn> 1) requiring registration 2) the ridiculous over-commercialization of every page on there 3) nobody knows anything 4) nothing really original
[04:49:09] <Jymmm> that's half the internet
[04:49:22] <SWPadnos> yeah - and 99% of IRC
[04:49:25] <SWPadnos> oops
[04:49:33] <SWPadnos> err - good night ;)
[04:49:37] <fenn> hah
[04:49:46] <SWPadnos> SWPadnos is now known as SWP_Away
[04:50:03] <wholepair> I had to register to get on yeah - the commercialisation is a drag - and I've run into some real jerks there - ha - I thought you went to bed
[04:50:25] <SWP_Away> I rarely get to bed when I first think it's a good idea
[04:50:31] <SWP_Away> see?
[04:50:59] <wholepair> how about the second time you think its a good idea - third fourth fifthe - and all of a sudden the sun is up
[04:54:04] <fenn> ok "nobody knows anything" is a bit harsh
[04:54:18] <fenn> i just got annoyed reading all the threads about open source software development
[04:55:44] <wholepair> yeah - there are some real long threads about that - I got pretty bored of them too.
[04:59:26] <wholepair> well - Im gonna take care of some errands before going to bed - i've got homework -yuck- I will finish my last e-stop configuration later - Im getting close to hooking my machine up to emc2 and getting away from turboCNC - I will have to put an ethernet line out to the shop.
[05:09:03] <Jymmm> you hatin tcnc?
[05:12:35] <wholepair> tcnc isn't smooth enough - I want to be able to turn off decelleration and cut big surfaces - tcnc does wierd things when you run big surface programs - I hardly ever hate - I ussualy just get really frustrated - I do like tcnc for its ease of sttup - it was a good learning experiance
[05:14:38] <wholepair> does emc2 have a decceleration overide like most industrial cnc mills?
[05:15:04] <fenn> you mean feed override?
[05:15:26] <fenn> i dont understand how you can turn off deceleration
[05:15:44] <fenn> constant surface speed maybe
[05:15:56] <fenn> or dont round the corners
[05:16:40] <wholepair> yeah maybe i mean constant surface speed - I think it is a G9 code
[05:17:28] <wholepair> you ussually turn it on before a hole lote of feed moves - I think rapid moves arent affected
[05:17:36] <wholepair> hole - whole
[05:17:51] <wholepair> lote - lot
[05:18:48] <fenn> "G9 Modal Deceleration Override On. " hmm
[05:19:03] <wholepair> i knew it!
[05:19:07] <fenn> what does it do exactly?
[05:19:14] <fenn> or what's the problem it fixes
[05:20:05] <fenn> G9 Set Consatnt Velocity Mode in Mach1
[05:21:24] <wholepair> it takes the jerkyness out - lets say your cutting a bunch of tangent arcs of a die profile - with g9 on you get a real nice clean surface from the side of your endmill - if g9 is off you sometimes get funky stuff on the sides of the surface your cutting - its only good for smooth curves - 90 degree turns are bad - complete switches in direction are bad
[05:22:28] <fenn> ok
[05:22:42] <fenn> in emc that would be g61 and g64 i think
[05:24:37] <wholepair> I need to find a gcode convention for the emc2 - have you programed alot of cnc machines - i've been a cnc programmer for -ever- i've delat with - fanuc - haas - okuma - anilam - bridgeport -
[05:24:40] <fenn> this should help you in translating different dialects of gcode: http://axis.unpy.net/files//gcode.html
[05:24:55] <wholepair> quick draw!
[05:25:05] <fenn> sorry i'm a mindreader
[05:25:36] <fenn> the g64 P option is pretty cool
[05:25:48] <fenn> emc won't deviate more than P units from the path
[05:26:00] <wholepair> that sounds cool
[05:33:29] <Jymmm> wholepair: I meant to ask you what tooling you used to machine those pieces, ballend?
[05:39:04] <wholepair> industrial machine - bridgeport discovery 308 - super ballscrews - real ones! - max rapid on the machine is 400 ipm - its scary - i ussualy keep it set at 100 ipm - its not a production envirement - I work in a university scientific instrument machine shop - we have alot of great machines - but the discovery is the only cnc mill we have in the shop
[05:39:55] <wholepair> - oh - you mean cutting tools - 1/8 four flute ball endmill
[05:40:15] <Jymmm> high step over?
[05:40:22] <Jymmm> to get the grooves?
[05:40:31] <Jymmm> /////////////////////
[05:41:46] <wholepair> I think the smallest stepover I used was 0.015" the biggest 0.030" - it's a trade off - one program I ran took 5 hours - I went to dinner and came back - I ran a haas once that had an auto off - that was great
[05:41:57] <fenn> ugh 400 ipm
[05:43:06] <Jymmm> heh, I really liked them... what were you doing with the ones you molded?
[05:43:35] <Jymmm> err casted
[05:44:15] <wholepair> can you play a .mov video - http://bikegeeks.com/candle.mov
[05:44:26] <Jymmm> yeah
[05:45:39] <Jymmm> Hmmm, something's weird with that
[05:46:55] <wholepair> my new ubuntu firefox doesnt like it - did you see the timelapse candle? I was just goofin around I guess - I had one caste in bell bronze - they make nice gongs.
[05:47:15] <Jymmm> the vertical lock on the video is fubar
[05:47:19] <fenn> do the waves actually act as a waveguide?
