#emc | Logs for 2010-09-24

Back
[00:03:30] <theorb> theorb is now known as theorbtwo
[00:06:05] <Jymmm> * Jymmm watches the tumbleweeds roll by
[00:07:29] <rooks> http://instanttumbleweed.com/
[00:32:07] <Endeavour> http://www.flickr.com/photos/thephoton/5018533085/ Here's the beginnings of my machine
[00:33:44] <Jymmm> you're kidding right?!
[00:34:00] <Jymmm> Should I look at the photo now?
[00:34:17] <Endeavour> ?
[00:38:15] <mikeggg> nice baseboards
[01:27:32] <ries_> ries_ is now known as ries
[01:45:49] <mikeggg> shit. it appears i've reached the power limit for the garage
[01:46:10] <Jymmm> dont forget the dryer outlet
[01:46:38] <mikeggg> heh. i didn't
[01:46:47] <mikeggg> that's actually against code I believe
[01:47:26] <Jymmm> so make it a plug then
[01:48:11] <mikeggg> I ran romex through the wall...
[01:49:01] <mikeggg> i'm not tripping the breaker either. but when the compressor trips on, the mill trips off..
[01:52:23] <Jymmm> then you're tripping the breaker... DUH
[01:53:47] <mikeggg> well, no. i meant the program stops running. EMC doesn't throw an error or anything...all my gear is still powered up. It just stops..
[01:54:26] <mikeggg> but yeah, re-reading it. that would be a DUH
[04:35:46] <elmo40> voltage drop.
[05:12:29] <Alexis_> avr with capability of 12Mbit/sec
[05:13:26] <Alexis_> sorry wrong posting channel
[05:15:07] <elmo40> this is neat. http://sourceforge.net/projects/openfactory3d/
[05:29:17] <mendelbuild> mendelbuild is now known as Mendel|mill
[05:38:44] <elmo42> elmo42 is now known as elmo40
[05:52:41] <Alexis_> Alexis_ is now known as Alexi5
[05:53:09] <Alexi5> Alexi5 is now known as Alexi5_
[05:59:32] <Alexi5_> its java
[06:04:55] <elmo40> ya, and?
[06:05:04] <elmo40> so is SweetHome3D
[06:11:21] <Alexi5_> any open-source cam?
[06:13:04] <awallin> Alexi5_: this is the best I've been able to produce so far http://code.google.com/p/opencamlib/
[06:13:19] <awallin> but there is no GUI, and support from heekscad is rudimentary
[06:20:15] <Alexi5_> opencam look very good...
[07:01:00] <elmo40> hrmm. EMC in virtualbox isn't all that great.
[07:01:07] <elmo40> when I close Axis it freezes
[07:01:19] <elmo40> come to think of it... it freezes on my installed machine, too!
[07:15:50] <Alexi5_> virtualbox... realtime not doing great at that state
[07:16:22] <elmo40> still, it runs
[07:16:29] <elmo40> just when I close Axis it goes to shit
[07:17:53] <Alexi5_> why virtualbox anyway?
[07:18:22] <micges_work> elmo40: Axis freeze on your machine? can you describe more about it?
[07:21:49] <elmo40> it says something about 'read only file system' or 'file system locked'
[07:21:52] <elmo40> something like that
[07:21:59] <elmo40> I can't open anything after I close axis
[07:22:13] <elmo40> not on it right now, the wife would kill me (3am here) ;)
[07:22:49] <elmo40> Alexi5_: because I am in Win7 atm, drawing in HeeksCNC and making tool paths. I want to 'see' them run with the simulation.
[07:25:16] <micges_work> elmo40: go to sleep !
[07:25:33] <elmo40> I should
[07:25:47] <elmo40> I want to make this part tomorrow, though... almost done drawing it.
[08:51:41] <MattyMatt> w00t. A axis works exactly how I want it. mill a tooth, turn work, mill another tooth
[09:00:54] <MattyMatt> what controlls the rotation speed in a G1 Annn move?
[09:02:19] <MattyMatt> does it use the feedrate at the tangent from 0,0,0 to the current X,Y,Z?
