#emc | Logs for 2008-02-06

[03:48:07] <gezar_> gezar_ is now known as gezar
[04:48:37] <ds2> wheee it is finished
[04:53:30] <gezar> what did you finish ds2?
[04:54:04] <ds2> work in plastic
[04:54:14] <ds2> let me put stuff on
[04:56:25] <renesis> where is mr cradek
[04:56:30] <renesis> cnc over usb
[04:56:33] <renesis> possible?
[04:56:37] <renesis> what options?
[04:57:05] <renesis> i figure bitbanging the actual step drivers or handling servo feedback would prob have to be offloaded to hardware
[04:57:45] <renesis> but yeah no idea if this exists already and maybe emc knows how to talk to it or what
[04:58:04] <SWPadnos> renesis, if you can live with 4ms cycle time (vs 1 ms by default), then USB might be usable
[04:58:18] <SWPadnos> if you can't live with that, then USB isn't usable
[04:58:33] <SWPadnos> there are no drivers for the various USB devices out there
[04:58:43] <renesis> ooh
[04:58:53] <ds2> gezar: this: http://www.ugcs.caltech.edu/~huny/clip.mpg
[04:58:57] <renesis> cuz i think it would be like half of a control unit, dont right
[04:59:18] <renesis> er, done right
[04:59:21] <SWPadnos> it is possible (but not implemented nor planned) to make a USB-connected device that does all PID and some of the trajectory planning, but that wouldn't be EMC
[04:59:43] <SWPadnos> (unless the device can run Linux, and the relevant portions of EMC are ported to it)
[04:59:51] <renesis> ha yeah, nuts
[05:00:10] <renesis> is there any emc/arm things?
[05:00:14] <SWPadnos> well, I shouldn't say it isn't planned - I am working (slowly) on an ARM board to replace the Rabbit on a G-Rex
[05:00:19] <SWPadnos> none at the moment ;)
[05:00:43] <renesis> interesting
[05:00:44] <SWPadnos> there are several USB devices, but none work with true feedback, much less to the PC
[05:01:01] <SWPadnos> they're all step/dir based, though some have feedback, it's unused for control
[05:01:04] <renesis> yeah i would think if it was servo usb, it would have to handle feedback itself
[05:01:13] <SWPadnos> yep
[05:01:20] <renesis> so emc would just be like, interpreting gcode, sending trajectories
[05:01:45] <renesis> so yeah, i didnt think it was really implimented, cuz it would have to do alot of shit emc does
[05:01:50] <SWPadnos> yes. there's a device called the ncPod which does that - they ported the EMC2 interpreter to Windows and made it output 1ms-timesliced positions
[05:01:56] <renesis> heheh, be like half-emc and the drivers
[05:02:22] <SWPadnos> sure, like all of CL if you want it, not to mention any HAL blocks you wanted to connect to its I/Os or encoder feedback
[05:02:26] <renesis> thats interesting
[05:02:37] <renesis> like, overclocked Eee might be nice to run a CNC from
[05:02:48] <SWPadnos> Eee?
[05:02:51] <renesis> just cuz so little, but full pc interface
[05:02:52] <gezar> ds2 thats pretty sweet
[05:03:05] <renesis> Eee PC, like a little $300-$500 laptop
[05:03:08] <ds2> gezar: got 2 days to spare ;)
[05:03:14] <renesis> kinda like a baby thinkpad, by asus
[05:03:26] <ds2> renesis: the PCIe bus might be usable for this
[05:03:41] <renesis> yeah it has slots
[05:03:49] <renesis> i think mine has one, alot have two tho
[05:03:50] <SWPadnos> then add $300-$500 for the USB thingy, and you might as well have bought a cheap desktop and a Mesa card
[05:03:53] <ds2> or a SDIO interface
[05:03:58] <renesis> because new gen eee have ssd on a pcie slot i guess
[05:04:11] <renesis> SD?
[05:04:16] <renesis> whats sdio
[05:04:17] <gezar> i want to find out of fagor is running emc
[05:04:24] <ds2> thought the eeePC has a SD slot?
[05:04:42] <ds2> if it only had a real CF interface....