[05:48:53] <wholepair> im not sure what affect they have - truthfully - a plane 1/2 inch plate sounds pretty good on its own - I don't understand plate ressonance fully -
[05:50:04] <fenn> yeah i have a 10 foot 2" square piece of CRS.. sounds awesome
[05:51:03] <fenn> i should record that some time
[05:52:12] <wholepair> http://sukeban.code404.com/visual/
[05:52:16] <Jymmm> Heh, I only get that when I'm cuttinf foam =)
[05:52:42] <wholepair> can I see some of your work - do you have a web page
[05:53:56] <fenn> erm, ah, um
[05:54:49] <Jymmm> wholepair: What fenn is trying to say is you put us to shame
[05:55:15] <fenn> http://fenn.dyndns.org/pub/camera/DCP_0605.JPG
[05:55:41] <fenn> its a 3d version of the spirograph algorithm
[05:55:45] <fenn> didnt quite turn out how i expected
[05:56:40] <Jymmm> http://static.flickr.com/33/61765799_f610fa36cb.jpg
[05:56:55] <fenn> there arent any good free/open source cad/cam tools or i'd probably have made a lot more
[05:57:52] <Jymmm> 404
[05:58:40] <wholepair> Jymm: are you a sign maker - Fenn: your image did'nt load
[05:58:48] <fenn> oops i turned off the server
[05:58:52] <Jymmm> lol
[05:59:34] <Jymmm> wholepair It seems to be goin that way at least =)
[05:59:49] <fenn> ok try it again
[06:01:26] <wholepair> thats a really wierd picture - im confused - do you have a camera next to you spindle
[06:01:44] <fenn> the spindle (a dremel tool) is on the left
[06:02:14] <fenn> http://fenn.dyndns.org/pub/camera/DCP_0604.JPG another view of the same
[06:03:25] <fenn> here's a lathe i made, and planning on adding little servos to http://fenn.dyndns.org/pub/camera/DCP_0258.JPG
[06:03:28] <wholepair> ahh - that one makes me less confused - what do you do with the foam when you are done? how do you right your code?
[06:03:45] <fenn> that was a python script originally written by jepler
[06:04:03] <fenn> i can cast aluminum into the foam if i like it
[06:04:17] <fenn> or lay up fiberglass over it
[06:04:30] <fenn> amazing how fast this year's gone by
[06:05:01] <wholepair> amazing how fast my life goes by - nice lathe - is that a concrete base ?
[06:07:07] <fenn> the original cast aluminum bed was too light
[06:07:17] <fenn> so i added 50 pounds of concrete
[06:07:21] <Jymmm> G'Night Folks!
[06:07:31] <fenn> dont let the bed bugs bite
[06:07:51] <wholepair> I played with python a bit when this PC was running XP - I don't know my way around ubunto yet to run any py code - I tried cradeks image to g_code script and a wierd cam package written by the fablab people I think
[06:08:33] <fenn> i want to gather all these scripts together and make a cam software suite
[06:08:45] <fenn> separately they're too hard to track down
[06:10:19] <wholepair> it's hard for me to want to learn anything new - I've been using Mastercam for a while now - I have a version of 8 that will install on XP
[06:11:25] <wholepair> when I want to do surfaces described by formulas I write simple programs in Qbasic, C++, and I tried Python a bit
[06:22:44] <wholepair> G'Night - thanks for everyones help!
[06:38:04] <fenn> http://www.cammhsm.com/ interesting stuff
[08:36:03] <chinamill> morning
[09:09:52] <giacus> morning
[09:12:00] <chinamill> morning giacus
[09:12:49] <giacus> hello chinamill
[09:13:06] <giacus> how the work proceed ?
[09:14:26] <chinamill> slow progress, but still progess
[09:14:44] <giacus> good
[09:45:06] <Bo^Dick> how do i declare an array as an input parameter to a function in the eagle script?
[09:48:42] <chinamill> beats me :(
[10:16:49] <chinamill> BD: did you find a referance manual for eagle script?
[10:22:18] <Bo^Dick> yeah but haven't found the answer
[10:25:31] <rayh> logger_aj, bookmark
[10:25:31] <rayh> See http://81.196.65.201/irc/irc.freenode.net:6667/emc/2006-04-21#T10-25-31
[12:35:49] <jepler> Bo^Dick: looking at the code I've written, it looks like I made any array used by more than one function a global.
[12:37:24] <CIA-4> 03rayh 07HEAD * 10emc2/tcl/bin/halconfig.tcl: Saves INI tuning. Saves a netlist. Doesn't tune from netlist yet.
[12:41:29] <CIA-4> 03compile-farm 07Ubuntu 5.10 (2.6.12-magma) * 10emc2head/: build PASSED
[12:52:44] <CIA-4> 03compile-farm 07BDI-2.18 (2.2.18-rtl3.0) * 10emc2head/: build PASSED
[12:57:07] <CIA-4> 03compile-farm 07BDI-TNG (2.4.18-rtai) * 10emc2head/: build PASSED
[12:59:31] <SkunkWorks> rayh: ?