[09:07:57] <MattyMatt> that's just idle curiosity. I dont have the nerve to do any coordinated 4 axis moves just yet, not while milling a gear on the motor shaft
[09:08:58] <MattyMatt> I need a 4 jaw chuck and a block of foam for this
[09:28:24] <ries_> ries_ is now known as ries
[09:53:32] <Alexi5_> zzz
[10:09:42] <alex_joni> MattyMatt: no, rotary axes are done at max vel, or synched to linear moves, so that the rotary (and the linear axes) start and stop at the same time
[11:32:48] <Fox_Muldr> I begin to like hal programming. now i can use my joystick to move all three axes and set the moving speed with the throttle on the joystick. and all this only starts when a button is pressed to be safe :)
[11:33:13] <Fox_Muldr> good that i have 4 axes on my joystick to do all this
[11:47:21] <Valen> Fox_Muldr are you nick mulder ?
[12:04:50] <Fox_Muldr> my nick is Fox_Mulder and it has always been this. not only mulder ;)
[12:05:02] <Fox_Muldr> so i think you mean another person
[12:14:10] <Valen> there is somebody on cnczone called that
[12:26:36] <skunkworks> tgif
[12:35:37] <cpresser> Fox_Muldr: since you know hot to handle hal_input... is it possible to use a joystick as an analog input, and read it via M66?
[12:35:48] <cpresser> ~s/hot/how/;
[13:32:41] <Fox_Muldr> cpresser: it is possible to use the joystick as an analog input. i doesn't test it in conjunction with and g-code. all axes reads values between -1 and +1 in 256/1024 steps (main axes has 10bit and neighbour axes has 8bit resolution in my joystick) which could easily be calculated to any other range needed. i calculated one axes to have values between 1...100 as a multiplier for the feed rate
[13:34:01] <cpresser> Fox_Muldr: i already read some info on hal_input and just ordered this device: http://leobodnar.com/products/BU0836A/
[13:35:21] <cpresser> my Goal is to use it as lenght-probe. I am engraving cast acrylic which has large thickness tolerances, so I want to measure them out and compensate :)
[13:36:11] <Fox_Muldr> cpresser: that sounds nice. it is quite easy to access all the inputs (analog and digital) from the attached usb joystick within hal
[13:38:38] <Fox_Muldr> i think 50$ is quite good for 8 analog axes and 32 buttons over usb. remains the question how fast it updates
[13:38:55] <cpresser> I am quite anxious now. I will start construction a potentiometer probe right now :)
[13:39:38] <Fox_Muldr> ah 500 updates/s
[13:40:11] <cpresser> thats okay for me, since i only want to measure at one discrete point
[13:41:25] <JT-Work> I use a floating probe with a micro switch to "measure" the height of the metal just before I fire off my plasma torch
[13:42:02] <atmega> what does your table look like (for the plasma torch)?
[13:42:26] <cpresser> so you are 'touching off' on the material?
[13:42:30] <JT-Work> http://s47.photobucket.com/albums/f163/johnplctech/
[13:42:33] <JT-Work> yes
[13:42:58] <JT-Work> actually the torch is floating with the micro switch to "touch off"
[13:43:00] <cpresser> are you doing this automated? via gcode?
[13:43:04] <JT-Work> yes
[13:44:29] <atmega> the work rests on the cross pieces?
[13:44:46] <JT-Work> on the slats yes
[13:45:43] <atmega> looks nice
[13:45:52] <JT-Work> thanks\
[13:46:04] <atmega> how thick can you cut? do the edges require finishing?
[13:46:49] <JT-Work> I usually don't cut much thicker than 1/2"
[13:47:10] <JT-Work> depends on what finish you need on the edges LOL
[13:47:39] <atmega> I don't think I've ever seen a 'raw' plasma cut edge of anything
[13:47:47] <atmega> or if I did, I didn't notice.
[13:49:18] <JT-Work> the edge can vary quite a bit depending on many things from smooth to ugly. On thinner materials the taper cannot be noticed,as you get thicker you can see the taper on the edge
[13:50:47] <atmega> is the taper predictable? does CAM software compensate?
[13:51:45] <JT-Work> compensate in what way?
