#emc | Logs for 2007-12-06

[00:00:27] <Dogfishguzzler> If I can do servos as cheap as the steppers even if they are used I would love the chance to tinker with them too.
[00:03:47] <skunkworks__> dmess: thanks - I missed that
[00:05:49] <Dogfishguzzler> SWPadnos: I am interesting in at least three of those. What are the part numbers so I can read about them and price what I need to complete a project with them? I gotta go right now but I'll be.
[00:06:06] <Dogfishguzzler> be back.
[00:10:35] <toastydeath> holy crap my microwave is programmable
[00:10:43] <toastydeath> it does like, dwells and power cycles and stuff
[00:10:53] <toastydeath> is this normal?
[00:31:07] <cradek> servos are faster and (can be) more powerful, but they do not make a small mill simpler...
[00:52:35] <Ziegler> heh
[00:59:48] <Dogfishguzzler> Anybody use emc with one of these. http://www.hobbytron.com/product574.html
[01:01:42] <jepler> looks like crap
[01:03:39] <cradek> not appropriate for cnc
[01:03:43] <jepler> two hints that it's crap: it only goes up to 15V and the motor voltage rating is the same as the input voltage rating. that means it is not a chooper-type stepper driver, which is the only kind worth having.
[01:04:04] <jepler> second hint: no real stepper driver has a built in clock generator that you adjust with a knob
[01:04:27] <jepler> third hint: there's not an obvious link to a datasheet to allow anyone to figure out more about it before buying
[01:04:39] <jepler> oh wait that's three hints
[01:04:52] <toastydeath> fourth hint: it's 34 bucks
[01:06:37] <jepler> you can get a xylotex stepper driver board for not much more than $34/axis, and that's a perfectly acceptable board for a low-end milling machine.
[01:08:35] <toastydeath> I STAND BY MY SNOBBERY
[01:10:47] <SWPadnos> Dogfishguzzler, I have a part number, but I'm not sure it'll help - I think these motors were meant for a large-ish printer
[01:11:31] <SWPadnos> oh wait, it says carousel drive, so it was likely fora tape magazine
[01:13:43] <SWPadnos> wow - a manual: http://yaskawa.com/site/dmservo.nsf/link2/TKUR-5ELJTH/$file/CHE-C241-5.pdf
[01:13:54] <SWPadnos> they're UGPMEM-09-CG11
[01:18:34] <Ziegler> Dogfishguzzler: if you want cheap entry cnc look at hobby cnc or xylotex
[02:03:58] <Dogfishguzzler> Ziegler: I saw the kits on hobbycnc.com and at some point probablly try them. Right now Im only experimenting with old floppy and printer motors. I'm googling chopper drives now.
[02:08:51] <Dogfishguzzler> Damn The haydon drives look nice but no price. I guess I know what that means.
[02:12:37] <Dogfishguzzler> jepler: Thanks for the xlotex hint, looking now.
[02:20:46] <Dogfishguzzler> SWPadnos:Interesting looking servos you got there. I still need to read a bit and find out what it will take make them run though.
[02:23:42] <SWPadnos> ok - take your time :)
[02:23:57] <SWPadnos> if you were thinking of Gecko stepper drives, their servo drives are the same price
[02:30:04] <SWPadnos> night all
[03:48:30] <jmkasunich> that 300 HP face mill video... wow
[03:50:10] <jmkasunich> 2-1/2 tons of chips per hour
[03:50:18] <jmkasunich> glad I don't have to clean up after it
[04:03:30] <fenn> i like how the push-broom they're shoveling chips with catches on fire
[04:05:57] <jmkasunich> ;-)
[04:08:13] <jmkasunich> fenn: I saw that last night you were talking about anti-backlash acme nuts
[04:08:48] <jmkasunich> have you tried tapping a nut using the same acme stock that you will use for the screw?
[04:09:46] <jmkasunich> I've done that (4140 steel screw, mill flutes, harden and temper with propane torch, grind relief with dremel, cut threads in bronze)
[04:10:05] <jmkasunich> and I got pretty low backlash without any "anti-backlash" tricks
[04:11:48] <jmkasunich> tap: http://www.metalworking.com/dropbox/_2001_retired_files/Worm05.jpg
[04:12:20] <jmkasunich> screw and nut (and a gear) http://www.metalworking.com/dropbox/_2001_retired_files/Worm03.jpg
[04:13:01] <jmkasunich> the tap is a two-stage affair, the first part was recut with a triangle threadform, to rough out the thread, then the second part cuts the final acme
[04:13:30] <jmkasunich> * jmkasunich stops talking to himself
[04:19:11] <cradek> that's cool. I might try something like that for my Z axis. I don't think one of those plastic nuts will handle all this weight
[04:19:31] <jmkasunich> I would suggest using something finer than the 5 tpi that I used
[04:19:38] <jmkasunich> tapping those nuts is a bitch!
