#emc | Logs for 2006-02-10

[01:13:17] <lerman_> lerman_ is now known as lerman
[01:27:46] <giacus> G night
[05:30:15] <CIA-8> 03swpadnos * 10emc2/src/emc/kinematics/tp.c:
[05:30:15] <CIA-8> Fix TP misblend problems with g0 reversals by reinstating parts of the old TP.
[05:30:15] <CIA-8> This makes 3D_chips run correctly for me using the default stepper_inch configuration.
[06:04:44] <SWPadnos> SWPadnos is now known as SWP_Away
[06:30:31] <lerman_> lerman_ is now known as lerman
[07:13:27] <orpheus> is servo-to-go the only servo controller supported by emc?
[08:07:01] <alex_joni> orpheus: no, there are quite a few
[08:07:23] <alex_joni> motenc, univpwm, m5i20, etc
[08:07:36] <alex_joni> but usually at this time of day it's pretty quiet in here
[08:07:49] <orpheus> ah, so i've seen now
[08:08:08] <alex_joni> ;-)
[08:08:13] <alex_joni> most folks from the US in here
[08:08:16] <alex_joni> where are you from?
[08:08:27] <orpheus> atlanta
[08:08:36] <orpheus> yes, it is 3:08
[08:08:47] <alex_joni> ouch :)
[08:08:53] <alex_joni> it's 10:13 over here
[08:08:54] <alex_joni> AM
[08:09:04] <alex_joni> Europe, btw
[08:09:11] <orpheus> i got a mill, and all i have to do now is convert it, somehow i wasted 3 hours sifting through ebay auctions
[08:10:16] <orpheus> i have no steppers, no drivers, and no magic control boxes, so it's kinda like, ...what do I wanna get?
[08:11:02] <orpheus> closed-loop servo? stepper motors? what kind of pc interface?
[08:11:15] <alex_joni> how much do you want to invest?
[08:11:19] <orpheus> thus i'm looking at ebay to get a feel f what's out there
[08:11:24] <orpheus> *for
[08:11:28] <alex_joni> close-loop servo is best (but expensive)
[08:11:34] <alex_joni> steppers is good (cheaper)
[08:11:40] <orpheus> i was hoping $200-300
[08:11:48] <alex_joni> ok, foget about servos then
[08:11:58] <alex_joni> you get about 1 axis for that money (if you are lucky)
[08:12:05] <orpheus> so i've noticed
[08:12:25] <alex_joni> so a 3 axis mill is about 1500$ at least (servo card, servos, motors, encoders, cables, etc.)
[08:12:38] <orpheus> it's a desktop mill, so do i really need closed-loop steppers?
[08:12:44] <alex_joni> steppers then, but I'm more a servo guy, so I can't really help you on that
[08:12:46] <alex_joni> no
[08:13:04] <alex_joni> you can get away with open-loop steppers on ANY machine, given configured ok
[08:13:10] <orpheus> heh
[08:13:27] <orpheus> t's just when things go wrong...
[08:14:06] <orpheus> so is my usderstanding correct in that you can wire a stepper driver straight to a parallel port?
[08:16:20] <orpheus> 1200 oz/inch... wow... tat would take your arm off
[08:22:59] <alex_joni> orpheus: yes, and yes
[08:23:07] <orpheus> heh
[08:23:28] <orpheus> yep, time to go to bed,
[08:23:32] <orpheus> thanks, btw
[09:07:34] <alex_joni> night
[13:12:50] <jepler> orpheus: "straight to the parallel port" is a bit of a simplification. I know firsthand that this board works well to drive stepper motors from step+direction signals, though: http://www.pminmo.com/l297-8/l297-8.htm
[13:13:01] <jepler> orpheus: parallel port -> breakout board -> one of these boards per axis
[13:50:46] <les_w> morning
[13:51:13] <les_w> I noted the list posts about TP errors in "3dchips"
[13:54:24] <les_w> was checking the math...seems ok
[13:55:38] <les_w> however it would be normal in a trapeziodal blend to not actually reach a waypoint
[13:55:54] <les_w> this would be magnified in a single axis reversal
[13:56:38] <les_w> there was hard code to go exact stop in trajectory angles greater than 90 degrees though
[13:56:47] <les_w> we may have disabled that bit
[13:57:39] <les_w> the large accel error is more disturbing
[14:11:44] <alex_joni> hi les
[14:23:29] <SWP_Away> SWP_Away is now known as SWPadnos
[14:23:41] <SWPadnos> hi guys
[14:24:07] <SWPadnos> les - it looks like reinstating some code that wad #ifdef'd out fixed the TP issues in 3D_Chips
[14:24:36] <cradek> hi
[14:24:40] <SWPadnos> hi
[14:24:50] <SWPadnos> I guess I found something :)
[14:24:54] <cradek> that's nice to know I guess
[14:25:16] <SWPadnos> I'm just not sure why the code was removed in the first place
[14:25:26] <cradek> because it made the typical misblends less bad
[14:25:54] <cradek> I mean, removing it did
[14:26:11] <SWPadnos> really? interesting
[14:26:52] <SWPadnos> so it improves typical misblends but screws up atypical ones
[14:26:58] <cradek> yeah, paul removed that and les saw better results, but nobody understood it.
[14:27:26] <cradek> well all the blends are always wrong.
[14:27:41] <cradek> even colinear (g1x1 g1x2 g1x3)
[14:29:57] <SWPadnos> well - I'll look at it a bit more today, once I've had coffee
[14:31:43] <alex_joni> hello guys
[14:31:46] <alex_joni> I'll be back later
[14:31:49] <alex_joni> going home now..
[14:31:53] <cradek> hi alex
[14:32:04] <SWPadnos> hi alex
[14:32:06] <SWPadnos> bye alex
[15:03:35] <les_w> sorry was away
[15:04:06] <les_w> yeah #ifdef ing that code out made a dramatic difference
[15:04:27] <les_w> I figured it might cause problems also
[15:04:42] <les_w> but the net effect was very good
[15:05:08] <les_w> with it in horrible misblending occurs at almost every point
[15:05:27] <SWPadnos> I'm trying to trace through it to see exactly what it does, but it hasn't sunk in yet
[15:05:51] <les_w> Now, it was put in to fix something....but it was in error
[15:06:03] <les_w> so removing it got rid of the error
[15:06:17] <les_w> but I guess there is still something to fix
[15:06:31] <SWPadnos> yeah, because putting it back fixed seomthing else ;)
[15:07:16] <les_w> with that bit in it's so bad it will tear a high speed machine apart
[15:07:43] <cradek> maybe we know what to do at fest now.
[15:07:55] <les_w> now it's better but has accel limit violations
[15:08:24] <SWPadnos> before fest would likely be best
[15:09:08] <les_w> let me check will's notes here
[15:11:31] <les_w> ach cvs is down for me
[15:11:36] <les_w> oh there it is
[15:13:20] <les_w> ok will's note specifically says that it was put in to make sure accel and vel limits were honored
[15:13:32] <SWPadnos> yep - that's why I put it back
[15:13:49] <les_w> so the observed 3dchips stuff makes sense
[15:13:54] <SWPadnos> I don't see exactly how it's doing this, but it did work
[15:14:26] <les_w> But with it in, horrible misblends occur
[15:14:44] <les_w> it's almost like exact stop mode even in g64
[15:14:54] <les_w> we did a lot of testing
[15:15:12] <SWPadnos> hmmm - I winder if there's a negative number happening somewhere
[15:15:15] <SWPadnos> wonder
[15:15:42] <les_w> for me I'll take good blends and allow accel limit violation....
[15:15:46] <cradek> I wonder if the trapezoids themselves are planned wrong
[15:16:20] <les_w> my machine can do about 200 in/sec^2
[15:16:45] <les_w> of course I want to have good blending and accel limit honored
[15:16:58] <les_w> but we could not figure it out
[15:17:21] <SWPadnos> heh
[15:17:23] <les_w> haha pretty much
[15:17:43] <SWPadnos> ./conigure --with-error=screw_the_limits
[15:17:58] <SWPadnos> ./conigure --with-error=screw_the_path
[15:18:03] <les_w> um...bbiam...chickens delivered!!
[15:18:04] <SWPadnos> ...
[15:18:35] <SWPadnos> of course, it would be more funny if I hadn't copied the misspelled 'configure'
[15:18:43] <SWPadnos> or maybe not
[15:31:46] <les_w> back...breeding chickens here
[15:32:02] <les_w> I think they can handle it from here
[15:32:15] <SWPadnos> err - OK, you can stop now ;)
[15:32:24] <les_w> haha
[15:32:40] <les_w> fantastic purebreed rooster
[15:33:42] <les_w> anyway...bottom line...Will's attempt to fix accel violations basically trashed the tp.
[15:33:48] <les_w> unfortunately
[15:34:32] <SWPadnos> hmmm
[15:34:59] <SWPadnos> I'll have to trace through tcRunCycle a bit to see what's really going on
[15:35:19] <les_w> good
[15:35:38] <les_w> half a dozen of us could not figure out a proper fix
[15:36:13] <SWPadnos> I probably won't either, though I know i can't if I don't look through tcRunCycle ;)
[15:36:13] <les_w> But will's code ddid such bad things that we figured it was best not there
[15:36:29] <les_w> well good to have a look anyway
[15:37:15] <SWPadnos> there are two questions - one is "what is the best way to blend?" and the other is "waht does the current code do?"
[15:37:31] <cradek> I took approach #1 when #2 failed
[15:37:37] <cradek> that's what simple_tp is about
[15:37:40] <SWPadnos> yep
[15:37:45] <cradek> I'm updating simple_tp today with the new changes
[15:37:54] <SWPadnos> so maybe adding some blend code to that would be the best route to take
[15:38:10] <cradek> maybe. It still plans trapezoids a bit wrongly and I could use some help with that.
[15:38:29] <cradek> once the trapezoids are right, I think it's not too hard to blend them
[15:38:48] <SWPadnos> ok. though we need to answer #1 before simple_tp can be made into not_so_simple_tp
[15:38:51] <les_w> math is simple enough.
[15:39:43] <cradek> yeah making some parabolas is not so hard
[15:39:49] <les_w> right
[15:44:26] <les_w> aw the hen is "not in the mood"
[15:44:38] <SWPadnos> that always happens
[15:44:45] <les_w> they need to get to know each other hahaha
[15:44:57] <les_w> I seldom had that problem.
[15:44:58] <SWPadnos> how many headaches can a person have before they should see a doctor?
[15:45:17] <les_w> ?? lots I guess
[15:45:27] <SWPadnos> heh - just kidding anyawy
[15:54:01] <les_w> hey...anyone ever bought from carmax?