[13:52:44] <atmega> like cutter radius compensation
[13:53:54] <JT-Work> yes, that is kerf width and it varies with the different nozzles used
[13:54:06] <Valen> you need to angle the cutter if you want a square edge
[13:54:16] <Valen> fancy waterjet places will do it if you ask them
[14:02:45] <alSMT> could someone test this link for enable the universe repository (for lyx-qt, libpth-dev on the wiki http://thread.gmane.org/gmane.linux.distributions.emc.devel/333/focus=340
[16:21:16] <Paragon39> Hello All could someone please supply me with a link to the git download on linuxcnc.org for Ubuntu 6.06?
[16:22:43] <cradek> can you ask for what you want in another way? your question does not make sense to me.
[16:23:11] <cradek> the git on linuxcnc.org has the emc2 source code
[16:23:39] <alex_joni> git source is not Ubuntu-version related
[16:24:56] <Paragon39> I was following this doc http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?git Para 1.1
[16:25:36] <cradek> oh I see, you mean you aleady have ubuntu 6.06, and you want git packages to install on it
[16:26:15] <cradek> you can find them in here: http://linuxcnc.org/emc2/dists/dapper/emc2.3/binary-i386/
[16:26:20] <Paragon39> Yes ... I used use CVS to get branchs etc but it's been a while and things have changed :-)
[16:26:24] <Jymmm> morning
[16:26:59] <Paragon39> Thanks Cradek
[16:30:12] <Paragon39> Oh on another note I wonder if you could help.... I currently have Xming running on Win XP desktop and I throw back an x session from the EMC server OK but when I run EMC or more specifically AXIS the screen goes crazy ie black untill I highlight a window then it displays then goes blank again. Is this a known issue and do you know of any resolvment?
[16:30:58] <cradek> yes, reinstall the windows xp system with a proper OS that gives a working X server
[16:31:17] <cradek> there are many free options to choose from
[16:31:23] <Paragon39> LOL somehow I knew you'd say that....
[16:32:09] <Paragon39> The X desktop works fine until I start EMC then it goes highwire! ?
[16:32:28] <cradek> AXIS uses OpenGL
[16:32:30] <cpresser> Paragon39: try another application that uses openGL
[16:32:49] <cpresser> glxgears for example. most likely XMing cant render openGL
[16:33:14] <Paragon39> Arrr that could be the reason... I couldnt get it to work with VNC either!
[16:33:31] <atmega> xming does gl
[16:35:13] <cpresser> I like to use M66 (analog input). is it possible to wire motion.analog-in-00 to input.N.abs-name-position(hal_input)?
[16:35:27] <alex_joni> http://www.youtube.com/watch?v=4XOa149EY_E
[16:35:39] <cpresser> is motion.analog-in-00 an input-pin?
[16:35:55] <Paragon39> The reason I would like to get this going is that I have a graphics design program located in my comfy computer room while the EMC desktop is located at the bottom of the garden. I have some g-code that I want to test on the EMC server in sim mode rather than having to run up and down stairs... if you get my meaning.
[16:36:35] <alex_joni> cpresser: analog-in-00 sure is an _in_put pin
[16:36:58] <cpresser> Paragon39: perhaps set up a virtual linux-system in your comfy-room
[16:37:32] <cpresser> alex_joni: thanks, so far I was not able to find it in the manual... I am still quite new to emc :)
[16:38:51] <cpresser> ah perfect, its is (float,in). matches the hal_input signal perfectly :)
[16:39:57] <Paragon39> cpresser: that was another option I looking at. Maybe using Sun (oopps Oracle) VirtualBox
[17:04:13] <IchGuckLive> Hi from Germany
[17:08:13] <Fox_Muldr> Paragon39: i use the latest ubuntu 10.04 live-cd installed in a vmware player (freeware) and it works fine. axis grumbles at start once about bad rtai timings but that was to be expected. for simulation even with hardware functions it works very good. :)
[17:09:46] <Fox_Muldr> Paragon39: i connected my usb joystick (attached to my pc) to the vmware and developed a little hal script to use my joystick in axis and even that worked very good ( to proof that hardware access works)
[17:14:24] <Jymmm> Has anyone seen a "thread lock" of sorts? It's sorta like lock-tight, but doesn't get hard and brittle, it's meant to be put on the full threads, sorta like teflon tape for pipe threads but in a liquid form.