[04:19:41] <cradek> uh yeah.
[04:19:45] <jmkasunich> 10 TPI would be much easier
[04:19:53] <cradek> my XY screws are 16
[04:20:27] <jmkasunich> I do however recommend the acme threaded rod from McMaster - that black coating you see on it is hard and smooth, and runs very nicely in a bronze nut
[04:20:41] <cradek> yeah that's what I used too
[04:32:28] <jmkasunich> count the cats and win a prize! http://bp3.blogger.com/_dqF7xsw7On8/R05NhLFnZkI/AAAAAAAAAF4/N2AOMJVHGZE/s320/Hidden+Pictures.JPG
[04:32:49] <jmkasunich> pretty sure its siz
[04:32:52] <jmkasunich> six
[04:32:57] <cradek> need more pixels
[04:33:04] <cradek> there are "enough" cats
[04:33:12] <cradek> maybe even "plenty"
[04:33:23] <jmkasunich> I think the prize might be one of the cats
[04:33:27] <cradek> an exact count of cats isn't possible once there are more than about 3
[04:33:37] <jmkasunich> only if they are awake
[04:33:45] <jmkasunich> if sleeping, the number gets a bit higher
[04:33:50] <cradek> maybe you've found a loophole
[04:34:34] <Jymmm> Looks like dog food to me
[04:34:54] <cradek> that's funny since you raise cat food
[04:35:12] <cradek> (my cat Muffin typed that)
[04:35:13] <Jymmm> fair enoug. lol
[04:35:38] <Jymmm> I'm againest killing for sport
[04:35:53] <cradek> so is Muffin I think
[04:36:14] <cradek> she only kills catnip logs and crunchies
[04:36:15] <Jymmm> I've never seen a cat kill for food
[04:36:31] <Jymmm> well under 100 lbs that is
[04:36:43] <cradek> outdoor cats don't exactly eat salads
[04:37:04] <Jymmm> I wouldn't know
[04:37:09] <cradek> (housecats are a different story)
[04:37:27] <Jymmm> I dont like cats, (only 2 I've ever met that were cool)
[04:37:39] <cradek> I wonder if killing "for sport" is purely a human peculiarity
[04:37:55] <Jymmm> No, cats do it too
[04:38:24] <cradek> I guess it's hard to say why they do it, not having ever been a cat
[04:38:35] <jmkasunich> cats are designed to kill for food
[04:38:52] <cradek> * cradek shivers at the use of 'designed'
[04:39:01] <cradek> let's change the subject :-/
[04:39:12] <jmkasunich> but they don't actually learn to do it right, and to eat the kills, unless they are taught (by observing mama cat) at a young age
[04:39:13] <Jymmm> how about breakout boards!!!
[04:39:24] <jmkasunich> so most domestic cats don't know how
[04:39:38] <jmkasunich> and play with the mice instead (I'm sure the mice don't consider it play)
[04:39:41] <Jymmm> Yeah, HOW BOUT THEM LAKERS!!!!!!!
[04:40:03] <jmkasunich> * jmkasunich pets the cat on his lap
[04:40:10] <jmkasunich> fat old thing couldn't catch a cold
[04:40:22] <cradek> old cats are the best
[04:40:34] <cradek> * cradek is hoping for a snow day tomorrow
[04:41:20] <Jymmm> cradek: so you can ditch work?
[04:41:31] <cradek> yeah, with a good excuse
[04:41:38] <Jymmm> there ya go =)
[04:41:46] <Jymmm> Hey, anyoen have a smartphone?
[04:41:55] <cradek> I don't even know what that is
[04:42:06] <Jymmm> cra a phoen with applications on it
[04:42:35] <cradek> guess not
[04:42:37] <Jymmm> usually windows mobile edition, or a cracked iphone with unix on it
[04:42:42] <cradek> my cell phone just makes calls
[04:42:50] <Jymmm> ah =)
[04:43:18] <cradek> I carried a Palm for a while, and it was handy, but it hurt my rear to sit on it all the time
[04:43:25] <cradek> wasn't worth it
[04:43:46] <Jymmm> I'm looking at a Motorola Rizr... quad band, camera, mp3 player, bluetooth, miniUSB and MicroSD card
[04:44:17] <Jymmm> http://reviews.cnet.com/cell-phones/motorola-rizr-z3-blue/4505-6454_7-31987125.html
[04:44:25] <jmkasunich> can you make phone calls on it?
