#emc | Logs for 2008-10-19

[00:00:01] <archivist> yes
[00:00:56] <archivist> the internal ref is not that good, use an external one if its allowed
[00:00:59] <maddash> so why is this MBedder guy mocking me?
[00:01:32] <maddash> yes, but ext.ref is more prone to noise from EMI (from cpu oscillator,etc)
[00:02:54] <archivist> mbedder is just an internet smartarse ignore
[00:03:13] <archivist> no a good external will be better
[00:05:03] <archivist> look up bandgap voltage reference
[00:08:34] <^^Kilroy> anyone seen Dalran lately?
[00:10:01] <maddash> ...and, I've been disconnected again.
[00:10:31] <archivist> we see "* maddash has quit (Remote closed the connection)" thats your end
[00:11:04] <maddash> yes, it's a price I pay for acquiring wi-fi from aircrack
[00:11:33] <archivist> wifi and timeouts are normally (Connection timed out))
[00:11:56] <maddash> brb,
[00:12:00] <archivist> ah they dont want users using
[00:12:05] <archivist> ah they dont want users using
[00:19:26] <archivist> Paragon, thin sheet inconel at the local scrap yard £10 per kilo
[00:19:51] <Paragon> Hello All... What apps are been used with emc2 to get the 3d simulation as seen in these vids? http://uk.youtube.com/watch?v=fWKYQUj5AOs&feature=related http://uk.youtube.com/watch?v=_tA-W7nTNSc&feature=related
[00:20:10] <Paragon> archivist: Realy?
[00:20:39] <Paragon> What thickness are they?
[00:20:43] <archivist> 645 iirc
[00:21:04] <Paragon> 645?
[00:21:09] <archivist> and looked about20 thou
[00:21:31] <archivist> its offcut
[00:21:56] <archivist> not sure if big enough pieces
[00:22:08] <Paragon> 20 thou would be too thin to work with I think
[00:22:26] <archivist> iirrc blender for the second vid
[00:22:49] <archivist> er third
[00:23:27] <archivist> bah late 2nd
[00:23:48] <Paragon> blender .... thanks archivist ... google'ing now
[00:24:40] <archivist> Paragon, ask jepler and cradek they were doing that a few months back
[00:25:14] <Paragon> Will do. Thanks again
[00:26:33] <Paragon> http://www.blender.org/ ?
[00:27:50] <archivist> look in the channel logs and you will find me asking the same question :) ish
[00:28:28] <Paragon> How does one do that again?
[00:28:48] <archivist> logger_emc: bookmark
[00:28:48] <archivist> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-10-19.txt
[00:29:21] <DanielFalck1> search for 'ray tracing' for that last video
[00:29:28] <archivist> will be around the date of those vids and the cnc workshop
[00:30:10] <Paragon> thanks guys
[00:39:20] <archivist> Paragon, model engineer exhibition in warwikshire this weekend till tuesday
[00:40:11] <Paragon> Love to be there but have on call commitments :-(
[00:41:08] <archivist> we have a works outing here, going monday
[00:42:04] <Paragon> Cool :-) let me know how it goes...
[00:42:39] <archivist> seems there is a lot more cnc (but probably mach3 spam)
[00:43:17] <Paragon> can't find any reference to blender in the logs. I'll speak to jepler. Mach3 yuk ...
[00:46:30] <jmkasunich> Paragon: the video that cuts the part out of a yellow block was made using PovRay
[00:46:47] <jmkasunich> the one that shows a 5-axis gantry mill is a vismach model - vismach is part of EMC
[00:47:06] <jmkasunich> vismach = MACHine VISualization
[00:47:24] <Paragon> vismach is part of EMC .... Have not seen it!
[00:48:02] <jmkasunich> it's well hidden
[00:48:17] <archivist> unhide it :)
[00:48:22] <jmkasunich> it is not a part of the program that you run when you type "EMC"
[00:48:31] <jmkasunich> it is a utility, like halscope
[00:48:39] <Paragon> vismach.py ?
[00:48:58] <jmkasunich> well, vismach.py is part of it
[00:49:22] <jmkasunich> you (or someone) needs to write a python program that uses the vismach library to define the "shape" of the machine
[00:50:04] <jmkasunich> we have some of those already
[00:50:16] <maddash> weee i'm back
[00:50:17] <Paragon> Cool
[00:50:40] <jmkasunich> I don't know where (or if) they are in the installed version, but in the source tree, they live in src/emc/usr_intf/axis/scripts
[00:50:47] <jmkasunich> and are named *gui.py
[00:51:16] <Paragon> thanks for the heads up jmkasunich
[00:51:29] <jmkasunich> we have models for a puma robot, a scara robot, a maho600 5-axis mill, and cradek's 5-axis MaxNC
[00:51:47] <jmkasunich> oh, and the 5-axis gantry you already saw
[00:51:51] <Paragon> I built from src so will take a look tomorrow when i am in the shop.
[00:52:06] <jmkasunich> and I'm in the process of making one for a horizontal boring mill - hope to commit it this weekend
[00:52:18] <Paragon> good stuff
[00:52:56] <jmkasunich> maddash: usually when a spec it "+/- n LSBs", it means "n increments of the least significant bit"
[00:53:21] <jmkasunich> if you have a 12 bit converter (4096 counts), and a 0-5V range, that works out to 1.25mV per LSB
[00:53:36] <jmkasunich> if they say +/- 5 LSBs, that means +/- 5 * 1.25mV
[00:54:23] <maddash> jmkasunich: huh? a bit has at most 1 increment
[00:54:27] <maddash> hang on, let me think
[00:55:03] <jmkasunich> 0x000 to 0x001 is one increment of the LSB
[00:55:09] <jmkasunich> 0x001 to 0x002 is another
[00:55:36] <jmkasunich> 0x03A to 0x03B is also one increment if the LSB
[00:55:57] <maddash> isn't the last bit always 2^0? so why not just say, +/-5 and leave out "LSb"?
[00:56:39] <maddash> 2^0 = 1
[00:57:13] <maddash> +/- k LSb = +/- k * 2^0 = +/- k
[00:57:15] <jmkasunich> 5 what? mV, volts? I suppose you could say "+/- 5 counts", but +/- 5 LSBs is the normal way of specing such things
[00:57:26] <maddash> oh, ok
[00:57:44] <anonimasu> * anonimasu sighs
[00:57:46] <anonimasu> stuck at physics again
[00:57:47] <anonimasu> :p
[00:57:57] <maddash> so the issue is figuring out what to call the units
[00:58:12] <jmkasunich> in conversation, I call em counts
[00:58:35] <jmkasunich> but when speaking of accuracy, most of the time you really want the answer in volts (or millivolts, or such)
[00:58:46] <maddash> when I see "+/- 123," I always think in terms of the basic resolution of the ADC (e.g., 2^10, 2^12)
[00:59:01] <maddash> right but the ansewr in mV depends on V_min and V_max
[00:59:09] <jmkasunich> exactly
[00:59:22] <maddash> but regardless of V_min/max, the absolute # alway s stays the same
[00:59:26] <archivist> dont ever confuse accuracy and resolution
[00:59:26] <jmkasunich> so they spec it in counts, and you convert to volts for the specifics of your situation
[00:59:44] <maddash> ugh, I'm getting mocked over semantics
[00:59:49] <jmkasunich> accuracy, precision, AND resolution ;-)
[01:00:00] <maddash> precision = resolution
[01:00:06] <maddash> accuracy = consistency
[01:00:40] <maddash> thanks, jmkasunich
[01:09:50] <maddash> hm, no mention of SWPadnos tonight
[01:10:01] <SWPadnos> ha!
[01:10:12] <toastydeath> hahahah
[01:11:50] <maddash> guess i got ahead of myself
[01:11:55] <SWPadnos> precision is more like repeatability (ie, how close together are several attempts to hit the same spot)
[01:12:07] <SWPadnos> accuracy is how close those attempts are to the theoretical number
[01:12:36] <SWPadnos> resolution is the coarseness of the attempts (ie, how close together two attempts can be)
[01:12:57] <maddash> so you're saying almost the vice versa of what I did: accuracy=delta(err), precision=consistency
[01:13:09] <SWPadnos> yep :)
[01:13:20] <SWPadnos> think about target practice:
[01:13:35] <archivist> accuracy encompasses the other errors
[01:13:45] <SWPadnos> someone who has a 3-shot group that's very small (ie, all the bullets go to the same place), has very good precision
[01:13:52] <maddash> i swapped the two -- so shameful
[01:13:56] <SWPadnos> if they happen to hit the bullseye, they also have good accuracy
[01:14:37] <SWPadnos> (resolution has no comparison to target shooting :) )
[01:14:44] <SWPadnos> s/to/in/
[01:15:03] <archivist> thickness of rings v score
[01:15:15] <SWPadnos> ok, sure
[01:15:39] <SWPadnos> but that's not quite right, since the actual thing being mesaured is hole psition
[01:15:51] <maddash> resolution in your trigger-happy example would be the fine/coarseness of the shooter's arm
[01:15:55] <SWPadnos> so the resolution is somewhere around the planck length or something
[01:16:13] <maddash> planck length != smallest length...
[01:16:17] <SWPadnos> maybe - for instance the notches to adjust the sight, but that's stretching it a bit IMO :)
[01:16:27] <SWPadnos> smallest measureable length in this universe
[01:19:37] <toastydeath> we have a resolution problem on two of our lathes
[01:19:50] <toastydeath> one lathe is accurate to a couple millionths, but can only resolve in increments of .0001"
[01:20:17] <jmkasunich> that sounds like a nice problem to have
[01:20:18] <toastydeath> the other lathe is accurate to about .0001" but resolves in nanometers
[01:20:32] <toastydeath> haha, it is kind of nice.
[01:20:33] <jmkasunich> its easier to improve resolution (new encoder, etc) than accuracy
[01:21:01] <toastydeath> in this case it's just the control that's causing the problem, but it's a good illustration of why resolution is in its own category of problems.
[01:23:35] <archivist> it amuses me when you ask accuracy and they reply with resolution
[01:24:02] <toastydeath> ?
[01:24:39] <archivist> they being people who dont understand
[01:25:07] <toastydeath> oh
[01:59:43] <JymmmEMC> moving target = resolution ???