[05:05:32] <renesis> oh
[05:05:34] <renesis> okay yeah
[05:05:40] <renesis> well thats usb
[05:05:42] <gezar> hmm, would it be possiable to find some way to use the sATA drive connector as a machine control?
[05:06:04] <renesis> its a integrated usb cardreader, you boot of it the same a usb mass-storage
[05:06:12] <ds2> hmmm
[05:06:16] <renesis> gezar: maybe
[05:06:21] <ds2> essentially machine control via SCSI?
[05:06:22] <renesis> but youd have to hack the driver
[05:06:28] <gezar> well yeah
[05:06:34] <gezar> but think about the io speeds
[05:06:43] <renesis> ds2: emulated scsi, for the SD and SSD(SATA) interface
[05:06:52] <renesis> but its usb and sata
[05:07:43] <gezar> you could definately get a retarded i/o
[05:08:01] <renesis> pci-e the best idea yet
[05:08:07] <renesis> that would work with any lappy
[05:08:15] <renesis> and latency prob next to nothing
[05:08:18] <ds2> or maybe there is a SPI or I2C bus on the internal connector
[05:08:26] <renesis> maybe fastest bus on pc?
[05:08:36] <ds2> it is not fast, it is RT response
[05:08:41] <renesis> yes, but who knows what the timing specs on that are
[05:09:03] <renesis> could be hella fast, maybe not, maybe hella fast but shit latency, dunno
[05:10:18] <ds2> latency isn't that horrible as long as it is predictable
[05:10:20] <renesis> swpadnos: ty for ncpod suggestion
[05:10:34] <renesis> well, my machine isnt even feedback based
[05:10:52] <renesis> so im not it even matters so much
[05:11:07] <SWPadnos> no problem. if you get one, I'd love to hear about all the problems :)
[05:11:12] <renesis> http://images.techtree.com/ttimages/story/86364_eeepc_600x600.jpg
[05:11:14] <SWPadnos> time for bed. good night
[05:11:16] <renesis> haha yeah fureal
[05:11:20] <renesis> swpadnos: thats it
[05:11:29] <renesis> the keyboards isnt that much smaller than standard
[05:11:37] <renesis> like 1" shorter on both sides
[05:11:37] <SWPadnos> not much bigger than the OLPC
[05:11:47] <renesis> its smaller, actually
[05:11:52] <renesis> olpc kinda bulky
[05:12:06] <renesis> olpc has better res
[05:12:23] <SWPadnos> one day I'll take mine apart, but until then. good night :)
[05:12:26] <renesis> this thing is 800x480 max, unless you use hacked i810 drivers
[05:12:31] <renesis> nite
[05:30:33] <fsdafsd> fsdafsd is now known as Unit41
[05:34:57] <Jymmmm> Ug
[05:41:01] <ds2> Jymmm: remember i was asking all the stuff about doing working plastics?
[06:28:28] <Jymmmm> ds2: sorta kinda
[06:53:06] <ds2> Jymmm: this is what it was for -
[06:53:28] <ds2> http://www.ugcs.caltech.edu/~huny/Lunarrat3.jpg
[06:53:31] <ds2> and in action
[06:53:41] <ds2> http://www.ugcs.caltech.edu/~huny/clip.mpg
[07:15:08] <fenn> hey i had that same idea
[07:15:55] <fenn> the 'layers' would have been different slices of the julia set
[07:25:24] <ds2> julia?
[07:25:33] <ds2> is that a kind of mandelbot set?
[07:25:50] <ds2> (the base was from a dollar store!)
[07:40:32] <fenn> yes its the general form - the mandelbrot set is where z=0
[08:03:51] <ds2> Hmmm
[12:13:21] <Guest999> why such an obscure forum, why not a php forum?
[12:33:31] <fenn> archivist: you were right, my homemade 1" to 1/2" impact adapter sheared off
[12:34:01] <fenn> the cheapie import set worked though
[12:34:19] <archivist> oo tome to get some cr vanadium steel
[12:34:23] <archivist> time
[12:34:58] <fenn> or at least something better than mild steel
[12:35:14] <archivist> ahem, yes
[12:36:17] <fenn> now i feel better about buying this massively oversized impact wrench
[12:36:43] <archivist> good tools have some power
[12:37:15] <fenn> had to lug the giant air compressor across town to use it
[12:40:29] <archivist> what are you taking to bits?