[04:44:41] <Jymmm> Yep =) and even use it in EU too
[04:44:51] <Jymmm> it's a sloder
[04:46:03] <eric_U> I didn't win an auction on ebay today, thought I would for sure
[04:46:56] <eric_U> bid with 15 seconds left, got outbid
[04:47:14] <Jymmm> eric_U: how many browser windows did you have open at the time?
[04:47:22] <eric_U> i just use one
[04:47:28] <Jymmm> that's the problem =)
[04:47:46] <Jymmm> how much did you get outbid for?
[04:47:57] <eric_U> I was going to get outbid anyway
[04:48:10] <Jymmm> oh, way over your limit huh?
[04:48:15] <eric_U> the whole idea is to bid your max with 15 seconds left
[04:48:31] <eric_U> and if someone else outbids you, it was inevitable
[04:48:53] <Jymmm> bull pucky! I HATE getting outbid over $1
[04:49:03] <eric_U> my mistake is that I would have bid higher, but I didn't think I was going to have to
[04:49:05] <Jymmm> or worse.... $0.01
[04:49:33] <cradek> ebay is trivially simple. near the end, you bid the maximum amount you are willing to pay.
[04:49:39] <eric_U> my strategy is to bid the highest amount that I wouldn't hate myself after
[04:50:02] <eric_U> so if it goes for $1 more, it's ok because I would have hated to pay that extra $
[04:50:08] <Jymmm> bit $0.01 more than that wouldn't kill you either, and you would kick yourself for it too
[04:50:08] <cradek> but, it's hard to do that, the inclination is to bid the maximum amount you hope you can get away with paying
[04:50:28] <eric_U> I'll tell you if that ever happens :)
[04:50:54] <cradek> if you bid the right amount, you're happy whether you win or lose!
[04:51:04] <Jymmm> what cradek said
[04:51:28] <eric_U> http://offer.ebay.com/ws/eBayISAPI.dll?ViewBids&item=200178995958
[04:51:44] <eric_U> I was outbid twice
[04:52:00] <Jymmm> eric_U: you'll probably get a 2nd bid notice
[04:52:06] <eric_U> second guy is kicking himself, but the high bidder probly bid $200
[04:52:21] <eric_U> not from these guys
[04:52:36] <Jymmm> what did you bid?
[04:52:43] <eric_U> 57.57
[04:52:54] <Jymmm> maybe.... not too shabby an amount
[04:52:56] <cradek> I hate those auction consignment guys
[04:53:31] <eric_U> I wrote these guys the other day because they were selling the sheet metal off some servo drives as a servo drive
[04:53:55] <cradek> yeah, what I said...
[04:54:03] <eric_U> told them their buyer would be unhappy. Never got an answer back
[04:54:04] <cradek> they're always 'masters of nothing'
[04:54:22] <eric_U> all the industrial equipment is sold by idiots
[04:54:29] <eric_U> almost all
[04:54:39] <cradek> glad they put their warranty policy for LCD monitors in this auction
[04:54:53] <eric_U> funny, they also say they test everything
[04:54:55] <cradek> and no toner cartridges are included with these servo drives
[04:55:38] <eric_U> the people I hate are the guys that strip off machines and throw away the cables
[04:56:16] <eric_U> most cables from servomotors cost at least $25 to make, and probably $100-200
[04:57:16] <eric_U> I'm watching an auction right now for a company that strips down machines and sells the accumulated cables
[04:57:22] <eric_U> but that's very rare
[04:59:17] <eric_U> "It seems ok visually, and it does not smell." http://cgi.ebay.com/Allen-Bradley-Servo-Motor-Alpha-Gear-Box-USED-Bosch_W0QQitemZ130180822453QQihZ003QQcategoryZ78195QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[05:00:25] <eric_U> gotta bid on that
[05:21:13] <Jymmm> cradek: what, as-is?
[05:22:03] <eric_U> jlmjvm: did you find me any 13 tooth timing pulleys for my bridgeport?