[15:54:07] <SWPadnos> nope
[15:54:37] <les_w> just wondering if their " no haggle price" is the bottom line
[15:55:24] <les_w> ...while the shop heats up
[15:56:05] <SWPadnos> hard to tell. check kbb.com and consumer reports for "invoce prices"
[15:56:20] <SWPadnos> or expected prices, or whatever they call them
[15:56:26] <les_w> I have to test and install those optos
[15:56:31] <les_w> I did check those
[15:56:39] <les_w> they are always too high
[15:57:38] <SWPadnos> ah - you're wondering if they'll haggle over the no-haggle price. good qquestion
[15:57:44] <les_w> yeah
[15:58:34] <CIA-8> 03cradek 07simple_tp * 10emc2/src/emc/kinematics/ (tc.h tp.c tp.h trajectory.h): update simple_tp to work again after recent changes
[16:00:01] <CIA-8> 03cradek 07simple_tp * 10emc2/src/emc/kinematics/prflib.h: unused
[16:39:13] <alex_joni> hello
[16:39:36] <SWPadnos> hello
[16:40:29] <les_w> ups man had a load.....piezos came!
[16:40:31] <les_w> and
[16:40:53] <les_w> rooster and hen are laying around having a smoke...
[16:40:54] <les_w> haha
[16:42:22] <SWPadnos> all Mr. UPS man brought me was new checks :(
[16:42:34] <alex_joni> written out on your name?
[16:43:13] <SWPadnos> yeah - as the payer, not payee
[16:43:25] <alex_joni> too bad.. the other way around is more fun :)
[16:43:34] <SWPadnos> it's so much better to be a funds sink rather than a source
[16:45:45] <alex_joni> anyways.. I've been to that friend of mine last night, to talk about the linuxcnc website
[16:46:22] <alex_joni> SWPadnos: not sure if you remember the thread on the users list some time ago
[16:47:08] <SWPadnos> not sure - I remember some discussion on IRC though
[16:47:31] <SWPadnos> ah - right, doing it over to look nice
[16:47:40] <alex_joni> well, he gave it to his designer to come up with a nice redesign
[16:47:41] <SWPadnos> and use php or the like
[16:47:49] <alex_joni> and afterwards we'll worry about implementation
[16:48:00] <alex_joni> although I kinda favor joomla
[16:48:03] <alex_joni> ever seen that?
[16:48:06] <SWPadnos> nope
[16:48:10] <giacus> nice
[16:48:10] <alex_joni> http://www.joomla.org/
[16:48:16] <giacus> cool
[16:48:17] <giacus> :)
[16:48:26] <alex_joni> SWPadnos: try it out, there's a demo there
[16:48:32] <SWPadnos> looking now
[16:49:16] <giacus> we use in in roboitalia
[16:49:21] <giacus> it
[16:50:04] <SWPadnos> I'm not looking at the code at this point, but I don't see it as that radically different (for the user) than things in php, or even slashcode
[16:50:15] <SWPadnos> but the site looks nice
[16:50:59] <alex_joni> SWPadnos: the idea is that joomla is a CMS
[16:51:05] <alex_joni> you don't need to worry about the code
[16:51:16] <alex_joni> it's a tool, and you can have it look almost anyhow
[16:51:33] <SWPadnos> yep - I understand that, but setting up the CMS to run how you want can be a real pain
[16:51:43] <SWPadnos> though there's an eventual payoff
[16:52:37] <alex_joni> right
[16:53:16] <SWPadnos> given that the preferred document for this group is lyx, I'm not sure how much traction you'll get for any web CMS ;)
[16:53:21] <SWPadnos> document format
[16:53:50] <giacus> http://www.docebolms.org/doceboCms/index.php
[16:53:53] <SWPadnos> hah -this is funny
[16:53:58] <alex_joni> did you see how you can add/modify documents?
[16:54:10] <SWPadnos> I get a bunch of smaps a day about "enlargement"
[16:54:14] <SWPadnos> tspams
[16:54:16] <SWPadnos> gah
[16:54:22] <alex_joni> spams
[16:54:28] <SWPadnos> anyway - here's a great subject line:
[16:54:39] <SWPadnos> "Your partner will thank you. yuck"
[16:55:01] <SWPadnos> they really should filter their random word lists
[16:55:02] <alex_joni> lol
[16:55:21] <alex_joni> probably that yuck is what makes it pass spamfilters ;)
[16:55:27] <SWPadnos> heh
[16:55:39] <bill2or3> don't worry, the one they send you tomorrow will have a better word.
[16:55:40] <SWPadnos> I actually don't have any auto-delete filters set up, just auto-mark
[16:55:47] <SWPadnos> I only get around 25-30/day
[16:56:50] <giacus> I also get about 100/day
[16:56:57] <giacus> 99 are porn :P
[16:56:59] <giacus> haha
[16:57:22] <alex_joni> I get around 6-10 on about 5 email accounts
[16:57:26] <SWPadnos> see - the porn ones are OK, I just don't need enlargement or longevity (or an ocean of sperm)
[16:57:37] <giacus> lol
[16:57:49] <SWPadnos> but I hardly get any porn spam these days
[16:58:20] <giacus> you shoud try hotmail for that :D
[16:59:22] <SWPadnos> I don't need it badly enough to do that ;)
[16:59:43] <CIA-8> 03jepler * 10emc2/src/ (54 files in 30 dirs): (log message trimmed)
[16:59:43] <CIA-8> Revamp the build system of emc2.
[16:59:43] <CIA-8> The make is "almost nonrecursive": All userspace targets are built from a single
[16:59:43] <CIA-8> invocation of make, and then all realtime targets are built from a second
[16:59:43] <CIA-8> invocation of make. For more information about why this is a good thing, read
[16:59:44] <CIA-8> http://www.pcug.org.au/~millerp/rmch/recu-make-cons-harm.html
[16:59:48] <CIA-8> On one test system, the time for a complete rebuild dropped 20% from 3:17 to
[17:00:01] <cradek> wooo
[17:00:16] <SWPadnos> to what - tell us, tell us!
[17:00:23] <jepler> 3:17 * 80%
[17:00:34] <SWPadnos> yeah -I can do the math, but I'm American!
[17:00:40] <jepler> 2:37
[17:00:44] <giacus> * giacus wave to jepler
[17:00:50] <jepler> hi giacus
[17:00:55] <giacus> hello
[17:00:58] <cradek> jepler: can I change an include file and get a correct build without cleaning now?
[17:00:58] <alex_joni> jepler: that sounds nice..
[17:01:09] <alex_joni> * alex_joni looks
[17:01:15] <cradek> jepler: that's more important than 20% speedup to me
[17:01:29] <bill2or3> I have what may be a totally stupid question.
[17:01:56] <jepler> cradek: That should be true for the userspace targets; I'm less sure for "kbuild"
[17:02:24] <bill2or3> if you're cutting a workpiece larger than the working area of your mill, can you unmount, move, and re-mount the workpiece to allow you to actually make it on a mill that's too small to do it all at once?
[17:02:25] <SWPadnos> kbuild will just issue a make in the given directory, after initializing kernel-related stuff
[17:02:47] <alex_joni> bill2or3: sure you can, you just need to make sure you are accurate enough on the mounting
[17:03:13] <bill2or3> ok cool. I'm trying to decide on a size for my DIY mill project.
[17:03:14] <bill2or3> that helps
[17:03:42] <bill2or3> I'm thinking I want a workspace of like 8" by 12".
[17:03:49] <CIA-8> 03jepler * 10emc2/src/.cvsignore: fix ignores to match new Makefiles
[17:03:50] <SWPadnos> but make sure that the workpiece isn't so big that the little mill can't move it, or that the miill tips over when it's in some position
[17:03:50] <cradek> often making a part takes several/many mountings
[17:04:00] <alex_joni> I always overengineer about 30-40% ;)
[17:04:14] <SWPadnos> I guess I'm a worse engineer - I go for 50-100%
[17:04:15] <bill2or3> heh.
[17:04:17] <SWPadnos> :)
[17:04:25] <alex_joni> SWPadnos: let's say more cautious
[17:04:30] <bill2or3> I just dont want to make something that takes over my garage.
[17:04:35] <SWPadnos> paranoid is fine
[17:04:40] <alex_joni> and eats you up one day
[17:04:41] <alex_joni> :D
[17:06:19] <jepler> wow that commit dropped a bunch of e-mails in my mailbox
[17:06:21] <alex_joni> jepler: one question
[17:06:30] <alex_joni> yeah, one email / directory
[17:06:38] <alex_joni> I am reading through the Makefile now
[17:06:58] <alex_joni> does this build rtapi.ko or rtai_rtapi.ko ?
[17:06:59] <alex_joni> # Subdirectory: rtapi
[17:07:00] <alex_joni> obj-$(CONFIG_RTAPI) += rtapi/$(RTPREFIX)_rtapi.o
[17:07:02] <SWPadnos> okly like 30 or so
[17:07:34] <jepler> alex_joni: hmm, good question
[17:07:52] <alex_joni> because it seems to me it's rtai_rtapi.ko
[17:07:53] <jepler> alex_joni: Looks like it's going to build rtapi/rtai_rtapi.ko which is probably wrong
[17:07:56] <SWPadnos> rtai_rtapi.ko
[17:07:58] <alex_joni> yes
[17:08:33] <alex_joni> make that obj-$(CONFIG_RTAPI) += rtapi/rtapi.o
[17:08:45] <jepler> yeah, I'm not rebuilding rtapi.ko
[17:09:10] <alex_joni> rtapi-objs := rtapi/$(CONFIG_RTAPI)_rtapi.o
[17:09:59] <jepler> I tested that emc would still start up, but I didn't catch that the old rtapi.ko was still there from the old build
[17:10:13] <alex_joni> ok, just struck me odd by reading the file..
[17:10:25] <jepler> (I mean $(TOP)/rtlib/rtapi.ko)
[17:11:11] <alex_joni> yeah I know
[17:11:36] <SWPadnos> hmmm - the axis binary should be deleted if there are changes to the headers it uses
[17:11:56] <SWPadnos> that would be easier to manage if axis were a part of emc2 ;)
[17:12:11] <alex_joni> jepler: I did some of the 2.6 port in the makefiles, but I'm a bit lost in this Makefile (lots more complicated than what I'm used to ;)
[17:12:20] <CIA-8> 03jepler * 10emc2/src/Makefile: build the rtapi kernel module with the right name
[17:12:21] <alex_joni> so I guess other people will be even more confused..
[17:12:35] <alex_joni> maybe you could hack up a short info
[17:12:46] <SWPadnos> I didn't get copies of the new files in the commit messages, just that they had been created
[17:13:00] <alex_joni> there are no new files
[17:13:06] <SWPadnos> submakefile?
[17:13:08] <jepler> alex_joni: there were a bunch of new files (Submakefiles)
[17:13:13] <alex_joni> ahh.. right
[17:14:53] <fenn> bleh.. joomla is all borked in konqueror
[17:15:08] <alex_joni> fenn: who cares about konqueror?
[17:15:17] <alex_joni> konqueror is borked by itself :D
[17:15:20] <alex_joni> kidding..
[17:15:44] <alex_joni> looks fine in firefox, opera, IE, lynx, w3m, over here
[17:16:32] <fenn> er, lynx doesnt really work right actually
[17:16:42] <alex_joni> almost
[17:16:43] <alex_joni> :D
[17:17:47] <CIA-8> 03jepler * 10emc2/src/Makefile: always use 'gcc' (assuming it's gcc 3 or better) to generate dependency information for "C" source files, because gcc-2.95 doesn't accept the right flags
[17:22:44] <alex_joni> jepler: I don't see the install rule anymore..