[12:43:13] <fenn> my car
[12:44:20] <BigJohnT> you going to run it with EMC now?
[12:44:34] <archivist> heh cars arent meant to need that much abuse to dissassemble
[12:44:41] <alex_joni> hope not
[12:45:24] <BigJohnT> you should do a brake job on an old VW!
[12:46:07] <archivist> mv old_vw > /dev/bin
[13:05:34] <BigJohnT> anyone know why the Synaptic Package Manager(Ubuntu 6.06) does not upgrade to Python 2.5 from 2.4?
[13:06:56] <SWPadnos> because 2.4 is the latest version released for 6.06
[13:07:36] <BigJohnT> ok, I see that I'm running 2.4.3 but there is a 2.4.4 so it must be after 6.06 I assume
[13:07:58] <SWPadnos> oh, maybe there will be a release sometime soon
[13:08:11] <SWPadnos> 6.06 is "bugfixes only" at this point, I believe
[13:08:32] <BigJohnT> ok, thanks
[13:13:11] <Xamusk> is it possible to run emc without the realtime stuff?
[13:13:51] <skunkworks_> yes - you can run simulator
[13:15:01] <skunkworks_> start here and then 3.4 for similator build
[13:15:02] <skunkworks_> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#On_Ubuntu_5_10_and_6_06_from_source
[13:15:15] <Xamusk> no, I don't really want the simulator
[13:15:18] <Xamusk> but the real thing
[13:15:29] <SWPadnos> you cannot run a machine without realtime
[13:15:41] <Xamusk> why not?
[13:16:08] <SWPadnos> well, for one thing, it wouldn't necessarily work all that we;;
[13:16:10] <SWPadnos> well
[13:16:53] <SWPadnos> consider what happens to motors when the computer decides to go off and do something else - a stepper will stall because it's not getting any more step pulses
[13:17:12] <SWPadnos> a servo will coast (or accelerate), and there will be no software counting of its position
[13:17:36] <skunkworks_> the question is - what are you trying to do?
[13:17:52] <Xamusk> that is considering it's running on a really slow system, but modern systems are really fast
[13:18:26] <SWPadnos> fast and predictable are not the same thing
[13:18:46] <Xamusk> I want to run emc in my ubuntu box without the troubles that come by installing realtime stuff
[13:19:06] <SWPadnos> heh, then you'd be trading some up-front trouble for a lot of later troubles
[13:19:38] <SWPadnos> have you tried the EMC2 LiveCD on the computer you want to use?
[13:19:42] <Xamusk> the problem is that this box is my primary box, not one used exclusively for emc
[13:20:08] <SWPadnos> the realtime kernel doesn't have much of an effect unless the RT modules are loaded
[13:20:29] <SWPadnos> with the exception that we don't have an SMP RT kernel package yet
[13:22:26] <Xamusk> so, is it possible to easily load and unload this module? some modules don't like to be unloaded
[13:22:48] <SWPadnos> for the most part, yes
[13:23:12] <SWPadnos> it is possible to have issues, but in normal operation the modules get loaded when you run EMC and unloaded when you exit
[13:23:34] <Xamusk> nice
[13:24:17] <Xamusk> I just hope it runs now
[13:25:05] <SWPadnos> the ubuntu-maintained realtime kernel may be good enough for some types of machine (servo systems only), and there have been some talks about making EMC2 run in userspace only, but since we need access to hardware, it isn't all that easy
[13:25:29] <Xamusk> I want to operate stepper motors
[13:26:41] <Xamusk> though the machine isn't built yet
[13:34:47] <BigJohnT> I can't tell any difference on this computer with or without EMC running
[13:35:36] <BigJohnT> * BigJohnT off to work
[13:38:57] <eric_U> with RTAI, you can do real time from user space
[13:39:03] <eric_U> still need a real time kernel
[13:39:52] <SWPadnos> I suspect it's not all that helpful, since we still have hardware drivers that really shuld have direct access to the hardware
[13:40:08] <SWPadnos> though CL and most other HAL modules would probably work in userspace RT
[13:41:50] <fenn> SWPadnos: see, this is a classic case for the USB open loop stepper thingy
[13:42:01] <SWPadnos> yep
[13:42:49] <fenn> how's that ARM project coming?