[02:00:26] <JymmmEMC> archivist: that's okey, I still don't understand =)
[02:00:54] <JymmmEMC> accurancy == repeatability I believe
[02:02:32] <SWPadnos> no, that's precision
[02:02:44] <SWPadnos> in math/science anyway
[02:04:17] <SWPadnos> I guess mathematically, precision is inversely proportional to the standard deviation of the samples
[02:04:50] <SWPadnos> accuracy is how close the measurements are to the "acutal" value
[02:05:06] <SWPadnos> resolution is the step size between possible measured values
[02:05:41] <SWPadnos> (so if I look at an analog voltmeter that has 1V ticks on it, and I estimate the tenths, then I have 0.1V resolution)
[02:07:41] <maddash> 40 minutes later, still orating over the same thing?
[02:08:01] <SWPadnos> no, JymmmEMC was late and needed the Cliff notes version :)
[02:08:08] <maddash> :)
[02:08:27] <maddash> so, SWPadnos, this pic32 looks to be the One
[02:08:34] <SWPadnos> cool
[02:09:13] <maddash> have you seen the specs? internal a/d runs @ 500ksps (compared to the shitty 52ksps using the ads7870 that rabbitsemi loves)
[02:09:22] <maddash> it's only 10-bits, unfortunately
[02:09:35] <SWPadnos> probably closer to 7 really
[02:09:38] <SWPadnos> maybe 8
[02:09:51] <maddash> whaaaa? why do you say that?
[02:10:12] <maddash> gain/offset errors?
[02:10:32] <SWPadnos> because every built-in ADC I've ever seen was actually about 2-3 bits worse than the spec sheet said
[02:10:37] <SWPadnos> and that goes for DSPs as well
[02:10:57] <maddash> hell, with this crazy quick sample rate, I can oversample and add 1-2 bits extra resolution
[02:11:19] <SWPadnos> have you checked to see if it's even supposed to be 10 bits at the highest sample rate?
[02:11:27] <maddash> yup
[02:11:32] <SWPadnos> usually they give faster rates with decreased resolution
[02:11:32] <maddash> nope
[02:11:34] <maddash> crap
[02:11:34] <SWPadnos> ok, cool
[02:11:42] <SWPadnos> heh
[02:14:05] <maddash> oh wait, i did
[02:14:05] <maddash> phew
[02:14:42] <maddash> yes, 10-bits @ 500ksps, but i have to use the internal adc vref-
[02:14:53] <maddash> erm external*
[02:19:19] <maddash> runs @ 1 cycle per instruction, 31 usable 32-bit registers
[02:19:47] <maddash> the official max speed is 80MHz, but someone in the forum clocked it up to 150 MHz
[02:27:48] <maddash> omg, so lonely
[02:29:19] <jmkasunich> boo!
[02:30:10] <davidf> hi
[02:30:56] <jmkasunich> hello
[02:31:31] <maddash> :(
[02:32:02] <SWPadnos> oh sorry, got distracted
[02:32:09] <maddash> crap, i think my laptop is giving off a "i'm burning up" smell
[02:32:20] <SWPadnos> sounds like the PIC32 is possibly undeserving of the PIC name
[02:32:53] <maddash> why? it's an interrupt controller (96 interrupt sources, 64 vectors !), and it's programmable...
[02:32:59] <davidf> jmkasunich, Hi. Man, I hate to bother you guys with this, but I'm kind of at an impass... I have a prob with dependencies on python2.4. I can't update or upgrade anything.
[02:33:13] <danbh_intrepid> davidf: I can still help ya
[02:33:37] <davidf> danbh_intrepid, you here too? :)
[02:33:47] <danbh_intrepid> heh, no, I followed you
[02:34:44] <jmkasunich> davidf: I don't know much about python dependencies, but for anyone to be able to help you, you will need to describe in detail what you are trying to do, and what is happening
[02:34:45] <davidf> OK. I thought you had given up...
[02:35:01] <jmkasunich> also, if you aren't using our regular distro, describe your environment
[02:35:39] <danbh_intrepid> davidf: I have a phone call, brb
[02:36:23] <davidf> jmkasunich, Maybe I should try elsewhere, 'cause this isn't an EMC problem really.
[02:36:50] <SWPadnos> oh, is it update-manager / apt?
[02:36:53] <jmkasunich> does it happen while trying to compile or run EMC?
[02:37:23] <davidf> SWPadnos, yes, jmkasunich , no, I have not tried that.
[02:37:48] <SWPadnos> ok, then the MOTU are probably better for answering :)
[02:37:53] <jmkasunich> oh, so it really is totally unrelated to EMC?
[02:38:33] <davidf> The deal is, I just tried to do a regular routine system update, and 132 programs can't upgrade because they depend on python2.4 and python2.4 is not configured.
[02:38:48] <SWPadnos> you manually installed 2.5?
[02:39:11] <davidf> SWPadnos, yes, that's right. I should probably not bug you guys with this one.
[02:39:26] <SWPadnos> you could try uninstalling 2.5, then upgrade again
[02:39:46] <SWPadnos> it should see that you have no python and install the one you need (unless there isn't one in the repos)
[02:40:20] <davidf> SWPadnos, No, I didn't read before I posted. it's not about emc at all.
[02:40:37] <SWPadnos> no biggie
[02:41:14] <davidf> Thing is, I do have python2.4 installed.
[02:41:18] <SWPadnos> oh
[02:41:40] <SWPadnos> well, there are apt options that they warn you against using unless you know what you're doing
[02:41:44] <SWPadnos> so you could try one of those :)
[02:41:55] <davidf> SWPadnos, LOL.
[02:42:01] <SWPadnos> (but I wouldn't know what I'm doing, so I can't tell you which one(s) )
[02:42:02] <SWPadnos> heh
[02:42:04] <davidf> SWPadnos, really.
[02:42:16] <davidf> SWPadnos, still lol.
[02:42:21] <SWPadnos> there's some option to fix dependencies
[02:42:30] <maddash> davidf: did you specify what distro this is?
[02:42:39] <SWPadnos> but it may fix like neuter vs. fix like repair - dunno
[02:42:49] <maddash> davidf: debian? ubun2?
[02:43:00] <davidf> ubuntu 6.06 LTS
[02:44:24] <maddash> mind pastbinning 'aptitude update ; aptitude upgrade' (as root)?
[02:45:21] <davidf> Well, I think it might be this... I have this GUI programming tool Boa Constructor.. I think long ago it broke when py2.3 became py2.4, and there was a workaround to change some file to use py2.3 files or something... but I can't for the life of me remember what foile I changed.
[02:45:36] <danbh_intrepid> maddash: its a dpkg problem
[02:45:39] <davidf> maddash, hmm? no comprede.
[02:45:46] <davidf> comprende
[02:46:33] <SWPadnos> apr-get remove --force-uninstall python2.4
[02:46:33] <maddash> davidf: then 'sudo aptitude purge python2.4'
[02:46:39] <SWPadnos> duh - apt-get ...
[02:46:50] <maddash> aptitude is cooler
[02:46:53] <maddash> :{P
[02:47:32] <davidf> I tried somnething like that, but got errors due to other programs dependent on py2.4 so sys would not allow removing it.
[02:47:57] <SWPadnos> sudo apt-get install --force-reinstall (or similar) python2.4
[02:47:59] <maddash> davidf: seriously, use aptitude to do it because it has automatic conflict resolution
[02:48:27] <maddash> davidf: including the option to purge the dependents of python2.4
[02:49:01] <davidf> does purge mean remove? cause that's about everything on this box...
[02:50:04] <SWPadnos> have you tried force-resinstall (or however it's spelled)
[02:50:06] <maddash> no, no, purge means removing the stuff under "remove" plus the pesky config files
[02:50:14] <SWPadnos> gah
[02:50:27] <maddash> Purge <package>: remove it and all its associated configuration and data files.
[02:50:37] <maddash> from 'man aptitude'
[02:50:56] <SWPadnos> what happens to the 162 packages that depend on the purged one?
[02:51:06] <jmkasunich> they break
[02:51:18] <davidf> ouch. no way.
[02:51:52] <maddash> whoops, didn't read that part
[02:52:11] <maddash> er
[02:52:13] <maddash> !bookmark
[02:52:41] <danbh_intrepid> davidf: try this : sudo dpkg --force-depends --purge python2.4-minimal python2.4 python2.4-dev python2.4-gdbm python2.4-tk && sudo apt-get clean && sudo apt-get -f install
[02:52:45] <maddash> '162 packages'? where did he mention this? I can't find it
[02:53:51] <SWPadnos> that was a fictitious but large number
[02:54:32] <maddash> hm, python2.4 is "installed but not configured"?
[02:55:08] <davidf> Here's what's going on... I have a firewall, and I've been hit about a thousand times in three days with a traceroute to my box. I wanted to install clamav, so (and I did,) but I need to update it before it will work. But I also did a system upgrade, and it found a gillion files to upgrade, but the first time one depends on py2.4, everything stops with errors. Is there some way to purge the upgrade list that is still stored?
[02:55:18] <davidf> it is 132 files. Yes.
[02:56:44] <davidf> I tried to apt-get py2.4 and install over the old one. Here's a pastebin of the errors....
[02:56:57] <davidf> http://pastebin.com/d2152bfa3
[02:57:00] <maddash> ah, finally, some output messages
[02:57:06] <davidf> Look at the bottom.
[02:58:53] <davidf> I get essentially the same thing with any update or upgrade. Also I can't remove py2.4 unless I remove practically every program on my computer. It's a catch 22.
[03:00:15] <maddash> http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=477145
[03:00:29] <maddash> you have that snake program. have you tried purging it first?
[03:00:50] <SWPadnos> davidf, did you try danbh_intrepid's suggestion?
[03:00:55] <davidf> No, but I could live without it.
[03:01:02] <davidf> which?
[03:01:09] <maddash> boa constrictor?
[03:01:11] <SWPadnos> [22:53:03]<danbh_intrepid>davidf: try this : sudo dpkg --force-depends --purge python2.4-minimal python2.4 python2.4-dev python2.4-gdbm python2.4-tk && sudo apt-get clean && sudo apt-get -f install
[03:02:20] <davidf> SWPadnos, yes, but that looks a bit differnt. I don't remember the force part.