[17:22:56] <alex_joni> maybe I'm looking in the wrong location ?
[17:27:18] <jepler> alex_joni: That's one of the notes in the log entry .. there's not an install target yet
[17:28:17] <jepler> but, argh, first I have to fix this compile problem that exists on bdi4 but didn't show up on ubuntu
[17:28:37] <alex_joni> what's that?
[17:28:42] <cradek> welcome to the wonderful world of backward compatibility
[17:28:46] <jepler> /usr/lib/gcc-lib/i486-linux/3.3.5/include/stdarg.h:54: warning: `va_copy' redefined
[17:28:47] <jepler> include/linux/compiler-gcc2.h:29: warning: this is the location of the previous definition
[17:29:01] <alex_joni> that's a warning :)
[17:29:07] <jepler> /usr/lib/gcc-lib/i486-linux/3.3.5/include/stdarg.h:43: parse error before `__gnuc_va_list'
[17:29:10] <alex_joni> since when do programmers care about warnings?
[17:29:14] <alex_joni> argh
[17:29:15] <alex_joni> :D
[17:29:24] <cradek> when they weren't there before you changed something
[17:30:43] <alex_joni> :(
[17:31:17] <jepler> gcc-2.95 -Wp,-MD,/home/emc/emc2/src/libnml/posemath/.sincos.o.d -nostdinc -isystem /usr/lib/gcc-lib/i486-linux/3.3.5/include
[17:31:26] <cradek> oops
[17:31:29] <jepler> this seems a bit suspect, to call gcc-2.95 ... -isystem ...3.3.5...
[17:31:35] <cradek> yeah, that doesn't seem right at all
[17:31:39] <jepler> but the -isystem isn't from my makefile
[17:32:19] <alex_joni> jepler: http://lists.samba.org/archive/distcc/2003q1/000651.html
[17:32:22] <cradek> NOSTDINC_FLAGS += -nostdinc -isystem $(shell $(CC) -print-file-name=include)
[17:32:37] <cradek> you must not have $(CC) set right?
[17:32:55] <cradek> % gcc-2.95 -print-file-name=include
[17:32:55] <cradek> /usr/lib/gcc-lib/i386-linux/2.95.4/include
[17:32:57] <jepler> then were did gcc-2.95 in the commandline come from?
[17:33:18] <alex_joni> jepler: what does Makefile.inc say?
[17:33:30] <jepler> cradek: yeah, maybe that fixes it
[17:33:30] <alex_joni> CC = @CC@
[17:34:33] <jepler> Missing rtapi.conf. Did you run 'make rtapi' ?
[17:34:49] <alex_joni> rtapi.conf was generated by one of the makefiles
[17:34:53] <jepler> oops
[17:34:55] <alex_joni> rtapi/Makefile iirc
[17:35:46] <alex_joni> jepler: http://cvs.sourceforge.net/viewcvs.py/emc/emc2/src/rtapi/Makefile?rev=1.39&view=markup
[17:37:46] <CIA-8> 03jepler * 10emc2/src/Makefile: Pass CC to kernel build
[17:37:57] <CIA-8> 03jepler * 10emc2/src/rtapi/Submakefile: Build rtpai.conf
[17:39:40] <cradek> whee look at it go
[17:41:32] <alex_joni> jepler's on a roll
[17:41:49] <jepler> it still doesn't work on bdi4
[17:43:47] <jepler> [1969250.367776] rtapi: Unknown symbol nano2count
[17:46:32] <SWPadnos> I got no errors on my bdi4 build
[17:46:38] <jepler> SWPadnos: does it run?
[17:46:46] <SWPadnos> well - I didn't try that :)
[17:46:53] <jepler> I get an error inserting rtapi.ko
[17:46:54] <SWPadnos> lemme install axis and see
[17:48:48] <SWPadnos> actually, I get about 30 or so unknown symbols from rtapi
[17:48:56] <jepler> yeah, me too
[17:49:03] <jepler> "scripts/realtime start" doesn't work
[17:49:13] <SWPadnos> right
[17:49:27] <jepler> shouldn't it insmod something before rtapi.ko?
[17:50:25] <alex_joni> jepler: adeos is the first
[17:50:55] <jepler> $ sh -x ./scripts/realtime start 2>&1 | grep -i adeos
[17:50:55] <jepler> $
[17:51:00] <jepler> maybe rtapi.conf is still incorrect
[17:51:27] <alex_joni> check .runinfo
[17:51:31] <alex_joni> rtlib/.runinfo
[17:51:37] <alex_joni> and let me know what's in there
[17:52:20] <SWPadnos> one byte
[17:52:20] <cradek> .runinfo isn't used anymore
[17:53:18] <jepler> # List of realtime kernel modules to be loaded
[17:53:18] <jepler> -MODULES=''
[17:53:18] <jepler> +MODULES='adeos rtai_hal rtai_ksched rtai_fifos rtai_shm rtai_sem rtai_math'
[17:53:19] <alex_joni> .runinfo is looked at by rtai-load I think
[17:53:27] <jepler> OK, I think I know what's wrong ..
[17:53:31] <cradek> rtai-load isn't used anymore
[17:53:37] <cradek> jepler: where is liblxrt?
[17:53:38] <alex_joni> cradek: ok
[17:53:44] <jepler> cradek: what's liblxrt?
[17:53:53] <cradek> I'm not sure
[17:54:08] <jepler> The empty list is from the new rtapi.conf
[17:54:53] <cradek> % nm /usr/lib/liblxrt.a |grep nano2count
[17:54:53] <cradek> 000016ac T nano2count
[17:54:53] <cradek> 000019b0 T nano2count_cpuid
[17:54:58] <alex_joni> yes, because RTAI_MOD doesn't exist anymore
[17:56:54] <alex_joni> that's odd..
[17:57:07] <jepler> yay, now it runs on bdi4
[17:57:13] <alex_joni> jepler: you missed something :)
[17:57:18] <alex_joni> 19:57 < cradek> jepler: where is liblxrt?
[17:57:22] <alex_joni> darn
[17:57:27] <alex_joni> # This is a list of the realtime modules provided by the OS
[17:57:27] <alex_joni> # that are to be loaded - Some will not be present on certain
[17:57:29] <alex_joni> # systems, but they still need to be listed if systems that do have
[17:57:29] <alex_joni> # them are to work.
[17:57:29] <alex_joni> RTAI_MOD="adeos rtai rt_mem_mgr rtai_sched rtai_fifos rtai_shm rtai_libm"
[17:57:32] <alex_joni> # For RTAI-3, we need to condsider using the RTAI scripts to load these modules.
[17:57:34] <alex_joni> RTAI3_MOD="adeos rtai_hal rtai_ksched rtai_fifos rtai_shm rtai_sem rtai_math"
[17:57:37] <alex_joni> RTAI3_RUN="emc2:ksched+fifos+shm+sem+math"
[17:57:39] <alex_joni> RTL_MOD="rtl mbuff rtl_time rtl_sched rtl_posixio rtl_fifo"
[17:57:42] <alex_joni> that was in rtapi/Makefile
[17:59:46] <CIA-8> 03jepler * 10emc2/src/rtapi/Submakefile: additional environment variables for rtapi.conf script
[18:02:06] <SWPadnos> hmm - I wasn't able to get it to work just by cvs up / make
[18:02:14] <cradek> yay, it runs for me
[18:02:16] <jepler> alex_joni: If not earlier, I'll try to be around on sunday to answer questions about the new Makefile+Submakefile system
[18:02:24] <SWPadnos> should I still configure or make clean (or make dep...)?
[18:02:27] <alex_joni> ok, does it still work on 2.4 ?
[18:02:30] <alex_joni> and 2.2 ?
[18:02:38] <cradek> alex_joni: haha, good one
[18:02:45] <alex_joni> cradek: I'm very serious
[18:02:49] <alex_joni> no joke
[18:02:49] <cradek> SWPadnos: rm scripts/rtapi.conf
[18:02:52] <SWPadnos> that's the crux of jmk's email
[18:02:54] <SWPadnos> ok
[18:02:55] <jepler> alex_joni: No, I don't have access to the systems to test that
[18:03:00] <cradek> I know, I know
[18:03:08] <alex_joni> jepler: that's not the best thing :(
[18:03:18] <cradek> alex_joni: that's the point of the compile farm
[18:03:28] <cradek> alex_joni: none of us have all the systems; that doesn't mean we should never change anything
[18:03:30] <alex_joni> ok, but is it supposed to work on that?
[18:03:38] <alex_joni> that's really the question..
[18:03:40] <cradek> yes
[18:03:47] <alex_joni> ok then
[18:03:47] <cradek> sorry, I shouldn't have been joking
[18:03:55] <SWPadnos> much better. now if only emc starup were sped up by the same amount ;)
[18:04:05] <alex_joni> hi steve
[18:10:04] <jepler> alex_joni: see my mailing list post
[18:10:27] <alex_joni> I don't see it..
[18:10:42] <alex_joni> * alex_joni looks if it ain't held back because you're not subscribed
[18:10:44] <SWPadnos> well - once it gets delivered :)
[18:11:18] <SWPadnos> no - he's allowed to send to that list - there was the git announcement
[18:13:54] <alex_joni> it just took a while
[18:14:01] <alex_joni> it eventually arrived
[18:14:10] <SWPadnos> yep
[18:14:21] <alex_joni> SWPadnos: you got it wrong about the IO stuff ;)
[18:14:30] <SWPadnos> which part?
[18:14:36] <alex_joni> check my latest email
[18:14:45] <SWPadnos> as soon as it arrives ;)
[18:14:51] <alex_joni> it did already
[18:14:57] <SWPadnos> not here
[18:15:04] <SWPadnos> if it was in response to my email
[18:15:08] <alex_joni> yes
[18:15:22] <alex_joni> we were talking about the string passed at insmod time:
[18:15:23] <alex_joni> insmod hal_stg dio="IIOO" (or loadrt hal_stg dio="IIOO")
[18:15:23] <alex_joni> that means the first 16 bits are input, the next 16 are output.
[18:15:36] <SWPadnos> yep
[18:15:52] <alex_joni> so basicly standardisation towards the user, not only inside the driver
[18:16:04] <SWPadnos> yes - that's exactly what I was talking about
[18:16:18] <SWPadnos> I think jmk did a quick driver for an I/O board for the Mazak
[18:16:18] <alex_joni> > Is there a reason why you would force a user to have half inputs and
[18:16:20] <alex_joni> > half outputs?
[18:16:39] <alex_joni> that's the functionality of the board
[18:16:50] <alex_joni> to allow 0..3 and 4..7 to be set independently
[18:16:59] <alex_joni> either as inputs or outputs
[18:17:01] <SWPadnos> well - hea said that it can be in+out or out+in
[18:17:10] <alex_joni> so user can do: io, oi, ii, oo
[18:17:34] <alex_joni> he did not
[18:17:36] <SWPadnos> you're thinking that he meant to use lower case for nibble-wide I/O specs?
