#emc | Logs for 2005-09-20

Back
[00:00:26] <paul_c> number of bug fixes and enhancements coming back from Universities... um.. some.
[00:00:54] <zwelch> "F.A." ?
[00:01:06] <paul_c> Much of the latest bug fixes are from the likes of jepler & cradek
[00:01:19] <zwelch> and "some"? :) useful quantitative datum that :)
[00:01:52] <paul_c> Difficult to know who is working with a Uni.
[00:02:43] <paul_c> I can probably point out a dozen fixes/enhancements from students.
[00:02:43] <zwelch> well, i'm thinking from the perspective of maintiner; you end up dealing with a patch flow of some "n" patches per {day,week,month,year} period
[00:03:10] <zwelch> what are the scope of those?
[00:03:31] <zwelch> (on average, plus examples of "best and worst" would be nice ;) )
[00:03:52] <paul_c> Hrmm... "n" patches in um.. err.. five years... Maybe seen three.
[00:04:14] <zwelch> eek
[00:04:52] <paul_c> Now go look at the number of "registered developers" and tally the number of commits.
[00:05:03] <zwelch> so, i need some more perspective.... how many users/developers attended the codefest this year?
[00:05:28] <paul_c> off the top of my head -
[00:05:35] <zwelch> (i'm not sure if i want to, if it's not going to be encouraging... and i'm starting to reckon it won't be)
[00:06:16] <paul_c> maybe six hardcore developers, three lurkers and a couple of pitas
[00:06:31] <zwelch> * zwelch laughs
[00:07:09] <zwelch> i can be a pita developer, as you may already be able to tell... i wonder where i'd get counted ;)
[00:07:42] <zwelch> so you mentioned NIST earlier; they still are active with this (or the underlying codebase)?
[00:07:53] <paul_c> lemme try to describe the scenario...
[00:08:56] <paul_c> A user with a real vintage installation bitching about (perceived) incorrect behaviour in the interpreter....
[00:09:22] <paul_c> Quite a bit of time spent working on identifying the issues....
[00:09:51] <paul_c> even more time spent tracking down a "real" bug...
[00:10:27] <paul_c> Now we say "code fixed. Please install & compile & report back".
[00:11:44] <paul_c> The guy responds with - I can't load the code on to the laptop because a) no CD drive, b) no network c) won't fit on a single floppy...
[00:11:59] <paul_c> "and I'm happy with turboCNC"
[00:12:17] <paul_c> That is a real PITA...
[00:12:41] <Phydbleep> USER = "Cheap Dickhead" ERROR..
[00:13:00] <zwelch> users... can't shoot 'em, can't debug them like good 'ol C code either
[00:13:02] <paul_c> IO error.
[00:13:17] <paul_c> as for NIST....
[00:13:19] <Phydbleep> paul_c: Idiot Online?
[00:13:31] <paul_c> summat like that.
[00:13:57] <Phydbleep> ID-10-T error. :)
[00:14:17] <paul_c> NIST (due to politics) had to distance it's self from EMC in an official capacity.
[00:14:40] <paul_c> but they still use it for some of their internal work
[00:14:56] <paul_c> and add stuff from time to time.
[00:15:37] <paul_c> Oh, abd the hosted the codeFest this year.
[00:17:26] <paul_c> and here's another one.
[00:17:35] <dave-e> with questions
[00:18:04] <paul_c> you wouldn't be here for any other reason would you ;-)
[00:18:09] <Phydbleep> dave-e: 42!
[00:18:22] <dave-e> nope...at least seldom
[00:18:36] <les> hi dave
[00:18:56] <dave-e> what was the name of the cute utility that gave modeline parameters.
[00:18:58] <dave-e> hi les
[00:19:16] <Phydbleep> dave-e: xvidtune?
[00:19:45] <dave-e> just got another flat screen to try out....is supposed to work at 60 Hz and doesn't ... darned
[00:20:04] <paul_c> BTW dave-e.... lemme introduce you to zwelch
[00:20:14] <paul_c> just down the road from you.
[00:20:14] <dave-e> ok
[00:20:18] <dave-e> like where
[00:20:38] <dave-e> down the road is a relative term
[00:20:41] <zwelch> * zwelch is in corvallis, or
[00:20:54] <dave-e> oh hey....at least i know where it is
[00:21:20] <zwelch> * zwelch waves his OSU Beavers penent aimlessly
[00:21:24] <Phydbleep> zwelch: You're down the road from my relatives... I'm sorry.. :)
[00:21:26] <paul_c> dave-e: gtf is what you are after I think..
[00:21:46] <dave-e> yes it is....now I'll screen print so i can remember
[00:21:58] <Jacky^> Phydbleep: :P
[00:22:01] <Jacky^> hi
[00:22:05] <Phydbleep> * Phydbleep moved to Albuquerque to escape them..
[00:22:12] <Phydbleep> Hi Jacky^ :)
[00:22:30] <zwelch> * zwelch grew up in Farmington, NM and went to school briefly at NMT in Socorro
[00:22:42] <paul_c> dave-e: Got a sharpie ?
[00:23:06] <dave-e> no why
[00:23:10] <Phydbleep> * Phydbleep has been having the summer from hell...
[00:23:12] <zwelch> but after learning that "socorro" means help in spanish, and that the city is also in "Socorro" county... i realized it was time for me to leave "Help! Help!", New Mexico ;)
[00:24:30] <dave-e> What I've seen of Albuquerque is ok
[00:24:38] <zwelch> * zwelch steps away from the microphone, sparing you from more of his off-topic introductory banter
[00:24:43] <dave-e> just a bit waarm forme
[00:24:47] <Phydbleep> ABQ is beyond help.. :)
[00:26:46] <dave-e> dry...abq to web-feet in
[00:26:51] <dave-e> Corvallis
[00:26:55] <les> hey dave, ray said he was going to put up some tkio tool change stuff on the wiki
[00:27:08] <dave-e> and ....
[00:27:13] <Phydbleep> Everybody agree that I should be OK welding a piece that's jigged in the lathe as long as I pull the power/data cables and ground directly to the piece?
[00:27:22] <les> and I'l read it when he does!
[00:27:23] <paul_c> No.
[00:27:40] <Phydbleep> paul_c: ?
[00:27:44] <zwelch> so regarding NIST's involvement; i understand the necessity for state agencies to distance themselves from public entities, obviously the need to keep everyone's codebases unified means that bits are still making their way across the supposed divide
[00:28:23] <paul_c> Phydbleep: In my opinion, welding should be done well away from a lathe.
[00:28:55] <zwelch> from what i would guess, their development is on the rcslib/nml bits more than the emc as a whole? what's their involvement with emc2?
[00:29:22] <paul_c> emc2 grew out of rcslib/emc
[00:29:38] <Phydbleep> paul_c: Got no choice, It's the only thing I have to use for a fixture and I need to tack it aligned to the lathe ways/chuck..
[00:30:00] <dave-e> I'm going to bail and try to get that lcd going. thanks paul
[00:30:05] <paul_c> Fred Proctor moved some of the canonical scripting stuff over during the codeFest (wats'n'all)
[00:30:13] <dave-e> good luck les...
[00:30:15] <dave-e> gone
[00:30:52] <paul_c> Phydbleep: Cover the bed & ways with something...
[00:31:01] <les> yeah
[00:31:06] <paul_c> sheets of ali for example
[00:31:18] <paul_c> stop the splatter sticking
[00:31:23] <Phydbleep> paul_c: I have bed protectors and I was just looking at 4 tacks to hold the thing together.
[00:31:38] <paul_c> stick, mig, or tig ?
[00:31:57] <Phydbleep> * Phydbleep is building an index head..
[00:32:08] <Phydbleep> Mig.. Flux core..
[00:32:28] <paul_c> Not as messy as stick then.
[00:32:33] <les> cover every square inch of that lathe heh
[00:33:21] <Phydbleep> * Phydbleep has several dead treadmill belts to sacrifice for the cause. :)
[00:34:48] <paul_c> clamp the earth to the work, not the bed or headstock.
[00:36:22] <les> why are you building an index head? fun?
[00:37:57] <les> Curious what your standard will be to test the index holes
[00:39:01] <les> I have an index plate on the cnc head to alow tilting and locking at various angles
[00:39:11] <les> allow
[00:39:12] <Phydbleep> Bicycle hubs.. I need 18/36 stops for spoke holes and I have a 72 tooth gear I bought for the lathe that gets little use.
[00:39:45] <les> I see
[00:40:13] <les> I used a rotary table to bore the plate holes
[00:40:44] <les> table is supposed to be good to 15 arc sec but I have never really tested it
[00:41:43] <les> I used as non-locking taper on the index holes and pin...about 10 degrees
[00:42:14] <zwelch> paul_c: okay, so i'm hacking on the configure.in right now
[00:42:25] <zwelch> there's a lot of stuff in there that i am going to clean up
[00:42:59] <zwelch> it will fix the with-rt options, add the disable-realtime option, and it will all make sense when you read the code ;)
[00:43:24] <Phydbleep> This is just going to be a 5 tooth lock.. I'll copy the gear profile/teeth for a 5 tooth span and use that for a lock/alignment peg.
[00:43:57] <les> that should average out the errors
[00:44:56] <Phydbleep> Should be 'close enough' for bicycle parts. :)
[00:45:12] <les> yeah
[00:49:56] <paul_c> well... I'm off for an early night.
[00:50:05] <les> night
[00:50:14] <les> arg just missed!
[00:50:59] <les> zwelch still there?
[00:51:12] <les> wonder what goy you interested in emc
[00:53:42] <zwelch> les: many factors
[00:54:22] <les> btw, I am a commercial user of emc. There aren't mant, but few
[00:54:26] <les> many
[00:54:50] <zwelch> cool
[00:55:21] <zwelch> i recall you also ran for a seat on the linuxcnc board
[00:55:21] <les> We typically are concerned with servo based systems only....speed and performance
[00:55:38] <zwelch> we == your company?
[00:55:43] <les> I was nominated...didn't ask
[00:55:46] <zwelch> ah :)
[00:55:48] <les> yes
[00:55:51] <zwelch> still, you didn't say no ;)
[00:56:18] <les> I am not a developer...but do a good bit of math. I am an aero engineer that designs machine tools.
[00:56:33] <zwelch> define "machine tools" please :)
[00:57:12] <les> So I limit my stuff here to talking mostly about trajectory planners, kinematics, servo loops, and the like.
[00:57:25] <les> cnc routers...high speed machining
[00:57:40] <zwelch> so you are using CNCs to build CNCs?
[00:58:29] <les> sometimes. I build stuff for clients, and am considering marketing a line of light industrial tools
[00:59:38] <les> Emc is the only reasonable performance servo control out there in the under $1000 bracket
[01:00:27] <les> It needs to be and can be better though
[01:01:07] <zwelch> how dependent is your business on EMC?
[01:01:42] <les> about 25% of income comes from running the router
[01:01:52] <les> that might change though
[01:02:08] <les> if we decide to market light industrial routers
[01:03:18] <Jacky^> :D
[01:03:34] <les> Software and hardwaree controllers cost thousands for competition
[01:03:55] <les> the next cheapest thing is wincnc/galil right now
[01:04:08] <zwelch> * zwelch is not familiar with that
[01:04:12] <zwelch> got URL?
[01:04:18] <les> about a kilobuck quan 10
[01:04:25] <les> looking
[01:04:48] <bosone> hi :) I'm italian physic student and i think emc is great. Today i try to understand the realtime/no-realtime process comunication,
[01:04:51] <zwelch> you know, i always feel silly asking
[01:05:10] <Jacky^> hay bosone :)
[01:05:12] <zwelch> google gave me galilmc.com, which is what i bet you were refering to
[01:05:35] <Jacky^> bosone: query ..
[01:05:38] <bosone> emc is great example of realtime linux power
[01:06:06] <zwelch> bosone: a-yup; what are you trying to understand exactly?
[01:06:43] <les> yes...that is dsp based control
[01:06:59] <les> they can now do about 10 kHz servo update
[01:07:36] <les> We are stuck at about 2 kHz because we are using a general purpose proccessor
[01:07:47] <les> might be able to improve that some
[01:08:00] <zwelch> what are the specs on your processor?
[01:08:33] <zwelch> (after all, the quickest way to improve a systems performance is to increase its clock rate/fundamental architecture ;) )
[01:08:45] <les> oh emc seems to max out at about 2kHz on any speed proc...
[01:09:05] <les> 200 mHz or 2 gig about the same I guess
[01:09:26] <zwelch> that seems irrational, but maybe there's a good reason
[01:09:58] <les> I'm not a developer, but I can see that caches and some proc architectures don't really help hard RT code
[01:10:17] <bosone> zwelch: i try to understand nml, is the emc way to interface rt?
[01:10:43] <les> I understand that RT kernal/user space comms is the reason
[01:11:29] <les> oops kernel heh
[01:13:14] <zwelch> bosone: i don't know; i'm just starting to learn about the system myself
[01:14:23] <les> anyway changing the nml message structure might improve that a lot
[01:14:43] <les> 2KHz is getting a little long in the tooth for this day and age
[01:15:03] <zwelch> * zwelch would think so
[01:15:27] <les> Works fine on an old Bridgeport, but not for modern tools
[01:15:54] <zwelch> what are the problems that the reduced frequencey causes?
[01:16:12] <bosone> I too, about 3 days ago. but is very complex. For now i'd like to write something like "extsmdromot.c" for a custom cheap encoder counter.
[01:16:38] <les> Well, sampling issues....if you are running at 500 or 1000 inches per minute
[01:16:57] <zwelch> so you have to restrict your feedrates
[01:16:59] <les> low bandwidth
[01:17:04] <les> yes
[01:17:13] <les> or
[01:17:23] <les> cut simple low bandwidth shapes
[01:17:34] <zwelch> heh :)
[01:18:07] <zwelch> tell that to an engineer, and i think they'll tell you to find a better system ;)
[01:18:12] <Jacky^> bosone: are you readin the query ?
[01:18:19] <Jacky^> reading*
[01:18:21] <les> Also, I need to look at an IIR biquad notch filter in the pid section.
[01:18:32] <les> he I am an engineer
[01:18:47] <zwelch> (okay, then a snooty engineer ;)
[01:19:03] <bosone> Jacky^: non vedi le risposte?
[01:19:14] <les> EMC's pid loses phase margin at first machine resonance poles
[01:19:15] <Jacky^> bosone: nope..
[01:19:35] <les> looses
[01:19:42] <zwelch> les: that's one jargon filled sentence; i don't grok
[01:20:17] <les> however IIR digital filters and faster servo rates are mutually exclusive
[01:20:30] <Jacky^> bosone: :\ i can't see your replys
[01:20:33] <les> Oh you beat me on jargon ;)
[01:21:27] <zwelch> refering to the earlier software discourse? ;)
[01:21:32] <bosone> Jacky^:i dont know
[01:21:34] <les> digital filters in code= double nested loop.
[01:21:38] <les> ha yeah
[01:21:52] <les> I write a little c but am not a programmer.
[01:21:56] <zwelch> i didn't say i didn't like it, did it? :)
[01:22:03] <les> heh
[01:22:08] <Jacky^> bosone: probably my irc client :(
[01:22:08] <zwelch> s/it/i/, natch
[01:22:32] <zwelch> * zwelch wonders if Jacky^'s nick having a ^ has anything to do with it
[01:22:39] <bosone> Jacky^: or my?
[01:22:43] <les> Now if you want to talk math...well I do that.
[01:22:50] <Jacky^> bosone: no idea..
[01:23:16] <les> you are not seeing bosone jacky?
[01:23:18] <Jacky^> zwelch: i like the ^ beacuse all people ask to me why i like it :P
[01:23:23] <zwelch> i'm not afraid of math, but i have a professor in my band to handle the tough stuff (he plays the banjo)
[01:23:46] <Jacky^> kidding ..
[01:23:48] <les> ha
[01:24:02] <Jacky^> mi irc client seem wont work fine..
[01:24:24] <zwelch> les: do you have a URL for your company?
[01:24:35] <les> Well for emc...I am performance oriented. A very high performance open source control would be a good thing.