[03:02:27] <davidf> It didn't work.
[03:02:32] <maddash> output?
[03:02:37] <SWPadnos> the force part is probably the important part ;)
[03:02:54] <maddash> it should have worked. --force-depends changes those dependency errs into warnings
[03:03:08] <danbh_intrepid> yeah, it kinda is. It allows you to remove the packages without removing anything else
[03:03:25] <davidf> I pasted it. But I'll try again. I am just afraid of losing all my emails, etc. It's got business data on it.
[03:03:51] <SWPadnos> if you can boot another computer, you'll have access to the data on this drive
[03:03:56] <SWPadnos> (small consolation, I know ;) )
[03:04:20] <maddash> was that an intended pun?
[03:04:40] <SWPadnos> hmmm. no
[03:04:59] <maddash> consol-ation
[03:05:14] <maddash> god, I have to stop using linux for a while
[03:05:28] <danbh_intrepid> davidf: you will only loose python, which should then reinstall. If that still doesn't work, then I would nuke the postinstall script, then try try again. And I would probably get rid of that boa module as someone suggested
[03:06:06] <davidf> ok, I'm getting brave, here goes...
[03:06:21] <maddash> paste the output
[03:06:28] <maddash> pastebin**
[03:07:09] <scutsxg> hi,is there anybody have experience running emc2 with pluto_step driver but with no hardeware.
[03:07:40] <SWPadnos> no pluto?
[03:07:44] <SWPadnos> or no machine?
[03:07:55] <scutsxg> my effect is very poor,the gui goes rigid,that means it doesnt go smoothly
[03:08:29] <SWPadnos> you can't use pluto without the pluto attached. the driver should refuse to load
[03:08:54] <SWPadnos> I don't know how long the update takes, but you should probably make sure you don't have the pluto read and write functions in a fast thread
[03:08:56] <scutsxg> no machine,i just want to port my hal driver with pluto_step as good example.
[03:09:13] <SWPadnos> hal driver for what?
[03:09:18] <jmkasunich> the pluto step driver _must_ have a pluto board to talk to
[03:09:37] <scutsxg> but i can run emc2 with pluto_step after some modification.
[03:10:01] <SWPadnos> sure, if you change it so it doesn't look for the board
[03:10:09] <SWPadnos> but it won't be all that useful in that case
[03:10:40] <SWPadnos> as sample code, it might be OK - if you're making a parallel-port-connected device
[03:11:27] <scutsxg> is it wrong way ?i found pluto_step is the most esay way to start my hal driver,i have designed a motion control card with pc104 bus,but hal driver is a little difficult .
[03:12:22] <SWPadnos> PC104 is ISA (unless it's PC104+, which is PCI), so maybe the mesa 5i20 driver would be easier to use
[03:12:42] <SWPadnos> unless you implemented the interface as a burst read/write to the same address (which is a bad way to do it IMO)
[03:12:56] <maddash> did davidf's system blow up?
[03:13:11] <scutsxg> i would like to found a good way to start,hal_skeleton.c is not very suitble for me
[03:14:05] <SWPadnos> hal_parport, hal_m5i20, hal_stg, hal_ax5214h ...
[03:14:35] <SWPadnos> parport and parport-based drivers are probably not good, if you designed your card with a memory map that the CPU can see
[03:14:48] <davidf> maddash, hang on...
[03:14:58] <davidf> not dead yet... :)
[03:15:13] <SWPadnos> if it's accessed with an address and data register, then the parport-connected devices may be good (also pico systems ppmc for that)
[03:16:41] <davidf> http://pastebin.com/d63c0bb3a Doesn't look good... I'll try a new sys update now...
[03:18:01] <scutsxg> yes,i found pluto-step is more simple,easy to start,and my key idea of motion board is give the two key words,the pulse number,and the pulse speed to fpga,and then
[03:18:23] <scutsxg> fpga give a defined speed and defined number pulses
[03:18:36] <SWPadnos> ok, so it may be a good start. you'll just need to remove most of rtapi_app_main so it'll load without giving an error
[03:18:49] <scutsxg> i think it is different from pluo_step and ppmc
[03:18:50] <SWPadnos> then add your own initialization/board detection instead
[03:19:10] <SWPadnos> mostly the same as the UPC from Pico Systems (which uses the PPMC driver I think)
[03:20:19] <scutsxg> now i can run emc2 with pluto_step driver (without hardware and machine),but the system goes very bad and some time unexpected realtime delay.
[03:20:52] <scutsxg> but emc2 goes well with hal_parport
[03:21:23] <scutsxg> so there is sth wrong,but didnt found it.:)
[03:21:44] <scutsxg> anyway thanks SWPadnos
[03:22:07] <SWPadnos> sure. good luck
[03:23:41] <maddash> doh!
[03:24:08] <maddash> davidf: have you tried removing that boa constructor package before 'sudo apt-get -f install'?
[03:26:02] <davidf> maddash, no I havn't. I just did the purge, and have 73 broken packages. sys says I need three pytonn pkgs, but I get errors when I try to get them. just a sec...
[03:31:58] <danbh_intrepid> davidf: try to get that boa package first
[03:32:26] <davidf> OK.
[03:35:59] <davidf> maddash, Yipes. If I uninstall it, synaptic says about a hundred programs need to be removed as well, and also py2.4 will be installed.
[03:36:14] <maddash> whaaa?
[03:36:23] <maddash> link to output?
[03:36:50] <maddash> or reuse the apt-get --force-depends command again, but use it on the boa
[03:38:11] <davidf> I don't have output. I used the synaptic pkg mgr, and the file list is in a dilog saying "mark selected packages?" (to be removed) Have to do it in a terminal.
[03:38:11] <danbh_intrepid> its dpkg --force-depends --purge
[03:38:11] <danbh_intrepid> and the package name at the end
[03:38:46] <davidf> hi danbh_intrepid , pls give exact cmd?
[03:38:58] <danbh_intrepid> whats the boa package name?
[03:39:16] <davidf> boa-constructor
[03:39:35] <danbh_intrepid> sudo dpkg --force-depends --purge boa-constructor
[03:43:29] <maddash> yeah, no kidding. stick with the term, synaptic can't do much for you here
[03:43:47] <maddash> damn snakes
[03:43:57] <maddash> "Snakes on my Linux!"
[03:48:13] <davidf> OK, I still need py2.4 pkgs, but now I think it's only three. Boa is bye-bye.
[03:48:37] <SWPadnos> http://cgi.ebay.com/Hardinge-HC-13-x-15-Precision-Lathe_W0QQitemZ130262856631QQcmdZViewItem?hash=item130262856631&_trkparms=39%3A1%7C66%3A2%7C65%3A1%7C240%3A1318&_trksid=p3286.c0.m14
[03:48:41] <SWPadnos> man that thing is clean
[03:48:54] <davidf> And, still 73 broken pkgs.
[03:49:36] <maddash> dude, paste the output so we know what's going on besides "it's only three"; we can only *guess* which one is not needed
[03:49:45] <danbh_intrepid> davidf: dont worry about things being broken. You just need to get python installed, then sudo apt-get install -f will fix all that
[03:50:08] <davidf> sorry, I didn't know either for a minute. ok...
[03:50:28] <maddash> SWPadnos: is it supposed to be shiny? photoshop?
[03:50:40] <SWPadnos> not photoshop, but possibly repainted
[03:50:51] <SWPadnos> shiny usually equates to new or less used
[03:52:40] <davidf> http://pastebin.com/d1608c25e
[03:53:02] <davidf> danbh_intrepid, maddash there ya go.
[03:53:48] <danbh_intrepid> davidf: do the apt-get install -f
[03:53:52] <danbh_intrepid> and paste that
[04:01:04] <davidf> danbh_intrepid, http://pastebin.com/d2ccf1730
[04:01:06] <maddash> davidf: python is a transitional package that automatically installs the latest version of pytthon (py2.5) for you
[04:01:22] <maddash> davidf: stop trying to install python and start installing **python2.4**
[04:02:02] <davidf> maddash, OK.
[04:03:10] <maddash> davidf: are you sure you killed off the boa-constructor package?
[04:03:30] <maddash> from lines 183-184 of your latest paste: "# Compiling /usr/lib/python2.4/boa/proxyroles.py ... # File "/usr/lib/python2.4/boa/proxyroles.py", line 16"
[04:03:34] <toastydeath> in soviet america, boa-constructor package kills YOU!
[04:03:38] <danbh_intrepid> maddash: but thats whats failing to install. I'm stumped on this one. Maybe just screw the package, nuke its post install script, and upgrade to hardy, hoping for the best (second worst case scenario. Worst case, copy /home somewhere and do a clean install)
[04:03:46] <maddash> 'soviet america'?
[04:04:16] <maddash> danbh_intrepid: "that" = "python2.4" in your last comment, right?
[04:04:29] <danbh_intrepid> maddash: yeah
[04:04:34] <maddash> davidf: dpkg -l|grep -i boa
[04:04:54] <davidf> http://pastebin.com/d5b41452
[04:05:06] <maddash> btw, it's easier to do something like, "dpkg -l | grep -i boa > output" and upload the file, "output" to pastebin
[04:05:42] <davidf> There are 3 files not removed (boa) they are: dpkg - warning: while removing boa-constructor, directory `/usr/share/boa-constructor/Docs/cache' not empty so not removed.
[04:05:42] <davidf> dpkg - warning: while removing boa-constructor, directory `/usr/share/boa-constructor/Docs' not empty so not removed.
[04:05:42] <davidf> dpkg - warning: while removing boa-constructor, directory `/usr/share/boa-constructor' not empty so not removed.
[04:06:27] <davidf> toastydeath, what do you know about this really?
[04:06:33] <maddash> so is boa installed or not? what's that dpkg -l command return?
[04:06:39] <danbh_intrepid> whats emc2-axis?
[04:06:57] <maddash> if it isn't, then figure out which package the file, "/usr/lib/python2.4/boa/proxyroles.py" belongs to by executing "dpkg -S /usr/lib/python2.4/boa/proxyroles.py"
[04:06:59] <davidf> danbh_intrepid, you just asked a mouthfull. LOL...