[12:59:49] <CIA-4> 03compile-farm 07BDI-4.20 (2.6.10-adeos) * 10emc2head/: build PASSED
[13:04:26] <rayh> Hey SkunkWorks
[13:06:59] <CIA-4> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10emc2head/: build PASSED
[13:09:59] <SkunkWorks> logger_aj: bookmark
[13:09:59] <SkunkWorks> See http://81.196.65.201/irc/irc.freenode.net:6667/emc/2006-04-21#T13-09-59
[13:10:27] <SkunkWorks> logger_aj: bookmark
[13:10:27] <SkunkWorks> See http://81.196.65.201/irc/irc.freenode.net:6667/emc/2006-04-21#T13-10-27
[13:11:22] <SkunkWorks> sorry got disconnected
[13:12:32] <SkunkWorks> Just to let you know - don't forget to shut off the ssid in your wireless :)
[13:13:18] <SkunkWorks> they had a neighbor email them saying - "hey - I see you have wireless - how did you get that"
[13:13:25] <SkunkWorks> Oops
[13:13:36] <rayh> Okay. Good thinking.
[13:14:24] <SkunkWorks> they could not connect because of the wep
[13:14:33] <rayh> And be sure that the airport uses real hardware addys
[13:14:52] <SkunkWorks> yes.
[13:15:00] <rayh> Got it.
[13:16:25] <rayh> How far were you getting with that?
[13:19:38] <SkunkWorks> I only tested it to around 300 yards
[13:19:52] <SkunkWorks> (almost antannaless distance)
[13:20:05] <rayh> Not bad at all.
[13:22:28] <jepler> 750GB hard drives announced. wow.
[13:27:48] <SWP_Away> SWP_Away is now known as SWPadnos
[13:28:22] <jepler> hi SWPadnos
[13:28:40] <SWPadnos> h ithere
[13:28:43] <SWPadnos> no - wait
[13:28:45] <SWPadnos> hi there
[13:31:17] <SWPadnos> jepler, who announced the 750G drive?
[13:31:26] <jepler> SWPadnos: seagate (seen on /.)
[13:31:35] <SWPadnos> ah - was just about to check there
[13:32:03] <jepler> somebody should put a bunch of those 1.8" hard drives in a single 3.5" enclosure, and call it an "all-in-one raid"
[13:32:11] <SWPadnos> heh
[13:32:28] <jepler> I'm not saying it's a good idea, I'm saying it would sell
[13:32:37] <SWPadnos> I think you could only fit 4 laptop drives into the 3.5" form factor
[13:33:20] <jepler> do it in a CDROM bay, then
[13:33:36] <SWPadnos> oh yeah - a 5.25" bay would hold many drives
[13:34:19] <SWPadnos> but 4 is enough for raid5 with a hot spare, or 480G of raid5 without the spare
[13:34:46] <SWPadnos> of course, I'm looking at 2,5" drives, not the 1.8's
[13:49:58] <Bo^Dick> is my concept a smart one? rotate an important group of parts in a board layout and log the number of wire crossings for each configuration
[13:58:02] <CIA-4> 03jepler 07HEAD * 10axis/TODO: new item
[13:59:27] <Bo^Dick> chinamill: my nick isn't registered. can you see my replies anyway?
[13:59:53] <Bo^Dick> chinamill: i could see your message
[14:00:20] <CIA-4> 03jepler 07HEAD * 10axis/scripts/axis.py:
[14:00:20] <CIA-4> fix accelerator keys in menu items (needed after re-organization)
[14:00:20] <CIA-4> make "view type" menu items into radiobuttons (yech, looks ugly, may revert)
[14:00:20] <CIA-4> 03jepler 07HEAD * 10axis/tcl/axis.tcl:
[14:00:20] <CIA-4> fix accelerator keys in menu items (needed after re-organization)
[14:00:20] <CIA-4> make "view type" menu items into radiobuttons (yech, looks ugly, may revert)
[14:01:58] <chinamill> BD: no
[14:02:51] <Bo^Dick> chinamill: uuhh. i think i'll have to register my nick in order to reply
[14:03:24] <Bo^Dick> chinamill: "DCC session closed" is what i get
[14:04:48] <Bo^Dick> chinamill: what's the parameters for the "/msg" command?
[14:05:13] <SWPadnos> msg <nickname> <message>
[14:05:40] <chinamill> BD: if you want to send a direct mess without DCC you must be regged with NickServ
[14:05:44] <SWPadnos> (you can find this with /help msg)
[14:06:37] <Bo^Dick> i've forgot what to write in order to register a nick
[14:06:55] <Bo^Dick> and "/help msg" doesn't work for me
[14:07:03] <SWPadnos> /msg nickserv help
[14:07:12] <SWPadnos> what software are you using?
[14:07:18] <Bo^Dick> mIRC
[14:07:31] <SWPadnos> strange
[14:07:48] <SWPadnos> did you do it in this channel, or somewhere else?
[14:08:16] <Bo^Dick> he didn't see my replies that's all
[14:08:29] <SWPadnos> I mean the /help msg command
[14:09:33] <Bo^Dick> no
[14:09:55] <Bo^Dick> i didn't do it in this channel window but in the main freenode window
[14:10:37] <SWPadnos> hmm - that should also have worked, I think. did you get an error message? (like "help is not a valid mIRC command")
[14:11:21] <Bo^Dick> it launched the help window but not about the "msg" command but for mp3 i think
[14:11:30] <SWPadnos> weird
[14:12:08] <Bo^Dick> chinamill: do you see my message now?