[17:15:36] <Jymmm> Hi skunky!
[17:15:55] <skunkworks> hey!
[17:16:00] <skunkworks> took 1/2 day
[17:16:10] <Jymmm> off?
[17:16:27] <skunkworks> yes
[17:16:55] <Jymmm> Got tired of work?
[17:17:11] <skunkworks> here and there
[17:17:13] <elmo40> who doesn't?
[17:17:22] <skunkworks> year end next week - going to be a pain.
[17:17:41] <Jymmm> ah
[17:17:58] <Jymmm> skunkworks: Remember to count by 10's =)
[17:18:01] <elmo40> Jymmm: there are different types of lock-tite
[17:18:06] <elmo40> red, blue and I think grey
[17:18:19] <Jymmm> elmo40: Which one does what I want?
[17:18:30] <Jymmm> red is the brittle stuff
[17:18:37] <Paragon39> Fox_Muldr: Thanks for the input!
[17:18:40] <elmo40> permanent, temporary and hydraulic (or something)
[17:19:27] <Jymmm> This is like teflon tape, but in liquid/paste form.
[17:19:51] <IchGuckLive> loctite 242
[17:20:18] <IchGuckLive> i use this for screw safety and hydraulik
[17:22:04] <elmo40> http://www.loctiteproducts.com/threadlockers.shtml
[17:22:09] <elmo40> more info there ;)
[17:22:26] <elmo40> their parent company is Henkel who also owns Dial (soaps) ;)
[17:22:37] <elmo40> German company.
[17:22:58] <Jymmm> 242 says prevents loosening from shock and vibration and that's what I'm looking for =)
[17:23:05] <Jymmm> IchGuckLive: TY
[17:23:16] <Anon9782345> you know
[17:23:21] <Anon9782345> I usually get a hammer
[17:23:26] <Anon9782345> and bash the threads
[17:23:31] <Anon9782345> where the nut is supposed to stop
[17:23:35] <Anon9782345> then thread the nut on
[17:23:45] <Anon9782345> That'll keep it in place aswell ;)
[17:24:15] <Jymmm> so will welding it
[17:24:19] <Anon9782345> true
[17:24:29] <IchGuckLive> Jymmm: B)
[17:24:36] <Anon9782345> I guess
[17:24:44] <Jymmm> Just because you can, doesn't mean you should.
[17:24:49] <Anon9782345> If you have a welder, so with Jymmm, else, go with the hammer.
[17:24:50] <elmo40> welds can break due to the rigidity of them. locktite won't 'break' since it is not a solid.
[17:25:04] <Anon9782345> you know
[17:25:10] <Anon9782345> Superglue works wonders aswell,
[17:25:17] <Anon9782345> its is viscous enough o get into tight spots
[17:25:19] <elmo40> plus it is lighter, removable and doesn't require electricity to apply ;)
[17:25:42] <Anon9782345> Why not simple superglue?
[17:25:49] <IchGuckLive> Anon9782345: but it is also loosen if you need it just head over 220°C
[17:26:22] <Jymmm> crazy glue dries brittle, I need it for vibration.
[17:26:55] <Anon9782345> Anyone ever worked with
[17:27:01] <Anon9782345> Thinned Resin?
[17:27:19] <Anon9782345> never used the loctite stuff
[17:27:41] <Anon9782345> Anon9782345 is now known as kb18951452
[17:27:54] <kb18951452> just remembered i was using an Anon tag.
[17:28:20] <kb18951452> Been watching the Anon attacks on the MPAA and RIAA
[17:28:25] <kb18951452> Yall heard about that?
[17:28:29] <IchGuckLive> loctite is industrial standard
[17:29:26] <kb18951452> The MPAA (Motion Picture Association of America) asked an Indian company to DDOS (Distributed Denial of Service) TDP (The Pirate Bay) a while back.
[17:29:41] <colin_> Has anyone out there had any luck with userspace python HAL components?
[17:29:59] <kb18951452> A DDOS is like......... its where you send so many requests to, say, open a web page that the web server can't do anything, and gets overwhelmed
[17:30:31] <kb18951452> Anyhow, a community who are involved with TPB put together an army and did the same thing right back
[17:30:39] <kb18951452> Took out severa website for a few days
[17:31:17] <kb18951452> They are still targeting sites.