[04:07:06] <maddash> danbh_intrepid: haha.
[04:07:12] <maddash> danbh_intrepid: this is #emc, btw...
[04:07:21] <danbh_intrepid> ah, lol
[04:07:37] <davidf> Its a cnc controller package.
[04:07:50] <maddash> danbh_intrepid: AXIS is that python-based GUI that cradek and jepler get high on
[04:07:58] <davidf> maddash, hang on re -l
[04:08:36] <maddash> * maddash waits for retribution from cradek or jepler
[04:11:03] <davidf> http://pastebin.com/d39eb4616
[04:11:20] <toastydeath> davidf: absolutely nothing!
[04:11:36] <toastydeath> i just saw an oppertunity and took it.
[04:11:55] <davidf> toastydeath, oh, good. very good. I feel better. LOL.
[04:12:06] <toastydeath> hahah
[04:12:09] <toastydeath> ^5
[04:12:19] <davidf> yipes, gimme a break. :) :)
[04:12:29] <toastydeath> also maddash
[04:12:29] <maddash> geez. so boa-constructor isn'tt installed. make sure that "/usr/lib/python2.4/boa/proxyroles.py" exists: "ls -l /usr/lib/python2.4/boa/proxyroles.py"
[04:12:44] <toastydeath> "soviet america" was a mention in Sinfest, as a play on the partially-nationlized banks
[04:12:55] <toastydeath> cuba calling bush "comrade" and whatnot
[04:13:03] <toastydeath> so now the joke is "in soviet america" rather than "in soviet russia"
[04:13:08] <maddash> toastydeath: good god.
[04:13:21] <SWPadnos> and then there's the Adam Ant song "U.S.S.A."
[04:13:22] <toastydeath> i am on the cutting edge of memes, you see
[04:13:24] <maddash> this would be so much easier if I were standing in front of your terminal
[04:14:13] <maddash> davidf: the ls -l command should output zero to one lines, so paste bin here
[04:14:19] <maddash> erm, paste**
[04:14:42] <davidf> dave@ubuntu:/$ "ls -l /usr/lib/python2.4/boa/proxyroles.py"
[04:14:43] <davidf> bash: ls -l /usr/lib/python2.4/boa/proxyroles.py: No such file or directory
[04:14:43] <davidf> dave@ubuntu:/$
[04:15:04] <maddash> davidf: "ls -l /usr/lib/python2.4/boa/" ?
[04:16:20] <maddash> if it exists, "dpkg -S ls -l /usr/lib/python2.4/boa"
[04:16:45] <davidf> that give hundreds of lines. too many for terminal to keep.
[04:17:13] <maddash> ok, so do "dpkg -S /usr/lib/python2.4/boa"
[04:17:18] <maddash> alternatively,
[04:17:20] <maddash> ok, so do "dpkg -S /usr/lib/python2.4/boa/*"
[04:17:35] <davidf> ls -l /usr/lib/python2.4/boa/ is what I was refering to..
[04:17:56] <maddash> "dpkg -S /usr/lib/python2.4/boa/*"
[04:18:45] <maddash> * maddash is doing a clumsy, clumsy job
[04:19:22] <maddash> "dpkg -S /usr/lib/python2.4/boa/* > dpkgS.output", and upload the file dpkgS.output to the pastebin site
[04:20:26] <maddash> er, use pastebin.ca to upload; pastebin.com won't let you
[04:22:02] <davidf> I got permission denied with that line and also with sudo.
[04:23:23] <maddash> "dpkg -S /usr/lib/python2.4/boa/* > ~/dpkgS.output"
[04:25:15] <maddash> hell, fuck it. just "sudo mv /usr/lib/python2.4 ${HOME}/python2.4"
[04:25:36] <danbh_intrepid> lol
[04:26:03] <maddash> seriously, python2.4 doesn't exist, so just "nuke it," as you suggested
[04:26:26] <maddash> s/doesn't exist/shouldn't be installed/g
[04:26:58] <maddash> davidf is like my human ssh link
[04:27:02] <maddash> to his box
[04:29:51] <davidf> OK, wtf?
[04:33:02] <davidf> maddash, are you still there?
[04:35:44] <maddash> yes
[04:36:55] <davidf> you're freakin me out a bit...
[04:37:54] <maddash> how appropriate, it's 2 weeks till halloween
[04:38:01] <maddash> how's the python2.4 going?
[04:38:34] <davidf> I haven't done anything since the ssh thing. :)
[04:38:58] <maddash> use the sudo mv command to remove python2.4 dir from /usr/lib incase there's some conflict with when you install python2.4
[04:39:01] <maddash> did you do it?
[04:39:58] <davidf> oh. so then I don't have to remove all the dependent files too?
[04:40:51] <maddash> what dependent files?
[04:41:37] <davidf> well, when I tried to remove it with pkg mgr, it said like a hundred dependent files would be removed as well.
[04:42:06] <maddash> you mean, a hundred dep PACKAGES, not files
[04:42:16] <maddash> just do the sudo mv command
[04:42:28] <maddash> then "dpkg -l|grep python2.4"
[04:42:31] <davidf> yes thats what I meant.
[04:42:51] <davidf> ok...
[04:43:20] <maddash> if there aren't any results from the dpkg -l command, then that means python2.4 package isn't installed
[04:44:05] <maddash> if python2.4 IS installed, 'sudo apt-get --force-depends --purge python2.4'
[04:44:18] <maddash> then, afterwards, 'sudo apt-get -f install'
[04:44:42] <maddash> let's try to fix this quickly, I have to sleep soon
[04:45:25] <davidf> ok, I did the dpkg -l|grep python2.4 and got a list but I was in my home dir. is that ok?
[04:45:46] <maddash> if you got a list, pastebin it
[04:47:10] <davidf> http://pastebin.com/dcf4e1a2
[04:48:43] <davidf> maddash, there it is.
[04:48:47] <maddash> you're on dapper drake, right?
[04:49:03] <davidf> 6.06 LTS
[04:49:37] <davidf> yes. Dapper drake.
[04:51:31] <maddash> execute: dpkg -l|grep python2.4|cut -d " " -f 3|tr "\n" " " > python2.4.olddeps
[04:51:37] <maddash> and paste the contents
[04:52:14] <maddash> of the file, python2.4.olddeps
[04:52:51] <davidf> permission denied, even with sudo
[04:52:54] <maddash> geez
[04:52:57] <maddash> are you in your home dir?
[04:53:03] <davidf> yes.
[04:53:10] <maddash> wtf
[04:53:19] <maddash> dpkg -l|grep python2.4|cut -d " " -f 3|tr "\n" " " > ${HOME}/python2.4.olddeps
[04:53:36] <maddash> and check out ~/python2.4.olddeps
[04:53:49] <davidf> ok that worked. just a sec...
[04:53:52] <maddash> you don't have write access to your own homedir?
[04:53:58] <maddash> you're not in your freakin' home dir
[04:54:05] <maddash> cd ~
[04:54:24] <maddash> geez
[04:54:52] <maddash> stop copying from the term by hand and just upload the file python2.4.olddeps directly, it'll save us both a lot of time
[04:55:10] <davidf> when I type cd space and shift- tilda, the key does not work.
[04:55:18] <maddash> gods
[04:55:25] <maddash> then, just "cd"
[04:55:26] <maddash> ok?
[04:55:30] <davidf> crazy
[04:55:58] <davidf> that worked.
[04:56:16] <maddash> then hurry up and upload the output, geez
[04:57:03] <maddash> * maddash hopes that jepler or cradek or alex_joni or jmkasunich or whoever owns linuxcnc.org won't kill maddash for filling the logs with this garbage
[04:57:27] <maddash> davidf: use pastebin.ca to upload the file directly
[04:57:36] <maddash> http://pastebin.ca/upload.php
[04:58:53] <SWPadnos> my log space runneth under
[04:59:05] <davidf> http://pastebin.com/dd459b5e
[04:59:22] <SWPadnos> (actualy it doesn't, you'd have to type fro a few centuries for that to happen)
[04:59:27] <SWPadnos> for
[05:00:50] <maddash> SWPadnos: thanks
[05:01:02] <SWPadnos> np. now I should be in bed. good night
[05:01:38] <davidf> http://pastebin.ca/1230752
[05:02:06] <davidf> good night swp. sorry.
[05:02:18] <davidf> ok now its on both sites.
[05:02:18] <maddash> ok ok.
[05:02:25] <maddash> pastebin.ca is enough.
[05:02:52] <maddash> run: cat ${HOME}/python2.4.olddeps|sudo xargs dpkg --force --purge
[05:03:09] <maddash> and paste the output, please
[05:03:52] <davidf> dave@ubuntu:~$ cat ${HOME}/python2.4.olddeps|sudo xargs dpkg --force --purge
[05:03:52] <davidf> dpkg: unknown force/refuse option `--purge'
[05:04:57] <maddash> oh geez
[05:05:04] <davidf> sorry.
[05:05:11] <maddash> cat ${HOME}/python2.4.olddeps|sudo xargs dpkg --force-depends --purge
[05:05:18] <maddash> that should do it
[05:05:58] <davidf> its running, removing...
[05:06:16] <maddash> finally
[05:06:17] <maddash> gods
[05:06:22] <maddash> paste the output please
[05:07:02] <maddash> this would've been resolved 2 hours ago if I had access to your term
[05:08:47] <davidf> http://pastebin.ca/1230756
[05:09:10] <davidf> well, no offense, but I'm glad you didn't. :)
[05:09:43] <maddash> that's all the outpu?t
[05:09:46] <davidf> wish there was a way to email a beer..
[05:10:34] <maddash> that's all the output?
[05:11:35] <maddash> helloo? what does dpkg say on the last few lines before exit?
[05:11:52] <davidf> looking.
[05:12:18] <davidf> Removing python2.4-gtk2 ...
[05:12:18] <davidf> dpkg: python2.4-htmlgen: dependency problems, but removing anyway as you request:
[05:12:18] <davidf> python-htmlgen depends on python2.4-htmlgen.
[05:12:18] <davidf> Removing python2.4-htmlgen ...