[13:42:57] <jepler> you mean, this is a perfect application for imaginary software?
[13:43:00] <SWPadnos> it's not, so far
[13:43:12] <fenn> jepler: not just software, but hardware too
[13:43:29] <SWPadnos> yes, if you consider a G100 driver imaginary (ie, they do exist, just not for EMC)
[13:44:01] <eric_U> I have so many applications for imaginary software, it's not even funny any more.
[13:44:37] <fenn> eric_U: you could start blogging about them and hope someone writes it for you
[13:44:53] <fenn> dear internet:
[13:45:17] <fenn> i would like 1 santa claus machine, capable of creating more santa claus machines
[13:45:42] <archivist> stuff them on rentacoder but no fee supplied
[13:46:34] <archivist> I wonder if any good code comes from rentacoder
[13:49:26] <fenn> sounds fun, like playing video games sorta
[14:08:00] <eric_U> you see a lot of people basically doing that
[14:08:17] <eric_U> academics in robotics particularly
[14:08:38] <eric_U> or they buy a new piece of equipment and announce the thing they want to prove with it
[14:12:23] <fenn> "in my thesis i am going to prove i can get someone else to write all the code for me"
[14:12:57] <eric_U> there was someone that had a program that wrote papers
[14:13:17] <eric_U> one got accepted for publication, but it wasn't in a refereed journal
[14:13:27] <fenn> awesome
[14:13:52] <fenn> it can be hard to tell sometimes..
[14:13:56] <eric_U> the idea was the field was so riddled with jargon, a computer could do it, it was one of the arts
[14:16:33] <fenn> comparative literature
[14:18:06] <archivist> It was available as an internet page somewhere so you could get your own BS written for you
[14:21:27] <fenn> hey it might have inspired some real papers even
[14:22:02] <fenn> link enough random concepts together and you'll hit something good eventually
[14:23:54] <fenn> i call it meme splicing
[14:26:53] <archivist> fenn http://pdos.csail.mit.edu/scigen/
[14:27:18] <archivist> shame one cant seed it
[14:30:47] <fenn> archivist just take a standard markov modeling program such as megahal and seed it with papers
[14:34:49] <skunkworks_> The Enhanced Machine Controller (EMC) videotape is complete. The video outlines the evolution of the NIST Real-time Control System (RCS) and highlights the implementation of EMC at a major automotive manu
[14:34:57] <skunkworks_> facturer
[14:35:18] <skunkworks_> wonder if you can still get that...
[14:35:31] <skunkworks_> http://nvl.nist.gov/pub/nistpubs/jres/100/6/j16nbr.pdf
[14:35:35] <skunkworks_> page 745
[14:35:44] <skunkworks_> Requests for copies of the videotape should be made to
[14:35:50] <skunkworks_> Tracy Becker, ext. 3452.
[14:35:53] <skunkworks_> :0
[14:35:55] <skunkworks_> :)
[14:36:07] <skunkworks_> (dated 95)
[14:54:11] <alex_joni> skunkworks_: they usually don't throw stuff out
[14:54:32] <alex_joni> skunkworks_: it can't hurt to call them :)
[15:04:48] <cradek> does anyone get today's xkcd?
[15:04:54] <alex_joni> bbl
[15:05:48] <fretless85_> hey guys
[15:07:13] <SWPadnos> cradek, um - other than not knowing what a/s/l ,means, I think I might get it
[15:07:46] <archivist> nearest for me is snake bite, but?
[15:08:02] <SWPadnos> well, the baslisk has the nick bslsk05 ...
[15:08:09] <SWPadnos> har har
[15:08:57] <archivist> never heard of a "baslisk"
[15:09:08] <SWPadnos> err - basilisk
[15:09:38] <skunkworks_> it is a snake
[15:09:58] <skunkworks_> (first I heard of it was in harry potter..)
[15:10:00] <cradek> ok, it helps to know that
[15:10:05] <archivist> "reputed to be king of serpents and said to have the power of causing death by a single glance. "
[15:10:14] <skunkworks_> you look into its eyes and you die.