[17:32:15] <kb18951452> They've opened a can of worms.
[17:32:51] <Jymmm> IchGuckLive: Do you think teflon tape would work too?
[17:33:01] <IchGuckLive> no
[17:33:10] <Jymmm> well, why the hell not?! lol
[17:33:41] <IchGuckLive> it is not fluid to fit the tread
[17:34:38] <IchGuckLive> as you may now the tread only clamps at 2 or 3 windings so you need to seel of a fluid
[17:39:10] <Jymmm> So 242 never dries 100% ?
[17:39:59] <IchGuckLive> it dries in 20min
[17:40:09] <IchGuckLive> but coudt be loosen
[17:42:26] <IchGuckLive> i use it to fix the water hole Stuff from molding
[17:43:03] <IchGuckLive> Ms 10x1 /14x1 into steel
[17:43:28] <IchGuckLive> it works also for GFK molding at 120°C
[17:43:43] <Paragon39> I have just installed the latests branch with git (EMC2-DEV) and am receiving the following error ... /home/emc/emc2-dev/scripts/emc: line 405: 11280 Segmentation fault axis-remote --ping
[17:43:45] <Paragon39> Cleanup done alloc: invalid block: 0x8118fc0: f0 8 ef
[17:43:56] <Paragon39> Any Ideas?
[17:44:09] <IchGuckLive> start the PC over and it will run
[17:44:26] <IchGuckLive> it is the first setup foult
[17:46:00] <Paragon39> IchGuckLive: I'll give it a go... Is this a normal feature on first install?
[17:46:49] <IchGuckLive> yes
[17:47:10] <Paragon39> Oh, OK thanks...
[17:48:24] <IchGuckLive> NP
[17:49:20] <Paragon39_> Paragon39_ is now known as Paragon39
[17:50:00] <Paragon39_> Paragon39_ is now known as Paragon39
[17:50:55] <Jymmm> alex_joni: That was pretty cool. I had to download the video to prevent the lag. Just wished the background music was lower.
[17:56:30] <IchGuckLive> lindsey Jaild till 22 okt B)
[18:11:08] <elmo40_> ok. i am on my linuxCNC box. i will run my part then close Axis and see if i can paste the error i get with the locked file system.
[18:13:38] <elmo40_> where do i set my G54?
[18:34:59] <elmo40_> ok, it froze.
[18:35:01] <elmo40_> now what?
[18:37:11] <alSMT> can i compile --enable-simulator without rt kernel I get this checking for glib... configure: error: no
[18:39:25] <elmo40_> i need a box around my mill :-P such a mess!
[18:43:20] <elmo40_> well, still surviving.
[18:43:24] <elmo40_> no lockup yet.
[18:43:37] <elmo40_> will leave it running while at work... maybe i have a hardware problem :(
[18:44:01] <cradek> when in doubt, run memtest
[18:49:18] <alex_joni> Jymmm: glad you liked it
[18:49:52] <alex_joni> Jymmm: a similar "tune" gets beaten every eastern over here
[18:49:58] <alex_joni> but much more monotonic
[18:53:15] <alSMT> now its complaining GTK2 missing what is the name of the package I need?
[18:53:28] <cpresser> gtk-dev or similar
[18:55:01] <cradek> you can see the list of build deps in the dsc file: http://linuxcnc.org/emc2/dists/hardy/emc2.4-sim/source/emc2_2.4.4.dsc
[18:55:18] <cradek> that link is right for 2.4 and hardy
[18:56:15] <alSMT> how about 10.04
[18:57:28] <alSMT> cradek, thanks
[19:01:46] <alex_joni> whacky idea: http://shweeb.com/index.php?m=image_gallery&g=1
[19:02:05] <alex_joni> alSMT: s/hardy/lucid/ for 10.04
[19:05:50] <cpresser> I want to use MDI_COMMAND = G53 G0 X0 Y0 Z0 within halui and add a pyvcp-button. however, it would be much nice to exectute two MDI-Commands, first G53 G0Z0; then G53 G0X0Y0; is this possible with one button?