[18:17:40] <alex_joni> YES
[18:17:47] <SWPadnos> ok - that's not how I read it
[18:17:59] <alex_joni> try to turn it upside down, and read again
[18:18:06] <SWPadnos> "I was thinking of using lower case "i" for bits 0-3 configured and inputs"
[18:18:14] <giacus> he didnt know
[18:18:18] <SWPadnos> and 4-7 as outputs - that weas truncated
[18:18:22] <giacus> who was singing this song ?
[18:18:34] <giacus> :P
[18:18:48] <alex_joni> SWPadnos: you might have a point..
[18:19:00] <SWPadnos> that's why I responded that way :)
[18:19:36] <SWPadnos> he had mentioned that "obviously he wanted half in and half out" - that's not obvious to me
[18:19:50] <SWPadnos> (well, maybe for him, but not for everyone)
[18:21:40] <alex_joni> SWPadnos: guess both emails were too vague for me ;)
[18:22:50] <SWPadnos> heh
[18:23:00] <giacus> it was 'I didnt know' .. if I remember well
[18:23:12] <giacus> who wah the band ?
[18:23:24] <SWPadnos> Benny Hill
[18:23:27] <giacus> * giacus hitting the head on the wall
[18:23:30] <giacus> nahhh
[18:23:38] <giacus> 80's
[18:23:41] <alex_joni> giacus: why don't you ask google?
[18:23:50] <giacus> tryng .. :(
[18:24:14] <giacus> too much titles result
[18:24:17] <SWPadnos> I still don't know what song you're talking about :)
[18:24:27] <SWPadnos> but Benny Hill seemed like a good guess anyway
[18:24:33] <giacus> I will find it !
[18:24:36] <giacus> :D
[18:24:55] <giacus> it was he didnt know
[18:25:02] <giacus> or she didnt know
[18:25:06] <giacus> or something..
[18:25:24] <giacus> was a very nice song :)
[18:25:31] <alex_joni> I don't know
[18:25:35] <alex_joni> ROFL
[18:25:45] <giacus> the leader was a male
[18:25:50] <SWPadnos> "nobody knew"
[18:25:52] <giacus> a band
[18:25:56] <giacus> nope..
[18:26:06] <SWPadnos> do you have some of the actual words?
[18:26:11] <SWPadnos> from the song
[18:26:16] <giacus> thinking ..
[18:26:33] <alex_joni> humm us the melody like this: mmm mmhmm hmmmhmmhm mmhmhmhm
[18:26:38] <alex_joni> lol
[18:26:40] <giacus> :)
[18:26:57] <giacus> was a band , im sure
[18:27:09] <giacus> was a sweet song
[18:27:11] <SWPadnos> fast or slow? "beat-box" or synth (or country, house, rock, rockabilly ...)
[18:27:17] <giacus> soft
[18:27:30] <SWPadnos> Kenny Rogers
[18:27:33] <giacus> nope
[18:27:39] <giacus> was a band
[18:27:40] <SWPadnos> no - wait, Lionel Ritchie
[18:27:45] <SWPadnos> oh - right.
[18:27:47] <SWPadnos> Air Supply
[18:28:08] <les_w> ach
[18:28:27] <giacus> les !
[18:28:32] <giacus> hahaha, ciao :)
[18:28:40] <les_w> hi giacus
[18:28:57] <giacus> do you remember that band ?
[18:29:11] <giacus> was the time of Frankie goes to Hollywood
[18:29:17] <giacus> maybe before
[18:29:27] <les_w> air supply?
[18:29:31] <les_w> bleh
[18:29:42] <les_w> haha
[18:29:48] <giacus> PHD !
[18:29:54] <giacus> maybe ..
[18:30:11] <giacus> no :(
[18:30:26] <les_w> I need to buy some cds
[18:30:33] <les_w> must drive a long way
[18:30:37] <giacus> was something like PHD
[18:30:43] <les_w> ?
[18:30:43] <giacus> the name of the band
[18:30:44] <les_w> hmm
[18:31:08] <giacus> p8d ..
[18:32:01] <les_w> I usually play either old 60's rock or jazz
[18:32:33] <les_w> did you like the songs I picked?
[18:32:42] <giacus> nope
[18:32:45] <giacus> ops
[18:32:48] <giacus> sorry :)
[18:32:51] <giacus> sure
[18:32:52] <les_w> haha
[18:32:55] <giacus> haha
[18:32:57] <giacus> :)
[18:33:00] <SWPadnos> so "The Who" and U2 aren't in the mix, huh
[18:33:13] <giacus> sure they was phd
[18:33:21] <giacus> now looking for the title
[18:33:27] <les_w> the who perhaps
[18:33:50] <SWPadnos> "Behind Blue Eyes" is a good song for good audio systems
[18:33:52] <giacus> I won't let you down, won't let you down again x3
[18:33:59] <giacus> do you remember that :P
[18:34:08] <giacus> http://www.risa.co.uk/sla/song.php?songid=15046
[18:34:19] <giacus> but wasnt what im searching
[18:34:20] <les_w> it was neat when I played a diana Krall song then k4ts sang the same song
[18:34:23] <SWPadnos> nope - just "Don't Bring Me Down" by ELO
[18:34:45] <giacus> uhmm im confusing
[18:34:54] <giacus> they was not the phd band then ..
[18:34:59] <SWPadnos> so, do Steely Dan or Donald Fagen fit in there?
[18:35:00] <giacus> mmm
[18:36:18] <SWPadnos> oh - of course there's "The Moody Blues", for good '60s rock (and leter)
[18:36:21] <SWPadnos> later
[18:36:23] <SWPadnos> that is
[18:37:53] <giacus> darn.. i must go for the dinner
[18:37:56] <giacus> later
[18:38:12] <giacus> giacus is now known as giacus_dinner
[18:38:50] <les_w> oh fyi swp....carmax does not haggle
[18:39:51] <les_w> ach forgot to order a fuseblock. All those boxes and still missing a part
[18:57:31] <SWPadnos> ok - I figured as much :)
[18:57:44] <SWPadnos> what size/type fuses?
[18:59:01] <alex_joni> * alex_joni has some 150Amps fuses lying around
[18:59:08] <alex_joni> the big ceramic type :D
[18:59:31] <alex_joni> they are meant 1 / phase
[19:00:20] <SWPadnos> air conditioner fuses
[19:00:50] <alex_joni> depends on the AC
[19:02:08] <SWPadnos> indeed
[19:02:54] <alex_joni> how do you rate your AC' over there? kW?
[19:03:06] <alex_joni> over here it's some silly BTU standard :D
[19:03:17] <SWPadnos> same here
[19:03:21] <SWPadnos> or tons
[19:03:29] <alex_joni> ROFL
[19:03:31] <SWPadnos> (whatever those are)
[19:03:41] <alex_joni> I wouldn't ever have thought US people use BTU
[19:03:44] <alex_joni> ROFLMAO
[19:03:55] <SWPadnos> yeah - it's so British ;)
[19:04:05] <alex_joni> BTU=British Thermal Units
[19:04:15] <SWPadnos> yep
[19:04:57] <SWPadnos> aha - one ton is 12000 BTU/hour
[19:05:21] <alex_joni> lol.. then I guess our office AC is 3 tons
[19:05:26] <SWPadnos> = 3.516 kW
[19:05:51] <giacus_dinner> giacus_dinner is now known as giacus
[19:05:53] <giacus> yeahh
[19:06:09] <giacus> damn it ! I found it in my old disc 45
[19:06:11] <giacus> :D
[19:06:24] <alex_joni> tell us
[19:06:28] <giacus> Ph.D - I didnt Know
[19:06:33] <giacus> :)
[19:06:38] <alex_joni> never heard
[19:06:49] <SWPadnos> never heard of either :)
[19:07:06] <giacus> * giacus scanning the cover.. :P
[19:07:39] <alex_joni> oh, it's italian
[19:07:46] <giacus> nope ..
[19:07:48] <alex_joni> http://www.netsoundsmusic.com/nsudsii/2/6006412/311/2.html
[19:08:01] <giacus> no no ..
[19:08:12] <alex_joni> yes yes
[19:08:17] <SWPadnos> why why
[19:08:23] <giacus> :)
[19:08:26] <alex_joni> google doesn't lie
[19:08:57] <giacus> hey guys are you watching The tutin olimpiadi
[19:09:02] <giacus> turin
[19:09:08] <SWPadnos> nope
[19:09:20] <SWPadnos> if I were there, I would be :)
[19:09:22] <giacus> look, its nice :)
[19:09:25] <giacus> right now
[19:11:04] <giacus> its starting ..
[19:11:52] <alex_joni> SWPadnos: care for a nice radio?
[19:11:59] <alex_joni> webstream I mean
[19:12:10] <SWPadnos> mot really, thanks :)
[19:12:13] <SWPadnos> not
[19:12:26] <alex_joni> found one that's really great
[19:12:30] <alex_joni> np
[19:13:00] <SWPadnos> any idea where to find the hole pattern for the X and Y axis bearing brackets on a Bridgeport series 1?
[19:13:59] <fenn> hmmm make tgz is no longer there..
[19:14:21] <giacus> http://digilander.libero.it/jackydgl0/PHD.jpg
[19:14:23] <giacus> wow
[19:14:24] <alex_joni> fenn: a few things aren't there yet
[19:14:29] <giacus> was that !
[19:15:07] <giacus> hard to find on the net
[19:15:10] <fenn> right-o
[19:15:41] <giacus> wonderful song :P
[19:16:22] <alex_joni> giacus: celentano now on the radio ;)
[19:16:28] <giacus> hehe
[19:16:35] <giacus> let me look at turin ;P
[19:16:39] <alex_joni> I think..
[19:16:47] <alex_joni> http://encoder2.rdsnet.ro:8000/live
[19:17:15] <giacus> :-)
[19:18:07] <giacus> http://www.torino2006.org/ENG/OlympicGames/home/index.html
[19:18:08] <giacus> :P
[19:20:49] <alex_joni> looks like joomla
[19:21:08] <giacus> dunno
[19:21:58] <giacus> Opening Ceremony with Sofia Loren, Yoko Ono and Susan Sarandon
[19:22:41] <alex_joni> hmm.. seems RTAI-3.3 got released last week
[19:22:44] <alex_joni> I missed that
[19:26:30] <giacus> wow
[19:34:56] <giacus> giacus is now known as giacus_turin_200
[19:35:01] <giacus_turin_200> giacus_turin_200 is now known as giacus_turin_06
[19:35:03] <giacus_turin_06> :P
[19:35:36] <alex_joni> http://media3.davesdaily.com/pictures7/219-hugedog.jpg
[19:35:56] <giacus_turin_06> haha
[19:36:04] <giacus_turin_06> an horse
[19:37:22] <alex_joni> then what's this: http://media3.davesdaily.com/pictures7/205-hugemuscles.jpg ?
[19:37:50] <cradek> ewwwww
[19:37:58] <giacus_turin_06> eeek
[19:38:02] <giacus_turin_06> :)
[19:38:46] <alex_joni> heh, chris do you have a lot of drive-in's in the US?
[19:38:54] <cradek> do you mean food or movie?