[15:10:18] <skunkworks_> right
[15:10:21] <cradek> ohhhhhhhhh
[15:10:25] <SWPadnos> http://en.wikipedia.org/wiki/Basilisk
[15:10:31] <archivist> from wikipedia
[15:10:39] <cradek> wow that's an obscure one
[15:10:43] <SWPadnos> heh
[15:10:44] <archivist> yup
[15:10:52] <SWPadnos> he dies from the smiley - that's a little bit funny
[15:14:37] <archivist> almost worksafe http://www.nickscipio.com/pod/2007/03/11_tushclock.html
[15:15:15] <archivist> but I thought cradek would apreciate the time reference
[15:15:28] <archivist> or clock
[15:15:54] <SWPadnos> icky poo
[15:16:03] <SWPadnos> no pun intended
[15:16:16] <fenn> epoch fail
[15:16:36] <cradek> epoch fail is pretty funny
[15:22:07] <fenn> someone should write a lolcats translator
[15:23:10] <fenn> the curious dialect from icanhascheesburger
[15:23:36] <archivist> that is an addictive site
[15:28:51] <skunkworks_> http://www.xkcd.com/378/
[15:28:53] <skunkworks_> funny
[15:29:37] <SWPadnos> http://www.xkcd.com/373/
[15:30:18] <SWPadnos> http://www.xkcd.com/371/
[15:32:55] <SWPadnos> http://www.xkcd.com/363/
[15:37:32] <fenn> let me take this opportunity to unleash a DDOS attack on #emc: http://www.contextfreeart.org/gallery/view.php?t=rand&num=25
[16:44:43] <gezar> how yall doing today?
[16:56:16] <archivist> * archivist is a happy bunny today, 6 gears made on the noo machine today and another on it already
[16:56:37] <gezar> you got photos?
[16:56:45] <gezar> and areyou using a 4th axis or?
[16:57:23] <archivist> www.archivist.info/cnc last link, 2 axis not done yet
[16:57:39] <archivist> 4th axis is running
[16:58:11] <archivist> still lots to finish off
[16:59:47] <archivist> www.archivist.info/gear stuff a number of teeth in box all others dontcare(at the moment!) program is then down loaded and run
[17:00:52] <gezar> so your just milling 1 tooth at a time right?
[17:01:01] <archivist> yup
[17:01:16] <archivist> same as the old hand method
[17:01:17] <gezar> if I make you a hobb would you be interested in trying to sync the axis together?
[17:01:46] <archivist> I have hobs
[17:01:53] <gezar> oh cool
[17:02:04] <archivist> and a hobbing machine
[17:02:11] <gezar> no way?
[17:02:27] <archivist> but yes I do want to play cnc hobbing
[17:02:44] <cradek> man encoder_ratio
[17:03:41] <archivist> I need anothe axis to tilt the hed for hobbing though
[17:03:49] <cradek> I haven't played with this but I'd like to sometime
[17:05:17] <gezar> any of you play with those tablet type pc's, im thiking one for school would be nice to have
[17:05:21] <archivist> the real question is start point as I will likely need to restart n times to get proper depth and it must remain in "gear"
[17:06:00] <gezar> as long as you dont remove them it would almost be like a threading operation using an index pulse right?
[17:06:33] <gezar> hmm, I dont know
[17:06:45] <archivist> well does it rewind the headstock to start or what
[17:06:58] <gezar> I have to go to my remaining 3 classes now, ill talk with you guys later
[17:07:35] <cradek> archivist: you talk as if someone has worked this all out already - I don't think anyone has, but I think the tools are there
[17:08:22] <archivist> cradek, Im talking as one whosed used a geared maching
[17:08:57] <archivist> the real fun comes in hobbed helicals
[17:09:09] <cradek> do you mean you'd take the gear out and put it in the clock to check depthing, or measure it some other way
[17:10:30] <archivist> on the hobbing machine it cycles and remains in gear, so it can be stopped and the head moved out the way to measure and then reset and rerun
[17:10:48] <archivist> never off the machine no
[17:11:29] <cradek> seems that fundamentally, electronic gearing can do everything mechanical gearing can do
[17:11:38] <cradek> including absolute positioning
[17:13:17] <tomp2> archivist: beautful music box/automata next to your weighted table http://www.archivist.info/cnc/stagethree/P1260003.JPG and it looks like it has one of those chinese windlass/spring arrangements
[17:17:49] <archivist> tomp2 www.archivist.info/clock/ for the whole thing
[17:19:27] <archivist> I am not the dork in the last pic!