[19:06:40] <cpresser> is it possible to have subsequent G-Codes in one line?
[19:11:44] <colin_> cradek: I'm trying to load a python HAL component in axis, but getting an error about waiting to become ready. I can load the component from the HAL command line just fine. Do you have any suggestions for what I might be doing wrong?
[19:12:24] <cradek> are you using loadusr -Wn with the right component name?
[19:14:06] <colin_> I was using -W only, don't think I had the n. Is it significant? I thought I was assigning a name in the python call to hal.component("foo")?
[19:14:26] <cradek> yes - then you have to wait for foo to come ready with -Wn foo
[19:14:35] <cradek> (I think)
[19:14:40] <cradek> I'm reading the halcmd manpage
[19:15:44] <colin_> Hmm ok, I'll give that a try. I'll be really happy if that fixes it as I think I have all the other parts just about ready...
[19:15:51] <cradek> cool
[19:22:26] <colin_> Oh yeah. I think that did it.
[19:36:35] <ckingsbury> cradek: while you're here, can you give me any insight into how Axis's functionality is split up between the Axis.py and Axis.tcl files? I want to hack around the GUI a little bit and I can muddle my way through Python and Tk well enough, but I had a hard time getting a sense of what role the tcl file plays in everything.
[19:37:27] <cradek> ckingsbury: not really. I struggle to understand it too. the tcl makes the gui, the python runs it
[19:38:08] <ckingsbury> Heh, fair enough. Am I right to understand there was a gui generator of some kind involved at some point?
[19:38:51] <cradek> yes at some point but I think axis.tcl has diverged from that a long time ago
[19:41:20] <ckingsbury> OK. So the best of both worlds, then.
[19:43:04] <ckingsbury> Do you think it would be possible to reuse the OpenGL preview component in another Python program? I wasn't sure how encapsulated that was with the rest of Axis
[19:45:24] <micges> ckingsbury: in master there is simple gui called gremlin, it use Axis opengl preview in gtk python application
[19:46:15] <ckingsbury> micges: Oh, cool. I hadn't heard of that before.
[19:46:58] <micges> beside that Axis in master was changed that gl preview reside in library so it can be resused
[19:50:18] <alex_joni> cpresser: use G28
[19:50:50] <alex_joni> there's an incantation that makes it go Z0 first, then X0Y0
[19:51:04] <alex_joni> it should be in the docs
[19:57:31] <cradek> alex_joni: not really
[19:57:43] <cradek> (not in one line of gcode)
[20:01:10] <alex_joni> hmm.. bummer, thought there is a go to home via a point
[20:01:33] <cradek> yes but the waypoint is not in g53
[20:02:05] <cpresser> but thanks for the idea :)
[20:02:33] <alex_joni> so G53 G28 Z0 is not in G53?
[20:02:47] <cpresser> right now I am trying to use comp with axis.2.joint-pos-cmd == 0 to disable movement when Z!=0
[20:02:54] <cradek> I think that's probably just an error
[20:03:39] <cradek> cpresser: that sounds pretty crazy - what problem exactly are you trying to solve?
[20:04:39] <cpresser> i want to have a button which clears the too (move to Zmax), then move to the machines origin
[20:05:08] <cradek> cpresser: you just can't easily do that
[20:05:29] <cpresser> ~s/too/tool/
[20:05:36] <cpresser> mmh.. why not?
[20:05:51] <alex_joni> what if with he uses 2 MDI's triggered by one HAL button?
[20:06:16] <alex_joni> one MDI to go to Z0, the next one (triggerd by in-pos) to go to x0y0
[20:06:46] <cradek> you can queue MDIs so you can just issue both of them, one after another
[20:07:06] <cpresser> so I need a delay? at least a slight one?