[05:23:13] <fenn> jmkasunich: i tapped the nut with a piece of the acme threaded rod i'm using, and at first it fit very nicely with no play, but then after i had put some oil on it and run back and forth five or ten times it started getting a bit more sloppy.. so now there's about 0.010" backlash
[05:24:12] <fenn> i think it tore the thread sidewalls as i was cutting, since it was a single pass tap
[05:24:35] <fenn> and then the burrs pushing on the thread is why it originally had no lash
[05:25:30] <fenn> ironically, if i had used a tap i think my anti-backlash scheme would have worked
[05:26:23] <toastydeath> how did you plan on removing backlash?
[05:26:41] <fenn> compressing the nut radially
[05:26:59] <toastydeath> a radially split nut or...
[05:27:03] <toastydeath> just compression on the nut
[05:27:09] <fenn> a split nut
[05:27:14] <toastydeath> oh
[05:27:24] <fenn> what happens is the flats on the threads are the first surface to make contact
[05:28:15] <fenn> so you get friction, but there's still slop on the thread flank
[05:29:56] <fenn> i may end up making a tap from scratch if ebay falls through
[13:17:38] <fenn> i bet martin held goes rock climbing so he can prove he's climbing the most efficient path
[13:27:18] <alex_joni> who?
[13:28:55] <fenn> a computational geometry professor whose proofs i'm wading through
[13:29:56] <fenn> the joke was intended for awallin
[13:39:03] <skunkworks_> fenn: I like your email to the list :)
[13:55:22] <alex_joni> darn, I wish I had a bigger CNC
[13:56:00] <alex_joni> http://imagebin.org/12227 <- would be cool to cut
[13:56:40] <skunkworks_> is the place your getting - big enough for a 'tinker' room?
[13:58:25] <fenn> he could always just build an extra house for the tools, seems to be a popular solution
[14:00:03] <skunkworks_> yes - it does
[14:08:00] <awallin> hi fenn
[14:13:39] <fenn> so, the voronoi stuff is fairly straight-forward after all, definitely not requiring 165 pages to explain
[14:14:22] <awallin> yeah, fairly straight-forward...
[14:15:46] <awallin> it's supposed to be O(n log(n) ) in contrast to the segment by segment things that are O(n^2). Don't know if it makes a difference in practice
[14:17:08] <fenn> well, personally i could care less how efficient the algorithm is, as long as it works
[14:17:31] <awallin> the segment by segment offset algorithms might be easier to understand and/or code
[14:20:31] <fenn> what are the parameters for controlling how your lawn-mower behaves?
[14:22:15] <awallin> it's very simpel right now. it evaluates moves in all directions (1...360 degrees), then determines which of these cut away a suitable amount of material, then picks one of these suitable moves (prefering continuing in the previous direction over sharp turns)
[14:22:19] <awallin> simple
[14:27:08] <fenn> i bet it would work better if you ran the algorithm "backward", starting with the part outline and ending up at the center
[14:27:31] <fenn> well, the offset outline
[14:28:30] <fenn> so for a square pocket you would get a square-ish spiral
[14:28:59] <fenn> instead of having to cut out the corners at the end
[14:29:01] <awallin> maybe, it's just a toy at the moment though... without an offset algorithm there's no way to cut anything precise with this pixel based approach
[14:29:33] <awallin> I think you want to run the simulation/tp-generation in the same direction as machining is done, i.e. 'time running forward'
[14:29:56] <fenn> i think including the concept of momentum in generating the tool path is very neat
[14:30:33] <awallin> yes, that's probably a good idea. also I'd like to have variable step lengths, now it just takes tiny steps all the time
[14:32:03] <awallin> anyway, it would be good to have a handle on the conventional zigzag and spiral pocketing strategies before playing some more with this adaptive/sim stuff
[14:32:12] <fenn> yep
[14:32:22] <fenn> i dont really see the point of zigzag though
[14:32:44] <fenn> besides 'that's how you do it on a manual mill'
[14:33:13] <awallin> that's probably true...
[14:41:33] <fenn> apparently inkscape does have a "slow but correct" offsetting algorithm: http://inkscape.modevia.com/doxygen/html/sp-offset_8cpp-source.php#l00615
[14:43:10] <awallin> the slow but correct algorithm should not be that hard, basically you need to determine line/arc intersections correctly and then be able to calculate closesti distances btw. lines/arcs
[14:43:21] <awallin> joining up the result into a reasonable
[14:43:29] <awallin> chain' or toolpath might be harder
[14:46:25] <awallin> just requires someone with a bit of inspiration to code it and test it...
[14:48:41] <cradek> do people really zigzag a manual mill? you'd be climb milling half the time which can be pretty bad while roughing
[14:49:27] <awallin> cradek: only zig is better.