[05:12:18] <davidf> dave@ubuntu:~$
[05:12:35] <maddash> er
[05:12:40] <davidf> yes it ended strangely
[05:13:12] <maddash> this should take less than a minute. run: cat ${HOME}/python2.4.olddeps|sudo xargs dpkg --force-depends --purge 2> ${HOME}/dpkg.output
[05:13:20] <maddash> and upload ~/dpkg.output to pastebin.ca
[05:13:37] <maddash> seriously, less than a minute.
[05:13:38] <davidf> ok.
[05:14:40] <maddash> .........
[05:15:55] <davidf> http://pastebin.ca/1230759
[05:16:31] <maddash> ok ok
[05:16:53] <maddash> run: sudo aptitude install python2.4
[05:17:21] <davidf> dave@ubuntu:~$ sudo aptitude install python2.4
[05:17:21] <davidf> sudo: aptitude: command not found
[05:17:25] <maddash> when it prompts you for a choice, copy and paste the output
[05:17:29] <davidf> apt-get?
[05:17:46] <maddash> sudo apt-get -f install
[05:17:58] <maddash> if it prompts you, select the choice of inaction, and copy/paste the output
[05:18:50] <davidf> oops it said continue? y/n and I said y.
[05:19:07] <maddash> gods
[05:19:32] <maddash> http://dictionary.reference.com/browse/inaction
[05:19:43] <davidf> yeah.
[05:19:50] <maddash> ctrl c
[05:19:53] <maddash> CTRL+C
[05:19:57] <maddash> if it's still running
[05:20:11] <maddash> did you abort it?
[05:20:35] <davidf> tried, just get a c.
[05:21:02] <maddash> what?
[05:21:04] <davidf> terms ignoring control c
[05:21:12] <maddash> ok, what is it doing now?
[05:21:35] <davidf> done, it was compiling & setting up, etc.
[05:21:39] <maddash> hell
[05:21:42] <maddash> paste output
[05:23:53] <davidf> http://pastebin.ca/1230760
[05:24:30] <maddash> shit
[05:24:33] <maddash> it worked
[05:24:39] <maddash> you're fixed, dude
[05:24:53] <davidf> maddash, my upgrade available icon is back. should I try it?
[05:24:55] <maddash> fuck, tomato juice all over kb
[05:25:02] <davidf> hahahaha
[05:25:09] <maddash> can't hit quote key
[05:25:20] <maddash> yes, upgrade to your hearts content
[05:25:20] <davidf> try a little vodka with that.
[05:25:34] <davidf> man, thank you.
[05:25:37] <maddash> "
[05:25:38] <maddash> """""""""""
[05:25:39] <maddash> ''''
[05:25:40] <maddash> '
[05:25:41] <maddash> '"""'''''''''''
[05:25:42] <maddash> '''''
[05:25:44] <maddash> ''
[05:25:47] <davidf> hahaha
[05:25:55] <maddash> oh crap
[05:26:02] <maddash> didn't defocus while cleaning
[05:26:05] <davidf> LOL
[05:26:23] <maddash> upgrade first, then thank me if it works
[05:26:28] <maddash> ' ;'/
[05:26:30] <maddash> '/;
[05:26:31] <davidf> ok. :)
[05:26:54] <maddash> SWPadnos: extra log is the tomato's fauly
[05:26:56] <maddash> fault*
[05:27:26] <maddash> btw, try staying away from the boa-constructor pkg
[05:27:36] <maddash> it's the one that started all of this
[05:27:47] <maddash> boas and pythons don't play nicely together
[05:28:22] <davidf> maddash, when I click the upgrade icon, it blinks but nothing happens.
[05:28:30] <maddash> /geez
[05:28:41] <maddash> sudo apt-get update
[05:28:46] <maddash> when that's finished,
[05:28:51] <maddash> sudo apt-get install aptitude
[05:28:54] <maddash> when that's finished,
[05:28:59] <maddash> sudo aptitude upgrade
[05:31:50] <maddash> davidf: you're in missouri?
[05:31:57] <davidf> 123 packages upgraded, 0 newly installed, 0 to remove and 2 not upgraded.
[05:31:57] <davidf> Need to get 169MB of archives. After unpacking 291kB will be used.
[05:31:57] <davidf> Do you want to continue? [Y/n/?]
[05:32:04] <davidf> do it?
[05:32:07] <maddash> geez
[05:32:08] <maddash> yes
[05:33:01] <maddash> gods
[05:33:48] <davidf> yes MO how did you know?
[05:33:50] <maddash> you're EST, too, like me
[05:33:58] <maddash> it's freakin' 1:30 AM here
[05:34:03] <davidf> no, cst
[05:34:08] <davidf> 12:30
[05:34:11] <maddash> I grabbed your ip from /whois, then ip geolocation
[05:34:22] <maddash> [sigh]
[05:34:44] <maddash> is aptitude finished yet
[05:34:52] <davidf> no.
[05:34:57] <maddash> btw, learn to use aptitude's GUI
[05:35:07] <maddash> 'sudo aptitude' gives you a pretty ncurses intf
[05:35:24] <maddash> on top of that you can use aptitude to do all sorts of tricks that apt-get does
[05:35:36] <maddash> unlike synaptic, which is useless on the command line
[05:36:15] <davidf> ok, i'll try it.
[05:36:54] <maddash> press F10 to bring up the menu in aptitude's gui, and f10 to close it
[05:36:57] <davidf> maddash, I'm really sorry you got into this. I do really appreciate how hard you worked on this for someone you don't even know.
[05:38:13] <maddash> :)
[05:38:44] <maddash> is aptitude done yet?
[05:38:55] <maddash> what pkgs is it upgrading?
[05:38:56] <davidf> so all this was really cause of boa?
[05:39:13] <davidf> open office. big.
[05:39:18] <maddash> just that?
[05:39:36] <maddash> what else besides openoffice?
[05:39:46] <davidf> oh no, sh*t loads.
[05:39:57] <maddash> hit ctrl+C
[05:40:08] <maddash> abort aptitude
[05:40:42] <maddash> when you've done that, retry: sudo aptitude upgrade
[05:40:48] <maddash> but this time, select "n"
[05:40:55] <maddash> and just paste the output of aptitude's prompt
[05:41:13] <davidf> ok, but oo is almost done, and a lot of them happened first.
[05:41:32] <maddash> geez
[05:41:45] <maddash> there's a "time remaining" in the lower right corner
[05:42:12] <maddash> if it says some number with "m", then you've got several minutes. I'm not waiting that long. ctrl+c
[05:43:42] <davidf> Fetched 169MB in 10m46s (261kB/s)
[05:43:42] <davidf> Extracting templates from packages: 100%
[05:43:42] <davidf> Preconfiguring packages ...
[05:43:42] <davidf> (Reading database ...
[05:43:47] <KimK_IA> maddash: thanks for your patience
[05:43:50] <maddash> [sigh]
[05:44:00] <maddash> KimK_IA: who are you?
[05:44:11] <maddash> KimK_IA: and more importantly, why are you thanking me?
[05:44:26] <davidf> looks like it's almost finished...
[05:44:31] <davidf> good question.
[05:44:33] <maddash> yeah, no kidding
[05:44:43] <maddash> KimK_IA: meh, welcome.
[05:44:59] <KimK_IA> just an observer on this occasion. thanking you because we all need help from time to time. I know I do.
[05:45:12] <maddash> so, I think before all this started, you needed to do something involving emc
[05:45:36] <maddash> 02:32:53 <maddash> why? it's an interrupt controller (96 interrupt sources, 64 vectors !), and it's programmable... 02:32:59 <davidf> jmkasunich, Hi. Man, I hate to bother you guys with this, but I'm kind of at an impass... I have a prob with dependencies on python2.4. I can't update or upgrade anything.
[05:45:59] <maddash> oh, so we're finished, then?
[05:46:36] <davidf> ??? what's all that?
[05:46:52] <maddash> that's your first post from over an hour ago
[05:47:18] <maddash> davidf: just for fun, try: aptitude show boa-constructor
[05:47:35] <davidf> no, I know, but were you telling Kim or me about doing something with emc? I use it...
[05:48:15] <maddash> davidf: talking to you
[05:48:37] <maddash> i thought you had some problem with python2.4 preventing you from using emc
[05:48:41] <maddash> bah
[05:48:48] <maddash> aptitude show boa-constructor
[05:48:53] <maddash> and tell me the version
[05:49:11] <davidf> I've been on here many times, for a few years. I'm not a developer though.
[05:49:13] <davidf> ok...
[05:49:42] <maddash> no, wait. forget it.
[05:49:50] <maddash> do all your upgrading first, leave boa alone.
[05:49:59] <maddash> you're upgrading to hardy, right?
[05:49:59] <davidf> term is still cranking...
[05:50:18] <maddash> aptitude hasn't finished yet?
[05:50:56] <davidf> no. And I'm just upgrading current OS dapper, at least I hope.
[05:51:28] <maddash> bah
[05:51:32] <maddash> you should be good to go
[05:51:51] <davidf> practically everything on this box depends on py2.4 so it's a lot of stuff I guess.
[05:52:13] <maddash> yeah, why so many snakes?
[05:52:20] <davidf> I think so. If you want to go, please don't worry about me.
[05:52:22] <maddash> seriously -- so many python pkgs
[05:52:54] <davidf> I don't know.
[05:53:21] <davidf> it's an older box, maybe just slow?
[05:53:35] <maddash> huh?
[05:53:45] <maddash> why do you have so many python-*s installed?
[05:54:05] <davidf> I mean maybe the length timewise is processor speed?
[05:54:14] <maddash> er
[05:54:18] <maddash> what's it say on the term?
[05:54:28] <davidf> It's done now.
[05:54:50] <davidf> It was upgrading and setting up a whole lot of pkgs.
[05:55:24] <davidf> here's the last few...
[05:55:31] <maddash> can you pastebin.ca the /var/log/aptitude file?
[05:55:33] <davidf> Setting up postfix (2.2.10-1ubuntu0.2) ...
[05:55:33] <davidf> Postfix was not set up. Start with
[05:55:33] <davidf> cp /usr/share/postfix/main.cf.debian /etc/postfix/main.cf
[05:55:33] <davidf> . If you need to make changes, edit
[05:55:33] <davidf> /etc/postfix/main.cf (and others) as needed. To view Postfix configuration
[05:55:33] <davidf> values, see postconf(1).