[01:24:48] <les> www.lmwatts.com
[01:24:56] <les> go to engineering link
[01:25:07] <les> small firm
[01:25:08] <bosone> Jacky^: is a my problem
[01:25:19] <Jacky^> bosone: ok, np
[01:26:02] <les> btw the logoplaque was carved with emc.
[01:26:28] <Jacky^> emc is great
[01:26:47] <bosone> Jacky^ : i'm working on servo project, chep, via parallel port and CPLD. I'll tell you.
[01:26:48] <Jacky^> i think is too young yet, but ..
[01:27:25] <Jacky^> bosone: nice, i was looking for some small circuit, just to try some dc motor on emc ..
[01:27:46] <Jacky^> but I understand servos drivring its hard :\
[01:27:57] <Jacky^> a lot of math is needed
[01:28:11] <les> it is not hard jacky
[01:28:33] <bosone> no, the pid is build in emc
[01:28:33] <Jacky^> les: i found something
[01:28:45] <Jacky^> like quadrature encoders, pid ..
[01:29:08] <Jacky^> but cant understand how to drive a small dc motor
[01:29:23] <les> encoder are not very expensive these days
[01:29:28] <Jacky^> i'm tryng to read more about
[01:29:49] <Jacky^> les: i know, are not expensive
[01:30:05] <Jacky^> but how to interface them to emc ?
[01:30:06] <zwelch> les: ah, right, i visited your cnc page last week while following various EMC links
[01:30:11] <les> drive a small motor with a full H bridge ic
[01:30:14] <les> oh ok
[01:30:37] <zwelch> you've got some good info up there
[01:30:43] <les> thanks
[01:30:56] <Jacky^> bosone: btw, i'm interesting in your project, but im absolute newbie :\
[01:31:31] <Jacky^> if you already have some doc would be a nice thing
[01:32:23] <bosone> Jacky^: emc standard motor output is a voltage, emulated for stepper, or direct to an adc for servos
[01:32:24] <Jacky^> coming back a bit.. I no need pid ?
[01:32:47] <bosone> no, the pid is in emc
[01:32:48] <les> jacky I am pulling out the dynomotion card to see the servo drive ic number
[01:33:10] <Jacky^> :(
[01:33:16] <Jacky^> understood ..
[01:33:58] <Jacky^> this mean i no need to use pic, microprocessor ?
[01:34:17] <Jacky^> * Jacky^ confused
[01:34:22] <les> you need a servo card
[01:34:31] <bosone> Jacky^: in real world the analog output from emc is converted in pwm and then drive the h-bridge
[01:34:37] <les> Soon the Motenc lite will come out
[01:34:40] <Jacky^> with setp-direction input right ?
[01:34:42] <les> 3 or 4 axis
[01:34:54] <les> no not step or direction
[01:35:08] <Jacky^> mmhh..
[01:35:15] <les> -10 to +10 analog or PWM
[01:35:27] <Jacky^> i read something abou duty cycle
[01:35:35] <les> so the computer and emc know the actual position
[01:35:36] <Jacky^> is that ?
[01:35:49] <les> not quite
[01:36:07] <les> in this case emc actually reads the encoder...not the amp
[01:36:43] <Jacky^> oh... this mean the uotput of encoder goes to parallel port ?
[01:36:49] <Jacky^> :\
[01:37:03] <les> it goes to cards like the motenc or stg
[01:37:14] <Jacky^> oh.. ok
[01:37:24] <les> which emc reads
[01:37:29] <Jacky^> so the driver take the control
[01:37:52] <Jacky^> mmhh.. too coplex for me
[01:37:53] <les> well the amp becomes a simple H bridge
[01:38:24] <les> no not complex.....what is complex is dealing with stepper mid band resonances!
[01:38:40] <Jacky^> I know ..
[01:39:18] <Jacky^> i've seen my stepper losting step at 10-30-40
[01:39:18] <zwelch> * zwelch squints his eyes at the fresh indecipherable jargon
[01:39:29] <Jacky^> and not at 15-25-35 ..
[01:39:31] <les> ??haha
[01:39:46] <Jacky^> strange curve ..
[01:39:54] <bosone> Jacky^: emc act a real-time closed loop for motor and encoder: http://home.att.net/~jmkasunich/EMC_Docs/EMC_Home.htm
[01:40:07] <les> right
[01:40:11] <zwelch> what is "mid band resonance" in the context of stepper motors?
[01:40:11] <Jacky^> bosone: thanks
[01:40:22] <Jacky^> is like a speaker..
[01:40:38] <Jacky^> or not ? les ?
[01:40:53] <les> zwelch, torsional and other resonances often cause slipped steps
[01:40:59] <les> yes like a speaker
[01:41:36] <les> We don't worry about it in machine tools, because they use servos.
[01:41:49] <les> But I use steppers in other things.
[01:42:36] <Jacky^> in some case steppers are usefull
[01:42:42] <les> yes
[01:42:45] <Jymmm> Jacky^ name 300 ?
[01:42:52] <les> they are great for printers!
[01:42:59] <Jymmm> that's 1
[01:43:01] <Jacky^> hi Jymmm , what ?
[01:43:13] <les> I use them for test rigs
[01:43:20] <Jymmm> ok 2
[01:43:29] <les> And I use them in the painting robot
[01:43:37] <zwelch> is anyone using linear motors with EMC?
[01:43:48] <Jymmm> I'll belive that when I see that photo les
[01:43:55] <les> Not to my knowledge
[01:44:06] <les> photo coming...some day...ha
[01:44:23] <zwelch> i was wondering how hard it would be to build one
[01:44:29] <Jymmm> les cough*BULLSHIT*cough
[01:44:34] <les> haha
[01:44:44] <Jacky^> lol
[01:44:44] <les> not hard...just expensive.
[01:45:15] <les> it's just a rolled out brushless drive...
[01:45:19] <zwelch> no, i mean the linear motors... from parts
[01:45:26] <zwelch> because, yeah, it's a very simply idea :)
[01:45:29] <les> I know
[01:45:32] <zwelch> s/ly/le/
[01:45:46] <les> yes simple conceptually
[01:46:03] <Jymmm> so it parting water.... conceptually
[01:46:07] <bosone> zwelch: stepper is simple but not real closed-loop
[01:46:11] <zwelch> heh :) yeah, darn reality and practice getting in the way of theory and concepts... again ;)
[01:46:12] <les> lots of coils and rare earth magnets I guess
[01:46:33] <Jymmm> les use mars magnets instead
[01:46:53] <zwelch> yeah, it's the "rare earth magnets" that has me concerned; i don't think i have any of those lying in my parts buckets
[01:47:01] <les> Understand that until a couple months ago emc had some serious trajectory problems that crept up at high speeds
[01:47:09] <les> that is now largely fixed
[01:47:19] <zwelch> well, i was seeing stuff like that just in the backplot
[01:47:28] <zwelch> * zwelch wonders if that's normal behavior though
[01:47:46] <les> every waypoint had a huge miscalculated accel spike
[01:48:03] <zwelch> nope; different
[01:48:11] <les> bad enough to tear apart a high speed machine
[01:48:23] <zwelch> eek :/
[01:48:40] <zwelch> it's hard to find beta-testers after bugs like that :/
[01:48:42] <les> like mine....10 KN force on the axes
[01:49:11] <Jymmm> fuck les, if you coudl just make a machien like I bought, less the fuckups, you'ld make a lot of money
[01:49:17] <zwelch> so how bad could the damage from that be?
[01:49:22] <les> Paul, NIST, and others came down to watch it on the big machine
[01:49:50] <les> imagine a 10 KN force jab 100 times a second...
[01:50:14] <les> how can I describe it....jackhammer?
[01:50:23] <zwelch> yeah, sounds fair
[01:50:42] <les> anyway Paul found it. pretty much fixed
[01:50:53] <les> after YEARS of work
[01:51:02] <les> elapsed time anyway
[01:51:47] <Jymmm> les : Man, make a table top machine! I'll beta test it
[01:52:15] <les> So now it is a fairly reasonable cubic sub interpolated trapezoidal path planner.
[01:52:28] <les> Albiet a little slow
[01:52:38] <les> Working on it Jymmm
[01:53:12] <Jymmm> les Eh, screw that chinese stuff... too much headache... light weight aluminum
[01:53:57] <Jymmm> les I have a lot of ideas in respect to 'erector set' it.
[01:54:03] <les> I talked to Bill Glenn, the founder of Glentek Warthog cnc routers.
[01:54:17] <les> We are getting a look at the chinese stuff.
[01:54:23] <les> It is bad...real bad
[01:54:34] <les> Horribly misengineered
[01:54:45] <Jacky^> les: seen the motenc card, nice, but expensive too
[01:54:53] <les> but...pretty good workmanship
[01:55:09] <les> the motenc lite will cost less jacky
[01:55:25] <Jacky^> looking
[01:55:39] <les> Anyway the chinese could make a good machine...but we need to design ity.
[01:55:45] <Jacky^> it seem the only way also for beginners..
[01:56:21] <les> I think an even lower cost card is being worked on
[01:57:05] <les> So we are talking about that...I design...Chinese make.
[01:57:28] <Jacky^> nice :)
[01:58:00] <les> Using emc is crucial to this product concept
[01:59:02] <zwelch> i would be interested in working on a such a project
[01:59:07] <les> I don't want to market "another stepper based $12,000 1X2 meter machine"
[01:59:19] <les> zwelch: good
[01:59:33] <zwelch> i've got lots of embedded linux experience
[01:59:50] <les> brb phone
[01:59:58] <Jacky^> zwelch: it seem ther's a lot of work
[02:00:06] <zwelch> yeah?
[02:00:14] <Jacky^> i'm sure
[02:00:45] <Jacky^> you just need to decide what
[02:01:36] <Jacky^> zwelch: what you mean for embedded ?
[02:01:46] <Jacky^> whats you experience
[02:02:22] <zwelch> well... i helped found (the original) Gentoo Embedded
[02:02:32] <zwelch> embedded == limited resources
[02:02:43] <Jacky^> like pda ?
[02:03:06] <Jacky^> i've a tiny pda, an ipaq 3970 with linux gpe running
[02:03:07] <zwelch> it usually also entails "single application', but, yes, the pda is kind of the "crown of embedded"
[02:03:30] <zwelch> i used to work for www.tdsway.com and did all of the original software on their very first handheld
[02:03:33] <Jacky^> cool
[02:04:03] <Jacky^> are the thing going well in the gentoo community ?
[02:04:20] <zwelch> during that time, i adapted ARM Linux to run on it in house, got bit by the Free Software bug, and quit to start my own Linux software development consultancy
[02:04:55] <Jacky^> nice
[02:05:08] <Jacky^> congrats :)
[02:05:20] <Jacky^> i'm a debian fan
[02:05:27] <bosone> zwelch: do you have RTAI or some other realtime-platform experience?
[02:05:30] <zwelch> well, you see... shortly after i helped found Gentoo Embedded, i led the attempt (formerly know as Zynot) to fork Gentoo into a new distribution (yes, that was me)
[02:06:06] <zwelch> bosone: i have experimented with it for years, but i have never had a reason to really use it (read as: no one has paid me to care about RT code)
[02:06:08] <Jacky^> :)
[02:07:21] <zwelch> emc will be the first application that i "want" (i.e. "don't need") to run, so i may start to care anyway
[02:07:50] <zwelch> however, i don't have any hardware (and don't foresee being able to afford any without finding some paying work that justifies the expenditures)
[02:08:30] <Jacky^> zwelch: oh.. are you a proprietary software developer ?
[02:08:44] <zwelch> no, i am 100% free and open source
[02:08:54] <Jacky^> ok
[02:09:01] <Jacky^> great
[02:09:26] <zwelch> i run nothing but linux on my systems, and i restrict "closed source" portions to less than 10% of any contract
[02:09:46] <Jacky^> good, i'm too
[02:10:10] <zwelch> (i.e. i respect some clients' desire to see the "crown jewels" of their IP protected, but i refuse to let the "supporting scaffolding" be closed up)
[02:10:34] <Jacky^> europe won a great battle with the europen parlment and software patence
[02:10:51] <Jacky^> i think the open source can have a great future here
[02:11:20] <zwelch> * zwelch hopes so
[02:11:36] <Jacky^> america its another world
[02:11:44] <Jacky^> Jymmm: right ? :P
[02:11:54] <Jacky^> but is changing too
[02:12:27] <zwelch> * zwelch shrugs
[02:12:41] <les> back
[02:14:02] <les> Well business was horrible for several years
[02:14:22] <les> first two years I made zero!
[02:14:25] <zwelch> * zwelch can't imagine why (<-- note abundant sarcasm)
[02:14:32] <zwelch> * zwelch started his company less than a month before 9/11
[02:14:39] <zwelch> hurray for bad timing :)
[02:14:51] <les> Engineers and engineering managers are not used to making zero.
[02:14:53] <les> but
[02:14:57] <Jacky^> les: zero i better than -20k :P
[02:15:03] <Jacky^> hehe
[02:15:04] <les> thing have popped pretty well
[02:15:07] <Jacky^> I know ..
[02:15:14] <Jacky^> its hard
[02:15:15] <zwelch> but if you made it for and by yourself, then it was work every penny? :)
[02:15:23] <les> heh
[02:15:39] <les> Well the engineering consulting side has done well
[02:15:45] <zwelch> * zwelch was happier making nothing those first two years, knowing the investment would eventually pay off
[02:16:06] <Jacky^> les: in dont know the exact word
[02:16:07] <les> and the production cnc supports 3 workers seasonally...about one quarter
[02:16:51] <Jacky^> but sometimes a bit of 'incoscience' is needed
[02:16:52] <les> I am in the process of upping the power on all the machines
[02:16:55] <Jacky^> in the life
[02:17:32] <les> Well when I had no work I just spent time rebuilding machine tools
[02:17:47] <Jacky^> les: good
[02:17:54] <les> and building the cnc...that was $20k and one year full time
[02:18:09] <Jacky^> rebuild machine tools isnt a work ?
[02:18:19] <les> si I made about -$40k that year!!!
[02:18:30] <Jacky^> hehe..
[02:18:36] <bosone> night all ;)
[02:18:48] <les> I like to rebuild machine tools...but usually you cannot do it in a business
[02:19:05] <les> it takes to much time
[02:19:11] <Jacky^> :)
[02:19:16] <les> But I had some time
[02:19:52] <les> So I did stuff like scrape in my surface grinder to +/- 1 micron....better than new!
[02:19:58] <Jacky^> right this evening a was talking with my cousin about to build slides with a lathe
[02:20:28] <Jacky^> i see some people on eaby ..
[02:20:34] <Jacky^> eaby*
[02:20:37] <Jacky^> ops
[02:20:40] <les> haha
[02:20:41] <Jacky^> ebay !
[02:20:47] <Jacky^> hehe
[02:21:07] <Jacky^> it should not be so difficult
[02:21:19] <Jacky^> you know :P
[02:21:38] <les> Well I was looking at linear guides quote from ABBATECH
[02:21:50] <Jacky^> up 1 mt lenght they have a very high cost ..
[02:22:03] <Jacky^> 2-3 mt
[02:22:07] <les> less than $100 per meter
[02:22:12] <les> pretty cheap!
[02:22:18] <Jacky^> uhm
[02:22:22] <les> haha
[02:22:24] <Jymmm> les Um... try $142/meter
[02:22:40] <les> not for me jymmm
[02:22:50] <Jymmm> les what size? 20mm
[02:22:57] <les> $90 or so for 20 mm
[02:23:05] <Jymmm> les what qty?
[02:23:11] <les> one!
[02:23:19] <Jymmm> les direct?
[02:23:23] <les> yes
[02:23:28] <Jymmm> les not bad
[02:23:41] <les> even beats HIWIN
[02:24:02] <les> the blocks are $52 each with light preload
[02:24:07] <Jymmm> les I think I was the one that told you about them
[02:24:21] <les> yes in fact you were Jymmm
[02:24:39] <les> So I called them
[02:25:16] <Jymmm> les I musta called him on a bad day =) He quoted me $52 for the slides, and $142/m but I was asking for ones twoss too.
[02:25:40] <les> The fact that I have an ITW fellowship helps I think. And I always mention that.....