[14:49:47] <awallin> esp. with the typical backlash
[14:49:54] <cradek> right
[14:50:19] <cradek> climb milling is good on cnc, bad on manual
[14:52:43] <cradek> awallin: for me the 'correct' pocketing algorithm didn't look that easy. The biggest problem is when the regions 'break up'
[14:54:17] <awallin> you would need a graph to keep track of the areas that 'break up', then you can machine them in some sensible order
[14:56:18] <cradek> I'm not saying you're wrong - 'not that hard' might be different for you than me
[14:56:32] <awallin> depends ofcourse on the definition of 'correct'. given a correct offset algorithm it wouldn't be that hard to make a path that will cut the pocket
[14:58:30] <cradek> correct enough for me is a pocket the right shape without a lot of wasted machine motion. arcs have to still be arcs (not split up) and the outermost path should be a mathematically precise solution (not approximated by some pixel based approach)
[14:59:07] <cradek> that was my goal anyway, I didn't do much other than read some papers and decide it was hard
[15:00:16] <awallin> ok. basically that would be a geometry based algorithm. the adaptive machining ideas could improve on this and you would get a path where you know the cutter engagement stays below some limit at all times
[15:02:44] <awallin> this CAM business is a bit like the EMC traj-planner. Most potential programmers won't touch 'hard' problems with a stick...
[15:03:41] <cradek> yes. also, most people who haven't tried to work on it or really study it say "it shouldn't be too hard"
[15:04:05] <cradek> and I did that too :-)
[15:04:56] <cradek> you have obviously looked at it more than that though
[15:05:03] <fenn> i'm still wondering how to solve the racing-line problem
[15:05:06] <awallin> yeah...
[15:05:41] <fenn> oh hey a software patent
[15:06:19] <awallin> well the O(n^2) line and arc segment offset algorithm is not conceptually hard. how much of a problem numerical inaccuracies will be is hard to tell. The offset algorithm would really be the stepping stone for any pocketing strategies.
[15:06:36] <awallin> fenn: what's the racing-line problem?
[15:07:04] <fenn> if you're driving a race car, where do you drive the car for the minimal time
[15:07:19] <fenn> given a certain accel, decel, track width, etc
[15:07:39] <fenn> same problem applies to moving a milling cutter within a tolerance zone
[15:08:02] <awallin> right. that's the trajectory planning thing
[15:08:17] <fenn> a part of it
[15:09:26] <fenn> its actually more complicated because of feedrate override and rotational axes
[15:10:55] <awallin> ok, so similar to off-line precalculated trajectories. then your accuracy won't improve even if you turn down feed override (since the path was calculated for full speed)
[16:14:56] <seb_> seb_ is now known as sebjames
[20:06:00] <skunkworks_> JymmmEMC: http://youtube.com/watch?v=4CfyE6rPwnA&feature=related
[20:45:31] <\x90> \x90 is now known as crepincdotcom
[20:45:40] <crepincdotcom> jepler: around?
[20:47:02] <jepler> crepincdotcom: yeah
[20:47:36] <crepincdotcom> jepler: my EMC box took a digger, along with my eagle scripts. any chance I could have your NC scripts for eagle again please?
[20:47:59] <jepler> crepincdotcom: http://unpy.net/cgi-bin/viewcvs.cgi/eagle/ulp/
[20:48:06] <jepler> sorry about your machine
[20:48:46] <crepincdotcom> ehn its not a big issue, all my homedir was backed up, i just forgot to back up the eagle dirs.
[20:48:56] <jepler> (I'm not familiar with the phrase "took a digger..")
[20:49:23] <crepincdotcom> haha. in the most literal sence, to "take a digger" is to be walking along and suddenly trip or fall in a funny manner
[20:49:37] <crepincdotcom> my EMC box was an old laptop, and the disk finally went
[21:10:06] <jepler> I wonder how the game console manufacturers keep employees of third-party game developers from including a backdoor to allow for homebrew software, and then leaking the backdoor onto the internet after the game has gone gold.
[21:11:07] <jepler> the individual programmers must be convinced that the backdoor could be traced to them and that they would suffer disastrous consequences
[21:11:08] <cradek> how do you know that's not where they have come from?
[21:11:41] <jepler> as far as I've been able to tell, there's no homebrew yet for the current-generation consoles (Wii, PS3, X360)
[21:13:24] <jepler> on the other hand, the disk reproduction protection has long since been broken for wii and I think for the others
[21:15:09] <jepler> (is legitimate software without license fees to the console makers a bigger threat to them than illegally copied software...?)