[14:19:50] <SkunkWorks> you could have 4 mini hot swap drives in the cd sized opening. That would be cool.
[14:19:59] <SkunkWorks> laptop drives
[14:20:12] <SWPadnos> they'd almost fit in a 3.5" form factor
[14:20:19] <SWPadnos> harder to swap though
[14:21:24] <SkunkWorks> cooling?
[14:21:51] <SkunkWorks> I suppose laptops are pretty packed. Although they usually dont have 4 togather
[14:22:33] <SWPadnos> they don't generally get much airflow in laptops, I think
[17:00:41] <Bo^Dick> question about the eagle script language. does anyone know what command that rotates a component. a already know how i move a component.
[17:03:32] <giacus> morning
[17:09:24] <Jymm> morning giacus
[17:10:34] <giacus> hello Jymm
[17:10:42] <rayh> Hi giacus. Ready for the weekend?
[17:11:02] <giacus> Hi rayh , almost ready :)
[17:12:26] <giacus> should be nice to pass a day in mountains here around
[17:12:47] <giacus> depend on the weather, actually isn't so nice
[18:10:08] <LerneaenHydra> random performance question: response time is very bad when running a program and trying to adjust feedrate for instance
[18:10:18] <LerneaenHydra> lag is something like 2-5 seconds for the display
[18:10:25] <LerneaenHydra> although the toolpath is 5fps or so
[18:24:40] <LerneaenHydra> jepler: how do you use the hal config utility?
[18:24:57] <LerneaenHydra> jepler: calibration I mean
[18:41:43] <LerneaenHydra> anyone else know how to use the calib. util?
[18:46:06] <cradek> I think only ray would know, it's very new
[19:14:40] <LerneaenHydra> cradek: is there a wiki page describing what the parameters from trajectory control/axes section and onwards mean?
[19:15:48] <cradek> give me an example of a parameter you're talking about
[19:17:54] <rayh> I'm here. But on the phone.
[19:19:47] <rayh> I can answer questions about halconfig and emccalib now.
[19:20:05] <rayh> Both of these are prototypes and have no written documentation.
[19:20:52] <rayh> I hope to have documentation for emcfest -- probably get linched iif not.
[19:26:55] <Jymm> NEW ROPE FOR SALE!!! NEW ROPE FOR SALE!!!
[19:27:39] <rayh> Jymm, meant to add CHEAP!
[19:28:07] <Jymm> rayh: Nah, high quality... wouldn't want it to break or anything
[19:28:42] <rayh> I am pretty heavy. My doctor thinks I should grow a few inches so I fit the "normal" chart.
[19:29:02] <Jymm> lol
[19:29:49] <Jymm> cnc corner chisel
[19:31:00] <LerneaenHydra> cradek: things like: linear units, angular units, def vel, max vel, probe index, probe polarity (in traj planner section). Also units, home, max_vel, stepgen max vel, cycle time, inout scale, min limix, max limit, ferror, min ferror, home_(all).
[19:31:13] <LerneaenHydra> there must be some documentation of what they all mean
[19:31:58] <cradek> I think the closest we have is the emc1 documentation of the ini
[19:32:17] <cradek> the homing stuff is described in the emc2 code notes pdf
[19:32:19] <LerneaenHydra> cradek: where is that?
[19:32:22] <cradek> ummm
[19:32:42] <cradek> alex found it the other day... on the wiki I think? did you try the search?
[19:32:54] <LerneaenHydra> not yet
[19:34:11] <cradek> most of that documentation is from before my time - we really need someone to step up and volunteer to update it
[19:34:34] <cradek> ideally, a non-programmer
[19:34:48] <rayh> LerneaenHydra, Did you have specific questions about either halconfig or calibration.
[19:35:13] <LerneaenHydra> yes, the calibration seems to be... hard to get to work
[19:35:58] <LerneaenHydra> it often prompts me when closing that hal parameter <parameter> has changed, do you really want to quit
[19:36:03] <rayh> What kind of system are you running, servo, stepper, ...
[19:36:09] <LerneaenHydra> stepper
[19:36:25] <rayh> Ah. There is not a lot to change in a stepper version.
[19:36:50] <LerneaenHydra> what is there to change, there must at least be max speed, accel, and scaling
[19:37:15] <rayh> Yep.
[19:37:42] <rayh> press apply then save. The quit thing is of no consequence.
[19:37:50] <rayh> I have to go back and fix it.
[19:38:08] <LerneaenHydra> oh, it seemed to me like a "changes unsaved, quit? yes/no"
[19:39:01] <rayh> Right the error is something like that. But if you have saved already you should press yes to quit.
[19:39:09] <LerneaenHydra> ok
[19:39:32] <LerneaenHydra> do the settings in traj control need to be changed?
[19:39:39] <LerneaenHydra> or are they for the feedback loop?
[19:39:40] <rayh> I didn't get save to update the list that it was working from.
[19:39:59] <rayh> * rayh looks.
[19:40:42] <rayh> Those parameters are all read so nothing to do there.
[19:41:05] <LerneaenHydra> so stepper speed changes don't entail a change there as well?
[19:41:45] <rayh> No they don't. Some of those variables need to be read from the ini file so that EMC sees them.
[19:44:09] <LerneaenHydra> do you need to reload EMC after changing the values?
[19:44:18] <rayh> Look at a running stepper_xx with halconfig
[19:44:26] <rayh> In the ini file yes you do.