[02:26:01] <les> I specify automation for a $10 billion dollar company
[02:26:02] <Jymmm> les His english was fairly broken.
[02:26:23] <Jacky^> Jymmm: like the mine ? :P
[02:26:24] <les> ...I think I get a little better prices even for small quantity
[02:26:32] <les> at least onthe first order
[02:26:57] <les> jacky your english is fine
[02:27:07] <les> better than my itialian
[02:27:07] <Jacky^> doh ... :)
[02:27:28] <Jymmm> les Well, I know what you want to build, but I think you still could do well on tabletop size.
[02:27:33] <Jacky^> oh.. that beacuse you wont came in italy :)
[02:27:42] <Jymmm> italy is icky!
[02:27:48] <Jacky^> hehe
[02:27:51] <Jymmm> =)
[02:27:56] <les> I would like to...closest I got was Nice.
[02:28:04] <les> Used to be Italuy!
[02:28:06] <Jacky^> italy was a beautiful place
[02:28:07] <les> oops
[02:28:20] <Jacky^> before 1 minister berlusconi... :(
[02:28:31] <Jacky^> but will change :)
[02:28:37] <les> heh
[02:28:47] <les> We have similar problems....
[02:28:56] <Jacky^> I know
[02:30:19] <les> Agh got a call from my aunt
[02:30:42] <Jacky^> good ?
[02:30:46] <les> "computer monitor won't come on"
[02:31:00] <les> She asked what it was.
[02:31:06] <Jacky^> the icons wont appear ??
[02:31:09] <les> I sais it was broken.
[02:31:09] <Jacky^> :)
[02:31:11] <les> hahah
[02:31:41] <les> Aw, sounds lile cmos mem got wiped. I'll go ther tommorow.
[02:31:50] <les> there
[02:32:13] <les> or just a bad connector or cord
[02:32:18] <les> or a card wiggle
[02:32:31] <Jacky^> yeah..
[02:33:34] <les> video prob on the MB, so cords or cmos I think.
[02:34:29] <les> Sometimes the backup battery fails...it is supposed to last the life of the machine.
[02:34:53] <Jacky^> in some case, i solved just removing the agp card and inserting again ..
[02:35:20] <les> yeah
[02:36:34] <les> Well otherwise...it's old. They need a new box.
[02:36:56] <les> I did their web site, and they have been getting a lot of sales on it.
[02:37:06] <les> They can afford a newe box.
[02:37:23] <les> want to see?
[02:37:42] <Jacky^> sure
[02:38:04] <les> http://www.broderickcrawfordart.com/
[02:38:38] <les> just a very simple site
[02:38:50] <les> so they can manage it themselves
[02:39:00] <les> I taught them how to ftp etc
[02:39:23] <les> and I set up credit card stuff for them
[02:39:30] <Jacky^> nice pictures
[02:39:47] <les> yes they are getting good prices
[02:40:00] <Jacky^> it's ok
[02:40:19] <les> the web hosting is $7/month
[02:41:03] <les> I bought them Sitespinner and showed them how to use it
[02:41:14] <les> so now they can do their own pages
[02:41:26] <Jacky^> yeah
[02:41:57] <Jacky^> i like simple sites
[02:42:12] <Jacky^> without java applet, banner, etc
[02:42:16] <les> yes I think it's important
[02:42:19] <Jacky^> just some nice picture
[02:42:24] <Jacky^> :)
[02:42:56] <les> and....no $100/month "management fee"
[02:43:19] <les> banners, applets, frames clutter up things
[02:43:26] <Jacky^> theorically the nice site shoulde be made in flash
[02:43:39] <les> well yes
[02:43:41] <Jacky^> but pratically, block notes it better :P
[02:44:10] <Jacky^> surfing pratically, peoples want clean pages
[02:44:14] <les> well Sitespinner writes nasty code but it works fine...and very easy to use
[02:44:14] <Jacky^> and fast
[02:45:10] <les> more than 10 seconds and you lose them
[02:45:19] <les> so fast is important
[02:45:39] <Jacky^> i like this guidelines for websites http://www.gnu.org/server/fsf-html-style-sheet.html
[02:45:52] <Jacky^> very simple
[02:46:57] <Jacky^> ah..
[02:47:04] <les> looking
[02:47:30] <Jacky^> i've a cousin that work for IVECO in north italy
[02:48:13] <Jacky^> he buid a part for trucks
[02:48:37] <les> right...I just googled it
[02:48:41] <Jacky^> i asked if some old machine will be removed
[02:48:55] <les> That is a good place to look
[02:48:59] <Jacky^> and i'm waiting for reply
[02:49:06] <Jacky^> Iveco ? yeah ..
[02:49:23] <Jacky^> he work the plastic parts
[02:49:28] <les> you know, most of my machine tools were FREE!!
[02:49:34] <Jacky^> :)
[02:50:18] <les> ITW has 600 divisions...so if a plant was changing equiptment I knew about it
[02:50:25] <les> oops
[02:50:44] <Jacky^> he will take some photos from there in some afternoon
[02:51:00] <les> They shut down a line because of labor cost...
[02:51:01] <Jacky^> i will show it tou you
[02:51:15] <les> and THROW AWAY the good tools as scrap!!
[02:51:23] <Jacky^> hehe
[02:51:30] <zwelch> * zwelch looks for directions to les's shop.. so he can break his knee caps for bragging about his good fortune ;)
[02:51:38] <les> haha
[02:52:09] <zwelch> but i believe you in your description of corporate insanity
[02:52:25] <zwelch> bookkeeping tricks should not make those kinds of practices common, and yet... :/
[02:52:41] <les> Well it's just deindusrialization
[02:53:12] <zwelch> * zwelch sighs
[02:53:20] <les> the tools are fully depreciated...I would just go to the plant manager and ask him if he wanted to make them go away
[02:53:48] <zwelch> yeah, i grok. i just wish i had the same kind of opportunities around here
[02:54:00] <les> boeing?
[02:54:15] <zwelch> HP is local; their printer division is still here
[02:54:32] <zwelch> used to be the home of their handheld division, before it was moved to australia in the late 90s
[02:54:44] <les> I did some stuff for HP in Sacramento a while back
[02:55:11] <les> shock mount analyssis for their server cabinets
[02:55:15] <les> oops
[02:55:35] <Jacky^> zwelch: HP Compaq are the same thing right ?
[02:55:38] <les> 5 foot drop
[02:55:43] <zwelch> Jacky^: they weren't back then
[02:55:53] <Jacky^> oh.. ok
[02:56:18] <zwelch> so, yah, that division had nothing to do with the ipaqs ;)
[02:56:45] <Jacky^> its new to me
[02:57:38] <zwelch> the folks i worked with at TDS were former HPers; the EE was the architect for the Omnibook 800 (iirc)... the one with the mouse that popped out of the right side and worked via fancy capacitice-resistance tricks
[02:58:11] <zwelch> it was mind boggling that HP cut that kind of talent out of their workforce
[02:58:22] <les> hmm
[02:58:33] <zwelch> * zwelch is always astounded by layoffs of engineering staff from big companies
[02:59:27] <les> Well I had pretty good luck...but finally decided I wanted to resign and start my own company
[02:59:44] <les> I still work for the old company as a consultant though.
[02:59:52] <Jacky^> les: good choice
[03:00:04] <les> I think so. I am happy.
[03:00:16] <zwelch> i would have, but i just couldn't/can't stand Windows anymore
[03:00:26] <les> And the old company now pays me twice as much!
[03:01:02] <zwelch> * zwelch focuses his envy back to rewriting the configure script
[03:01:49] <les> I made a bid to get a Linux based manufacturing automation system in a large plant but didn't get very far
[03:02:12] <Jacky^> its hard to start a business, everywere
[03:02:17] <les> yes
[03:02:26] <les> I have done it twicenow.
[03:02:52] <les> I started a division at the old company. I worked myself into a hospital!!
[03:02:53] <zwelch> les: the thing i've discovered is that most businesses don't care about Linux; they want a good solution that's cost effective. the fact that Linux powers the solution only matters to (us) zealots.
[03:03:52] <les> Some engineering staff (non linux) worry about any open source solution
[03:04:05] <les> even though their code is not open
[03:04:19] <les> I guess I am talking about emc
[03:04:45] <zwelch> if that's true, i'd wager it's because they understand the economy of maintenance costs
[03:04:53] <les> yeah
[03:05:09] <Jacky^> i think Linux can developed a t different levels, Red hat, Suse are good examples
[03:05:20] <les> yes
[03:06:01] <zwelch> i've found that some companies that can't equate the production of software with the production of tangible goods; they only see that the "reproduction" costs are "near zero"
[03:06:15] <les> Well my intent is to make sure there are some maintenance costs!!! But less than doze.
[03:06:44] <zwelch> well, "'doze" offers a whole other range of disadvantages over open source
[03:06:57] <les> Again. I'm not a developer. But I sell things!
[03:07:21] <zwelch> personally, i find being in complete control of the software running on my production systems to be under my complete control - and that i have the ability to audit every line of it
[03:07:40] <zwelch> companies can understand that ;)
[03:08:27] <les> Well, long ago, I set up my line with pcs running c programs in dos. This was before Linux.
[03:08:40] <zwelch> i can't really imagine ever trusting a windows machine to do hard real-time
[03:08:44] <les> It worked well.
[03:08:56] <zwelch> dos != windows....
[03:09:05] <Jacky^> yeah
[03:09:06] <les> right
[03:09:17] <les> dos could be soft RT
[03:09:45] <les> My programs were simple, but lots of math.
[03:09:46] <Jacky^> Jymmm tried something..
[03:10:25] <Jymmm> * Jymmm tried a lot of things
[03:10:33] <Jacky^> hehehe
[03:10:43] <les> You see, the only formal training I had in programming was FORTRAN.
[03:10:50] <Jymmm> right now, my patients are being tried
[03:11:09] <les> for what , murder?
[03:11:17] <les> doc?
[03:11:25] <Jymmm> les at least you didnt cay the C___L word
[03:11:35] <zwelch> * zwelch giggles
[03:11:40] <les> hahaha
[03:12:29] <Jacky^> Jymmm: give me your wishes ..
[03:13:33] <Jacky^> after 2 week of hard work, finally i get my machine up and running !
[03:13:37] <les> I know that...he wishes he was cutting $1000 signs right now!
[03:13:49] <les> wood signs
[03:13:54] <Jacky^> tomorrow.. i will start milling :P
[03:13:59] <les> really
[03:14:05] <Jacky^> :)
[03:14:42] <Jacky^> les: i don't know why.. but my circuit work !!
[03:14:46] <les> I got quotes from Colombo today for ATC spindles
[03:14:57] <Jacky^> lol
[03:15:11] <les> I have to make sure with ray that emc can do the plc functions
[03:15:26] <Jacky^> here's the pcb http://www.roboitalia.com/modules.php?name=coppermine&file=displayimage&album=lastup&cat=12775&pos=0
[03:15:29] <Jacky^> :P
[03:15:57] <Jacky^> fast optos 6n137 .. woow
[03:16:08] <les> 4 KW ATC electrospindle = $5400
[03:16:46] <les> where is Colombo? Milan?
[03:16:55] <Jacky^> no.. Genova
[03:17:01] <les> oh ok
[03:17:04] <Jacky^> theres a nice aquarium there
[03:17:14] <Jacky^> like the pacific aquarium
[03:17:28] <Jacky^> maybe better ..
[03:17:43] <Jacky^> ah !
[03:17:43] <les> There price is a little higher than HSD...not sure which is best
[03:17:46] <Jacky^> doh..
[03:17:56] <Jacky^> you mean.. colombo industry
[03:18:11] <les> yeah Colombo or HSD
[03:18:30] <Jymmm> Jacky^ two wekks?! Try two months!
[03:18:36] <Jacky^> Jymmm: :\
[03:18:57] <les> Time does seem to fly messing with cnc
[03:18:58] <Jymmm> les you hit it right on the head!
[03:19:16] <Jacky^> les: mmm cant find it
[03:19:23] <les> HSD?
[03:19:30] <Jacky^> nothing ..
[03:19:33] <Jymmm> les More like.... Time flies when messing with clueless ones
[03:19:49] <Jacky^> probably noth italy
[03:19:53] <Jacky^> no idea..
[03:19:57] <les> it is
[03:20:02] <les> let me check
[03:20:05] <Jymmm> les They sent me the wrong part THREE TIMES, emailed them and they are making it for the 4th time.
[03:20:13] <les> blah
[03:20:42] <Jymmm> les I have a box of parts they have made, each is numbered in how many times it's been replaced becasue they made HUGE human errors
[03:20:45] <Jacky^> Jymmm: give the phone number !
[03:20:54] <Jacky^> i will call immediatly
[03:21:17] <Jymmm> Jacky^ oh, I'm tired of calling... I must have at least 400 minutes of being on the phone.
[03:21:21] <les> http://www.cncsupport.co.uk/hsd_spindles.htm
[03:21:25] <Jymmm> maybe 800 minutes
[03:21:28] <Jacky^> Jymmm: really..
[03:21:30] <les> it is italian
[03:21:42] <Jymmm> Jacky^ Yes, and about 50 emails too.
[03:21:44] <Jacky^> let me know
[03:21:48] <les> that is the uk distributor
[03:21:57] <les> came up first in google
[03:23:39] <Jacky^> googling
[03:24:02] <les> I got my quote from HSD USA in florida
[03:24:51] <Jacky^> Jymmm: got a phone number ?
[03:25:05] <les> somewhere...let me check
[03:25:10] <Jacky^> ok
[03:25:40] <les> http://www.hsdusa.com/eng/home.htm
[03:28:07] <les> HSD 4kW electrospindle=$3800
[03:28:18] <les> colombo=5400
[03:28:25] <les> Perske much more
[03:28:38] <Jacky^> mmmmm
[03:28:44] <les> colombo is best known
[03:29:53] <les> anyway, you should feel better if you have to spend a bit on cnc stuff
[03:30:06] <les> you know others have to spend much more!
[03:30:14] <les> but this is industrial use
[03:32:14] <Jacky^> the only one I found in italy seem to be this: http://www.colomboimpianti.it/
[03:32:26] <Jacky^> i don't know if is the same ..
[03:33:17] <Jacky^> seem Milan
[03:35:34] <les> http://www.pdscolombo.com/brochure.htm
[03:36:54] <Jacky^> Carate Brianza, yes, is Milan
[03:36:57] <les> link is at the bottom
[03:37:26] <les> nice looking routers , huh?
[03:37:45] <Jacky^> http://www.elettromeccanicagcolombo.com/
[03:37:49] <les> right
[03:38:12] <Jacky^> ok
[03:38:29] <Jacky^> theres phone and fax
[03:39:07] <Jacky^> Jymmm: you ordered from there ?
[03:39:23] <les> Colombo price is high, but I think good quality and service
[03:39:42] <Jacky^> les: never tried
[03:39:51] <Jacky^> to be honest
[03:40:12] <les> I don't really know either...hsd or colombo
[03:40:28] <les> salesmen all say theirs is best od course
[03:40:35] <Jymmm> bastards
[03:40:38] <les> of course
[03:40:49] <Jacky^> Jymmm: what the prob ?
[03:40:57] <les> salesmen
[03:40:58] <Jymmm> human stupidity
[03:40:59] <les> but
[03:41:17] <les> jymmm is about to become one too...hahaha
[03:41:32] <Jymmm> what a salesman?
[03:41:38] <les> yeah.
[03:41:54] <Jymmm> maybye but I have ethics adn I'm not clueless
[03:42:08] <les> heh ok
[03:42:38] <Jymmm> and I take pride in my work
[03:42:48] <zwelch> * zwelch thinks that combination might instantly disqualify you
[03:42:56] <zwelch> ;)
[03:43:03] <Jymmm> it might, but I gotta try anyway
[03:43:06] <les> I like to think my stuff is good
[03:43:23] <zwelch> on topic, how many people here have built the emc2 package from source?
[03:43:28] <Jacky^> Jymmm: if you want i can call they
[03:43:32] <les> But I must admit I tend to think my stuff is the best on the planet
[03:43:33] <Jymmm> zwelch I dont have to embelish my wares.