[19:38:59] <alex_joni> I only heard of drive-through till now
[19:39:00] <alex_joni> http://media3.davesdaily.com/pictures7/197-drivein.jpg
[19:39:11] <fenn> that is totally a balloon suit
[19:39:32] <cradek> alex_joni: ok I get the joke
[19:40:41] <alex_joni> some nice pics at davesdaily ;)
[19:41:32] <fenn> http://outhouserag.typepad.com/photos/uncategorized/huge_muscles.jpg
[19:42:13] <giacus_turin_06> hahaha maded with gimp ?
[19:42:48] <alex_joni> hahaha http://media3.davesdaily.com/pictures7/172-canyoufindthekid.jpg
[19:43:32] <giacus_turin_06> he's a good fork :)
[19:44:04] <giacus_turin_06> denmark ..
[19:44:17] <giacus_turin_06> alex_joni: are you in the olimpic games ?
[19:45:04] <alex_joni> no, I'm in bed
[19:45:08] <alex_joni> I'm tired today
[19:46:52] <giacus_turin_06> your country is strong :D
[19:47:13] <alex_joni> seems there are quite a few drive-in's in the US (http://media3.davesdaily.com/pictures7/147-drivein.jpg)
[19:47:22] <alex_joni> giacus_turin_06: oh, thought you were asking me :D
[19:47:33] <giacus_turin_06> :-))
[19:52:28] <alex_joni> lol, now that's a workstation http://media3.davesdaily.com/pictures7/1130-workstation.jpg
[19:53:01] <giacus_turin_06> nice laptop
[19:54:05] <SWPadnos> for some reason, I can't view the images you link
[19:54:10] <SWPadnos> I get "hotlink.gif" instead
[19:54:50] <alex_joni> SWPadnos: huh?
[19:54:53] <alex_joni> http://media3.davesdaily.com/pictures7/90-pissedofftree.jpg
[19:55:12] <SWPadnos> the davesdaily.com images - I get a blank image called hotlink.gif
[19:55:25] <SWPadnos> http://media3.davesdaily.com/hotlink.gif is what I get
[19:55:48] <alex_joni> try www instead of media3, remove the 7 from pictures7, and make the ending htm
[19:55:59] <alex_joni> http://www.davesdaily.com/pictures/90-pissedofftree.htm
[19:56:17] <SWPadnos> yep - that works
[19:58:13] <alex_joni> SWPadnos: any woods / bears around where you are?
[19:58:42] <SWPadnos> yes, both
[19:59:07] <alex_joni> http://www.davesdaily.com/pictures/68-bear.htm <- might be a bit much for weak hearted ;)
[20:00:25] <SWPadnos> indeed
[20:03:42] <fenn> people who keep bears for pets should be sent to jail!!!
[20:04:29] <alex_joni> http://www.davesdaily.com/pictures/22-bush.htm
[20:07:12] <SWPadnos> that one, I like ;)
[20:08:01] <fenn> http://outhouserag.typepad.com/photos/uncategorized/love_your_job.gif
[20:24:14] <ValarQ> fenn: they are (at least here)
[20:24:33] <alex_joni> hi ValarQ
[20:24:36] <fenn> i was joking.. obviously the bear wasnt the guy's pet..
[20:24:47] <ValarQ> alex_joni: hiya
[20:24:55] <alex_joni> what's up captain?
[20:25:18] <ValarQ> heh, i was just about to comment that you missed the false title :)
[20:25:25] <alex_joni> no way
[20:25:26] <alex_joni> :D
[20:26:08] <ValarQ> and i even got a debianversion named after my title... :P
[20:26:18] <alex_joni> ?
[20:27:23] <ValarQ> have you replaced the crap in the cvs tree yet?
[20:27:38] <ValarQ> i saw some discution about it on the mailinglist
[20:28:24] <SWPadnos> hey - forgot it was there ;)
[20:28:35] <SWPadnos> I thought it had been added, but didn't remember where
[20:28:52] <alex_joni> hal/utils/halgui?
[20:28:59] <ValarQ> yeah
[20:29:12] <alex_joni> my memory seems to serve me right.. still
[20:29:37] <SWPadnos> hmmm - it needs gnome
[20:29:43] <SWPadnos> or at least gnomecanvas
[20:29:48] <ValarQ> yeah, should be tk...
[20:29:50] <alex_joni> SWPadnos: add it ;)
[20:29:57] <fenn> i played around with it alittle but never did anything with it..
[20:30:01] <fenn> swp check the wiki page..
[20:30:12] <SWPadnos> could do, and may ;)
[20:30:20] <ValarQ> requires pretty late versions of the pygtk libraries as well
[20:30:33] <ValarQ> a tk interface would be much better
[20:30:42] <fenn> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?Crapahalic
[20:30:45] <ValarQ> especially since emc uses it
[20:31:52] <alex_joni> ok, I decided I need one of these: http://www.hp.com/products1/servers/scalableservers/superdome/index.html
[20:32:00] <SWPadnos> funny - I went to write a C Din ubuntu, and I get an estimate of 1195432 hours, 45 minutes, 15 seconds ;)
[20:32:03] <SWPadnos> CD in
[20:32:19] <SWPadnos> like the extra 15 seconds is gonna matter ;)
[20:32:51] <ValarQ> :)
[20:33:22] <alex_joni> lol, Operating temperature: 20�C to 30�C
[20:33:33] <ValarQ> alex_joni: runs Hockey-Pux :P
[20:33:43] <SWPadnos> only 1 TB RAM, come on
[20:33:54] <alex_joni> SWPadnos: it's for me, homebox
[20:34:00] <SWPadnos> wimpy
[20:34:01] <alex_joni> I figure it's enough for starters
[20:34:07] <alex_joni> upgrade later
[20:34:24] <SWPadnos> of course, you could control a lot of robots with 192 M5I20 cards ;)
[20:34:55] <alex_joni> hotswappable robots :D
[20:35:07] <SWPadnos> hotswappable factories
[20:35:39] <ValarQ> i would settle with one of these: http://www.sgi.com/products/servers/altix/4000/configs.html
[20:36:05] <SWPadnos> this: http://www.tyan.com/products/html/thunderk8qsdpro.html
[20:36:20] <ValarQ> * ValarQ is looking at the 4700 model
[20:36:25] <SWPadnos> with the daughterboard, and fully loaded with RAM and Opteron 880 (or better) would be just fine with me
[20:36:57] <ValarQ> up to 128TB RAM per system :D
[20:37:06] <SWPadnos> Meg, I think ;)
[20:37:13] <SWPadnos> I mena Gig :)
[20:37:15] <SWPadnos> mean
[20:37:33] <ValarQ> nope :)
[20:37:40] <SWPadnos> oh cool - you can still spend $2500+ on an Opteron chip
[20:37:47] <SWPadnos> oh - the SGI
[20:38:20] <ValarQ> yep, their latest monster
[20:38:53] <SWPadnos> only 48G per module though
[20:39:26] <SWPadnos> that Tyna board can actually take 64G on the motherboard, and another 4 on the daughtercard (with an additional 4 dual-core CPUs)
[20:39:29] <SWPadnos> Tyan - damn
[20:39:44] <alex_joni> damn Tyna
[20:39:46] <alex_joni> :)
[20:39:58] <SWPadnos> friggin tynas - a bear should eat them all
[20:40:13] <ValarQ> :)
[20:40:39] <ValarQ> * ValarQ just got his new x64 crapmachine
[20:40:53] <SWPadnos> what did you get?
[20:41:09] <ValarQ> some new pc crap
[20:41:29] <ValarQ> an Athlon64 X2
[20:41:38] <jepler> why are you calling it crap?
[20:41:51] <ValarQ> jepler: it's build upon intels architecture
[20:42:17] <jepler> with enough transistors, everything is fast///
[20:42:29] <ValarQ> yeah
[20:42:42] <SWPadnos> for ValarQ, everything is crap ;)
[20:42:51] <ValarQ> SWPadnos: almost :)
[20:42:57] <alex_joni> skit
[20:43:03] <ValarQ> SWPadnos: that 4700 isn't ;)
[20:43:23] <SWPadnos> well - it's crap that you (probably) can't afford it then ;)
[20:43:31] <alex_joni> jepler: I think modern processors are a mockery
[20:43:40] <ValarQ> alex_joni: i can hear you have taken a more "mastering swedish" lessons :)
[20:43:50] <ValarQ> SWPadnos: thats true
[20:43:51] <alex_joni> ValarQ: no, that was memory actually
[20:44:06] <ValarQ> alex_joni: i don't belive you
[20:44:20] <ValarQ> alex_joni: i guess you're hooked on that stuff by now :P
[20:44:40] <alex_joni> jepler: having about 15-20% leaking currents from the power consumption..
[20:44:50] <alex_joni> ValarQ: not really
[20:45:05] <ValarQ> alex_joni: ok, had enough of it in lund?
[20:45:27] <ValarQ> alex_joni: remember that not all of us speak as weird as those Skåningar
[20:45:48] <jepler> A few months I chose not to get an X2 for my new machine .. I wonder if it was the right decision.
[20:45:53] <jepler> ValarQ: what motherboard chipset in your machine?
[20:46:04] <ValarQ> jepler: nforce
[20:46:06] <SWPadnos> it was if you didn't spend more than $150 or so on the CPU you chose
[20:46:22] <ValarQ> i hope it works with my adaptec raidcard :/
[20:46:33] <SWPadnos> though the shift to DDR2 late this or early next year may throw a wrench in the works
[20:47:15] <jepler> SWPadnos: for the CPU itself? It was Sempron 64 300+, socket 754. $89 at the time.
[20:47:35] <SWPadnos> well, then you probably saved enough to be able to buy an A64 later
[20:47:56] <jepler> No X2s for this socket, though, so I get stuck buying a new motherboard too
[20:48:02] <SWPadnos> yep
[20:48:18] <alex_joni> * alex_joni still runs an 1600 AthlonXP
[20:48:22] <ValarQ> can't wait to test my machine...
[20:48:22] <jepler> ValarQ: Is nforce an okay chipset with linux these days?
[20:48:28] <ValarQ> jepler: dunno
[20:48:29] <alex_joni> and I really see no need for faster ;)
[20:48:33] <SWPadnos> me, too - I'm on an Athlon XP 1800 right now
[20:48:37] <ValarQ> jepler: my current works pretty ok
[20:48:39] <jepler> alex_joni: I replaced a 350MHz machine
[20:48:46] <SWPadnos> though I hope the new software works well on the Opteron as well
[20:48:48] <alex_joni> that's 1600MHz, right SWP?
[20:48:55] <SWPadnos> 1533, I think
[20:49:02] <alex_joni> ok
[20:49:06] <ValarQ> i have some assemblyjob to do on my new machine, but i must get sober first :(
[20:49:17] <jepler> er, the Sempron I bought was "3000+", 1.8GHz core speed.
[20:49:21] <SWPadnos> it may not be an XP - it's a PC133 / KT133 motherboard
[20:49:23] <alex_joni> try to stick the CPU directly into the PCI
[20:49:41] <ValarQ> alex_joni: ah, thats a well known trick :)
[20:49:52] <alex_joni> btw, did you read about watercooling?
[20:50:30] <ValarQ> nope, where?