[19:44:33] <LerneaenHydra> and changing the values seems to cause a joint 0 follow error
[19:45:03] <rayh> For the scale variable that is true unless the axis in question is at 0.0000
[19:45:10] <LerneaenHydra> I set them to scale 1, maxvel 1, max accel 0.1
[19:45:37] <rayh> Scale changes will do that.
[19:46:16] <LerneaenHydra> how do you change scale so that it is set correctly?
[19:46:24] <rayh> There also seemed to be a problem with reversing direction using a minus sign
[19:46:55] <LerneaenHydra> if I have a scale of 10 (10 pulses of the motor gives 1mm of travel)how do I set that up correctly?
[19:47:15] <rayh> I start by computing turns per inch, times belt ratio, times motor steps per rotation.
[19:47:38] <rayh> times microstepping from the drive.
[19:47:53] <LerneaenHydra> in my current setup I want a scale of say 10 (running bare on the table)
[19:48:05] <LerneaenHydra> I have set steptype to 2 (grey code)
[19:48:12] <LerneaenHydra> so I don't have microstepping
[19:48:35] <rayh> okay. you want 10 steps per mm?
[19:48:41] <LerneaenHydra> yes
[19:48:51] <rayh> Then scale is 10
[19:49:17] <LerneaenHydra> and maxspeed of 1 (that should be 1mm/s, right?) and max accel of 0.1 (0.1mm/s^2, right?) which will give a visible accel
[19:49:36] <rayh> Yes
[19:49:49] <LerneaenHydra> I've set all those values to that and EMC gives a joint 0 following error
[19:50:01] <LerneaenHydra> while running g0 x5
[19:50:02] <rayh> I use sound for testing the accel.
[19:50:08] <LerneaenHydra> (the motor I have is X)
[19:51:01] <rayh> Sounds about right. You can increase maxaccel and maxvel as you get a understanding of how the axis acts.
[19:51:24] <LerneaenHydra> yes but it's not working, EMC spits an error message at me
[19:51:27] <rayh> * rayh tests
[19:51:56] <LerneaenHydra> joint 0 following error it says
[19:52:57] <LerneaenHydra> could ferror be too small?
[19:53:50] <rayh> Yep. Ferror and Min_ferror should be multiplied by 25.4
[19:54:06] <rayh> Have to make those changes in the ini and restart.
[19:54:08] <LerneaenHydra> sounds very much like a metric/imperial thing
[19:54:28] <rayh> In fact they are unitless.
[19:54:53] <rayh> but when mm is the unit 0.001 is fairly small.
[19:55:09] <LerneaenHydra> I am running from a stepper_mm.ini and they are currently set at ferror1.27 and min_ferror 0.254
[19:55:52] <rayh> * rayh tests with mm
[19:56:57] <LerneaenHydra> I still get a joint 0 following error when I've changed steps/mm to 200, maxvel to 1 and maxaccel to 1
[20:00:35] <rayh> I set INPUT_SCALE to 10, MAXVEL 1 and MAXACCEL 1 and g0 x5 worked.
[20:01:03] <LerneaenHydra> strange. I had an otherwise standard stepper_mm.ini
[20:01:10] <LerneaenHydra> running in AXIS
[20:02:11] <LerneaenHydra> when I set maxaccel to 520 it works again
[20:02:35] <LerneaenHydra> that and speed to 35
[20:02:36] <rayh> You working on per axis settings?
[20:02:40] <LerneaenHydra> yes
[20:02:55] <LerneaenHydra> where else can I set that up?
[20:03:05] <rayh> I know nothing of the axis interface. I was using the standard configuration.
[20:03:14] <LerneaenHydra> ok
[20:03:23] <LerneaenHydra> I doubt that the interface has anything to do with this
[20:03:41] <rayh> The trajectory planner [TRAJ] has accel and vel settings as well.
[20:04:09] <LerneaenHydra> are those global? when does TRAJ have priority over the induvidual and vice versa?
[20:04:56] <rayh> Traj defines the big path in user space in terms of its values.
[20:05:34] <rayh> Then hacks those up into segments according to it's loop parameter.
[20:05:39] <LerneaenHydra> big path in user terms.. that did't say much
[20:06:41] <LerneaenHydra> or rather, I didn't understand
[20:06:49] <rayh> For coordinated motion it plans all axes at once.
[20:07:01] <LerneaenHydra> like x10y5?
[20:07:15] <rayh> so g1 f10 x1 y2 z-3
[20:07:27] <rayh> all gets planned out in user space.
[20:07:29] <SWPadnos> I believe G code is always coordinated, with the exception that G0 isn't guaranteed to be
[20:07:34] <K4ts> hello
[20:07:40] <LerneaenHydra> uses traj settings instead of per axis?
[20:07:51] <LerneaenHydra> and per axis is only for a single-axis move?
[20:07:54] <rayh> Um, not true in the old emc.
[20:08:03] <LerneaenHydra> that sounds like a strange set-up
[20:08:20] <SWPadnos> no, the per-axis limits are the fastesta particular axis can go, the TRAJ limits are the most the TPP will ever plan for
[20:08:45] <SWPadnos> consider a machine that can go 10m/s on X, but only 5 m/s on Y
[20:08:46] <LerneaenHydra> so per-axis is hardware limitations, and traj is the max planned speeds?