[03:43:47] <zwelch> * zwelch is just poking fun
[03:43:57] <Jacky^> and ask what is happening
[03:44:10] <les> zwellch I have to stick with emc1...emc2 is experimental and not for production
[03:44:24] <zwelch> do you build it from source?
[03:44:32] <les> yeah
[03:44:54] <Jymmm> Jacky^ Not gonna do any good; at this point in the game talk is cheap.
[03:44:56] <zwelch> which rt do you use?
[03:45:06] <les> uh...RTAI
[03:45:13] <zwelch> okay, 3.x?
[03:45:37] <Jacky^> Jymmm: as you want, let me know
[03:45:40] <Jymmm> Jacky^: It's not a matter of communications, it's purely a matter of human error, LOTS AND LOTS OF HUMAN ERROR and ****ZERO**** QA
[03:45:44] <zwelch> les: doesn't matter really
[03:45:58] <Jacky^> Jymmm: i believe you ..
[03:46:06] <zwelch> when you build from source, do you specify --with-rtlinux or let it figure it out?
[03:46:14] <les> I forgot...not a LINUX guy....last build was when I altered some source in TP.c...I just used Paul's makefiles
[03:46:15] <zwelch> * zwelch is trying to improve the configure script
[03:46:17] <les> and rays
[03:46:28] <zwelch> ?
[03:46:41] <les> ray henry
[03:46:47] <Jymmm> Jacky^: They sent me the EXACT same and wrong replacement part three times. I'll have to see if they get it right the fourth time. and this was today too.
[03:47:00] <Jacky^> ouch
[03:47:01] <zwelch> les: okay, right, nevermind..
[03:47:07] <zwelch> configure is only emc2 :)
[03:47:11] <Jymmm> Jacky^ AND they have to machine each part to send to me.
[03:47:40] <Jacky^> :\
[03:47:43] <les> I just run the file and watch the gibberish on the screen...I don't know what's going on
[03:47:45] <zwelch> and, if you're representitive of the userbase, the build system behavior of emc2 is probably very much open to mutation
[03:47:53] <les> this ain't borland turbo c!
[03:48:02] <Jacky^> Jymmm: let me know how will end..
[03:48:02] <Jymmm> les yes it is
[03:48:20] <zwelch> heh, yeah, it can be a little obfuscated, but the principles haven't changed
[03:48:28] <les> right
[03:48:31] <Jymmm> Jacky^ Probably me making a 400 mile drive to them and return it in person with a shotgun in hand.
[03:48:43] <les> I do change source though...if it's math
[03:48:49] <zwelch> * zwelch nods
[03:48:51] <Jacky^> Jymmm: nahh
[03:48:56] <Jacky^> let me know and
[03:49:09] <Jacky^> i've a cousin near Milan, in brescia
[03:49:18] <Jacky^> he can go there
[03:49:18] <les> but linux and all...clueless
[03:49:47] <zwelch> yeah, i've spent years beating my brain against it; most if it has sunk in, but not without a little pain along the way
[03:49:52] <Jacky^> ok.. to late for me, 4:30 am :\
[03:49:56] <les> I can get around and find file and run scripts...that's about all
[03:49:56] <Jacky^> gnight
[03:50:03] <les> night
[03:50:09] <Jymmm> Jacky^ thanks
[03:50:11] <Jymmm> night
[03:50:17] <Jacky^> Jymmm: np,
[03:50:21] <Jacky^> let me know
[03:50:23] <Jacky^> bye
[03:50:33] <Jymmm> Jacky^ No need for 'The Family' yet
[03:51:45] <les> I generally stay away from the latest and greatest build...I need something rock solid
[03:51:51] <les> emc1 is that
[03:51:56] <les> it never crashes
[03:52:06] <Jymmm> O_o
[03:52:07] <les> and almost always compiles ok
[03:52:15] <Jymmm> les wanna bet?
[03:52:40] <les> I have never had a lock up in what...four years?
[03:52:45] <Jymmm> ok, not crash, but I've gotten it to lock up hard
[03:52:58] <Jymmm> kill -9 wouldn't even stop it
[03:53:18] <les> I can force it by setting servo times too fast
[03:53:19] <Jymmm> I tell ya, I dint find bugs, bugs find me
[03:54:08] <les> I often leave the controller on for weeks
[03:54:59] <les> Now EMC2 might get that way later, but I cannot push aroung 1000 lb+ at high speed with something experimental
[03:56:11] <les> I am all in favor of emc2 though, because much of the emc1 source is virtually undecypherable
[03:57:50] <zwelch> cool
[03:58:12] <les> really, it took YEARS to find the trajectory bug
[03:58:44] <les> and keep in mind the author of emc has come here from NIST!
[03:58:50] <cradek> it only takes years to find a bug if nobody cares
[03:58:56] <les> right
[03:59:02] <zwelch> the critical capacity of a production system warrants special precautions be taken with "stable" releases to ensure such bugs either never reach the customer or, once found, are never allowed to regress
[03:59:31] <les> well I was the only one pushing big iron at high speed for a while I think
[03:59:39] <zwelch> any the whole "many eyeballs.." saying only works with N eyeballs and N gets *really* large
[04:00:03] <les> heh
[04:00:27] <Jymmm> cradek: Does that mean you want to know how I got axis to hard lock the sytem ?
[04:00:52] <zwelch> les: should i read that last remark as "the only one willing to crash and rebuild a big/expensive machine"? :)
[04:01:28] <les> heh...lots of safety stuff...but lots of nevousness too!!
[04:02:03] <zwelch> * zwelch thinks the cost/benefit of CNC machine software control R&D might not look good to many businesses, if you don't have any "spare" machines ;)
[04:02:08] <les> and the non RT sim doesn't cut it as far as actual moving iron stuff
[04:02:09] <Jymmm> les thats what 1kw resistors are for
[04:02:51] <les> I have some of those!
[04:03:22] <zwelch> * zwelch has N of them in this very room! (where N is probably surprisingly large :/ )
[04:03:32] <les> 5.6KW of servo can make you mighty nervous.
[04:03:59] <zwelch> * zwelch wishes his analog electronics skills were better
[04:04:25] <les> yes 1 kW resistors...I think they are called "hair dryers"
[04:04:39] <zwelch> i can design all of the software digital components for a handheld computer, but analog is whole other (mysterious, to me) world
[04:05:00] <zwelch> * zwelch hopes everyone mentally adds the "and" where appropriate above
[04:05:10] <les> I do a lot of analog
[04:05:31] <zwelch> well, i can design everything that does 1/0 :)
[04:05:33] <les> I used to be an Audio r&d engineer
[04:05:53] <les> well hey...if it's fast enough...it's analog!
[04:06:51] <zwelch> okay, so i should qualify my statement: i can do (relatively low-speed) digital design ;)
[04:06:55] <zwelch> a "slow" handheld
[04:07:12] <zwelch> * zwelch knows things get crazy at high speeds
[04:07:19] <les> Tell you what...you do that Linux stuff and I'll deal with the complex conjugate pole pairs!
[04:07:27] <zwelch> * zwelch winks
[04:07:33] <les> heh
[04:07:53] <les> well off to bed for me
[04:07:57] <zwelch> * zwelch waves
[04:08:04] <les> nice to meet you zwelch
[04:08:16] <les> night jymmm
[05:44:25] <zwelch> * zwelch sighs at the cruft in the build system
[05:45:00] <zwelch> i'm cleaning up things as I go, but there is code in there for BDITNG that really shouldn't be
[08:47:03] <alex_joni> morning all
[08:47:15] <Jymmm> woof
[08:47:28] <alex_joni> yeah.. but don
[08:47:34] <alex_joni> don't get too excited
[08:48:09] <alex_joni> * alex_joni got home last night around 03 am :(
[08:48:24] <Jymmm> what was her name?
[08:48:35] <alex_joni> now why would I tell you
[08:48:36] <alex_joni> :D
[08:48:48] <alex_joni> anyways.. was away lately a lot
[08:48:49] <Jymmm> so I know who to avoid!
[08:49:01] <alex_joni> btw.. packing is iminent
[08:49:08] <alex_joni> * alex_joni leaves again today
[08:49:15] <Jymmm> to?
[08:49:21] <alex_joni> some town in .ro
[08:49:33] <Jymmm> ah
[08:49:36] <alex_joni> there's a big meeting this week
[08:49:56] <alex_joni> spraying / painting stuff
[08:50:06] <alex_joni> next week maybe to a customer
[08:50:17] <alex_joni> and after that I'll be available (maybe)
[08:52:06] <Jymmm> sounds icky
[08:52:17] <alex_joni> well .. not that bad
[08:52:24] <alex_joni> last week was nice
[08:52:28] <alex_joni> was in germany
[08:52:39] <alex_joni> and a couple of days austria
[08:53:26] <Jymmm> what do you get me?
[08:53:28] <Jymmm> did
[08:53:50] <alex_joni> I could tell you.. but then I would have to kill you
[08:54:12] <Jymmm> then dont tell me, just give it to me instead
[08:55:42] <alex_joni> heh
[08:55:42] <alex_joni> ok
[08:55:53] <alex_joni> on my next visit to the us
[08:56:42] <Jymmm> let me guess... US Customs has you on the "Oh hell no he aint getting in at all" list.
[08:57:30] <alex_joni> not me ;)
[08:57:34] <alex_joni> but the stuff I got you
[08:58:05] <Jymmm> liar
[08:58:34] <alex_joni> ;;)
[09:01:24] <Jymmm> just looking for shype hardware
[09:01:28] <Jymmm> skype
[09:07:48] <Jymmm> ok alex, I'm gonna call it a night. chat at ya later
[09:08:06] <alex_joni> night
[09:46:50] <zwelch> well, i've made some progress on getting emc2 to build for non-realtime systems
[09:47:00] <alex_joni> zwelch: cool
[09:47:18] <zwelch> the configure script and makefiles now progress to building src/rtapi/sim_rtapi.c
[09:47:30] <zwelch> but... the code has bit rot
[09:47:55] <zwelch> i.e. the API defined by .../rtapi.h is not what is implemented in sim_rtapi.c
[09:48:28] <zwelch> i've started to try and hack together something, but... i need to look at the other modules and figure out what magic they're doing
[09:49:04] <zwelch> (seeing as how the old code has *explicit* references to MAGIC constants, i know i don't want to be monkeying around in there without a good firm grip on a clue)
[09:49:29] <alex_joni> zwelch: I recomend bugging jmk about this
[09:49:44] <zwelch> but at least the build system has reached a point where it can be actively developed
[09:50:02] <zwelch> i.e. --disable-realtime is the default if you don't have any installed
[09:50:57] <zwelch> the unfortunate side effect is that a) the changes were "non-trivial" and b) i have no idea if the RT options still work ;)
[09:51:45] <alex_joni> define a) ;)
[09:51:57] <zwelch> the patch is big
[09:52:05] <zwelch> and "non-obvious"
[09:52:10] <alex_joni> I see.. touching a lot of common emc2 code?
[09:52:18] <alex_joni> or only rt-related stuff?
[09:52:20] <zwelch> no, it's the configure.in file
[09:52:25] <zwelch> build system related
[09:52:29] <alex_joni> bugger that ;)
[09:52:34] <alex_joni> as paul_c might say
[09:52:36] <zwelch> nothing that affects use, just compiling
[09:52:47] <zwelch> well, the current build system bites
[09:53:03] <alex_joni> the configure.in has been looked at by a few people (jmk, paul_c, myself, etc)
[09:53:08] <alex_joni> but most won't touch it..
[09:53:17] <alex_joni> so it's not smthg you should be very concearned
[09:53:24] <zwelch> i'm getting more and more tempted to start from scratch
[09:53:33] <zwelch> * zwelch did already in some places
[09:53:43] <alex_joni> kernel detection?
[09:53:48] <zwelch> the line count should go down
[09:53:52] <zwelch> * zwelch hopes
[09:53:54] <alex_joni> heh
[09:54:04] <alex_joni> that would be too good
[09:54:27] <zwelch> i've not messed with the kernel detection stuff, other than to make it work with the simulater (it uses /usr/src/linux, which should always point to *some* valid kernel source)
[09:54:32] <alex_joni> but.. as it's smthg not everyone touches (e.g. 2-3 people actually need to know what's in there)
[09:54:35] <zwelch> it was enough to pass my tests
[09:54:58] <alex_joni> I don't think the effort to start over is justified
[09:55:04] <zwelch> right, it's not important... except it's also the first thing that people experience when building it on a new platform
[09:55:16] <alex_joni> right ;)
[09:55:19] <alex_joni> let them bother
[09:55:24] <alex_joni> kidding
[09:55:28] <zwelch> it's critical in that case; which means widespread adoption is directly linked to having a painless build system
[09:55:30] <zwelch> :)
[09:55:46] <alex_joni> it's a nice idea..but I think there are a lot more urgent areas that need addressing
[09:55:51] <zwelch> * zwelch shrugs
[09:56:00] <zwelch> it all depends on your perspective, i suppose :)
[09:56:00] <alex_joni> dito
[09:56:08] <alex_joni> yeah
[09:56:19] <alex_joni> anyways.. I really need to finish the STG driver
[09:56:29] <alex_joni> unfortunately I can't seem to find time for that :(
[09:56:58] <alex_joni> if you want you can mail me the configure.in, I would like to look at it
[09:57:16] <zwelch> sure
[09:57:22] <zwelch> * zwelch wonders what the STG driver is
[09:57:29] <alex_joni> alex DOT joni AT robcon DOT ro
[09:57:36] <alex_joni> a driver for a servo card
[09:57:39] <alex_joni> pretty old card
[09:57:45] <zwelch> k, i'll send a tar shortly
[09:57:54] <zwelch> is the manufacturer still producing it?
[10:00:07] <zwelch> (also, can one find them on ebay at a reasonable/reduced price? ;) )
[10:01:31] <alex_joni> I think so
[10:02:04] <zwelch> and taht answers which question?
[10:02:20] <alex_joni> the thing is .. there are a lot of users who use the stg with emc1
[10:02:29] <alex_joni> and they need the driver to make the switch
[10:02:40] <alex_joni> I think you can find some stg boards on ebay
[10:02:48] <alex_joni> but only once in a while
[10:03:00] <alex_joni> there is a new one still in production
[10:03:02] <alex_joni> stg2
[10:03:08] <alex_joni> not very different from the stg1
[10:03:16] <alex_joni> and the driver will support both (one day)
[10:04:35] <zwelch> presumably you have an stg1?
[10:04:52] <zwelch> * zwelch knows from experience that it's awfully hard to debug a driver without hardware ;)
[10:04:54] <alex_joni> I got one ;)
[10:05:01] <alex_joni> eventually :D
[10:05:05] <alex_joni> thanks to mshaver
[10:05:18] <alex_joni> he sent me one to help me debug the driver
[10:05:32] <alex_joni> it's almost there.. still got to test the encoder interface and it should be ready
[10:05:45] <alex_joni> and some cleaning up in the sources.. but that's easy
[10:06:05] <alex_joni> check out the emc2 CVS .. it's in there (under hal/drivers/hal_stg.c & .h)
[10:10:06] <zwelch> okay, i'm about to send you a tar of four patches
[10:10:16] <zwelch> the three small ones are the key bits
[10:10:24] <zwelch> the big one is a diff of my whole repo
[10:10:38] <zwelch> (including configure and sim_rtapi.c changes)
[10:10:53] <zwelch> they were made with cvs diff, so they suck as far as patches go
[10:10:57] <zwelch> * zwelch shrugs
[10:13:16] <zwelch> sent... let me know what i've borked :)
[10:13:25] <zwelch> (which rt are you using, btw?)
[10:14:32] <alex_joni> rtai-3.2 iirc ;)
[10:14:39] <alex_joni> last I worked on bdi-4.20
[10:22:27] <alex_joni> * alex_joni goes away
[10:22:32] <alex_joni> catch you later.. I hope
[10:26:30] <zwelch> * zwelch just realized there probably bugs in the new configure code path for rtai
[10:26:41] <zwelch> too late to tell alex though ;)
[10:46:41] <zwelch> * zwelch waves to paul
[10:46:53] <paul_c> You still up ?