[20:50:34] <SWPadnos> heh - watercooling is why I originally started the CNC thing ;)
[20:50:53] <alex_joni> * alex_joni looks for an URI
[20:51:10] <SWPadnos> silentpcreview.com or somewhere like that
[20:52:29] <alex_joni> no, got a better link.. only if I would find it :(
[20:53:37] <SWPadnos> www.wawter-cooling.com?
[20:53:41] <SWPadnos> water, that is
[20:53:50] <les_w> what does a p4 dissipate now? couple hundred watts?
[20:54:01] <SWPadnos> 140 to 220 or so
[20:54:03] <SWPadnos> I think
[20:54:18] <les_w> sounds about right
[20:54:24] <SWPadnos> the mobile ones are much better
[20:54:36] <les_w> I would imagine
[20:54:51] <SWPadnos> the Opterons are 55, 95, or 130W
[20:55:17] <SWPadnos> per chip, so the dual-core chips are the same as single-core for the same power rating
[20:55:21] <les_w> makes me think of the amps in the music room
[20:55:33] <SWPadnos> heh - 7.1 CPU channels ;)
[20:55:34] <les_w> hmm 3500w @65% efficiency?
[20:55:41] <SWPadnos> 5kw
[20:56:38] <les_w> it's a good heater. Course I don't run it at 3500 (much)
[20:56:50] <SWPadnos> have to be next door to be able to hear
[20:57:26] <les_w> I am still not quite ready for PWM audio....
[20:57:35] <les_w> it's getting better though
[20:57:52] <SWPadnos> the hybrid amps are pretty good
[20:57:58] <les_w> yeah
[20:58:02] <SWPadnos> class A+D, I think
[20:58:03] <les_w> good specs anyway
[20:58:28] <SWPadnos> well - the "sound" won't be the same as tubes for a while ;)
[20:58:46] <alex_joni> ValarQ: FOUND IT
[20:58:48] <alex_joni> http://www.avforums.com/forums/showthread.php?t=56924
[20:58:56] <les_w> I had the fender crankin in the music room last night
[20:59:03] <les_w> 6l6 williamson
[20:59:11] <SWPadnos> heh - that one
[20:59:21] <jepler> somebody posted that thread awhile back. pretty funny.
[21:00:38] <alex_joni> yeah I did..
[21:02:40] <les_w> I had an optically pumped Yag laser that ran submersed in DI water
[21:04:11] <alex_joni> without kidding, I've seen movies of PC's running submerged in oil
[21:04:20] <les_w> Hey I ought to try a tube of that "silicon"
[21:05:31] <les_w> I just ordered some high Tg conductive epoxy...60 bucks an ounce.
[21:06:39] <les_w> even though materials cost under $5
[21:06:53] <les_w> Can make it later in house
[21:06:56] <SWPadnos> not unlike diamonds
[21:07:13] <les_w> heh yeah
[21:07:46] <les_w> Well we have formulated and used thousands of kilos of my silver screen printing ink
[21:08:26] <les_w> It's good stuff...well under 20 milliohms/square/mil
[21:09:09] <SWPadnos> in any direction?
[21:09:28] <les_w> I still have the two german astra 3 roll mills I developed the stuff with
[21:09:36] <les_w> bought em from moto
[21:09:42] <les_w> yes any direction
[21:10:13] <SWPadnos> do you get an oxidation layer, or will it conduct if you lay two sheets on top of each other?
[21:10:15] <les_w> it's quite anisotropic
[21:10:48] <les_w> silver oxide is conductive....
[21:11:07] <SWPadnos> as conductive as silver?
[21:11:10] <les_w> but in the heater product it is sealed between two layers of PET
[21:11:41] <alex_joni> " Look at it this way, you're famous...You may not realize it, but half of the computer enthuasist and overclocking community is linking to this thread right now...=)"
[21:11:45] <alex_joni> lol
[21:11:45] <les_w> no not as conductive....butas a membrane switch is can function well for years
[21:12:31] <SWPadnos> I'm thinking board interconnect (not a specific application though)
[21:12:58] <les_w> I have some twenty year old parts that still work fine
[21:13:03] <SWPadnos> or a flex cable connector - clamp it down, and you're done
[21:13:16] <jacky^> jacky^ is now known as giacus_paltk
[21:13:24] <les_w> we have to watch silver migration issues closely on boards
[21:14:33] <les_w> in some extreme cases like the minn kota trolling motor control I actually used a slot in the board between elements
[21:14:47] <les_w> stops migration issues for sure!
[21:17:59] <les_w> also use the material in the GM int wiper control, ge flat pack light dimmer, and rockwell water meter
[21:18:12] <les_w> so have some pretty good history on it
[21:21:16] <SWPadnos> cool
[21:21:24] <SWPadnos> (worry - wife arrived home from work)
[21:21:29] <SWPadnos> sorry, not worry ;)
[21:21:54] <les_w> well....4:30 friday here...I say let the weekend begin
[21:22:23] <SWPadnos> I agree, even though it's only 4:22 here ;)
[21:22:32] <les_w> I'm gonna go turn off the phase converter
[21:22:46] <SWPadnos> that sounds pretty final
[21:23:16] <les_w> ok...keep my watch to 1 second...let's see...4:23:!%MARK
[21:23:22] <les_w> OOPS
[21:23:24] <les_w> HAHA
[21:23:31] <SWPadnos> 4:23:30
[21:23:46] <les_w> THAT WAS 4:22:15
[21:23:47] <SWPadnos> we have atomic clocks in basically every room ;)
[21:24:14] <les_w> I just set my watch to nist every few days
[21:24:18] <SWPadnos> REALLY - IT DIDN'T LOOK IT TO ME!
[21:24:22] <SWPadnos> heh
[21:25:35] <les_w> how about smoke detectors
[21:25:54] <SWPadnos> good idea
[21:25:54] <les_w> what's in those things...americium or polonium?
[21:26:05] <SWPadnos> dunno
[21:26:56] <les_w> I think americium....alpha emitter. polonium is beta IIRC
[21:30:05] <SWPadnos> heh - I never really studdied nukular chemistree
[21:30:27] <SWPadnos> I could ask my mother though
[21:30:30] <les_w> haha
[21:30:56] <SWPadnos> it's nice having P-chem PHd s for parents
[21:31:36] <fenn> hmmm..
[21:31:38] <les_w> just looking at that new rooster....what colors...brilliant reds, snow white ear tufts, huge mettalic green tail
[21:31:45] <les_w> show bird.
[21:31:55] <SWPadnos> dammit - he'sa peacock ;)
[21:32:01] <les_w> just about
[21:32:33] <fenn> * fenn is getting aruacanas in a couple days
[21:32:36] <les_w> I understand he is a rescued purebreed fighting cock
[21:33:08] <les_w> what's an aruacana?
[21:33:36] <les_w> I gotta take some pictures of this bird
[21:33:53] <les_w> they are free-range...just out in the yard
[21:33:54] <fenn> an easter egg chicken.. all the eggs are different colors, green, white, yellow, brown, gray
[21:34:01] <les_w> oh neat
[21:34:04] <fenn> and the chickens are all different colors too
[21:34:32] <les_w> I should get some very nice chiks in a while
[21:34:39] <fenn> sorry araucanas
[21:34:43] <les_w> if a dog doesn't eat the rooster
[21:35:35] <les_w> Usually those fighting cocks can take care of themselves
[21:37:25] <les_w> But i know little about chickens...except they like to peck paul corner.
[21:37:43] <dmessier> hello all
[21:38:22] <alex_joni> bonne soir
[21:38:23] <dmessier> landing gear aint all its cracked up to be.... ITS MORE
[21:38:31] <dmessier> soir??
[21:38:53] <alex_joni> fenn: thought those are araucana not aruacana ?
[21:38:58] <alex_joni> 23:42 < dmessier> soir??
[21:39:01] <fenn> alex_joni: right
[21:39:56] <alex_joni> and they're not really laying lots of colours
[21:40:04] <alex_joni> but some blue-ish kind of eggs
[21:40:32] <dmessier> chickens??
[21:40:32] <alex_joni> otoh, they should be very yummy I heard
[21:40:40] <alex_joni> easter chickens
[21:41:06] <dmessier> name of the breed of bird pls??
[21:41:23] <les_w> I may have new barnyard politics. Hen just followed the cat around....but now a rooster.
[21:41:33] <les_w> I see rooster/cat issues.
[21:41:35] <les_w> haha
[21:41:53] <dmessier> me too.. some one's gonna get hurt
[21:42:12] <dmessier> get it on video though
[21:42:17] <les_w> well the spurs on the rooster are small
[21:42:20] <les_w> right now
[21:42:37] <dmessier> operant words .. "right now"
[21:42:43] <les_w> haha
[21:43:10] <les_w> they go for the eyes too
[21:43:30] <dmessier> i had a ferret and a siemese male that were buddies
[21:43:47] <les_w> ah...possible 6 inches of snow tonight here....fairly rare
[21:44:22] <dmessier> with 2 parrokette's .... really...
[21:44:36] <les_w> I believe it
[21:44:47] <les_w> the cat and chicken slept together
[21:45:10] <alex_joni> this is another classic: http://www.datadocktorn.nu/us_frag1.php
[21:45:26] <dmessier> i have pictures of the cat and ferret where you cant tell where one ends and the other begins
[21:45:51] <dmessier> there are just too many feet
[21:46:06] <les_w> heh
[21:46:22] <les_w> funny defrg alex
[21:47:16] <les_w> ah...possible ice here too. No salt, no plows. Steep grades. Better check gas for the generator
[21:49:36] <giacus_paltk> * giacus_paltk plaing a great guitar solo :P
[21:50:30] <les_w> I restrung some of the guitars last night
[21:50:49] <les_w> also need strings for the sitar...have to order those though
[21:51:38] <les_w> my playing is very rusty.
[21:51:51] <giacus_paltk> :))
[21:52:18] <ValarQ> alex_joni: nice url
[21:55:02] <dmessier> kudos to DR. narton
[21:56:28] <ValarQ> alex_joni: i have to counter with crunchly ;)
[21:56:49] <alex_joni> crunchly?
[21:56:53] <ValarQ> alex_joni: http://www.catb.org/~esr/jargon/html/graphics/74-12-25.png
[21:56:57] <dmessier> i picked up my sons guitar the other night and picke some wicked open chord stuff... we were BOTH surprised...
[21:57:17] <les_w> good?
[21:57:43] <les_w> the more beers I have the better I get of course
[21:58:19] <ValarQ> alex_joni: here is the first if you want to read them all: http://www.catb.org/~esr/jargon/html/B/bells-and-whistles.html#crunchly73-05-28
[21:58:20] <SWPadnos> beer goggles, meet beer headphones
[21:58:54] <CIA-8> 03jepler * 10emc2/src/ (Makefile Submakefile.skel): explain things a little bit
[21:59:05] <dmessier> beer.... lower my expectations...
[21:59:08] <Jymmm> les_w Forget to mention... Freud bits SUCK!
[21:59:16] <les_w> I kind of did a major cleanup of some pretty old guitars....steelwooled and oiled the fingerboards....buffed up the body....etc
[21:59:27] <SWPadnos> Jymmm, how about their saw blades?