[20:09:25] <SWPadnos> and consider that the TP is limited to 20 m/s
[20:09:34] <LerneaenHydra> m or mm?
[20:09:40] <LerneaenHydra> TP = traj?
[20:09:55] <SWPadnos> program a move in X at 20 m/s, and the TP will limit to the amx X vel of 10 m/s
[20:10:05] <SWPadnos> meters per second, TP = trajectory planner = traj
[20:10:19] <LerneaenHydra> amx?
[20:10:21] <SWPadnos> s/amx/max/
[20:10:27] <SWPadnos> max
[20:10:29] <LerneaenHydra> ok
[20:10:42] <SWPadnos> program a 20m/s Y move, and the TP will limit to the Y max of 5 m/s
[20:10:42] <LerneaenHydra> assuming X was set to max 10 in the hardware limit?
[20:10:47] <SWPadnos> yes
[20:11:10] <SWPadnos> now, consider a move at a 45 degree angle, at 20 m/s
[20:11:14] <LerneaenHydra> then what is the point of having the TP section in the INI if it never goes over the hardware limits?
[20:11:20] <LerneaenHydra> sqrt(2) I guess
[20:11:31] <LerneaenHydra> 14-ish m/s
[20:11:33] <cradek> hi guys
[20:11:36] <LerneaenHydra> if they are both 10
[20:11:37] <SWPadnos> X can move 10, Y can move 5, so they can move a combined rate of ~11.3 m/s
[20:11:41] <SWPadnos> right
[20:11:45] <SWPadnos> hi cradek
[20:11:49] <LerneaenHydra> hello there
[20:11:55] <SWPadnos> hopefully I'm not describing future TP features ;)
[20:12:01] <cradek> what are we arguing about?
[20:12:30] <SWPadnos> pointing out the functionality of separate axis and traj limits (accel and vel)
[20:12:41] <LerneaenHydra> I'm having trouble figureing out what the settings in TP are for in the stepper_mm.ini when there are already settings for max_speed in the per-axis settings
[20:12:42] <cradek> ah
[20:13:05] <cradek> the ini has both tooltip limits [TRAJ] and axis limits [AXIS_*]
[20:13:18] <SWPadnos> theoretically, the TP max vel should be set to sqrt(maxX^2 + maxY^2 + maxZ^2)
[20:13:47] <SWPadnos> or at least it could be set to that, since that'e the highest tooltip speed that the machine can attain
[20:14:10] <LerneaenHydra> but why have the TPmaxvel setting as a value one can change? why not have it as sqrt(maxX^2 + maxY^2 + maxZ^2) always?
[20:14:25] <SWPadnos> you may want to limit it for other reasons
[20:14:42] <SWPadnos> like spindle power / speed (or something)
[20:14:47] <LerneaenHydra> hmm.. ok
[20:14:58] <LerneaenHydra> but setting it to a value over sqrt(maxX^2 + maxY^2 + maxZ^2) is a no-no?
[20:15:14] <SWPadnos> I don't think so, since it will still honor the per axis limits
[20:15:16] <LerneaenHydra> or will that just make the per axis settings act as the cap
[20:15:20] <SWPadnos> yes ;)
[20:15:32] <LerneaenHydra> so for teh moment I can set them to insane values (100000 or so)
[20:15:59] <LerneaenHydra> and if the spindle bogs down becuase of too high feeds or something I can limit it there
[20:16:39] <LerneaenHydra> what does default vel/acc do? when does one ever have a move with "default" feed values?
[20:17:30] <cradek> default vel is the initial jog velocity (or the jog velocity at 100% FO in axis)
[20:17:38] <cradek> I don't thing default acceleration does anything
[20:17:41] <LerneaenHydra> ah, ok
[20:18:02] <LerneaenHydra> however the per axis will still be applied?
[20:18:10] <LerneaenHydra> *per axis max vel/acc
[20:18:48] <cradek> per axis is always applied
[20:19:13] <rayh> There is a gotcha's when it comes to using individual axis max vel and tkemc.
[20:19:24] <LerneaenHydra> ok. what is the difference between max_velocity and stepgen_maxvel?
[20:19:41] <rayh> It does not show the correct jog slider endpoint.
[20:21:01] <cradek> stepgen maxvel has to be set a little higher than axis maxvel because of irritating programming issues
[20:21:57] <rayh> a bit of overhead is needed by stepgen.
[20:22:16] <rayh> a few steps tend to pile up during accel
[20:22:26] <LerneaenHydra> cradek: irritating program issues? so it's like a temporary hack-ish parameter? why doesn't EMC calculate a suitble value?
[20:22:48] <rayh> That's the irritating part.
[20:23:12] <LerneaenHydra> so the maxvel one sees when running the motor is maxvel?
[20:23:13] <rayh> They are parameters in different parts of the system.
[20:23:20] <LerneaenHydra> argh.
[20:23:34] <rayh> so the easy way is to use two of em.
[20:23:59] <cradek> there are differences of opinion about how that should be done, so we make the user deal with it
[20:24:10] <rayh> halcmd does not do math yet.
[20:24:16] <LerneaenHydra> ah..
[20:24:19] <rayh> some think it ought not to do math.
[20:24:25] <LerneaenHydra> bloat?
[20:25:27] <cradek> it wouldn't need to be in halcmd
[20:25:35] <LerneaenHydra> ok
[20:25:47] <rayh> Right. Some think that stepgen ought not to do creative math.