[10:46:57] <zwelch> i've managed to get the configure changes done
[10:47:15] <zwelch> and (as i had guessed) i reached a new stumbling block
[10:47:32] <paul_c> gonna mail the patches to the dev list ?
[10:47:36] <zwelch> <zwelch> the configure script and makefiles now progress to building src/rtapi/sim_rtapi.c
[10:47:37] <zwelch> <zwelch> but... the code has bit rot
[10:47:37] <zwelch> <zwelch> i.e. the API defined by .../rtapi.h is not what is implemented in sim_rtapi.c
[10:47:37] <zwelch> <zwelch> i've started to try and hack together something, but... i need to look at the other modules and figure out what magic they're doing
[10:47:37] <zwelch> <zwelch> (seeing as how the old code has *explicit* references to MAGIC constants, i know i don't want to be monkeying around in there without a good firm grip on a clue)
[10:48:32] <paul_c> gimme an example of a magic..
[10:48:48] <zwelch> i sent them to alex_joni a little bit ago, but i realized shortly after that there was still work to do
[10:49:07] <zwelch> okay, Let Me Quoteth From The Code:
[10:49:30] <zwelch> /* label as a valid task structure */
[10:49:30] <zwelch> /*! \todo FIXME - end of non-threadsafe window */
[10:49:30] <zwelch> task->magic = TASK_MAGIC;
[10:49:35] <zwelch> ;-p
[10:50:28] <zwelch> * zwelch isn't really afraid of such magic, but the API *has* changed ;)
[10:51:04] <zwelch> i'll need at least another day to figure out how to get the module in shape
[10:51:10] <paul_c> *_MAGIC is only used inside sim*.c
[10:52:07] <paul_c> and I've bugged JMK on a couple of occasions to clean up the code - Too much is duplicated between usr/RT & RTAI/RTLinux
[10:52:24] <zwelch> i was wondering that myself
[10:52:36] <paul_c> oh, and the use of _t suffixs really gets on my TITs
[10:52:37] <zwelch> i would probably have got around to suggesting that myself
[10:52:47] <zwelch> but... that's just a matter of a new common .c file, nothing big
[10:52:55] <zwelch> yeah? :)
[10:53:04] <zwelch> * zwelch shrugs
[10:53:21] <zwelch> i glanced at the CodingStyle, btw ...
[10:53:28] <zwelch> ... yes, people *do* read that file
[10:53:31] <zwelch> ;)
[10:53:52] <zwelch> (strange people that probably wear tin-foil hats on more than rare occasions....)
[10:54:19] <paul_c> Quoting from B. Stroustrup - The use of _t suffix should be reserved for system libraries.
[10:54:22] <zwelch> * zwelch wonders if paul_c remembers making that comment in the commit log in may ;)
[10:54:47] <zwelch> paul_c: interesting bit of idiomatic lore, that... i believe it, though i suppose
[10:54:58] <zwelch> define "system" library though... bwuhahahhaa
[10:55:12] <paul_c> Go read some of the hal driver code - and I do remember writing that comment.
[10:55:12] <zwelch> * zwelch loves playing devil's advocate just for fun - watch out
[10:55:39] <zwelch> * zwelch has had to read (And deal with) the very worst code
[10:55:52] <zwelch> usually, that means rewriting everything that can't be read until it can
[10:55:59] <paul_c> /op #emc paul_c
[10:56:12] <zwelch> (... after drinking several pints of guiness, mind)
[10:56:31] <paul_c> * paul_c hands zwelch a pint of old engine oil.
[11:04:30] <zwelch> anyway, i'm going to head off to bed; i'll see how far i can get tomorrow looking at the rtapi stuff
[11:04:57] <zwelch> i'll even start looking at putting together rtai and rtlinux kernels to use for testing those portions
[11:05:17] <zwelch> * zwelch is guessing the rtai and rtlinux patchsets stand little chance of playing nice together though
[11:05:40] <paul_c> mail the patches over & I can test some of it.
[11:06:21] <zwelch> okay, i'll put together a patch real quick
[11:18:45] <zwelch> right then
[11:18:55] <zwelch> where shall i send it? :)
[11:19:34] <paul_c> [nolog] bdi-emc at ntlworld.com
[11:19:56] <zwelch> hmm, i suppose there's no point discussing the changes i made now
[11:20:10] <zwelch> they are substantial
[11:20:10] <paul_c> you heading to bed ?
[11:20:15] <zwelch> yeah, i need to sleep
[11:20:22] <zwelch> and the scope is rather large
[11:20:33] <zwelch> lots of little details (but with software, that makes a world of difference)
[11:20:48] <zwelch> i'll just be curious to know how close i came to not breaking anything ;)
[11:21:15] <paul_c> to be honest, I'm not unduly bothered about changes to the build (config & makefile)
[11:21:18] <zwelch> oh, i also wanted to talk about removing the BDI special case :)
[11:21:31] <paul_c> as long as they are easy to understand and maintain.
[11:21:34] <zwelch> that's *got* to be solved outside of the configure script :)
[11:21:48] <zwelch> well, i'd like to think the script is more understandable than it was
[11:21:59] <zwelch> and i'm far from finished with what i *could* do
[11:22:19] <paul_c> Certain areas of the core code base, I will fight anyone on....
[11:22:35] <zwelch> e.g. the configure script needs a --with-dietlibc option, if i'm not mistaken
[11:23:21] <paul_c> the dietlib requirement is a special case for RT linkage.
[11:23:42] <zwelch> right, but it needs to be made user visible
[11:24:17] <zwelch> moreover, couldn't the same be done with the uClibc libm?
[11:24:33] <paul_c> I would disagree... But we can discuss the reasons for/against later.
[11:26:18] <paul_c> libm from uClibc (or dietlib) is all that the test is interested in.
[11:27:29] <zwelch> well, i would say that the current check needs to be rewritten to avoid a recursive "find"
[11:27:46] <zwelch> i removed the massive /usr scan entirely
[11:28:40] <zwelch> it stands to reason that either the *-config scripts need to be in the path, or the directory location needs to be specified using --with-*
[11:29:31] <zwelch> for the future, i was wondering if the rtlinux and rtai packages install .m4 scripts that we could use in the configure script to "outsource" the checks for those packages
[11:30:01] <zwelch> if not, we should look at factoring out our RT checks into reusable macros and sending them upstream to those maintainers for that very purpose
[11:30:14] <paul_c> dunno about rtlinux - Haven't touched that since RTAI support worked....
[11:30:32] <paul_c> RTAI does not ship with m4 macros.
[11:30:37] <zwelch> similarly, there are m4 libraries for,e.g., GTK
[11:30:49] <zwelch> * zwelch didn't look at that chunk of config code though
[11:31:09] <zwelch> adding a .m4 file isn't too big a thing
[11:31:34] <paul_c> afaik both rtai & rtlinux use a realtime-config to export assorted parameters.
[11:33:45] <zwelch> right, my patch works over some of that logic
[11:33:46] <zwelch> there are {rtai,realtime,rtl}-config, from what i can see
[11:33:46] <zwelch> (not having installed the packages, i only can go by what the code tells me ;) )
[11:33:46] <zwelch> but there could be AC_CHECK_FOR_{RTAI,RTLINUX} macros that encapsulate the bulk of the code for interacting with them and setting up RT* variables
[11:33:46] <paul_c> I suspect getting m4 into all RT variants & versions would be an up hill fight.
[11:34:03] <zwelch> well, it has to start with the package maintainers, and i can't imagine them turning down a patch ;)
[11:34:17] <zwelch> and it'd only be "future" versions
[11:34:34] <zwelch> but the future our versions could simply use the macros, even to detect the older versions
[11:34:55] <paul_c> Would you want m4 in all the older versions of RTAI or just the current dev branches ?
[11:34:59] <zwelch> (as we've got the code to do that now, we're just talking about moving it in the RT package and having them ship it)
[11:35:22] <zwelch> this is a "for the future" thing, and only for our conveinece
[11:35:48] <zwelch> it would also "future proof" us, to some extent, from future RT variants (e.g. fusion)
[11:36:14] <paul_c> tell you what.. Mail the rtai list with your suggestion (keeping the emc name out of the picture) and we'll see what the response is.
[11:36:14] <zwelch> i've structured my initial changes with such (fusion) in mind
[11:36:42] <zwelch> nah, if i know much, their response will be "show us the code"
[11:36:52] <zwelch> * zwelch would rather just send them a patch, after factoring out the appropriate code
[11:37:17] <zwelch> speaking of which, did you get my patch? :)
[11:37:38] <zwelch> * zwelch hates sending unfinished patches, but you asked for it ;)
[11:37:41] <paul_c> eeks... 18.7K
[11:37:47] <zwelch> what?
[11:38:00] <zwelch> i told you it's rough :)
[11:38:35] <paul_c> I'll apply and look it over while you get some sleep.
[11:38:42] <zwelch> i need to clean it up so the diff is optimal; for me, style conformity comes after functional comformity ;)
[11:39:43] <zwelch> k; i'll catch you on the flip-side
[11:40:25] <zwelch> * zwelch goes into super stealth lurk mode, where no one can see him... (oops... is this thing still on....)
[11:48:42] <paul_c> Most of the changes proposed look OK....
[13:56:12] <les> good morning all
[14:10:23] <jepler> hi les
[14:12:56] <les> hi jeff
[14:13:28] <les> reviewing patents today
[14:13:30] <les> blech
[14:14:35] <les> ...technical advisor in a corporate aquisition
[14:14:41] <les> pays the bills.
[14:18:10] <paul_c> edit the name @ the top & replace it with mine ;-f
[14:18:24] <les> oh, ok!
[14:18:27] <les> heh
[14:18:53] <les> oh and have to go fix michelle's computer
[14:19:01] <les> you remember her...
[14:19:08] <CIA-8> 03paul_c * 10emc2/src/configure.in:
[14:19:09] <CIA-8> AC_MSG_NOTICE throws an error with autoconf2.13 - Minimum requirement should be
[14:19:09] <CIA-8> version 2.53. This means anyone using RH7.3 or earlier will *have* to upgrade if
[14:19:09] <CIA-8> they want to regenerate configure from configure.in. N.B. This does NOT mean an
[14:19:09] <CIA-8> upgrade is required to run configure.
[14:19:16] <paul_c> your great aunt..
[14:19:29] <les> " my monitor does not work"
[14:19:37] <les> how do you know?
[14:19:44] <paul_c> "Is it plugged in"
[14:19:51] <les> "it says so on the screen"
[14:20:27] <les> ah....bad shutdown and trashed drivers....heh
[14:20:33] <paul_c> She's not a blonde is she ?
[14:20:42] <les> yes...remember?
[14:20:47] <les> haha
[14:21:14] <les> oh also...no backup whatsoever
[14:21:20] <les> no burner
[14:21:38] <les> I have to show her how to back up on the web host
[14:21:39] <paul_c> Install Linux for her.
[14:21:51] <les> I think I got them 30 gig or something on there
[14:22:13] <les> heh...yeah bring a bdi....
[14:23:54] <les> I have to go and get a nut driver...might have to do the old card wiggle.
[15:17:51] <vq_> vq_ is now known as ValarQ
[16:00:58] <les> oh that repair job was rough....turn off and turn on.
[16:01:00] <les> haha
[16:09:58] <mrallen> i'm having a tough time finding a supplier for greenfield end mills .. they were recommended to me without pointers
[16:15:55] <les> greenfield is now kennametal...try that
[16:16:06] <les> I assume you need thread mills?
[16:22:37] <mrallen> just standard end mills - replacements for the 3/8" double-ended ones from sherline -- which are wearing out on me
[17:08:49] <les> sorry was away....3/8 duble 4 flute hss end mills are about $9 each in MSC for good import ones
[17:08:54] <les> double
[17:09:59] <les> singles are about 5 bucks
[17:10:20] <les> I use them a lot
[17:10:26] <les> they work fine
[17:11:33] <paul_c> One thing to bear in mind when machining wood - Some have a high silica content which is very tough on cutters.
[17:12:02] <les> some do paul...teak comes to mind
[17:12:42] <les> metal cutting end mills have the wrong rake and clearance for most wood cutting too
[17:14:01] <les> although they will work in a pinch
[17:15:07] <les> In sets the metalworking Ti coated end mils end up costing me $3 or so each...I don't even bother sharpening them!
[17:15:43] <les> I do sharpen the carbide ones though.
[17:17:45] <les> oh paul....great aunt's computer....turn off and on fixed.
[17:18:01] <les> There was some file corruption though.
[17:18:32] <les> Some config stufff...musta locked up.
[17:18:44] <les> heh
[17:19:50] <anonimasu> iab
[17:19:56] <les> hi
[17:20:24] <anonimasu> hello
[17:20:36] <anonimasu> * anonimasu has been riding racecar today
[17:20:45] <les> I am writing patent disclosure right now and I am bored out of my gourd
[17:20:48] <anonimasu> :/
[17:20:54] <anonimasu> 0-100 in 2.3 sec
[17:20:59] <les> racecar...neat
[17:21:03] <anonimasu> yeah
[17:21:10] <les> ????? yikes
[17:21:13] <paul_c> road car... OTT
[17:21:26] <anonimasu> paul_c: what?
[17:21:38] <anonimasu> it's for the largest class of rallycross there is..
[17:21:53] <paul_c> 0-100KPH for a road car is well over the top in performance.
[17:22:10] <anonimasu> 0-100kmh.. that is..
[17:22:13] <anonimasu> yeah
[17:22:19] <anonimasu> kills any WRC car easily :D
[17:22:32] <anonimasu> * anonimasu is kind of in extacy still
[17:22:44] <les> I guess
[17:22:50] <les> how many g is that?
[17:22:56] <anonimasu> I have no idea
[17:23:05] <les> (too lazy to calculate)
[17:23:08] <anonimasu> on the long first corner straight, before the driver(which is handicapped hit the brakes)
[17:23:17] <anonimasu> the speed were about 170kmh
[17:23:30] <anonimasu> ;)
[17:23:31] <les> wow
[17:23:45] <anonimasu> that's what's even more cool
[17:25:12] <anonimasu> it's like there was no end to how much power that engine would deliver
[17:25:45] <anonimasu> even though it was awd.. it would slide :)
[17:25:59] <anonimasu> les: are your patents going through?
[17:26:43] <les> 1.23 g
[17:26:54] <les> yeah I guess
[17:27:09] <les> it's just boring to write it up
[17:27:31] <anonimasu> :)
[17:27:45] <anonimasu> les: is that a good acceleration?
[17:27:48] <anonimasu> ^_^
[17:28:10] <les> so you were accelerating down the road 25% faster than falling!
[17:28:23] <les> i'd say that was good
[17:28:53] <anonimasu> nice
[17:29:37] <les> and 125% of your body weight pressing you back in the seat
[17:30:34] <les> 123%
[17:30:40] <les> heh
[17:30:51] <anonimasu> I am amazed that the gearbox lasts..
[17:30:58] <anonimasu> for a start like that
[17:32:08] <les> yeah
[17:32:42] <les> I am so bored...I am going to go out and do something
[17:32:47] <anonimasu> hehe
[17:32:53] <les> drive the tractor or something
[17:33:00] <anonimasu> lol
[17:33:02] <les> push over some trees?
[17:33:05] <anonimasu> is it still shiny?
[17:33:20] <anonimasu> les: is the storm comming over there?
[17:33:22] <les> sure. All my stuff stays shiny forever
[17:33:30] <les> no not here
[17:33:34] <anonimasu> nice
[17:33:43] <anonimasu> I thought you had storms all the time
[17:33:51] <les> 25c, no wind, no clouds
[17:34:11] <les> The weather broke...it's fall weather now
[17:34:20] <anonimasu> nice
[17:34:31] <anonimasu> or well boring
[17:34:36] <les> yeah
[17:34:37] <anonimasu> * anonimasu dosent like fall too much
[17:34:43] <anonimasu> I like the winter much much more
[17:35:51] <les> well I'm off...finished this boring stuff. TRACTOR!