[21:59:32] <les_w> And I built these instruments
[21:59:39] <Jymmm> SWPadnos no clue.
[21:59:44] <dmessier> cool
[22:00:06] <les_w> what happened? Freud is generally middle of the road...usually not the best but far from bad
[22:00:15] <dmessier> hats off to the instument maker
[22:00:36] <Jymmm> les_w chattered like crazy and the twist was very loose - compared to CMT
[22:00:52] <les_w> I used to do that dmess. Guitar makers still come and use my cnc for pearl inlay though.
[22:01:13] <dmessier> be the mntor you can be
[22:01:18] <les_w> chatter?
[22:01:20] <dmessier> mentor
[22:01:35] <les_w> ah
[22:01:42] <les_w> slipping in the collet?
[22:01:44] <SWPadnos> wrong feed rate / SFM
[22:02:02] <dmessier> smaller tool nose rad
[22:02:14] <Jymmm> les_w Nope. but toss in the CMT bit.... nice and quiet.
[22:02:19] <dmessier> lower doc
[22:02:27] <les_w> My CMT bits perform well
[22:02:47] <les_w> I have never used freud in the cnc
[22:02:51] <Jymmm> les_w If I was to guess, I'd say not balanced.
[22:02:54] <les_w> but have in manual routers
[22:03:18] <les_w> hmm
[22:03:27] <dmessier> i do damage control analysis
[22:03:34] <les_w> balance would not screech
[22:03:51] <les_w> not enough clearance angle would
[22:04:18] <Jymmm> les_w Well, it didn't have much of a twist to it that's for sure.
[22:04:18] <les_w> what kind of bit?
[22:04:22] <dmessier> drop the rpm... leave the feedrates the same.. 10 % steps.... find where it works
[22:04:26] <Jymmm> les_w 1/4" downcut
[22:04:35] <les_w> cutting what?
[22:04:44] <Jymmm> les_w ipe
[22:04:49] <les_w> ok
[22:05:04] <Jymmm> hell, even douglas fir
[22:05:17] <Jymmm> but swap out for the CMT bit. nice.
[22:05:51] <les_w> the Freud must be misground
[22:05:57] <les_w> or dull
[22:06:02] <les_w> or damaged
[22:06:06] <dmessier> torque seems high enuf rigidity ssue... how long a protrusion from the spindle face
[22:06:43] <les_w> dmess spiral downcuts have negative rake on the bottom
[22:07:30] <dmessier> not so good for low rigidity machines
[22:07:42] <dmessier> or setups...
[22:07:44] <les_w> perhaps not
[22:08:08] <les_w> I don't use them in hand routers
[22:08:20] <les_w> and rigidity is not an issue in the cnc
[22:09:06] <dmessier> in fin ops i like to keep tools short .. hold the part as best i can... and use high sheer & rake cutters.. screw tool life.. make god parts
[22:10:29] <Jymmm> god parts?
[22:10:40] <Jymmm> Palying Diety are we?
[22:10:45] <Jymmm> Playing Diety are we?
[22:10:47] <les_w> I use both up and downcut...but usually downcut for heavy work. It's stable. Won't suck the work up or the tool down.
[22:11:34] <les_w> god likes high shear angles and rake, of course.
[22:11:44] <Jymmm> lol
[22:12:22] <anonimasu> haha
[22:12:53] <dmessier> sorry.. its a good complex
[22:13:00] <les_w> haha
[22:13:23] <dmessier> i thought i was wrong once but i was mistaken..
[22:14:21] <dmessier> cuttin IS a specialty of mine... i like chips
[22:14:53] <dmessier> its the sound of money in the morning
[22:16:26] <fenn> if you do climb milling, do you ever have to worry about wood splitting?
[22:16:46] <dmessier> weve recently gotten involve in a research projec with McMaster University to investigate material abuse and compressive layer induced by machining
[22:16:53] <les_w> with climb milling splitting is reduced dramatically
[22:17:12] <les_w> but can't do it on a hand router safely
[22:17:30] <dmessier> operant word.... safely
[22:17:35] <alex_joni> * alex_joni goes to bed
[22:17:38] <alex_joni> night all
[22:17:53] <dmessier> bonne nuit
[22:17:59] <dmessier> beau reve
[22:18:06] <les_w> yeah climb stuf is for cnc routers only
[22:18:27] <dmessier> this CNC stuff RULES
[22:18:36] <fenn> what if you have a router table with a fence?
[22:18:48] <fenn> and only do climb cutting on finish passes
[22:18:50] <les_w> I wouldn't
[22:18:55] <dmessier> climb on bord
[22:19:06] <les_w> I climb mill on the shaper
[22:19:20] <les_w> but it is an industrial unit with power feed
[22:19:45] <dmessier> locking gibs??
[22:20:14] <les_w> without power feed sooner or later it will grab something, take off, and suck you in
[22:20:42] <dmessier> there is a climb mill and down mill lever??
[22:20:52] <les_w> well it kicks forward rather than back
[22:21:24] <les_w> fun one is slots...one side climb...the other not
[22:21:45] <dmessier> i agree... but some double nut machines... lock against each other to climb mill
[22:22:20] <les_w> well with any king of positive power feed it's perfectly safe
[22:22:24] <les_w> phone
[22:23:20] <dmessier> i like plunge roughing.... climb the last .025" efficiently... and rock &* roll
[22:25:48] <dmessier> or ultra high efficiency feed mills... 0.12" feed per TOOTH
[22:26:32] <fenn> in what, jello?
[22:27:08] <dmessier> 400 rpm 4 flute...0.03" doc.... actual results....375 ipm feed... and data starvation ffrom the dnc
[22:27:39] <fenn> wah .03" depth isnt much to brag about
[22:27:47] <dmessier> 4140 heat treated to 45-49 Rc
[22:28:26] <CIA-8> 03jepler * 10emc2/src/ (3 files in 3 dirs): cleanups. use addprefix instead of patsubst when appropriate
[22:28:59] <dmessier> no part ... just CHIPS... Canadian Machine Tool Sholw 2000
[22:29:31] <dmessier> 375 ipm on a 12 inch extension.... I BRAG
[22:29:49] <dmessier> 53 mm cutter
[22:30:09] <fenn> that's 0.037" per tooth
[22:30:38] <dmessier> i made 1 ton of chipe in 3 days...
[22:30:40] <fenn> still 12" is pretty long
[22:31:20] <dmessier> 50 HP machine... gantry..cat 50
[22:31:39] <les_w> wow
[22:31:45] <les_w> what's a chipe?
[22:32:09] <dmessier> ppl would come by and ask... What are you makin'.... we'd say chips
[22:32:46] <dmessier> chips
[22:32:46] <fenn> it's a type of bird, apparently
[22:32:51] <les_w> I guess I wlll see some outrageous routing at IWF 2006
[22:33:27] <les_w> cutting 3/4 MDF at 2000 ips etc
[22:33:32] <dmessier> lok up hitatchi tool les... high feed milling
[22:33:37] <dmessier> look
[22:33:45] <les_w> k
[22:33:56] <les_w> well I try as much as I can
[22:34:36] <les_w> chips are money it's true
[22:34:46] <les_w> make em fast
[22:35:15] <dmessier> steel is now a HIGH commodity
[22:35:34] <dmessier> turn it over if you can
[22:35:55] <les_w> Looks like this year we are doing fewer but higher end turkey calls...so I have some machine time to sell
[22:36:17] <les_w> We may do a run of some furniture item
[22:36:29] <dmessier> high feed milling is desighnrd ti drive ALL the forces up the spindle
[22:37:17] <les_w> well, my machine can push 2,200 lb....but I don't think the spindle would like it
[22:37:20] <giacus_paltk> wow
[22:37:25] <giacus_paltk> damn bass
[22:37:26] <giacus_paltk> :P
[22:37:47] <les_w> better bass in my music room!
[22:37:51] <giacus_paltk> * giacus_paltk got a hit in the soul
[22:37:51] <les_w> I promise
[22:37:53] <giacus_paltk> hhahahaha
[22:37:56] <giacus_paltk> :D
[22:38:09] <dmessier> for wood.. i would tool up on pcd tools.. 1 of all your finishers... and forget it unlesss you screw up
[22:39:10] <les_w> I was going to try PCD...read some stuff and talked to some people and the consensus was micrograin carbide was more cost effective
[22:39:33] <les_w> So I just set up diamond sharpening in the shop
[22:39:44] <les_w> for the spiral carbide
[22:40:01] <les_w> we can do about 5 before we lose the secondary bevel
[22:40:25] <dmessier> you HAVE to knok its virgin wood.. no nails & there is NO set up error allowed
[22:40:48] <CIA-8> 03cradek * 10emc2/src/ (5 files in 2 dirs): (log message trimmed)
[22:40:48] <CIA-8> * Use the module helper even when running in place. The remaining
[22:40:48] <CIA-8> piece of the puzzle is to fix halcmd to not require setuid (by also
[22:40:48] <CIA-8> making it call the module helper.) This will make the module helper
[22:40:48] <CIA-8> the only setuid part of emc, which is the final goal.
[22:40:48] <CIA-8> * make it so './configure; make; make install' will give a working
[22:40:52] <CIA-8> system in /usr/local (using the module helper.) This is how every
[22:42:51] <dmessier> prove outs carbide... find the clearance clarence....then R & R
[22:43:11] <les_w> We would get about 3 hrs from a 1/2 spiral carbide with solids....but that was due to the emc TP
[22:43:20] <les_w> it trashes tools pretty fast
[22:43:41] <les_w> it's some better now
[22:45:25] <dmessier> how much better??
[22:45:27] <les_w> when feed slows down drastically at waypoints in g64 the resulting near zero chip lot gobbles up $60 bits fast
[22:45:44] <les_w> twice as good with the tp mod
[22:45:56] <les_w> we were talking about that today
[22:46:27] <les_w> but there are side effects to the code mod...check the list
[22:46:36] <dmessier> so it is in accel / decel...that she pucked??
[22:47:17] <les_w> yeah. When we fixed the stopping at waypoints...it now can violate max accel
[22:47:26] <dmessier> is it poaaibly overbuffering??
[22:47:52] <les_w> basically a fix was coded in 2000 to address accel violations
[22:47:55] <dmessier> possibaly
[22:48:07] <les_w> but it trashed the whole contouring thing
[22:48:31] <les_w> so we took out the "fix" last year
[22:49:09] <les_w> contouring is now smooth, but it may accel more than the maximum value in the .ini
[22:49:21] <les_w> for me, it's a no brainer
[22:49:34] <les_w> let it violate the max accel
[22:49:46] <les_w> if the path is now smooth
[22:50:29] <dmessier> 6is there a kinematic file available... add kin vel limits to continue to monitor if your gonna throw the machine into next week
[22:50:33] <les_w> I'ts not overbuffering...it's a math error.
[22:50:47] <SWPadnos> you can't violate max_accel with stepgen, by design
[22:50:55] <les_w> I see
[22:50:57] <SWPadnos> that's why there's an error
[22:51:11] <les_w> I am using servo emc1.