[05:55:35] <davidf> After modifying main.cf, be sure to run '/etc/init.d/postfix reload'.
[05:55:37] <davidf> Setting up kdelibs-bin (3.5.2-0ubuntu18.6) ...
[05:55:39] <davidf> Setting up kdelibs4c2a (3.5.2-0ubuntu18.6) ...
[05:55:41] <davidf> Setting up kdelibs4-dev (3.5.2-0ubuntu18.6) ...
[05:55:43] <davidf> Setting up khelpcenter (3.5.2-0ubuntu27.4) ...
[05:55:43] <maddash> oh god
[05:56:01] <maddash> are you done?
[05:56:12] <davidf> term is finished.
[05:57:16] <maddash> ..
[05:57:20] <maddash> /var/log/aptitude ===> pastebin.ca
[05:57:36] <davidf> i have it coppied. pastbining...
[05:58:11] <davidf> http://pastebin.ca/1230774
[05:59:04] <maddash> ok, so what do you want to do now?
[05:59:24] <davidf> Just got an error, Boa-constructor icon not found... I like that.
[05:59:46] <maddash> you're using kde, which I abhor and hence know nothing about
[06:00:01] <maddash> there should be some specific command to update your KDE menus, though
[06:00:02] <davidf> Now? probably let you get some z's
[06:00:27] <davidf> But I want to upgrade my clamav which is why I started all this.
[06:00:48] <maddash> er, through the ubuntu repos?
[06:00:58] <davidf> I had a thousand or so hits on my firewall, traceroute to my box.
[06:01:01] <maddash> clamav wasn't listed among the things to upgrade
[06:01:44] <davidf> I loaded it with pkg mgr, but it wont run till it is updated.
[06:01:58] <davidf> data files I think.
[06:02:03] <maddash> ok, so can you upgrade it or not?
[06:02:29] <davidf> I mean, it runs, but with error messages about outdated files.
[06:02:46] <davidf> I haven't figured out how yet.
[06:03:12] <maddash> it's likely that the "outdated files" are the data files, not the clamav binaries
[06:03:32] <maddash> ok, if there's nothing else, I have to pass out now
[06:03:43] <davidf> right. I realize that. I went to the site but the docs are poor and confusing.
[06:03:59] <davidf> LOL. You deserve it.
[06:04:11] <davidf> Are you here a lot?
[06:04:49] <davidf> I hope I can help you somehow sometime.
[06:04:55] <maddash> no class for the next few days, so i will be
[06:05:03] <maddash> here a lot
[06:05:28] <maddash> SWPadnos: sry about the log size
[06:07:36] <davidf> SWPadnos, me too. But thanks. I tried elsewhere first.
[12:42:03] <BigJohnT> Dallur: how goes the boat building?
[13:48:46] <fragalot_> cpu hogs. I has it.
[13:48:47] <fragalot_> :(
[13:58:31] <Kohlswa> im trying to find information on "spindle encoder"/"spindle sync"/"spindle index"/"spindle pulse" in the documentation and the wiki but for some reason cant seem to find anything?
[13:59:06] <fragalot_> Have you looked behind the couch?
[14:00:53] <Kohlswa> found something in emc2 code notes from 2004
[14:03:10] <Kohlswa> nevermind *sight*
[14:03:40] <SWPadnos> Kohlswa, you may want to look at the documentation for the "encoder" HAL module
[14:05:58] <Kohlswa> thank you, will do
[14:06:17] <SWPadnos> are you using EMC2 2.2.x or are you using a compiled version from source?
[14:08:01] <Kohlswa> well i beleave with ubuntu update its now the regular 2.2.6 ?
[14:08:19] <SWPadnos> yep, if you've installed packages only, then that's what you have
[14:08:38] <SWPadnos> are you trying to use a spindle with a single index pulse?
[14:09:16] <Kohlswa> yes, useing a inductive proximity sensor
[14:10:04] <SWPadnos> hmmm - ok. hopefully it's a good sensor :) (otherwise you may have variation in the spindle angle with respect to when the index is detected)
[14:10:41] <SWPadnos> I believe that 1PPR threading is only available in the CVS "pre-2.3" version, but I'll check
[14:11:39] <Kohlswa> I'm useing a spindle sync schematic I've found for the HobbyCNC board, useing a AE1-AP-2A sensor.
[14:15:52] <Kohlswa> I've found the encoder section in the hal documentation, thanks.
[14:22:44] <SWPadnos> ok, the 1PPR support isn't released yet, it's only in the latest source
[14:26:30] <Kohlswa> ah, ok. reading up on pulses, seems one pulse per rev isnt popular.
[14:26:46] <SWPadnos> it "seems to work", but shouldn't be very good :)
[14:27:08] <SWPadnos> it's dependent on you taking light cuts relative to the spindle power/stability
[14:28:05] <SWPadnos> it seems that it can work well, as long as you never take heavy cuts, and you use enough light cutting/spring passes at the end
[14:28:57] <Kohlswa> *nods* since thats the way around it i guess ill have to live with it for now.
[14:29:25] <fragalot_> fragalot_ is now known as fragalot
[14:29:34] <SWPadnos> skill and technique can do a lot - we don't use stone tools any more now do we? :)
[14:30:18] <Kohlswa> adding more triggering trips to the spindle aparently isnt going to help ether since it needs one signal per rev ?
[14:30:37] <SWPadnos> oh no - if you can put an encoder on the spindle, that will work very nicely
[14:30:42] <SWPadnos> that's the preferred method
[14:31:21] <SWPadnos> it's just that low resolution encoders (like 1 count per rev) don't work well
[14:32:17] <Kohlswa> well i ment add one more pin 180degess from the first so the sensor trips twice per rev
[14:32:21] <jepler> 4PPR won't work much better than 1PPR (i.e., not at all in 2.2.6)
[14:32:33] <SWPadnos> heh - right. something like 100 would be good :)
[14:32:46] <SWPadnos> if there's a gear you can use that prox switch on, that might work
[14:32:50] <SWPadnos> but you still need index
[14:37:31] <jepler> if you don't mind doing just a little bit of commandline stuff (as well as testing experimental code), it looks like the new encoder component can be easily added to 2.2.6. step 1: install emc2-dev using package manager. step 2: download http://emergent.unpy.net/index.cgi-files/sandbox/encoder23.c . step 3: in a terminal, go to the directory where you downloaded it, and type "sudo comp --install encoder23.c". step 4: modify your hal files to use 'loa
[14:38:06] <Kohlswa> hmm, is threading with a proper encoder supported ?
[14:38:11] <SWPadnos> yes
[14:38:38] <Kohlswa> japler, your mesage was cut short after step 4.
[14:38:42] <SWPadnos> there's a lot that can be done with a proper encoder that can't be done with 1PPR
[14:38:55] <jepler> step 4: modify your hal files to use 'loadrt encoder23' (pin names are the same, starting with encoder., but there's a new pin position-interpolated available)
[14:39:50] <jepler> (oh, and in step 3 you'll be prompted for your password)
[14:40:14] <jepler> (oh, and in step 1 select the package build-essential too)
[14:40:57] <Kohlswa> copied and saved for later, thank you.
[14:42:35] <jepler> when it turns out I left something else out of the instructions, come back and ask
[14:42:46] <jepler> that file will remain at that location for at least a month, and probably much longer
[14:43:23] <roberth> latest CVS when jogging axis the indicator on axis doesn not move on screen, just updated my cvs today
[14:43:58] <Kohlswa> how come threading seems to so be experimental in emc ?
[14:43:59] <SWPadnos> do your motors move?
[14:44:06] <SWPadnos> threading is quite established
[14:44:07] <jepler> roberth: prior to today, whens't he last time you updated?
[14:44:14] <jepler> Kohlswa: it's not; 1PPR threading is experimental, though.
[14:44:16] <roberth> few weeks old
[14:44:29] <SWPadnos> but doing it with sub-optimal 1PPR hardware is
[14:44:33] <roberth> yes SWPadnos they do move, if i switch back to 2.6.6 it works ok
[14:44:38] <SWPadnos> ok :)
[14:44:44] <SWPadnos> (had to check you know)
[14:45:02] <jepler> roberth: run emc from the terminal and see if there are some error messages being printed. if there are, put them on pastebin.ca and tell us the URL
[14:45:13] <roberth> also its not obaying the max vel it seems, if u change value its using something else,
[14:45:32] <roberth> max accel, not vel sorry
[14:46:05] <Kohlswa> ok, but 1ppr seems to be the normal approach in other softwear like turbocnc and mach ?
[14:46:16] <roberth> jepler, i couldt see any errors erlyer, let me go check again tho
[14:46:33] <jepler> Kohlswa: so I hear
[14:46:43] <roberth> KimK_IA, maybe they not real cnc controllers ;)
[14:48:29] <roberth> i think 1ppr seems ok on small cuts etc, but on something big where want to take some nice big cuts on abar, doing css/feed per rev u start to see limits of a 1ppr feedback, and also pitch error on screw cutting over long distance etc
[14:50:58] <SWPadnos> 1ppr is the easier way if you don't normally use any feedback. since EMC has the ability to use real feedback, we didn't bother to add 1PPR support, since it's inferior in every way except cost
[14:51:36] <SWPadnos> then someone came along and wanted 1PPR support badly enough that they paid someone to add it
[14:51:48] <SWPadnos> so now you (and we) get the benefits :)
[14:51:59] <Kohlswa> roberth, *nods* well thats a level above what i whant out of this machine. what spindle encoders do people use with emc to get good results then?
[14:52:05] <robh> jepler, heres a pastbin, http://pastebin.ca/1230997
[14:52:33] <robh> i cant test out x y z tho, only A axis at moment, so maybe only a rotaional problem in axis?
[14:52:56] <SWPadnos> hey - got embedded systems with good RT performance there?
[14:54:10] <robh> sorry
[14:54:20] <SWPadnos> I noticed the leadwell name
[14:54:34] <SWPadnos> I may have it confused with another manufacturer though
[14:54:36] <robh> aaah, sorry yea, my RT performce is on wiki added it for others
[14:54:39] <Kohlswa> with a proper spindle index setup are you able to cut multiple start threads ?