[17:36:00] <anonimasu> les: laters
[17:36:02] <anonimasu> :)
[17:36:03] <les> k
[17:36:05] <anonimasu> good luck
[17:49:29] <mrallen> what are good end mills for wood?
[17:56:05] <mrallen> axis crashed out on me -- a lot of python object errors
[17:57:36] <cradek> make sure you save all the output
[17:57:50] <mrallen> i will - have to run at the moment - back in a bit
[17:57:52] <cradek> and remember what you were doing right when it crashed
[17:58:03] <paul_c> apt-get install --reinstall emc-modules-2.6.10-adeos
[17:58:06] <mrallen> it crashed my part and i noticed because of the noies
[17:58:15] <cradek> ouch
[17:58:21] <mrallen> back in 45
[18:17:54] <Jymmm> Morning Ladies
[18:38:43] <les> just taking a break from bush hogging
[18:39:48] <les> oh and for mrallen...when he comes back....for wood a spiral carbide 2 flute router bit would be good....
[18:40:08] <les> one would prob last you a lifetime.
[18:40:31] <les> well back to the tractor.
[18:43:16] <Jymmm> what in gawds green earth is bush hogging?
[18:43:52] <Jymmm> * Jymmm thinks it must be some hillbilly thing!
[18:44:31] <paul_c> you really don't wanna know..
[18:44:47] <Jymmm> I don't think I do =)
[18:44:48] <paul_c> in most states, it would be a crime.
[18:45:43] <Jymmm> and illegal in 39 countries
[18:45:56] <paul_c> 119
[18:47:14] <paul_c> You might get away with it in Australia..
[18:47:40] <Jymmm> You can get away with lots of things in .au
[18:55:46] <mrallen> here's what happened as near as i can tell:
[18:56:11] <mrallen> i've had axis open for 3 days running 4 different parts. this morning i was running the same part i ran last night -- same session -- same loaded file
[18:56:24] <mrallen> i watched it start to cut just fine -- first layer every thing looks OK
[18:56:38] <mrallen> i come back in 20 minutes and the Z is 1.5 inches up and cutting air
[18:56:47] <mrallen> 'odd' i say to myself
[18:57:07] <mrallen> the display shows the tool cutting the right spot, but the Z is off. the part is 1/2 done
[18:57:18] <mrallen> so i proceed to do something really stupid
[18:57:55] <mrallen> i stop the job, rehome the X,Y and Z and re-run the job.. again it starts off just fine and i walk away
[18:58:10] <mrallen> 10 minutes later i hear a godawful noise and run in to stop it
[18:58:42] <mrallen> the Z axis this time has plunged down about 0.75" below where it shows it's cutting on the display
[18:59:03] <mrallen> the console shows python errors but it's overrun the console buffer
[18:59:12] <mrallen> next time i will log the console to a file
[19:00:51] <mrallen> les: dowcut or upcut? i'm looking on mscdirect. about 1000 options. too much for a newbie like me. i need guidance.
[19:01:09] <les> hi just got back in.
[19:01:21] <les> what do you want to do?
[19:01:28] <les> make a mortice?
[19:01:32] <mrallen> avbout the router bit for wood
[19:01:43] <les> yes...but what kind of cut?
[19:02:06] <mrallen> flat
[19:02:40] <les> spiral downcut prevents splintering of the top surface (the spindle side)
[19:02:58] <Jymmm> les : I didn't know that
[19:03:13] <les> spiral up cut prevents splintering if the cut is all the way through on the other side
[19:03:31] <les> upcut has better chip removal too
[19:03:44] <mrallen> aha. which cut direction? LH or RH
[19:03:45] <Jymmm> define chip removal por favor
[19:04:06] <les> We almost always use downcut...because it forces the work away from the spindle
[19:04:21] <les> upcut pulls the chips out
[19:04:28] <Jymmm> ah
[19:04:32] <les> downcut tends to push them down
[19:04:44] <mrallen> http://www1.mscdirect.com/CGI/NNSRIT?PMPXNO=9144742&PMT4NO=1484400
[19:04:50] <mrallen> will that work with the sherline?
[19:05:08] <mrallen> i do not see a flat spot
[19:05:22] <les> looking
[19:06:44] <les> yeah that sort of thing
[19:06:56] <mrallen> right handed or left handed?
[19:06:59] <les> must check rotation direction of your spindle
[19:07:04] <mrallen> clockwise
[19:07:17] <les> facing the spindle?
[19:07:27] <mrallen> looking down
[19:07:32] <les> ok
[19:07:36] <les> let's see
[19:07:55] <mrallen> regarding the flat spot (or lack thereof) ...
[19:08:54] <les> that on on the page seems to be the wrong way for you...and as far as the flat spot...that is called a Weldon shank
[19:09:09] <les> not used on routers due to balance issues
[19:09:51] <les> but you can mount with a set screw...no burr will kick up...it's too hard
[19:09:52] <mrallen> are the stock sherline end mills left or right handed?
[19:10:13] <les> gosh I don't know...don't have one
[19:10:27] <mrallen> cradek: did you see what i wrote above?
[19:10:27] <anonimasu> hehe
[19:10:30] <Jymmm> les would never be seen with a sherline!
[19:10:36] <les> let me run to the shop and get some of mine here
[19:10:42] <anonimasu> Jymmm: probably at one point
[19:12:48] <alex_joni> good evening gents
[19:12:52] <les> ok my downcuts are clockwise looking down
[19:13:06] <anonimasu> evening alex
[19:13:13] <les> I don't care much for the terms right hand and left hand rotation
[19:13:20] <alex_joni> hey anders
[19:13:21] <alex_joni> hello les
[19:13:26] <les> hi alex
[19:13:42] <alex_joni> * alex_joni got home last night
[19:13:48] <mrallen> les: the web site wants to know left or right. it's not more descriptive than that
[19:14:28] <les> mrallen you might try sites like whiteside, amana, and cmt
[19:14:34] <les> better prices too
[19:15:07] <mrallen> les: links? google is not very illuminating without more specifics
[19:15:14] <les> just a sec
[19:15:26] <alex_joni> anything new lately?
[19:16:30] <les> http://www.amanatool.com/bits-fv/46100.html
[19:17:06] <les> not for me alex. I did boring patent disclosures in the morning, then bush hogged.
[19:17:25] <les> If you want to know what that is ask Paul or Jymmm
[19:17:33] <les> heh
[19:17:34] <mrallen> les: thanks
[19:17:36] <alex_joni> alex_joni is now known as alex_joni_afk
[19:17:51] <les> yw
[19:21:26] <les> These bits are not cheap, but for you it will prob stay sharp forever.
[19:21:47] <les> For us, they last 4 hours...then we diamond resharpen
[19:23:16] <les> Now that emc can go faster, we hope to have higher chip load and longer sharpen intervals
[19:24:42] <Jymmm> I hate that term 'chip load'
[19:25:14] <Jymmm> les so what do you use motly for signage.... up or down cut?
[19:25:18] <Jymmm> mostly
[19:25:20] <anonimasu> heh
[19:25:25] <anonimasu> Jymmm: chipload is a nice thing
[19:25:41] <Jymmm> anonimasu Just the term itself
[19:25:55] <anonimasu> Jymmm: why?
[19:25:57] <les> down for area clearance
[19:26:06] <anonimasu> Jymmm: tried cutting anything?
[19:26:09] <Jymmm> clearnace?
[19:26:17] <les> keeps the work from being sucked up toward the cutter
[19:26:42] <anonimasu> or why are you referring to chipload as a bad thing?
[19:26:44] <les> yeah...um...clearnace haha
[19:26:48] <Jymmm> les Ah, I liked the fact that it's a cleaner surface cut.
[19:27:08] <les> yes they cut beautifully on wood and plastics
[19:27:16] <Jymmm> * Jymmm smacks anonimasu with a pot of coffee!
[19:27:39] <anonimasu> Jymmm: Why are you slapping me?
[19:27:46] <anonimasu> and why is chipload something you hate?
[19:27:49] <Jymmm> anonimasu so you'll wake up
[19:28:06] <Jymmm> anonimasu <Jymmm> anonimasu Just the term itself
[19:28:27] <les> like left and right hand rotation with me?
[19:28:37] <anonimasu> what?!
[19:28:41] <Jymmm> les I came up with a idea for a tool changer for my router
[19:28:50] <les> that is a term I don't like
[19:29:08] <les> jymmm...me too...spend $6000
[19:29:11] <Jymmm> les would you prefer CW CCW ?
[19:29:18] <les> yeah
[19:29:24] <Jymmm> les Oh no, maybe $500
[19:29:33] <les> ok...
[19:30:00] <les> trained mokey with collet wrenches?
[19:30:08] <Jymmm> les very simple principal, doubt would work for your big honking 10hp spindles
[19:30:08] <les> monkey
[19:30:22] <Jymmm> les nah, that's what we have anonimasu for.
[19:30:29] <les> you know bush hogging makes me mis spell
[19:30:40] <Jymmm> damn hillbilly!
[19:30:49] <les> hahaha
[19:32:25] <anonimasu> Jymmm: still waiting for parts?
[19:32:36] <Jymmm> les based on the concept of this http://www.copshop.com/Merchant2/graphics/00000001/keychain.jpg
[19:32:47] <les> looking
[19:33:04] <Jymmm> anonimasu The part I rcvd Monday was the EXACT same (and wrong ) part I was replacing, so yes.
[19:33:11] <les> ???
[19:33:57] <Jymmm> les It's a pull-apart keychain. You push in on the top and they release. there are two ball bearing in the tube that keeps them togther
[19:34:58] <Jymmm> les very simular to how airline fittings are
[19:35:09] <Jymmm> http://www.blackburn.edu/Store/admin/images/closeup_images/pull_apart_keychain.jpg
[19:35:11] <les> yes...
[19:35:42] <les> I am still debating atc vs palletizing
[19:35:56] <les> must do a time/motion study
[19:35:57] <Jymmm> American brand padlocks use ball bearings too, seems very effective.
[19:36:38] <Jymmm> les you dont get it, do you?
[19:36:42] <Jymmm> my idea that is
[19:36:48] <les> jacobs tried to make some kind of snap thing...a router quick change
[19:36:52] <les> but it slipped
[19:36:58] <Jymmm> slipped, how?
[19:37:03] <les> they took it off the market I think
[19:37:21] <les> bit slipped out while cutting
[19:37:46] <Jymmm> Ah, well I couldn't see how that would happen here.
[19:38:02] <les> with a ball detent?
[19:38:15] <Jymmm> yep
[19:38:40] <les> I tighten the crap out of my collets...slipped bits are real, real bad for cnc!
[19:39:22] <les> also, the collets last only a couple weeks
[19:39:35] <Jymmm> you mena your collets?
[19:39:38] <Jymmm> mean
[19:39:54] <les> the PC ones yes
[19:40:11] <Jymmm> PC == Pretty Crappy?
[19:40:33] <les> good tools, I am just beating them up too much
[19:40:56] <Jymmm> why?
[19:41:15] <les> running them on a cnc production machine
[19:41:18] <Jymmm> is it abuse? underrated? poor performance?
[19:41:38] <Jymmm> oh... using handtools on a automated machine you mena
[19:41:59] <Jymmm> err manual tools
[19:42:14] <les> actually they go pretty well. Turned thousands of pounds of cherry into sawdust, and the bearingd and brushes are still ok!
[19:42:48] <les> the PC 3 1/4 router motor.
[19:42:54] <Jymmm> les btw... my "Craftsman Prof 2.24HP variable speed soft start" router is really a Bosch.
[19:43:34] <les> Fine for signs...just not for continuous production like the calls
[19:44:18] <les> The Colombo router I was quoted on is 72 dba at 1 meter
[19:44:27] <les> even the 10 hp ones!
[19:44:38] <les> that is super quiet
[19:45:20] <Jymmm> I had a question about that.... If I have gcode to make a sign that's 24" x 24", then later on I need to make the same sign but only 12" x 12". Is there a scaling method within gcode somewhere or would I need to regenerate the gcode all over again?
[19:45:35] <les> hmmm
[19:45:56] <les> I scale it in the cam
[19:46:15] <les> you could play with the ini scale factors too
[19:46:15] <Jymmm> Ok, so gcode is absolute then.
[19:46:36] <Jymmm> like a PDF
[19:46:44] <cradek> Jymmm: if you create it yourself, you can use a # var as scale
[19:46:52] <cradek> ttt does that
[19:47:02] <les> hmm
[19:47:23] <cradek> #1=1.0
[19:47:37] <cradek> g1x[1.234*#1]y[2.345*#1]
[19:47:40] <Jymmm> cradek: Ok, so if on the 24" sign I used a .25" cutter, I could also setup a var to say I'm using a 0.125" cutter instead as well?
[19:47:40] <cradek> or whatever the format is
[19:48:13] <cradek> that depends completely on what your gcode does
[19:48:17] <cradek> it's a machining question, not a gcode question
[19:48:33] <anonimasu> depends on stepover and stuff..
[19:48:39] <anonimasu> if you are just carving single line fonts yes..
[19:48:40] <Jymmm> cradek: Well I'm talking scaling a sign from 24" sq to 12" square
[19:48:42] <jepler> EMC has "cutter radius compensation"
[19:48:57] <cradek> but you don't understand
[19:49:05] <cradek> the answer to your question depends on how the sign is cut
[19:49:18] <cradek> if you're carving out the inside or outside of letters vs. a line font vs. ??
[19:49:19] <Jymmm> * Jymmm thinks cradek and jepler are married or at least living together.... they both show up at thae same time =)
[19:49:54] <les> heh
[19:50:20] <Jymmm> cradek: Ok, so gcode is more like pdf in the fact that it's a final uneditable output (so to speak)?
[19:50:24] <les> for area clearance with signs I almost always climb cut, inside or out
[19:51:01] <jepler> Jymmm: we just work together
[19:51:27] <Jymmm> jepler =)
[19:52:04] <Jymmm> cradek I'm just trying to understand the limitations of gcode and how far I can manipulate it is all.
[19:52:24] <cradek> sounds like you haven't worked with it at all yet
[19:52:30] <cradek> maybe you should start by writing some by hand
[19:52:46] <Jymmm> cradek getting there =)
[19:53:07] <Jymmm> I was just wondering if it's more like PS or PDF, sounds more like PDF to me.
[19:53:54] <Jymmm> you can do some neat things in PS or SVG
[19:54:24] <Jymmm> SVG even has animation capabilities, but we won't go there.
[19:59:08] <les> get a cam demonstration program, draw some stuff, and look at the code it makes
[19:59:15] <les> fastest way to learn
[19:59:31] <les> again millwrite 2000 is great
[19:59:37] <les> 30 day free trial?
[20:00:58] <les> Anyway...having bush hogged everything in sight, I now need to go to walmart and get office supplies
[20:01:06] <les> I hate stores and shopping
[20:01:55] <paul_c> Use the tractor & hayter
[20:02:49] <les> he yeah
[20:02:55] <les> http://members.aol.com/m9685123/M1.htm
[20:03:23] <les> later
[20:06:34] <Jymmm> laters les
[20:07:00] <Jymmm> cradek Eh, I was just thinking I might be able to automate some things is all.
[20:21:33] <alex_joni_afk> alex_joni_afk is now known as alex_joni
[20:22:51] <Jymmm> hey alex_joni!!!
[20:23:08] <alex_joni> 'lo
[20:23:57] <Jymmm> alex_joni all packed and ready to go?
[20:24:16] <alex_joni> not really
[20:24:26] <alex_joni> still toying with the idea of packing
[20:24:36] <alex_joni> I didn't even finish unpacking ;)
[20:25:00] <Jymmm> alex_joni well spray some disinfectant in the suitcase and go on with life
[20:25:18] <alex_joni> heh
[20:25:24] <alex_joni> I need clean shirts
[20:25:48] <Jymmm> tuen them inside out
[20:25:50] <Jymmm> turn
[20:26:00] <paul_c> wear jumpers
[20:26:21] <paul_c> or use the birthday suit.
[20:26:41] <Jymmm> paul_c want him to scare everyone? It's not halloween yet
[20:27:01] <alex_joni> right Jymmm
[20:27:32] <Jymmm> alex_joni Oh come on now... just pulling your chain a bit =)
[20:28:01] <Jymmm> Why to I alwys comfuse tkemc with minimill?!