[20:07:08] <cradek> that doesn't qualify as "easily"
[20:07:18] <alex_joni> ah, so you can just run both
[20:07:20] <cradek> try it and let us know :-)
[20:07:36] <cpresser> I am just getting into HAL, but i will try :)
[20:07:55] <cpresser> perhaps I will come back later with more specific questions :)
[20:09:24] <cpresser> for not, i think loading a timer and setting the G00Z0-MDI-Command with the button. the timer then runs the G00X0Y0-mdi-command
[20:11:16] <cradek> touchy has an input for a "go up" button
[20:15:29] <micges> cradek: I'm testing now and it seems that g28 doesn't work correclty
[20:15:57] <cradek> heh that depends on how you define "correctly"
[20:16:36] <cradek> the way our g28 works is a compromise that sort of includes how g28 works on various controls
[20:16:57] <micges> I issued 100 to #5163, then in axis in mdi I issued 'g28 Z100'
[20:17:29] <micges> according to docs there should be move in z axis to 100 and then move to X0Y0Z100
[20:17:30] <cradek> that should go to Z100 in your current work system, then Z100 in the unoffset system
[20:17:52] <cradek> then the docs are wrong
[20:18:02] <cradek> looking...
[20:18:39] <cradek> yeah the docs are wrong
[20:19:00] <cradek> if you specify axis words, only the specified axes are moved
[20:20:37] <micges> imo docs are right, and there can be much more achieved with it
[20:20:58] <micges> like going to z max and then to x0y0z100 in one gcode line
[20:21:27] <cradek> no, the waypoint is always in the work coordinate system, so even according to the docs, it would not go to z max for the waypoint
[20:21:36] <cradek> http://git.linuxcnc.org/gitweb?p=emc2.git;a=blob;f=src/emc/rs274ngc/interp_convert.cc;h=d913f7d810f0e6d92d49d9b3c03e5a26bb8220a6;hb=e3ac9c55085a34ba702eb21319562d34ce210440
[20:21:47] <cradek> see the comment at convert_home
[20:22:17] <cradek> emc1 did what the docs say - it was unlike any other control in the world
[20:22:59] <cradek> now we do what everyone else does (people program G91 G28 Z0 to move Z up)
[20:23:36] <cradek> 2391 // if any axes are specified, home only those axes after the waypoint
[20:23:37] <cradek> 2392 // (both fanuc & haas, contrary to emc historical operation)
[20:24:30] <micges> I see
[20:24:33] <micges> thanks
[20:25:08] <cradek> I think G28 works stupidly, but I researched it fully and made it do the least surprising/most customary thing
[20:26:39] <cradek> G0 G53 Z0 is better because it doesn't switch modes, but people are used to writing G91 G28 Z0 for the same thing (yuck)
[20:27:19] <micges> I didn't even imagine that G91 G28 Z0 works
[20:27:42] <cradek> it's apparently what you have to write on fanuc and haas for "move Z up all the way"
[20:28:01] <cradek> if you forget the G91, it means "crash Z into something and then move it up all the way" haha
[20:28:54] <micges> heh right
[20:30:08] <cradek> was it you who was working on g53 + probing?
[20:30:44] <micges> yes
[20:30:56] <cradek> how's that going?
[20:31:17] <micges> disabling TLO before probe works so I didn't move further ;)
[20:31:25] <cradek> darn
[20:32:24] <micges> sadly machine is waiting already 3 weeks for installing (packaged) so I don't even have ability for some play with it
[20:35:10] <micges> cradek: what use you have in mind with G53 probing?
[20:35:42] <cradek> same as you - stuff like tool length measuring
[20:36:51] <micges> I see
[20:39:29] <micges> I should build material probe and play with it
[20:39:46] <Jymmm> whats a material probe?
[20:40:42] <micges> probe in tool holder that can be used to see Z level of material
[20:40:57] <micges> s/see/check/
[20:41:02] <Jymmm> ah
[20:41:42] <Jymmm> http://www.181.fm/player/
[20:46:50] <micges> cradek: what do you think about adding g28.2 that will work like docs describes now?
[20:48:35] <cradek> I'd rather see someone fix halui so it can send a sequence of mdi commands
[20:48:57] <micges> ok
[20:49:43] <cradek> trying to write arbitrarily complex series of motions in one line of gcode is doomed - and adding gcodes to the interp to do those complex things in one line seems like the wrong approach
[20:50:09] <cradek> next guy might want Z, then X, then Y for some reason
[20:50:20] <cradek> third guy has a lathe and wants X then Z
[20:50:56] <alex_joni> o-call?