[17:20:15] <archivist> he is a "customer" who happens to infest here now and again
[17:20:47] <tomp2> ooh! its a clock with exotic chimes. nice work!
[17:21:20] <archivist> we added the dial, moon, date and night silent
[17:21:32] <archivist> and fitted to the case
[17:24:52] <tomp2> i guess the widget at top front is moon phase, beautiful.
[17:25:01] <skunkworks_> archivist: that was quick :)
[17:25:23] <archivist> what was quick
[17:26:11] <skunkworks_> getting your machine to the point of working.
[17:26:19] <skunkworks_> (seems like just last week.....)
[17:27:41] <archivist> well lots not done! column only clamped to base, two axis not got steppers yet
[17:37:22] <skunkworks_> where is your site being hosted? in house?
[17:37:28] <skunkworks_> * skunkworks_ could look I guess
[17:37:33] <archivist> electronics er vunerable to swarf!!!
[17:37:41] <archivist> in house
[17:38:09] <archivist> actually in portacabin outside
[17:38:54] <skunkworks_> heh
[17:39:07] <archivist> skunkworks, no silly isp restrictions doing it myself
[17:41:12] <skunkworks_> heh.. makes sense. I pay 48 dollars a year for 2GB space and 35GB/month transfer. I am sure there are better ones out there.
[17:43:05] <archivist> I have a number of subdomains on archivist.info I can add when I feel like it
[17:43:45] <archivist> its on an adsl line so speed not that good
[17:45:30] <archivist> I also run the faq bot in #mysql so stuff stays up 24/7
[17:46:13] <archivist> really fun when people mistake for human
[18:05:23] <renesis> http://www.geckodrive.com/product.cfm?pid=17
[18:05:26] <renesis> can emc do that?
[18:06:42] <SWPadnos> that's the G100 I was saying we have no driver for :)
[18:06:51] <renesis> oh neat, ok
[18:07:10] <renesis> but theres maybe emc code for the ncpod thing?
[18:07:14] <SWPadnos> or G-Rex I guess
[18:07:28] <SWPadnos> the NCPod software uses the EMC2 interpreter
[18:07:33] <SWPadnos> that's the PC side
[18:07:45] <SWPadnos> the embedded side is proprietary
[18:08:38] <SWPadnos> well, it uses a modified EMC2 interpreter and trajectory planner actually - they spit out positions that the USB device is supposed to hit every millisecond
[18:08:59] <SWPadnos> since USB has good bandwidth, you can spit out a bunch of waypoints in a very short time
[18:09:00] <skunkworks_> There is really no need.
[18:09:09] <skunkworks_> with emc2
[18:09:13] <skunkworks_> imho
[18:09:25] <SWPadnos> if you have a computer with only USB expansion options, then there is a "need"
[18:09:40] <SWPadnos> (need for a better computer IMO, but that's a different story)
[18:10:55] <SWPadnos> ok - time to thaw the hands. bbiab
[18:14:27] <Vq^> btw of io, what is a good IO-card to use with emc2?
[18:14:49] <SWPadnos> digital only?
[18:14:56] <Vq^> yes
[18:15:20] <SWPadnos> some parport card would do it, or the AX5214 (I think) 8255 card, or a Mesa 5i20
[18:16:11] <Vq^> i assume most parport cards is as direct as those built in then
[18:16:48] <SWPadnos> they're accessed the same way, if that's what you're asking
[18:17:02] <SWPadnos> sometimes PCI parport cards are a little faster than built-in actually
[19:01:41] <Xamusk> how easy would it be to make a driver for a custom usb controller?
[19:02:33] <ds2> host or device?
[19:03:26] <Xamusk> device
[19:03:38] <ds2> shouldn't be too hard
[19:03:41] <Xamusk> I meant an usb motor controller
[19:04:03] <ds2> it would depend on how much hardware assistance is there
[19:04:26] <Xamusk> I would do the hardware myself
[19:04:30] <ds2> on one extreme, you could have an AVR bit banging it all the way to a full blow device controller with DMA support
[19:05:07] <Xamusk> I was imagining something like the CDC/IO
[19:09:03] <Xamusk> http://www.recursion.jp/avrcdc/
[19:09:06] <skunkworks_> there was a huge discussion on the emc-user list about using usb.. http://thread.gmane.org/gmane.linux.distributions.emc.devel/390/focus=400
[19:11:21] <Xamusk> I'll take a look
[19:13:34] <skunkworks_> It probably isn't what you want to hear.