[20:28:03] <alex_joni> heh
[20:28:20] <alex_joni> anyways.. was away a bit lately
[20:28:24] <alex_joni> and it'll continue :(
[20:28:25] <Jymmm> Haven't done it in a while... but I LOVE scarring the shit out of ppl
[20:28:46] <Jymmm> you get better reactions from the adults than you do the kids.
[20:32:15] <Jymmm> I'll blend in to the surroundings just behind where the parents wait for their kids to get candy and just step out very quietly behind them. They are either talking with each other or focused on their kids. When they turn to leave I'm standing right there with 7 ft sicle in hand, glow-in-the-dark gloves/makeup, and the most hidous maks I can find. Scare the SHIT out of them every time. and all I'm doing is standing there... I dont make a jerk reaction or
[20:33:10] <Jymmm> I guess it doens't hurt that I'm 6'4" plus wearing black boots either =)
[20:35:01] <alex_joni> heh
[20:37:21] <paul_c> http://www.zatarain.com/products/product.php/34/Dirty_Rice/Dirty_Brown_Rice - Just add water...
[20:39:49] <cradek> man, none of this stuff looks good
[20:41:22] <cradek> pureed shrimp?
[20:41:27] <cradek> ugh.
[20:50:18] <alex_joni> brb
[20:53:16] <zwelch> * zwelch yawns and peers at paul_c
[20:53:51] <zwelch> did i break anything?
[20:56:38] <paul_c> yup
[20:58:06] <paul_c> holy bovine...
[20:58:31] <paul_c> Fusion latency is still up around 40-45uSec
[21:01:21] <zwelch> paul_c: so should i start crawling through every line, or do i get a clue?
[21:01:53] <Jymmm> * Jymmm hands zwelch his knee pads...
[21:02:14] <paul_c> um... sorry... Was looking at something else.
[21:02:17] <Jymmm> NO CLUES FOR YOU!
[21:03:21] <paul_c> Unless you plan on writing a kernel space sim, don't want to be using kernel headers in usr space.
[21:04:58] <zwelch> i don't see where i am pulling in the kernel headers right now
[21:05:12] <zwelch> i.e. how did my changes break *that*? :)
[21:05:59] <paul_c> configure is looking for /usr/src/linux/include/linux/version.h
[21:06:21] <zwelch> well, that was my work around for the configure script
[21:06:33] <zwelch> yes, i that does need to be checked only if rt is enabled
[21:06:49] <zwelch> that broke your compile?
[21:06:58] <paul_c> broke configure
[21:07:06] <zwelch> configure/compile... whatever ;
[21:07:13] <zwelch> * zwelch sighs
[21:07:33] <zwelch> that is easy enough to remove
[21:07:42] <paul_c> symantics, but it narrows the breakage to a specific area
[21:25:23] <zwelch> paul_c: well, my fix for that one is a simple if to wrap around that chunk of the configure script, only running it if building the sim
[21:28:44] <paul_c> the other minor prob is the automatic search for the RT configs
[21:29:32] <mrallen> cradek: http://mrallen.com/axis.error
[21:30:11] <zwelch> paul_c: did i bork the new code? or did the old code find other (strange) installations
[21:30:36] <paul_c> the search doesn't.
[21:31:21] <zwelch> so a) there's a syntax bug, the commands don't run or 2) there's a semantic bug, and the script isn't found
[21:31:23] <paul_c> i.e. ./configure without any options should find rtai-config if it is present in the usual locations.
[21:31:37] <zwelch> and the "usual location" is where?
[21:31:51] <alex_joni> zwelch: a list of common locations
[21:31:58] <alex_joni> taken from various rtai flavours
[21:32:04] <alex_joni> like: /usr/realtime
[21:32:12] <paul_c> The RTAI default is /usr/realtime
[21:32:44] <paul_c> but it could be /usr/lib/realtime, /usr/bin, or a number of other places.
[21:32:47] <alex_joni> but also other posibilities under /usr usually (like /usr/src/rtai , /usr/local/realtime, etc)
[21:32:49] <zwelch> well, okay, i can deal with a list of choices... what i could not (and will not) deal with is a "find" command in my /usr directory -- when i know it will not find it
[21:33:12] <zwelch> right now, it looks for *-config in your path
[21:33:24] <alex_joni> that's not always the case
[21:33:28] <zwelch> that's the only "probe" it does, unless you specify an explicit directory
[21:33:31] <alex_joni> having *-config in the path
[21:33:32] <zwelch> right; i've got that now
[21:33:49] <zwelch> but i will not let the script simply "find" through /usr
[21:33:53] <zwelch> that's obscene
[21:34:03] <alex_joni> yet it works great ;)
[21:34:09] <zwelch> not if you don't have RT
[21:34:14] <zwelch> it takes minutes
[21:34:19] <alex_joni> so?
[21:34:37] <zwelch> that's unacceptable for someone like me - a developer - who is running that script over and over and over
[21:34:53] <alex_joni> supply your own flag (--no-realtime)
[21:34:54] <zwelch> i'd rather take a bullet to the head than continue to work on this project with that code in it
[21:34:59] <zwelch> i did
[21:35:07] <zwelch> but what do you do if you supply nothing?
[21:35:13] <alex_joni> sorry, didn't look at the patch yet
[21:35:14] <zwelch> and you want it to "just figure it out"?
[21:35:15] <Jymmm> In TkEMC, can comments follow code? bind blah # this is blah
[21:35:19] <paul_c> A couple of assumptions are made.... 1) A supported realtime extension is being used.
[21:35:29] <zwelch> false
[21:35:40] <zwelch> that's a completely groundless assumption
[21:35:52] <zwelch> or at least, i strongly believe it is :)
[21:35:56] <alex_joni> those were taken at the time when configure was created (or at least that was it's purpose)
[21:36:30] <paul_c> EMC expects hard realtime to schedual the RT tasks.
[21:37:04] <zwelch> okay, so then why am i trying to build a non-RT simulator?
[21:37:23] <alex_joni> you can do that
[21:37:36] <alex_joni> that is actually encouraged, and it will be a nice thing to have
[21:37:51] <zwelch> well, that's where assumption #1 falls flat on its face
[21:38:03] <alex_joni> but.. if you get that working ;) I think there's no problem in reconsidering the above assumptions
[21:38:19] <zwelch> and that's the angle that ive taken in rewriting the build scripts
[21:38:35] <alex_joni> try the actual code first
[21:38:40] <alex_joni> * alex_joni has a hunch ;)
[21:38:47] <zwelch> "actualy code"?
[21:38:57] <zwelch> i.e. RT?
[21:39:16] <alex_joni> if you didn't like the configure.in, you might easily run away screaming after looking at some emc specific code
[21:39:35] <paul_c> alex_joni: Don't put the guy off...
[21:39:42] <alex_joni> I'm not
[21:39:43] <zwelch> well, if you're saying its crap, then why doesn't it get rewritten? :)
[21:39:57] <alex_joni> zwelch: don't want to go there
[21:39:59] <zwelch> ;)
[21:40:09] <alex_joni> would take too much time.. :)
[21:40:16] <alex_joni> we had these conversations a few times
[21:40:33] <paul_c> 'cos too many people have said its crap, but none have done jack shit about it.
[21:40:42] <zwelch> well, i'm guessing that once someone produces a patch to rewrite those bits, the conversation will lead to a conclusion
[21:40:49] <paul_c> like the troll over there ->
[21:40:56] <zwelch> until there's a solution, anyone can talk
[21:40:59] <alex_joni> over there?
[21:41:16] <alex_joni> anyways..
[21:41:22] <zwelch> * zwelch just plans to write as much code as he can, reducing the crap quotient as i go
[21:42:01] <alex_joni> zwelch: to put it nicely, there is a big bunch of crap code (which - as strange as it may seem - actually works)
[21:42:20] <zwelch> oh, i grok that; we're talking developer perspective
[21:42:29] <zwelch> just because it's ugly doesn't mean it won't work ;)
[21:42:30] <alex_joni> the thing is.. once you start changing things it's like a lavine
[21:42:38] <alex_joni> not sure if that's the right word
[21:42:50] <zwelch> yes, well.. maybe i'll call my rewrites the emc3 branch ;)
[21:42:57] <alex_joni> anyways... the task you took simply starts to grow and grow
[21:43:02] <alex_joni> till you fail ;)
[21:43:06] <zwelch> hah
[21:43:12] <zwelch> if you believe that, then you have already lost
[21:43:21] <alex_joni> well.. I didn't believe that
[21:43:24] <paul_c> don't need another 'riggin' tree to go pissing up.
[21:43:26] <zwelch> if you do not believe you will win, you won't ;)
[21:43:42] <zwelch> "riggin"?
[21:43:53] <alex_joni> don't mind him ;)
[21:43:57] <zwelch> ah... the 's were important
[21:44:07] <alex_joni> uk people have funny words :D
[21:44:18] <zwelch> funny ways of abbreviating them, that's for sure ;)
[21:44:32] <alex_joni> yeah
[21:45:02] <paul_c> lemme put it another way then...
[21:45:17] <paul_c> Currently we have the rcslib/emc tree9s)
[21:45:23] <paul_c> Currently we have the rcslib/emc tree(s)
[21:45:31] <paul_c> the emc2 tree
[21:45:55] <paul_c> there was very nearly an EBo-cos -its-my-way tree
[21:46:07] <alex_joni> heh.. yeah (that was fun)
[21:46:20] <paul_c> and now you propose an emc3 tree....
[21:46:21] <alex_joni> and don't forget the bdi-tree :P
[21:46:37] <alex_joni> oh.. 'scuse me.. the bdi-branch
[21:46:43] <paul_c> bdi-4 is a branch (a dead one)
[21:47:03] <Jymmm> * Jymmm calls in a tree surgeon
[21:47:15] <zwelch> i joke about emc3
[21:47:23] <zwelch> * zwelch has read the history of the project
[21:47:25] <paul_c> * paul_c cuts Jymmm's phone line
[21:47:29] <zwelch> but... innovate or be forked
[21:47:33] <zwelch> * zwelch shrugs
[21:47:37] <Yuga> hey all
[21:47:43] <Jymmm> * Jymmm wonders "What Line?"
[21:48:13] <alex_joni> paul_c: they don't have phone lines over there. when they say call, they actually mean shout
[21:48:31] <alex_joni> a big fellow like Jymmm can shout about 3-4 villages away
[21:48:39] <Jymmm> 30-40
[21:48:51] <paul_c> or scare the hell out the area
[21:48:52] <Jymmm> smoke signals
[21:49:07] <Jymmm> bongo drums...
[21:49:34] <Jymmm> * Jymmm runs thru the logs to remember what paul_c said yesterday
[21:50:52] <alex_joni> zwelch: have you met Aunt tillie yet?
[21:51:17] <paul_c> that dumb old bat needs putting out in a field
[21:52:19] <paul_c> then get Les to bush hog her
[21:52:32] <alex_joni> lol
[21:53:10] <Jymmm> kinky
[21:54:02] <Jymmm> paul_c it was a typo
[21:54:49] <Jymmm> hmmm... diff has some intelligence to it... cool
[21:54:56] <paul_c> anyways, back to trees....
[21:55:50] <paul_c> The more "trees" the more it will confuse potential users... How do they know which one to use.
[21:56:08] <paul_c> or which is being actively developed.
[21:56:36] <zwelch> heh
[21:57:04] <zwelch> i tend to agree
[21:57:12] <mrallen> newbie sherline question... which way to i crank on the end-mill holder to loosen it? i've only taken it off twice before. now it's good and tight and i don't want to tighten it further
[21:57:32] <zwelch> although, you should understand the rather unique perspective i take on open source development
[21:57:51] <paul_c> As you look down from the top, clockwise to undo
[21:58:11] <mrallen> dang. then it's really tight.
[21:58:40] <paul_c> or are you refering to the drawbar at the top ?
[21:59:16] <mrallen> no
[21:59:22] <paul_c> zwelch: what is your perspective on FOSS ?
[21:59:46] <paul_c> mrallen: 'k, then you've done it up too tight.
[22:00:02] <zwelch> with open source, it is up to a project's management to ensure that contributors are able to actively merge their patches with the "main branch" in a timely manner; in my world view, a fork happens everytime a developer makes changes to a package that get put into production *before* those changes are part of mainline
[22:00:41] <Jymmm> * Jymmm hands mrallen an impact wrench, blow torch, and bottle of tequilia. Use each as you see fit, in any order you like
[22:00:52] <zwelch> in other words, the chance of a fork occuring is roughly proportional to the time it takes to get patches mainlines
[22:01:11] <zwelch> if you have someone cranking out code faster than you can merge it, you almost certainly will end up with a fork
[22:01:33] <zwelch> is that a fair assessment?
[22:01:43] <alex_joni> don't think that'll ever happen here
[22:01:53] <alex_joni> someone cranking out code fast enough
[22:01:56] <alex_joni> :)
[22:02:25] <paul_c> by "fork", do you mean a branch within the existing tree, or a totally new tree (either on the same site, or off site) ?
[22:02:56] <zwelch> paul_c: that would depend entirely on how civil things go down
[22:03:12] <zwelch> obviously, the other factor is interpersonal
[22:03:18] <zwelch> can the developers even get along?
[22:03:31] <zwelch> that's a wholly different (and uncontrollable, in my view) problem
[22:04:07] <alex_joni> yet sometimes a funny thing to watch
[22:04:08] <zwelch> personally, i have seen the absolutely worst results of a fork
[22:04:24] <paul_c> I don't have a problem with branches in the existing tree - In fact, it is something I would encourage....
[22:04:25] <zwelch> after all, i led the zynot fork of gentoo
[22:04:48] <paul_c> It is easier to merge a branch than a totally separate tree.
[22:05:53] <zwelch> well, i believe its important to design a project with the expectation that it will be forked
[22:06:11] <zwelch> thus, when the codebase is taken and re-released, you can fork it right back into your tree
[22:06:39] <alex_joni> * alex_joni leaves for a while
[22:06:49] <zwelch> "plan for the worst".... www.guft.org
[22:08:16] <zwelch> * zwelch has been using the prototype of guft for a while now, even though the project isn't due for completion until 2012
[22:11:28] <paul_c> alex_joni: What happened with fenn & his emc3 rewrite ?
[22:11:55] <alex_joni> hmmm.. smthg like robin & his nml rewrite
[22:12:32] <paul_c> did you ever see any of that code ?
[22:12:35] <jepler> zwelch: interesting story about gentoo (still in google cache0. btw at my day job we make very important use of distcc to cross-compile our software, though I don't know if we benefit directly from any of your work.
[22:13:06] <zwelch> jepler: cool.. i know i helped "push it over the edge" in terms of critical mass of interest and proof of concept code
[22:13:17] <zwelch> as far as i know, most of my contributions were systematically wiped out
[22:13:39] <zwelch> i.e. it looks like they had various stooges rewrite all of the bits that had my copyright in them
[22:13:42] <zwelch> * zwelch sighs
[22:14:08] <zwelch> (btw, i am talking about distcc, above... ;) )
[22:14:36] <zwelch> * zwelch curses google cache
[22:14:52] <zwelch> sometimes, you just want to take down a page and have it go *away*
[22:15:08] <zwelch> * zwelch giggles nervously
[22:15:40] <jepler> if you hadn't told me the magic term "zynot" I wouldn't have been able to find it
[22:16:11] <zwelch> ;) nah, i still stand by what i wrote; though, i'd have done a lot different had i known then what i know now....
[22:16:57] <zwelch> yeah, i wouldn't have mentioned it by name if i were truly ashamed of my past; i know that anyone who cares can immediately find the facts with google ;)
[22:18:09] <zwelch> guft is what came out of it, but also is the biggest lessons of all: forking happens all the time when developers patch FOSS code and deploy it before waiting for their patch to reach mainline
[22:19:39] <zwelch> to prevent absolute chaos, every developer needs to work both to get their changes mainlines, but to ensure that they work with everyone else trying to do the same thing in order to minimize the release cycle time (without sacrificing quality)
[22:20:53] <zwelch> the linux kernel is the shining example of this: forks are rampant, but they aren't divergant because everyone is always working to push their patches closer to Linus's tree
[22:21:07] <jepler> that still doesn't eliminate the forks that happen because some developer's ideas are simply bad.