[14:54:50] <robh> nah im converting my leadwell mcv into emc
[14:54:54] <jepler> robh: you are expecting the cone to rotate when you move A?
[14:55:04] <robh> ye as in 2.6.6 did
[14:55:30] <robh> notice MDI looks for axis to be home now before taking MDI command nice touch added btw
[14:55:48] <jepler> robh: ok, hold on -- I think I know what's going on
[14:56:06] <SWPadnos> Kohlswa, with a properly matched spindle/feedback setup (ie, one where the feedback is good enough to indicate when speed changes occur), you can cut multi-start threads
[14:57:05] <SWPadnos> you do this by offsetting the Z start position - if you're doing 5TPI and you want a second start, you'd move 0.1 inch further away before starting the threading pass (0.1" = 1/2 turn at 5TPI)
[14:57:18] <robh> also the MAX_ACCELERATION = is not read or something it seems also, as i have nice long ramp as tho it was reading MAX_VELOCITY or something
[14:57:26] <jepler> robh: I think you need to add [DISPLAY]GEOMETRY=XYZA in your inifile to get the same behavior as 2.2
[14:57:45] <jepler> robh: in 2.3, axis has more flexibility as to how to display rotary axes
[14:58:26] <SWPadnos> hmmm. I thought leadwell was a computer manufacturer - I guess not
[14:58:46] <jepler> (for instance, depending how your A axis is set up, GEOMETRY=XAYZ may be more appropriate)
[14:59:26] <robh> i see, sort of telling it which plain to display in?
[15:00:01] <jepler> there's a little bit of documentation for GEOMETRY= here: http://linuxcnc.org/docs/devel/html/config_ini_config.html#sub:AXIS-Interface
[15:02:02] <robh> jepler, that sorting the problem out
[15:02:11] <JymmmEMC> SWPadnos: Bundwell (Kaypro Compatables bka CP/M)
[15:02:13] <robh> sorted
[15:02:16] <JymmmEMC> BONDWELL
[15:02:26] <SWPadnos> yeah, I remember them
[15:03:34] <JymmmEMC> SWPadnos: I wish I still had mine... lugable... 7" monochrome (amber) and two 5.25" fdd
[15:03:45] <SWPadnos> 37 pounds
[15:03:50] <SWPadnos> thise were the days
[15:03:53] <SWPadnos> those
[15:04:24] <jepler> robh: now about your A-axis acceleration. is this something you think recently changed too?
[15:05:01] <robh> yea, worked ok before, where now seems to limit to max_vel or something
[15:07:42] <jepler> is this while jogging or while running gcode?
[15:09:36] <robh> jogging only, mdi or code it obays
[15:10:11] <jepler> if you have a [TRAJ]MAX_ACCELERATION line in your inifile, remove it.
[15:11:49] <robh> still same
[15:15:59] <jepler> same with [TRAJ]DEFAULT_ACCELERATION
[15:17:27] <robh> jepler, that got it
[15:18:31] <jepler> I'm glad, but unlike the other thing I don't remember it changing recently
[15:20:27] <robh> bin awhile since iv had chance to play around with it all,
[15:24:36] <robh> out of question, what is the inpostion window with emc? before the next axis command will start to move
[15:26:16] <jepler> you can specify it with G64 P-
[15:27:09] <jepler> if you don't specify it (G64 or G64 P0) then it's "best feedrate possible, so long as each move specified in the part program is touched at at least one point"
[15:30:18] <robh> like on our fanucs there a parameter u can set, defins at what value next axis move can start in op direction etc
[15:33:20] <jepler> That sounds similar to what G64 P- does. If you program G64 P.01 then blending will not take you more than .01 from the programmed path.
[15:34:32] <jepler> here's a thumbnail (original image lost) of what specifying different P values does: http://wiki.linuxcnc.org/uploads/g64-3-small.png (same path three times, different G64 P- values)
[15:39:59] <jepler> bbl
[15:42:08] <robh> jepler, ill have a play with it once i can machine up and running fully still heaps to do on it so little time
[15:42:22] <robh> who wants to have a UK EMC fest lol
[15:45:52] <fragalot> me, but you have to hold it in belgium.
[15:47:28] <jepler> I may have American-level knowledge of geography, but what part of the UK is belgium in?
[15:47:52] <fragalot> jepler: both in europe.
[15:49:03] <robh> fragalot, u pay transport ill bring machine ;)
[15:49:50] <fragalot> rent-a-truck ?
[15:50:47] <robh> depends which machine we brings lol have a HNC needs retro fitting, and this leadwell
[15:51:12] <fragalot> * fragalot is so proud of himself right now
[15:51:17] <robh> fragalot, be cheaper u come here plenty rooms spare :)
[15:51:38] <fragalot> I just reinstalled Gentoo on my laptop.. managed to get it to boot in under 7 seconds (from grub to xdm)
[15:51:59] <fragalot> and yes, all services are started.
[15:59:22] <robh> just wish they would port solidworks onto linux, then i could get rid of crappy windows at work :)
[16:00:01] <fragalot> :'( it's faster to reboot now than it is to use the hybernate feature.
[16:00:11] <fragalot> robh: Agreed.
[16:00:48] <robh> mind u then ud need cam too, live and hope i guess
[16:55:57] <KimK_IA> roberth: Sorry I missed your earlier msg, I stepped away for a minute. Yes, I have no use for 1 PPR spindle "encoders" either. Except that they make real good "actual spindle RPM" counters. And that's all.
[16:57:00] <jmkasunich> 1ppr threading was added for the benefit of people who simply can't come up with a better encoder
[16:57:22] <jmkasunich> regular encoder based threading has been part of emc2 for at two or three years
[16:57:32] <roberth> yea, would be nice to take say Z mark from spindle incoder and display a routh spindel speed to operator so you could see on screen what spindel speed u was realy doing to the programed one
[16:57:52] <jmkasunich> you can do that with any spindle encoder
[16:58:08] <jmkasunich> doesn't need to be the Z mark, I use the actual encoder tracks
[16:58:33] <jmkasunich> on my lathe, I'm software counting the encoder with our encoder hal module - it has a velocity output
[16:59:03] <jmkasunich> that output is in revs per second, so I run it thru a scale block to multiply by 60
[16:59:03] <roberth> aaah cool
[16:59:17] <jmkasunich> and then send it to a pyvcp widget that gives me a visual and numerical indication next to the axis preview window
[16:59:23] <roberth> just its alot of pules to count at 6000rpm if u have say 2000 count incoder
[16:59:39] <jmkasunich> right - if you have that many counts, you need hardware encoder counting
[16:59:56] <roberth> i have the mesa cards
[16:59:58] <jmkasunich> assuming you are using some hardware encoder counter for threading, you can still get the speed from that
[17:00:28] <jmkasunich> I'm not sure if the mesa driver exports a velocity pin, but if it doesn't, you can send the position pin thru a ddt block to convert to a velocity
[17:03:22] <roberth> one thing i will be looking to do is have a way to show a distance to go for each indivial axis
[17:05:59] <KimK_IA> emc2 doesn't have distance to go? I never looked. Wow, that's scary. I like my feedrate limiter knob and my distance to go display.
[17:06:52] <roberth> it has one, but it shows for which ever axis is active i belive how it works as its a 1 dtg only, not like comcerial where u have 1 for each axis shown, along with all active Gcodes etc
[17:07:35] <roberth> would be nice to slot them down the side in axis sim window next/opsit the machine/actual values
[17:07:56] <fragalot> http://omploader.org/vdW4w Epic.
[17:10:20] <jmkasunich> DTG is not per axis
[17:10:30] <jmkasunich> it is the distance along the move
[17:10:56] <jmkasunich> so if you are making a 45 degree cut that is 1" long (which means X and Y both move 0.707"), the DTG will start at 1.0 and count down
[17:11:20] <jmkasunich> if you are cutting a complete circle 1" in diameter, I'm about 99% sure that DTG will start at 3.14" and count down
[17:11:58] <KimK_IA> ha, thanks, you anticipated me. 180's and 360's were going to be my next question
[17:12:13] <roberth> aaah im withya
[17:13:49] <roberth> i was just saying would be nice to see a dtg per axis, as in program prove out its very handy for first runs when your machine is doing 10m or what ever in rapid
[17:14:40] <jmkasunich> I'm not sure I see the benefit, can you explain?
[17:14:53] <KimK_IA> btw, in the event you find a way to add multi-axis DTG to emc2/axis, you might like to know that Dynapath has a test mode where the DTG display is borrowed to show following error
[17:14:54] <jmkasunich> if you are doing a single axis rapid, EMC's DTG = that axis DTG
[17:15:12] <roberth> if u have 2 or 3 axis coming in at once
[17:15:13] <jmkasunich> if you are going diagonally, it shows the total - why would you want the individuals?
[17:16:18] <jmkasunich> KimK_IA: EMC2 has following error available as a HAL pin (or maybe its a parameter) If it is a pin, you can route it to PyVCP indicators and have it displayed all the time
[17:17:08] <KimK_IA> jmkasunich: the benefit is when you are debugging a program for the first time, you can slow or stop motion with the feedrate knob, and look at DTG to predict what is going to happen next (to avoid unforseen disasters)
[17:17:41] <KimK_IA> jmkasunich: thanks, the HAL pin should do nicely
[17:18:04] <roberth> makes a very quick visual check
[17:18:25] <jmkasunich> and you want to make your measurements parallel to the axes, rather than parallel to the direction of the move?
[17:18:32] <roberth> ill check out the hal pins thx, and ill write up a wiget
[17:21:26] <KimK_IA> Once I stop motion and look away (at a part drawing? the program listing?) when I come back, what angle was it moving again? Or just look so much x left, oh, y is done, so much z left. And on rapids you can't count on the angle anyway once it's arrived.
[17:21:35] <roberth> if im moving 10mm into ajob, i want dtg start at 10 count down as it does now infact, so i can see iv only got 5mm to go say, is my clamp clear or what ever or have i got program totaly wrong lol
[17:23:47] <KimK_IA> there's nothing wrong with having the diagonal distance too, if there's room for it.