[19:15:56] <Xamusk> no, really it isn't
[19:17:01] <Xamusk> actually, for me, it would be best to implement the g-code interpreter directly in the controller chip, though it would probably get too big or too complex
[19:17:35] <SWPadnos> yes
[19:18:03] <Xamusk> it would solve both the realtime need and latency problems
[19:18:18] <SWPadnos> well, not entirely
[19:18:24] <SWPadnos> it makes the realtime a little less "real"
[19:18:43] <SWPadnos> consider someone dragging the feedrate override slider on a GUI
[19:18:52] <SWPadnos> how long does it take to actually reduce the speed?
[19:19:16] <SWPadnos> do you need to flush the buffer (on the device), or can the FO command jump the queue?
[19:19:29] <SWPadnos> what if that change sthe path (due to tolerance modes)
[19:20:17] <SWPadnos> if you implement all of EMC on the controller (might be possible, though any microcontroller will probably have a hard time with the floating point math), and send NML mesasges, that could work
[19:20:47] <SWPadnos> all of the motion and IO controllers, that is - the interp might be able to be on the PC, with the CANON "interpreter" on the hardware
[19:26:20] <Xamusk> I never really got emc to work properly, though I have dealt a little with TurboCNC, for DOS
[19:27:22] <Xamusk> unfortunately, TurboCNC isn't open source, so I can't look at the code, but I imagine it would be small
[19:28:43] <Xamusk> but that's what I was imagining: an interpreter in the PC for the heavy-lifting, and an "intermediate" protocol for PC-microcontroller communications
[19:30:58] <Xamusk> though I don't know emc internally to know if it could be used for that
[19:31:31] <SWPadnos> it could, but there are a lot of floating-point calculations, even at the lower levels
[19:31:52] <SWPadnos> look up the "canonical machining interface" or similar in the EMC documentation
[19:32:06] <SWPadnos> that's the intermediary "language" for specifying motion
[19:32:32] <SWPadnos> there's also NML which is the communications scheme between the GUI, motion controller, and I/O controller
[20:18:50] <skunkworks_> shouldn't you be in boston?
[20:19:06] <SWPadnos> no - rescheduling
[20:19:33] <SWPadnos> strangely, you can't renew your passport in person unless you have a ticket for travel within the next 14 days
[20:21:36] <alex_joni> huh, that's odd
[20:21:42] <alex_joni> even if it's expired?
[20:21:47] <SWPadnos> yes
[20:21:55] <alex_joni> so what if you need to travel tomorrow?
[20:22:12] <SWPadnos> tomorrow is within the next 14 days :)
[20:22:30] <alex_joni> so you'd fly today to boston? and then to new yorK?
[20:22:31] <SWPadnos> you can renew by mail, and there's the option of paying an expedite fee (which is supposed to make the turnaround time be ~3-4 weeks or less)
[20:22:47] <alex_joni> SWPadnos: sounds like a PITA
[20:22:50] <SWPadnos> it is
[20:22:54] <skunkworks_> * skunkworks_ is glad his passport is not expired
[20:22:59] <alex_joni> I went to change my passport at around 9am
[20:23:07] <alex_joni> had the replacement in my hand around 12am
[20:23:26] <skunkworks_> whatever - we could squash ro... ;)
[20:23:37] <alex_joni> not without a passport you can't :P
[20:23:42] <SWPadnos> especially if you don't live in either the city where your passport needs to be renewed (4 hour drive each way) or the city where you get the work visa (6 hour drive or flight + wwaiting + transportation into the city)
[20:23:42] <skunkworks_> heh
[20:23:44] <cradek> hahaha
[20:24:05] <alex_joni> cradek: what's so funny?