[22:21:21] <zwelch> darwin provided the answer to that problem
[22:21:29] <zwelch> survival of the fitest, brother
[22:22:17] <zwelch> but again, look at linux: there are lots of perfectly valid alternatives that do survive
[22:22:29] <zwelch> now, not every project can support such a diverse ecology of forks
[22:22:48] <zwelch> but every project affords their developers room to fork, by way of their revision control systems
[22:24:07] <zwelch> my theory is that - in order to harness a project's maximum parallel development potential - the number of forks should be maximized to the extent of available resources
[22:25:47] <zwelch> it's impossible unless you have a system like GUFT to manage the numerous processes, code bases, documentation, support infrastructure, and related project resources and activities
[22:27:19] <paul_c> does that not assume there is a sizable active group of developers ?
[22:27:36] <zwelch> with the linux kernel, GUFT devolves into an ad hoc system of individual networks of trust; with processes and mechanics differing at each level
[22:27:42] <zwelch> not necessarily
[22:28:19] <zwelch> my forking theories also encapsulate the problems of dealing with a single project having a 1.x and 2.x branches, where critical bug fixes in 2.x absolutely must be backported
[22:28:43] <zwelch> moreover, any single project should enable each developer to maintain their own "Experimental" fork
[22:28:56] <zwelch> otherwise, you are potentially limiting the amount of innovation that may be attempted
[22:29:22] <zwelch> of course, "cvs co" can be said an effective tool for creating a fresh fork :)
[22:29:42] <zwelch> * zwelch has obviously loosened the definition of "forking" quite a bit
[22:30:21] <zwelch> but with that, I added the key qualifier that the fork only happens if the developer puts the changes into a "production environment"
[22:30:29] <zwelch> otherwise, yeah, it's just development :)
[22:30:56] <zwelch> but once it's in production, that developer must port those features forward if their patch doesn't get mainlines; they have forked the code
[22:31:17] <zwelch> * zwelch keeps typoing 'mainlined' as 'mainlines' sorry
[22:31:24] <alex_joni> still forking?
[22:31:25] <alex_joni> :)
[22:31:32] <zwelch> * zwelch was just winding down
[22:31:40] <alex_joni> * alex_joni hands zwelch a bigger fork
[22:32:09] <zwelch> * zwelch also assumes that he'll get a virtual knife in the ribs if he rants too loudly or insanely
[22:32:19] <alex_joni> nope
[22:32:33] <zwelch> * zwelch chuckles
[22:32:38] <alex_joni> this is an educated crowd
[22:32:40] <paul_c> just a boot.
[22:32:44] <zwelch> right, right... you lot like to feed the trolls ;)
[22:32:58] <paul_c> nah...
[22:33:04] <alex_joni> we only have boots, wet kippers, trouts, and other blunt objects that don't leave marks
[22:33:25] <alex_joni> especially paul_c seems to have a nice arsenal of such objects
[22:33:29] <zwelch> * zwelch hopes his forking theories are interesting, if provacative
[22:34:43] <Jymmm> wth is a kipper?
[22:34:51] <zwelch> anyway, i hope it was relatively on-topic to the issue of EMC trees
[22:34:57] <Jymmm> must be some Euro Trash thing =)
[22:35:13] <zwelch> paul_c: too much perspective for you? :)
[22:36:07] <zwelch> anyway, the thing that i've been hearing hear about emc is that a number of different shops are using it to do their own thing
[22:36:16] <zwelch> to me, that's a bunch of forks
[22:37:36] <paul_c> One thing you need to ask these shops is "How did they install EMC"
[22:37:48] <zwelch> the amount of changes they make to the core may be zero, but being able to "re-fork" from the handful that have made their own bug fixes or improvements
[22:38:28] <zwelch> moreover, i see an "easily forkable EMC" as being a key ingredient to getting more active developers
[22:38:56] <zwelch> and, sure, i will be that most install from the BDI (or some other binary distribution method)
[22:43:21] <paul_c> attracting developers is not an issue - Hell, we got 50 already....
[22:43:58] <paul_c> for a single application, that is a fairly large number considering the size of the code base.
[22:44:05] <alex_joni> getting them to do smthg is hard ;)
[22:44:18] <Jymmm> lol @ alex_joni
[22:45:08] <Jymmm> Question.... ESC -vs- F1 (estop), whats the difference?
[22:45:28] <jepler> Jymmm: estop will do stuff like shut off servo amps, escape won't.
[22:45:36] <jepler> escape just stops running the program
[22:45:44] <Jymmm> jepler stop or pauses?
[22:45:47] <jepler> stops
[22:46:10] <Jymmm> ok, thanks. I'll map NUMPAD_Enter to F1 instead
[22:46:29] <jepler> looks like escape leaves the spindle running too
[22:47:02] <Jymmm> F1 sounds safest
[22:50:00] <zwelch> paul_c: while he may have been joking, i agree with alex_joni :/
[22:51:09] <zwelch> moreover, getting them to all work together, toward, and on a unified plan of action can be more challenging than herding cats with a segway
[22:51:32] <paul_c> no, Alex is quite correct. Getting the majority of the 50 to do anything, even respond to emails, is virtually impossible.
[22:52:10] <alex_joni> too bad it ain't a perfect world
[22:52:28] <zwelch> "the world is not fair because you allow it to be"
[22:52:37] <alex_joni> heh
[22:55:39] <zwelch> is linuxcnc incorporated?
[22:56:02] <AchiestDragon> robin_z: you there ??
[22:56:09] <zwelch> * zwelch scratches his head
[22:56:37] <zwelch> you have a board of directors, which implies (to some extent) that an organization exists behind it
[22:59:00] <alex_joni> it's more a guiding comitee
[22:59:31] <zwelch> what are they guiding?
[22:59:33] <zwelch> :)
[22:59:39] <paul_c> cats.
[22:59:47] <zwelch> * zwelch cackles uproariously
[22:59:48] <alex_joni> with wet kippers
[23:00:01] <alex_joni> and fishes
[23:00:09] <zwelch> so, where's the money?
[23:00:21] <alex_joni> missed an 'k'
[23:00:52] <zwelch> i.e. who pays the costs of the linuxcnc domain, hosting, and bandwidth?
[23:01:04] <alex_joni> steve does ;)
[23:01:25] <paul_c> not all of it...
[23:01:32] <alex_joni> for linuxcnc.org (hosting & bandwidth)
[23:01:39] <alex_joni> not domain, afaik
[23:01:49] <paul_c> Ray, Dan, Matt - They all have their hand on it.
[23:02:09] <zwelch> so there's no Linux CNC, Inc.?
[23:02:13] <alex_joni> zwelch: nice people :)
[23:02:20] <paul_c> nope.
[23:03:06] <zwelch> * zwelch guesses that there probably is little interest in such either
[23:03:35] <zwelch> well, that's interesting
[23:04:05] <paul_c> there is already a commercial linuxCNCoutfit
[23:04:49] <zwelch> well, i was thinking more along the lines of "non-profit"
[23:05:08] <zwelch> i.e. the "profits" go back into the organization instead of anyone's individual pockets
[23:05:42] <paul_c> http://www.writerscritique.com/items/7163388369.html <- wt....
[23:06:53] <paul_c> the cheeky bugger.... claiming copyright on the Chips background...
[23:07:31] <zwelch> yeah, and so what are their resourceS?
[23:07:45] <zwelch> what are resources of this community?
[23:07:54] <zwelch> (and where do you think they are getting their bits...?)
[23:08:15] <zwelch> do you think this community could compete?
[23:08:27] <paul_c> looks like the main source is Bubwieser
[23:08:42] <zwelch> * zwelch thinks it should: sell the BDI CD from the linuxcnc.org site :)
[23:09:17] <zwelch> but to do that "above the boards", i think a corporate identity will be required
[23:09:42] <paul_c> and who gets the profits from such a venture ?
[23:09:58] <zwelch> profits? you'll be lucky if the revenues pay for hosting/bandwidth, etc :)
[23:10:09] <zwelch> if it does, then you have a board right?
[23:10:14] <zwelch> steer the boat, guys :)
[23:10:25] <zwelch> but without $$, there's no wind in the sails
[23:10:49] <zwelch> or rather, no incentive for your 50 AWOL developers to contribute
[23:11:24] <paul_c> even with money in the pot, you'd not get all 50 contributing.
[23:11:35] <zwelch> * zwelch would suggest using "extra" revenues to create a "small gift giveaway" in return for contributions
[23:12:09] <zwelch> by this, i mean - get sponsors (e.g. sherline) - to either provide at low cost (or free) some thing to give to active contributors
[23:13:00] <zwelch> (board members, employees, and other individuals associated with the new corporation are not eligable; offer void where prohibited... ;) )
[23:13:56] <zwelch> i would have to think that sherline would back a non-profit organization aimed at developing, promoting, and providing education about EMC
[23:14:17] <zwelch> (particularly if it grew out directly of the linuxcnc.org community)
[23:14:22] <zwelch> * zwelch shrugs
[23:14:38] <zwelch> have i mentioned that i tend to think "too big" sometimes/always? ;)
[23:14:50] <alex_joni> * alex_joni thinks...
[23:15:14] <alex_joni> * alex_joni thinks... it's late
[23:15:14] <alex_joni> zwelch: you didn't have to
[23:15:14] <alex_joni> :D
[23:15:42] <zwelch> * zwelch just wanted everyone to know that he is aware of the "problem"
[23:15:56] <paul_c> didn't we cover this ground last night ?
[23:17:49] <zwelch> i don't think i presented the same perspective
[23:18:19] <alex_joni> does it change that often?
[23:20:07] <zwelch> these last suggestions present a pro-active means of "getting from here to there", where "there" means "money to pay for things that need to be done" and "growing a large active community of contributors"
[23:20:43] <alex_joni> ain't gonna happen
[23:20:45] <zwelch> alex_joni: only when i find reason(s) necessitating such adjustments be made
[23:21:17] <zwelch> alex_joni: your defeatist attitude is only encouraging me :)
[23:21:39] <alex_joni> darn, you discovered my hidden goal
[23:21:40] <alex_joni> :D
[23:21:41] <zwelch> give me *reasons* to believe you, and i will consider their merit :)
[23:21:55] <alex_joni> I only want to stir ;)
[23:22:04] <zwelch> * zwelch pets the troll
[23:22:18] <alex_joni> * alex_joni does the same
[23:22:34] <alex_joni> but the troll is awfully quiet tonight..
[23:22:38] <alex_joni> might be sick
[23:22:58] <alex_joni> ok.. I'm calling it a day
[23:23:23] <alex_joni> 3h sleep last night, 18h ago ;)
[23:23:37] <an0n_> an0n_ is now known as anonimasu
[23:30:57] <zwelch> paul_c: so i've fixed up the configure script
[23:31:38] <zwelch> although i still need to improve the default probing
[23:32:24] <alex_joni> zwelch: there's a branch called auto_configure_0_2 iirc
[23:32:30] <alex_joni> not used, you can play in there safely
[23:33:13] <paul_c> or use HEAD, no one else is using that branch...
[23:33:34] <alex_joni> buggeroff
[23:33:47] <paul_c> language Timothy
[23:34:01] <zwelch> will it suffice to simply look for the *-config script under /usr/realtime and, if not found there, the path?
[23:34:19] <alex_joni> I wouldn't think so..
[23:35:04] <LawrenceG> http://focus.ti.com/lit/ds/symlink/opa549.pdf anyone looked at these for a mid sized servo amp?
[23:35:21] <LawrenceG> they give away 5 free samples
[23:35:48] <LawrenceG> 10 if you dont mind the alternate package
[23:36:08] <zwelch> the user always has the alternative to select the rtai directory, if it's not in a standard path
[23:36:29] <zwelch> but find /usr is not going back into my version of the script
[23:36:33] <alex_joni> anyways... really getting late
[23:36:36] <alex_joni> night guys
[23:36:45] <alex_joni> zwelch: once again, welcome to the club
[23:36:56] <zwelch> * zwelch waves to alex_joni
[23:37:04] <LawrenceG> goodnight
[23:37:11] <paul_c> LawrenceG: 60V @ 10A max is pretty small...
[23:37:16] <alex_joni> it really is a bunch of crazy people, but it's a nice company
[23:37:17] <anonimasu> hm yeah
[23:37:53] <zwelch> hahah, someone needs to add something to emc.sourceforge.net :)
[23:37:57] <LawrenceG> paul_c, yep.... but on the order for a lot of desktop stuff
[23:39:22] <LawrenceG> paul_c, was thing of using them as a motor driver with the mesa card
[23:39:27] <zwelch> * zwelch would suggest a simple redirect to sourceforge.net/projects/emc/
[23:39:50] <LawrenceG> s/thing/thinking
[23:42:52] <les> hi lawrence
[23:42:58] <les> and hello zwelch
[23:43:20] <les> just back from um...digging holes
[23:44:37] <anonimasu> hehe
[23:45:20] <les> I wanted to level up some areas
[23:45:45] <les> spirited discussion I see from reading back
[23:46:19] <Yuga> hey there les.. how u doing?
[23:46:31] <les> I am ok
[23:46:47] <Yuga> good good... long time no c
[23:46:55] <les> yeah
[23:47:15] <les> we have been working more on the china cnc router thing
[23:47:18] <Yuga> friggin telkom here in sa has been doing shit with the adsl
[23:47:51] <les> seems their workmanship is ok, but the design engineering is horrible
[23:50:10] <les> I was just doing patent disclosures...got very bored around lunch and gave it up to go out and do farmer stuff on tractors
[23:51:06] <les> I hate having to use words like "multiplicity"," preferred embodiment" and "prior art"
[23:51:18] <paul_c> eeuww... php - Is that anything like html ?
[23:51:28] <Yuga> nopes
[23:51:32] <Yuga> totaly different
[23:52:02] <les> worse paul...it's lawyer talk
[23:52:35] <les> it was easier in chicago...the would just send paralegals down and take pictures of my whiteboard
[23:53:11] <les> and record stuff on little pocket audio recorders
[23:53:14] <paul_c> les: nope - You need "duplicity", "Double standards", "snake in the grass", etc, when talking lawyers
[23:53:23] <les> heh
[23:53:48] <Jymmm> paul_c: PHP is a script language primarily focused on web usage
[23:54:27] <les> oh well I am getting paid well to write this crap
[23:54:57] <Jymmm> les Where's my 10% of that 10,000,000 USD ?!
[23:54:58] <les> but it is like being paid well to hold your head in a bucket of SH#t
[23:55:01] <paul_c> * paul_c makes a note to bill Les for consultancy fee.
[23:55:38] <les> WONDERED WHEN YOU WOULD ASK...HAHA
[23:55:45] <les> oops cap lock
[23:56:09] <les> bleh
[23:56:25] <Jymmm> les Gonna have Jacky^ send over his "Friends" to collect! They'll have much use for that traactor of your if you dont pay up!
[23:56:37] <les> haha
[23:56:44] <Yuga> les... u have any luck with that spindle?
[23:57:15] <les> I got some more quotes...the ebay one seemed fishy and I did not go for it
[23:57:26] <les> so I am looking at about $5000
[23:57:32] <Yuga> aaarrrg
[23:57:40] <Yuga> not that great
[23:57:57] <les> need some meetings with customer before I pull the trigger on a $5k spindle
[23:57:59] <Jymmm> Anyone have an idea why KP_8 bindings aren't beign accepted? resorting to KP_Up instead
[23:58:24] <Yuga> les... how's your new project coming along? the glue cnc?
[23:58:41] <Jymmm> Yuga just ask les for pictures
[23:59:03] <Yuga> Jymmm... i would... but he is ignoring my private msg's :>
[23:59:32] <Jymmm> Nah, he's not ignoring them, just isn't seeing them
[23:59:42] <les> glue...? I have a compressed air generator contract...and a technical advisor to a planned aquisition contract
[23:59:47] <Yuga> * Yuga hand's les's his glasses
[23:59:51] <les> can't say a word about it though