[20:26:49] <rayh> I'm not doubt oversimplifying the issue.
[20:26:55] <rayh> no doubt
[20:27:16] <rayh> it's no doubt to late in the day for me to type.
[20:30:12] <cradek> the issue with stepgen maxvel is that there's a compromise that must be made - some think emc should make it itself, some think the user should have full control over it
[20:30:22] <LerneaenHydra> is there any way to get EMC to start with a pre-selected config, so yo don't always have to select one?
[20:30:50] <LerneaenHydra> if there is any purpose in having control oneself than I agree with the say it's set up now
[20:30:59] <cradek> yes, it's described in the man page (man emc)
[20:31:12] <cradek> there are several ways to start it
[20:32:28] <cradek> most likely you want `emc /path/to/myconfig/some.ini'
[20:32:47] <LerneaenHydra> yes that sounds very suitable
[20:33:04] <cradek> dried coffee must be conductive
[20:33:29] <LerneaenHydra> :|
[20:37:12] <rayh> cola is conductive when dry
[20:37:51] <rayh> Guy I knew trashed a 30K video switcher control panel that way.
[20:38:00] <LerneaenHydra> not fun.
[20:42:05] <LerneaenHydra> when running a program similar to spiral.ngc
[20:42:43] <LerneaenHydra> I get strange behavior, the accel seems to be messing up the signal, as the pulses length jumps around
[20:42:57] <LerneaenHydra> a g3 doesn't do that
[20:43:20] <cradek> a g3 is going to be much smoother than blended g1s
[20:43:34] <cradek> the spiral is there because it's a good test case
[20:43:55] <LerneaenHydra> yes I know that it will be smother but the behaviour is much more jerky than it need be
[20:44:17] <LerneaenHydra> I get "jerks" that are 10-20% of the step period
[20:44:30] <LerneaenHydra> by jerk i mean that the step length changes
[20:44:42] <LerneaenHydra> (I'm running in stepping mode 2,2,2)
[20:48:23] <SkunkWorks> LerneaenHydra: are you running actual hardware?
[20:48:48] <LerneaenHydra> yes
[20:48:59] <LerneaenHydra> measuring pulses with a phyical osciliscope
[20:49:53] <giacus> later
[20:50:22] <SkunkWorks> what proccessor/speed computer are you running?
[20:51:33] <LerneaenHydra> 800mhz PIII
[20:51:45] <LerneaenHydra> coppermine, socket-a like socket
[20:51:49] <SkunkWorks> what base period?
[20:51:58] <LerneaenHydra> 20k ns
[20:53:27] <SkunkWorks> I really don't have anything to add - just wondering what kind of setup you where seeing this on. can you actually detect the modulating pulse train in the hardware?
[20:54:02] <cradek> you may be able to reduce it by reducing your TRAJ period, but it's not going to be perfect
[20:54:38] <LerneaenHydra> an external osciliscope and I can hear it
[20:54:45] <LerneaenHydra> ok
[20:54:45] <SkunkWorks> (I have been goofing around with emc2 for a few months on a big gantry and have not noticed a real performance issue)
[20:55:01] <SkunkWorks> on a 450mhz II
[20:55:58] <SkunkWorks> 2540 steps per inch.
[20:56:56] <SkunkWorks> (10000 steps per inch Z)
[20:57:21] <LerneaenHydra> ok, ATM this is 200steps/mm
[20:59:09] <SkunkWorks> 5080 steps per inch. <- need a refference :)
[20:59:49] <SkunkWorks> I have been meaning to play around with a faster computer - might have to take one home for the weekend ;)
[21:00:48] <LerneaenHydra> I can record a video clip of the behaviour if it would help
[21:15:59] <LerneaenHydra> can one to subroutines in the gcode for emc, like goto line x?
[21:16:16] <cradek> yes, see "O words" on the wiki
[21:17:12] <giacus> ping giacus
[21:17:18] <giacus> am I online ?
[21:17:22] <giacus> O_o
[21:17:39] <K4ts> ah ha
[21:17:44] <giacus> damn wifi connection
[21:18:04] <giacus> who is K4ts ?
[21:18:12] <giacus> /whois K4ts
[21:18:16] <giacus> :D
[21:18:22] <giacus> where you from ?
[21:18:29] <K4ts> naples
[21:18:33] <K4ts> ih ih
[21:19:09] <giacus> sad sad ..
[21:19:09] <giacus> too sad peoples around :(
[21:19:46] <K4ts> rayh: you sleep?
[21:22:25] <SkunkWorks> later
[21:22:35] <giacus> K4ts: yes
[21:22:49] <giacus> there are 4 am
[21:22:59] <giacus> :P
[21:23:16] <giacus> there is the fuse
[21:23:52] <giacus> they are 6-8 hours forward
[21:24:26] <rayh> Hi K4ts I was away testing emc.
[21:25:11] <K4ts> well
[21:25:13] <K4ts> sorry
[21:28:46] <LerneaenHydra> cradek: is there some place where normal users can write down things that they beleive should be implemented in upcoming versions of EMC and gui frontends?