[22:51:16] <SWPadnos> and I never saw it with the USC, since it can go so far above the ini max_accel
[22:51:51] <SWPadnos> but stepgen has a hard physical limit on speed, and it also is coded to never exceed the max
[22:52:09] <SWPadnos> imagine what would happen with backlash compensation if that limit were ignored
[22:52:20] <les_w> well we need to fix it for sure. In the near term I'll take the accel violations since my machine can do 0.5g in x, and several g in y and z
[22:52:33] <dmessier> steppers are limited by the generator
[22:52:49] <SWPadnos> sure - you can have people set the max_accel and max_vel limits in the ini to be well under the actual machine limits
[22:53:00] <les_w> yeah
[22:53:01] <SWPadnos> and then let the stepgen / pid / whatever go over that
[22:53:08] <SWPadnos> but you need 2x to 3x headroom
[22:53:27] <SWPadnos> I'm sure you'd rather not set the limits to 1/3 your machine's capabilities, just to account for a bad TP
[22:53:37] <dmessier> overhwad aint cheap
[22:53:44] <dmessier> overhead
[22:53:46] <SWPadnos> which only shows up on high speed reversals, for the most part
[22:53:58] <les_w> accel violations don't hurt me...but the nasty spike we had before burns up $60 cutters real fast
[22:54:10] <dmessier> just wher all the comps come on too
[22:54:17] <SWPadnos> (the offending code in 3d_chips is a g0z10 followed by a g0z-55)
[22:54:28] <dmessier> backlash / gradient /
[22:54:30] <les_w> yeah
[22:54:35] <SWPadnos> that's you, with your overdesigned machine, and servos
[22:54:40] <les_w> got a 3x accel violation
[22:54:43] <SWPadnos> not the average emc user ...
[22:54:52] <les_w> right swp I am not typical
[22:54:57] <SWPadnos> the best fix is to fix the TP
[22:55:02] <les_w> yeah
[22:55:19] <SWPadnos> or leave that OLD_CODE define out, and you'll get the new code with the overloads ;)
[22:55:33] <SWPadnos> but smoother operation
[22:56:06] <dmessier> from a read.. that dumpster diving 65 inches... DOWN.... unless you have 75" on Z... its NOT GOOD TO GO
[22:56:16] <les_w> We tried so hard to do the fix that will intended...without trashing the tp
[22:56:22] <les_w> but we could not
[22:56:32] <SWPadnos> do you have a little time to discuss TP issues?
[22:56:38] <les_w> yeah
[22:56:52] <SWPadnos> ok - give me a few minutes to finish an email, and I'll get on that subject with you
[22:56:57] <les_w> k
[22:58:00] <les_w> I am going to check on my 98 yo grandma for a few minutes...we may get snowed in tonight
[22:58:03] <les_w> back in a few
[22:58:34] <dmessier> les... im not a coder ( these days) but i know systems...
[22:58:48] <dmessier> any help ... im in
[22:59:49] <SWPadnos> part of the discussion will be "what's the best thing to do for blending"
[22:59:57] <SWPadnos> so you can definitely help there, I think
[23:01:04] <dmessier> what are the current tp/blend algorithms based on.... or are there any
[23:03:13] <SWPadnos> I've been trying to figure that out ;)
[23:05:01] <dmessier> average the last 3 points.... lookahead the next 3 pts put it the middle.. upon an acute condition... evealuated on poit to poin... mode.. "wae nned a sharp corner"
[23:05:24] <SWPadnos> nope
[23:05:30] <SWPadnos> that's not it ;)
[23:05:39] <dmessier> can we incorporate b-splines>>
[23:06:24] <dmessier> i'm playing the devils advocate
[23:06:51] <dmessier> it COULD change the world...
[23:07:19] <dmessier> knots & weights
[23:09:59] <dmessier> vector & time
[23:10:33] <SWPadnos> phone
[23:11:05] <fenn> dmessier: it seems to be more like "whoever wants to write it, will you please write it already cause we know what we want it to do"
[23:18:05] <les_w> back. Checked on my grandma and an employee. get snowed in in the mountains and you could starve or freeze.
[23:19:12] <bill2or3> * bill2or3 won his stepper drivers, $50each
[23:19:13] <bill2or3> yay
[23:19:24] <les_w> cngrats
[23:19:50] <bill2or3> that just leaves figuring out what I'm doing for rails/bearings
[23:19:58] <les_w> abba!
[23:20:24] <bill2or3> abba?
[23:20:40] <les_w> low cost high precision linear rails
[23:20:54] <les_w> google...
[23:20:59] <bill2or3> your ideas intrigue me and I wish to subscribe to your newsletter.
[23:21:10] <les_w> haha
[23:21:19] <les_w> well, I design machines.
[23:21:58] <fenn> dont listen to him, he just talks on the phone/irc all day
[23:22:20] <les_w> newsletter....you know, that's a damn good idea
[23:22:40] <les_w> yeah I get paid for talking on the phone like teenage girls.
[23:22:58] <fenn> we call that "spam" in more enlightened parts of the world
[23:23:13] <les_w> heh...yeah
[23:23:32] <bill2or3> who actually has a price list for abba rails?
[23:23:44] <les_w> well people welcome advertising if it has useful content
[23:23:53] <les_w> I have some quotes.
[23:24:07] <les_w> about $100/meter for 20 mm
[23:24:16] <les_w> trucks are about $50
[23:24:21] <fenn> i would pay money for a service that gave me the actual prices of various vendors..
[23:24:22] <bill2or3> we have like 15+ mail servers here because of what all the advertisers consider "useful"
[23:24:23] <les_w> good deal
[23:24:30] <bill2or3> hmm
[23:24:32] <les_w> unless you get luvky on ebay
[23:24:48] <bill2or3> "truck" is the bearing-part that slides on the rail?
[23:24:57] <les_w> yeah
[23:25:05] <bill2or3> and I need *8* of those.
[23:25:06] <bill2or3> ouch
[23:25:26] <les_w> well, it's about one third to one half the price of most...
[23:25:52] <fenn> stop! drawerslide! woo oo oo oo ooh oh ooh oh
[23:25:53] <bill2or3> what do you think about fixedxy spindle on a gantry over an X-Y table, compared to a moving-X spindle over a "y table" ?
[23:26:00] <bill2or3> I haven't decided which to do yet.
[23:26:18] <les_w> it's as good as I can get for new stuff
[23:26:29] <bill2or3> you do need 2 trucks per rail right?
[23:26:44] <bill2or3> they're not so wide that you can use one-per...?
[23:26:47] <les_w> moving gantry has advantages...but costs much more
[23:26:59] <fenn> what are the advantages of a moving gantry?
[23:27:06] <les_w> 2 trucks per rail yes
[23:27:07] <bill2or3> seems like it'd be a lot harder to keep rigid
[23:27:12] <fenn> less mass to sling around?
[23:27:17] <bill2or3> moving gantry machine is narrower, too.
[23:27:48] <les_w> yeah....moving mass does not change with work....so servos can be tuned much tighter
[23:27:55] <fenn> * fenn is stuck in "make it as rigid as possible" mode
[23:28:09] <les_w> that's a good mode to be in
[23:28:14] <bill2or3> yeah, I probablly will use a fixed gantry
[23:28:44] <les_w> unless it is a big machine fixed gantry is the way to go
[23:29:37] <bill2or3> it's gonna be small, 18" x 24" I think.
[23:29:51] <bill2or3> maybe 4-5" of Z travel
[23:30:11] <SWPadnos> phone still
[23:30:32] <SWPadnos> it's my mother - I'm asking her about this problem
[23:31:26] <les_w> np
[23:34:06] <les_w> basically in machine tools maximize SQRT (stiffness/mass)
[23:34:49] <les_w> BTW as a bulk property that ratio is just about identical for aluminm and steel
[23:35:01] <les_w> steel is cheaper though.
[23:35:18] <fenn> steel has better damping
[23:35:24] <Jymmm> I have some (actual dimension) 1" x 5.5" I'd like to glue up, but dont have a jointer/planer, am I screwed ?
[23:35:57] <fenn> jymmm just face mill it :)
[23:36:36] <Jymmm> fenn ....and 10 hours later....
[23:36:39] <Jymmm> lol
[23:36:57] <Jymmm> 240 sq inches.... yuck
[23:37:40] <Jymmm> I've never glued up anything, so nfc.
[23:38:05] <Jymmm> anything I should watch out for?
[23:39:10] <fenn> use lots of clamps
[23:39:16] <les_w> phone
[23:39:16] <fenn> and lots of glue
[23:39:37] <Jymmm> I have soem gorilla glue
[23:39:57] <fenn> that will foam and expand to fill cracks.. fyi
[23:40:09] <Jymmm> is that a good or bad thing?
[23:40:20] <fenn> probably a good thing
[23:40:48] <Jymmm> This wood is pretty fiberous if that makes a difference
[23:41:05] <Jymmm> Mangaris
[23:41:10] <fenn> http://www.woodworking.com/ww101glue.cfm
[23:41:47] <SWPadnos> ok - I'm back now
[23:41:50] <SWPadnos> phew
[23:44:42] <fenn> btw jymmm gorilla glue is cyanoacrylate and gives off toxic fumes when burned..
[23:44:51] <fenn> er, crap not cyanoacrylate
[23:45:07] <fenn> its polyurethane but gives off cyanide when burned
[23:45:29] <fenn> not enough to kill you but..
[23:46:24] <Jymmm> I hear ya... but hopefully what I make isn't going to be burned either =)
[23:46:32] <SWPadnos> hmmm - it looks like there's an uninitialized variable in that OLD_CODE
[23:48:27] <SWPadnos> maybe not
[23:52:12] <les_w> ahhh skype to mexico...goverment messes with it I am told...lost connection
[23:53:37] <les_w> anyway glue...titebond III or gorilla
[23:54:00] <les_w> gorilla might not be good for you...it is moisture cure and you are in a desert
[23:54:18] <fenn> you can spray water on the wood before applying the glue
[23:54:21] <Jymmm> les_w Californai isn't a desert yet =)
[23:54:31] <les_w> some parts
[23:54:41] <les_w> yeah fenn spray water
[23:54:53] <les_w> and what variable is not initialized?
[23:56:01] <les_w> I glue up panels with titebond II...they have stood up to several years outdoors np
[23:56:38] <SWPadnos> well - the velCart variable is initialized to 0, then gets added to each time through the loop, which makes no sense to me
[23:56:55] <SWPadnos> it's done after it's used though, so it looked wrong at first
[23:57:18] <les_w> oh
[23:57:59] <les_w> so it's an integration?
[23:58:20] <SWPadnos> what doesn't make sense to me is why this calculation is done for the entire tpRunQueue every time
[23:58:43] <SWPadnos> I do like the comment at the beginning of the function though:
[23:58:56] <SWPadnos> /* what the hell does this do? It is one of the key functions in
[23:58:57] <SWPadnos> the whole planner, and it doesn't have any overall comments :-(
[23:58:59] <SWPadnos> */
[23:59:03] <les_w> haha
[23:59:57] <les_w> well myself, paul , jon E., Dave, Alex, jmk could not figure it out.