[20:50:57] <cradek> fixing mdi of 'O<file> call' would be an even better approach
[20:51:02] <cradek> heh
[20:51:04] <alex_joni> heh
[20:51:10] <micges> heh
[20:51:20] <micges> cradek: I see you point
[20:51:27] <cradek> micges: thanks for volunteering to fix O<file> call! :-)
[20:52:01] <micges> hehe I hate O code parts in interp ;)
[20:52:28] <cradek> yeah unfortunately it's broken deep in task
[20:53:15] <micges> in task? didn't heard of that
[20:53:52] <cradek> yeah try putting m66 or g38.2 or anything else that waits for IO in a file sub and use O<file> call
[20:54:02] <cradek> it's like it doesn't know to wait for things that need waiting
[20:54:09] <cradek> even tool change I think
[20:54:23] <cradek> if you put just a bunch of motion moves, I think it works
[20:55:15] <micges> oh, this one I know
[20:55:28] <cradek> bbl, time to go home (weekend!)
[20:56:56] <micges> here is weekend already about 7h
[20:56:57] <micges> bbl
[21:03:24] <alex_joni> it's saturday here
[21:08:29] <alSMT> what lib could this be checking for glBegin in -lGL... no
[21:08:43] <alex_joni> libGl ? :P
[21:09:15] <alex_joni> mesa something (the package name)
[21:10:14] <alSMT> i put the ones from the depends in i thought i re check
[21:11:14] <alSMT> i missed one
[21:13:53] <micges> alex_joni: wee need a road map for emc developers, don't you think?
[21:14:26] <alex_joni> road map?
[21:16:12] <micges> some directions on about what could be changed/added/fixed in emc
[21:16:44] <alex_joni> micges: not sure that works
[21:17:00] <alex_joni> there were a couple of attempts, but in the end everyone does what he likes
[21:17:06] <alex_joni> most of the time only that ;)
[21:17:14] <alex_joni> s/likes/likes or needs/
[21:17:47] <micges> I see
[21:18:46] <celeron55> maybe somebody would look at the list and realize he needs something mentioned in there, and would do it
[21:19:05] <Jymmm> alex_joni: I likes boobs and needs a beers =)
[21:19:26] <alex_joni> drink enough beers and you'll grow some
[21:19:31] <alex_joni> if you haven't already
[21:20:18] <Jymmm> lol
[21:21:26] <Jymmm> Women think men are complicated. They just don't it that us men are really simple. We just want a beer and to see something nakkid.
[21:30:37] <alSMT> damn axis still sag faults
[21:36:23] <micges> alSMT: can you pastebin.com error?
[21:36:28] <micges> and dmesg
[21:37:30] <alSMT> it seem to be libcrypto.so
[21:38:11] <micges> ah I remember
[21:38:44] <micges> what system?
[21:45:27] <tom3p> on the mesa 7i37TA boards, do ppl break it out to 4 sets of [ 8pin & 4pin] ? (8 for enc & power, 4 for Vcmd&Ena )
[21:47:43] <willeo6709> my mesa boards will be on the way next week, but I think you can pin them out however makes most sense to you
[21:47:47] <alSMT> brb
[21:47:49] <willeo6709> just write it down...
[21:48:56] <tom3p> i was thinking that encoder cables and control cables are always(?) different
[21:53:10] <willeo6709> i don't think they haveto be
[22:10:55] <alSMT> micges, dell intel 1 ghz p4 onboard video http://pastebin.com/sshnWYvM
[22:13:15] <alSMT> tkemc runs
[22:15:03] <micges> I wonder if line '[ 35.912020] [drm:drm_reclaim_locked_buffers] *ERROR* reclaim_buffers_locked() deadlock. Please rework this' have something in common
[22:18:56] <cradek> I'd try switching to vesa driver and use software rendering
[22:19:05] <cradek> I agree those drm warnings look really suspicious
[22:19:32] <tom3p> gimbal using ring primitive in vismach http://imagebin.ca/view/xTuLgg.html
[22:19:41] <tom3p> a lot of PMK need gimbal
[22:20:44] <cradek> cool!
[22:25:12] <alSMT> i'll give it wirl
[23:26:45] <Mendel|mill> im looking for some kind of compability list with the EMC2 software
[23:26:56] <Mendel|mill> Endeavour:)