[17:25:04] <roberth> as said if each dtg was opsit the current axis values plenty of room to fit them into the sim window where axis values already are :)
[17:25:22] <KimK_IA> and as always, dtg in arcs/circles may get larger before it gets smaller
[17:26:14] <cradek> so it's distance to endpoint, but it's customarily (incorrectly) called distance to go
[17:27:18] <roberth> cradek, i guess ur right ya, as u askign for the distance to the end of move
[17:27:23] <KimK_IA> yes, especially in arcs/circles. (You programming guys are always thinking)
[17:28:05] <cradek> yes, that kind of thing makes us uncomfortable :-)
[17:28:23] <roberth> just we all say dtg as on prg/chk button it shows it as dtg on menu, well on our fanucs and siemens
[17:28:40] <cradek> I understand
[17:28:40] <KimK_IA> cradek: no, we don't expect you to "pi" your way around a circle, lol
[17:29:48] <cradek> the "problem" with our current DTG is I implemented it when someone told me he wanted DTG
[17:30:05] <cradek> so, I wrote the thing the words mean
[17:30:39] <cradek> I'd be happy to review a patch if someone contributes individual-axes-distance-to-endpoint aka "DTG"
[17:30:39] <jmkasunich> and the internals really are DTG - emc's trajectory planner computes the length of the path, whatever its shape or angle, and then moves along that path
[17:30:44] <KimK_IA> Google maps: What is the distance from New York to Los Angeles to New York? Answer: zero!
[17:30:56] <cradek> KimK_IA: haha
[17:30:59] <jmkasunich> so the distance remaining to travel along that path is used internally
[17:31:25] <jmkasunich> we could call the new one(s) DTE - distance to endpoint
[17:32:01] <KimK_IA> do you mean DTG per axis and DTE on the diagonal?
[17:32:06] <cradek> DTEPATTA: distance to endpoint projected along the three axes
[17:32:17] <jmkasunich> KimK_IA: no, exactly the opposite
[17:32:19] <KimK_IA> lol, no we need a longer name
[17:32:22] <roberth> if i could program i would sure as hell help write the dte/dtg all i can program is a machine tool lol
[17:32:31] <jmkasunich> DTG is the actual distance to be traveled along the path
[17:32:49] <jmkasunich> DTE is the distance between where we are now and the endpoint, regardless of the path
[17:33:09] <jmkasunich> DTG per axis has no meaning, DTE can be computed both globally and per-axis
[17:34:22] <KimK_IA> OK but in the industry DTG is per axis isn't it? Someone with multi-control experience help me here. If we said our DTG is different from everybody else's DTG, isn't that bad?
[17:34:56] <KimK_IA> Nobody is using DTE as far as I know
[17:34:59] <roberth> if we do a 45, on dtg/dte it should show how far the X and Y has to go to reach end of the 45
[17:35:10] <jmkasunich> says you
[17:35:19] <cradek> heh
[17:35:32] <jmkasunich> (not being rude, my point is that everybody might have their own idea as to what DTG should mean)
[17:35:41] <roberth> KimK_IA, i think its still called dtg on main controls as they have no other displayed, and by nature it easy to say, thats the distance to go till we reach end of motion on that axis
[17:36:09] <cradek> but but ... it's not
[17:36:13] <cradek> but but but
[17:36:18] <roberth> i wunder what it says in fanuc manual under dtg heading
[17:36:22] <cradek> it's the distance(s) from here to the endpoint
[17:36:31] <roberth> correct
[17:36:53] <roberth> its one of them names i feel never been changed and say u righly say should be dte
[17:37:04] <jmkasunich> if I'm driving from NY to LA and back, and I'm in Kansas, my DTG should be higher if I'm still going towards LA than if I'm on my way back already
[17:37:09] <cradek> I'm bold enough to say everyone in the whole world is using the wrong words
[17:37:39] <jmkasunich> and I agree with cradek
[17:37:50] <cradek> we can, but don't need to, also use the wrong words
[17:38:04] <cradek> * cradek <- pedant
[17:38:25] <roberth> if ur going from point A to B, and ur half way, its 50mm long, ur dtg is 25mm
[17:38:37] <KimK_IA> cradek: you may be right. But have you met any Indians lately? And were they from Arizona or Bombay?
[17:38:38] <jmkasunich> not if I'm cutting a circle
[17:39:19] <jmkasunich> most of the indians I see are from bombay
[17:39:26] <jmkasunich> the ones from AZ want to be called native americans
[17:39:28] <cradek> ahh, KimK_IA is a "words don't have any meaning except what we ascribe" anti-pedant
[17:39:31] <jmkasunich> they're pedants too
[17:40:20] <jmkasunich> both DTG and DTE accurately describe two different things
[17:40:36] <jmkasunich> IF (and only if) you are talking about a straight line, there is no difference
[17:41:06] <KimK_IA> No, just trying to go with the popular usage. Even if technically incorrect. And if we had displays of both the diagonal distance and the Manhattan distance, well, that's a point for emc2!
[17:41:07] <jmkasunich> for a curve, there certainly is a difference, and I'd rather we use the words accuratly
[17:41:08] <cradek> (the functionality is more important than what we call it; until we have the functionality, who cares?)
[17:41:20] <jmkasunich> right
[17:41:48] <cradek> for ancient controls that could only do one quadrant arcs, this was less of an issue
[17:41:52] <jmkasunich> except lets make it clear that we don't want to replace our display of actual distance to be traveled along the path
[17:42:00] <KimK_IA> OK, Manhattan distance is not exactly correct either. I mean individual distances.
[17:42:01] <cradek> at least the numbers never went up then back down
[17:42:25] <jmkasunich> we want to _add_ a new display of distance from current location to endpoint, per axis and total
[17:42:26] <cradek> by the time controls could do full circles, it was too late to change the name
[17:42:49] <KimK_IA> Hey, it's already written, and it's the hardest one, how often does that happen!? I say have them both.
[17:43:10] <jmkasunich> actually, its not the hardest one, its the easiest one
[17:43:29] <jmkasunich> the trajectory planner needed to calculate the DTG along the path as part of computing the path
[17:43:47] <jmkasunich> it has no reason (other than display) to calculate the other distances
[17:44:02] <KimK_IA> Really? three differences (subtracts) harder than two rectangular to polars?
[17:44:11] <jmkasunich> EMC does not think in terms of "how much do I need to move this axis"
[17:44:23] <jmkasunich> it thinks of "how far do I have to go along the path"
[17:44:24] <robh> change of comps, yea in arc its a dte with in the arc as the X and Y move in out of the arc move
[17:44:48] <jmkasunich> and then once if figures out a spot on the path to be at a certain moment, it translates that spot on the path to spots on each axis
[17:45:17] <KimK_IA> * KimK_IA realizes rect to polar on works on linear move. Bah!
[17:45:27] <KimK_IA> on > only
[17:45:46] <KimK_IA> only work on -- I give up
[17:46:00] <jmkasunich> heh
[17:46:06] <cradek> bbl
[17:46:07] <jmkasunich> gotta run for a while....
[17:47:41] <KimK_IA> well, whaddya think, roberth, robh? (Hey, are you the same guy?)
[17:48:00] <robh> dte should be how far to archive the distance move, arc are hard yes as ur moving in and out of move
[17:48:16] <robh> KimK_IA, yea, this other comp with emc on it
[17:50:02] <KimK_IA> Let's see Fanuc chat on IRC! (Oh, wait, they'd consider us risky, never mind)
[17:50:12] <robh> so in arc gte will go to + & - values
[17:50:37] <robh> lol
[17:51:04] <robh> im quite happy emc will take 99% of my fanuc programs and run them with no eddits, only thos with sub programs dont work
[17:51:17] <robh> that is milling wise
[17:51:19] <KimK_IA> I still think we should follow the most popular usage. Maybe a little research is in order?
[17:51:38] <KimK_IA> and what do you have to do to edit them to work?
[17:51:57] <robh> yea, ill see what the fanuc mills do 2morow at work with arcs, see if any mention in manual but i dout it be in manul much info
[17:52:13] <robh> take program number off emc dosnt know what 0xx is at start
[17:52:33] <robh> O
[17:52:43] <robh> bbl food time here staving :)
[17:52:52] <KimK_IA> thanks for the chat
[17:53:45] <robh> thx for backup ;)
[17:54:00] <KimK_IA> haha, np
[18:22:34] <roberth> KimK_IA , haha i see hass (or hassel ;) ) call it pos-to-go lol
[18:33:34] <maddash> whahaha!!! 500ksps with a 72MHz uc, no EMI !!
[18:38:05] <KimK_IA> roberth: ha, might as well have three opinions (four on the way, probably?)
[18:39:47] <roberth> KimK_IA, Heidenhain would be other odd one i bet they like there own things
[18:40:27] <anonimasu> hm..
[18:40:33] <anonimasu> heidenhain does great stuff..
[18:40:34] <anonimasu> :)
[18:40:34] <maddash> SWPadnos: did I ever tell you about the rabbit's A/D EMI problem
[18:40:37] <maddash> ?
[18:42:10] <anonimasu> *yawn*
[18:43:00] <alex_joni> good night all
[18:43:14] <maddash> alex_joni: what? did you switch time zones ?
[18:49:18] <dmess> hi all
[19:02:43] <maddash> ACH
[22:03:18] <maddash> haha, oscilloscope equalizer RULES!
[22:03:40] <maddash> pic32 ftw!
[22:07:03] <renesis> eh
[22:07:17] <renesis> arm > u
[22:13:29] <maddash> heh
[22:13:53] <maddash> now i need to make a simple 1 R 1C lowpass filter for my pwm output
[22:25:50] <maddash> my lowpass filter works -- i see voltage change on my scope. it's so beautiful!
[22:37:06] <KimK_IA> hey, congrats
[22:41:50] <maddash> :D
[23:04:15] <maddash> maddash is now known as cr-_-
[23:04:27] <cr-_-> cr-_- is now known as d-_-b
[23:06:53] <d-_-b> d-_-b is now known as i-_-i
[23:08:08] <i-_-i> i-_-i is now known as w-_-w
[23:08:17] <w-_-w> w-_-w is now known as wO_Ow
[23:57:36] <JymmmEMC> * JymmmEMC so hates M$, yet it pays the bills *sigh*