[21:29:16] <LerneaenHydra> for example I think that a press-up key to run previous line a-la bash would be very nice to have in MDI programminf
[21:29:34] <cradek> LerneaenHydra: on sourceforge, there is a bug tracker and feature request tracker
[21:29:57] <cradek> I believe there are categories for each of the guis
[21:30:02] <cradek> that is the place to put that kind of thing
[21:30:05] <LerneaenHydra> ok
[21:32:25] <rayh> There are also future pages at the wiki
[21:39:51] <giacus> rayh: are you doing a tester ??!
[21:39:54] <giacus> :(
[21:40:46] <giacus> that's not good ..
[21:40:50] <rayh> Nope, building one.
[21:41:01] <giacus> oh.. ok then :D
[21:42:48] <giacus> if you need a tester, you got at least two here
[21:42:54] <giacus> I'm the first
[21:43:07] <giacus> and K4ts should be the second
[21:43:17] <rayh> I'll keep that in mind.
[21:43:21] <giacus> K
[21:43:40] <giacus> whan you need to stress something, please let we know
[21:43:42] <giacus> :D
[21:43:46] <giacus> when*
[21:44:01] <rayh> You got it.
[21:44:06] <giacus> haha
[21:44:29] <cradek> we need you to stress test the latest TESTING release made on Apr 9
[21:44:38] <cradek> it will be emc 2.0.0 very soon
[21:44:54] <giacus> well, we'll do it hehehe
[21:45:04] <giacus> ;P
[21:46:11] <giacus> since I'm pretty familiar with emc yet ..
[21:46:24] <giacus> best thing is to let K4ts to try run it
[21:46:30] <giacus> and see what happen
[21:46:36] <giacus> :D
[21:46:54] <giacus> if a bug exist .. I'm sure she'll found it !
[21:47:42] <giacus> so, K4ts be ready for the next gcode file stress test :P
[21:48:55] <giacus> rayh: she's a professional stress-testing ..
[21:49:12] <giacus> don't think she's a newbie
[21:49:20] <giacus> tester
[21:49:35] <giacus> I tried her
[21:50:23] <rayh> There are a precious few who can stress most anything.
[21:50:55] <giacus> uhm
[21:51:50] <giacus> rayh: you can believe me as K4ts can stress, there are no peoples
[21:53:05] <giacus> she's a debugger star*
[21:53:09] <rayh> * rayh doesn't know how to respond.
[21:53:16] <giacus> hehe
[21:53:30] <giacus> K4ts: is reading righ know ..
[21:53:30] <K4ts> rayh: is very bad
[21:53:36] <giacus> :)
[21:53:40] <giacus> why ?
[21:53:46] <giacus> ins't true ?
[21:53:58] <giacus> who is bad ?
[21:54:02] <giacus> rayh: ?
[21:54:15] <giacus> * giacus agreed with K4ts rayh is bad ..
[21:54:20] <giacus> hehe
[21:54:28] <giacus> :D
[21:55:21] <K4ts> no ryw giacus is bad
[21:55:26] <K4ts> rayh:
[21:55:38] <giacus> * giacus wonder a wonderful life without womans around :)
[21:55:46] <rayh> okay.
[21:56:10] <giacus> K4ts: yeah
[21:56:19] <rayh> I don't know. I like the women in my life.
[21:56:45] <rayh> * rayh has four granddaughters.
[21:57:14] <giacus> hehe nice :)
[21:57:25] <giacus> I'm too
[21:58:10] <giacus> but I wonder why womans are so clever as 'stress-testers' :(
[21:58:19] <giacus> and we not
[21:58:26] <giacus> nature s weird
[21:58:54] <rayh> Probably has something to do with multitasking. Women are much better at it than men.
[21:59:05] <giacus> I believe in DNA
[21:59:12] <rayh> on average that is.
[21:59:32] <giacus> talking in a scientific mode
[21:59:48] <giacus> there's something in the DNA
[21:59:59] <giacus> that's is very interesting ..
[22:00:15] <giacus> could be open source !
[22:00:17] <giacus> heheh
[22:00:39] <giacus> but yes
[22:00:45] <giacus> womans are wonderful
[22:00:49] <giacus> :)
[22:00:58] <rayh> some of us would say it's all in the genes (jeans).
[22:01:08] <giacus> hehe
[22:10:27] <rayh> websys, Hi Bob. Nice price for students.
[22:10:59] <websys> well they expressed the interest
[22:11:14] <websys> more users the better
[22:11:27] <rayh> Gonna have to go back to school;)
[22:11:45] <rayh> Dave sure seems happy with it now that he's gotten into it a bit.
[22:12:05] <websys> been meaning to get back to that threading program you wanted but we're knee deep in a ton of projects
[22:12:11] <rayh> Will Synergy be represented at cnc-workshop this year?
[22:12:26] <rayh> np
[22:12:39] <websys> Talked to Todd this AM about at qt int for emc
[22:12:47] <rayh> Todd sent me some good stuff for threading.
[22:13:02] <websys> I'm going to try but so many things are happening at once
[22:13:14] <rayh> I know that feeling.
[22:13:36] <rayh> Way behind on the EMC2 handouts for the workshop
[22:13:57] <websys> I guess feast is always better than famine
[22:14:48] <rayh> When you have time you don't have money, when money no time.
[22:14:57] <websys> yup
[22:15:40] <websys> speaking of time I was just about to bug out for the weekend - just connected so I could log what was going on
[22:15:56] <rayh> See you.
[22:27:25] <giacus> * giacus goes to bed
[22:27:28] <giacus> night