[20:24:06] <alex_joni> :P
[20:24:12] <cradek> you
[20:24:22] <SWPadnos> actually, my passport doesn't expire until August, but the rules for getting an India work visa are that the paessport must have at least 6 months remaining
[20:24:34] <alex_joni> * alex_joni hopes his argument keeps standing in the eventuality
[20:24:43] <SWPadnos> I've got 6 months and a couple of weeks at the moment ;)
[20:25:03] <skunkworks_> I really didn't know you needed a us visa to work in india.. Is that for tax purposes?
[20:25:20] <SWPadnos> but, then the visa expires as soon as my passport expires, rather than 5 or 10 years later, so it makes sense to get the passport renewed first so I don't have to get another visa the next time a job comes along in India
[20:25:24] <alex_joni> he needs an indian visa :)
[20:25:29] <SWPadnos> you need an Indian visa ;)
[20:25:33] <skunkworks_> ah - that makes more sense.
[20:25:42] <alex_joni> SWPadnos: that should be native american visa
[20:25:44] <alex_joni> lol :P
[20:25:46] <SWPadnos> hence the visit to the Indian consulate in New York
[20:25:48] <SWPadnos> heh
[20:26:14] <skunkworks_> so... alex... your passport is up-to-date for this summer?
[20:26:25] <alex_joni> skunkworks_: no, I need a turist visa for the us
[20:26:25] <SWPadnos> no - he needs a US visitor visa to come here
[20:26:34] <alex_joni> and I need to travel for that
[20:26:39] <SWPadnos> maybe we could meet in Canada ;)
[20:26:55] <SWPadnos> or do that EuroFest somebody mentioned a long time ago
[20:26:56] <alex_joni> * alex_joni ponders jumping the fence from mexico
[20:27:02] <alex_joni> :P
[20:27:17] <alex_joni> skunkworks_: that's not the real problem though
[20:27:32] <alex_joni> summer is already booked solid :/
[20:27:59] <skunkworks_> aww
[20:29:25] <skunkworks_> maybe we could mark you as a terrorist.. they we would just have to pick you up at guantanamo.. although the trip there might be rough. (and we might not be able to get you out)
[20:31:24] <skunkworks_> ok - that might have been in bad taste
[20:51:32] <alex_joni> skunkworks_: not really
[20:51:39] <alex_joni> only my laptop crapped out in disgust :D
[20:51:52] <skunkworks_> heh
[20:57:00] <skunkworks_> http://www.wedin.com/news.html
[21:05:55] <alex_joni> skunkworks_: I saw some bigger ones
[21:08:53] <skunkworks_> what on?
[21:09:09] <alex_joni> a big tilting table
[21:09:15] <skunkworks_> cool
[21:10:08] <alex_joni> skunkworks_: similar to a teeterboard
[22:14:36] <alex_joni> away
[22:14:44] <alex_joni> +/
[22:15:09] <skunkworks_> when did alex get a cat?
[22:15:18] <alex_joni> cat?
[22:15:30] <alex_joni> no cats here :)
[22:15:37] <skunkworks_> looked like random keyboard input
[22:15:46] <alex_joni> no.. just a typo
[22:15:52] <alex_joni> meant to type '/away'
[22:15:56] <alex_joni> but forgot the /
[22:15:57] <skunkworks_> ah ;)
[22:16:10] <skunkworks_> * skunkworks_ never used that command..
[22:16:15] <skunkworks_> Imagine that
[22:16:29] <alex_joni> I use it to mark when I'm not here
[22:16:39] <alex_joni> '/away bbl or something'
[22:16:56] <alex_joni> then when I come back, '/away' will give all messages which have my name in them
[22:45:16] <skunkworks_> bbl
[23:12:15] <BigJohnT> * BigJohnT bbl
[23:13:32] <alex_joni> good night all
[23:13:47] <cradek> goodnight
[23:31:52] <skunkworks> wow - did anyone have tornadoes near them?
[23:32:03] <alex_joni> not me :D
[23:32:14] <skunkworks> we suprisingly missed the 18 inches of snow south of us
[23:32:19] <skunkworks> That is good. :)
[23:32:24] <alex_joni> skunkworks: ever..
[23:32:38] <skunkworks> your not missing anything ;)
[23:36:15] <alex_joni> I probably would after one
[23:38:43] <skunkworks> heh
[23:39:06] <skunkworks> I haven't seen one in person. There was some close ones this winter even.