[21:16:05] <cradek> it may cost them christmas sales
[21:16:11] <cradek> probably every buys SOME games
[21:17:09] <cradek> also, I bet they're all soon going to be making their money with online nickle-and-dime stuff. I bet someone screwing that up for them is a huge threat
[21:17:42] <jepler> if billybob software could sell his own games on CD-R or DVD-R for $5-$20 it would create a big pressure on legitimate titles to come down from their introduction price of $60
[21:18:02] <cradek> that's true
[21:18:04] <jepler> I have heard that there's now an "opening weekend" effect for new video games
[21:18:36] <jepler> just like for movies, the majority of sales are during (or before, with pre-orders) the first few days of availability
[21:18:44] <cradek> huh
[21:21:32] <jepler> "Apparently Konami is saying they need to sell a million units of Metal Gear Solid 4 on opening day to get into the black."
[21:23:41] <alex_joni> good night all
[21:23:49] <jepler> 'night alex_joni
[22:02:52] <anonimasu> -_-
[22:03:34] <anonimasu> jepler: if billybob had a team of 60 developers working for a year the game he developed would cost $60.
[22:05:32] <jepler> anonimasu: yeah it would be a different world -- the games would mostly be shit because of their low budget, not despite it.
[22:05:53] <anonimasu> thats the big difference
[22:05:56] <anonimasu> :)
[22:06:19] <anonimasu> I wouldnt think anyone would spend lots of cash on keeping lots of developer making a game if there was a chance it wouldnt pay off in the end
[22:06:57] <anonimasu> jepler: though I wish there was a market for thoose kind of games
[22:07:07] <anonimasu> it'd open for more independent small buisnesses
[22:09:31] <jepler> as I see it, it is fundamentally anticompetitive to keep small developers out of the console development business.
[22:10:17] <anonimasu> well, there is a market.. obviously
[22:10:29] <anonimasu> but as they cant develop for the consoles it dosent exist quite yet.
[22:10:33] <toastydeath> well you have to define "small developer"
[22:10:45] <anonimasu> say 10 guys coding on their sparetime
[22:10:57] <toastydeath> i'd call that 10 guys coding in their spare time
[22:11:10] <anonimasu> eh?
[22:11:20] <toastydeath> if those ten guys had venture capital, a business plan, and salaries to devote 8 hrs a day
[22:11:32] <toastydeath> i'd call that a small developer
[22:11:36] <toastydeath> er small studio
[22:11:49] <anonimasu> yeah
[22:12:41] <jepler> if you think that is necessary to develop software, plese quit using linux and emc right now.
[22:12:46] <toastydeath> lol
[22:12:57] <toastydeath> all the major contributors to linux get paid to do so
[22:13:07] <jepler> you don't get my point
[22:13:11] <toastydeath> i do get your point
[22:13:14] <toastydeath> and i'm completely refuting it
[22:13:32] <anonimasu> eh?
[22:13:45] <toastydeath> yes, plenty of quality, task-specific software can be developed by people in spare time
[22:13:54] <toastydeath> emc is a pretty good testiment to that
[22:14:19] <jepler> and luckily for us, we don't have to beg somebody or sign onerous contracts to be able to develop software on PCs
[22:14:31] <jepler> on game consoles it's different: you can't be an independent game developer
[22:14:31] <toastydeath> but compared to say, someone like Fanuc or Haas
[22:14:39] <SWPadnos> youhave obviously left software such as Apache, FireFox, and MySQL out of your analysis
[22:14:42] <jepler> because the console makers are engaged in this anticompetitive practice
[22:15:24] <toastydeath> which practice
[22:15:43] <SWPadnos> well, you have to use their development kit, which generally costs 5-6 gifures, for one
[22:15:48] <SWPadnos> figures, that is
[22:16:14] <anonimasu> mysql is comercial if you've missed that.
[22:16:27] <toastydeath> and apache and firefox both have paid programmers
[22:16:40] <jepler> I don't have any first-hand knowledge of just what agreement you have to enter with the console manufacturer, but there's simply no way that I can decide one day "I want to write software for the PS3" and do it without their blessing.
[22:17:00] <jepler> and it's more than "you have to pay for a dev kit"
[22:17:10] <toastydeath> right, because they lose money on the console
[22:17:20] <toastydeath> and don't want people making money on their loss without recouping losses
[22:17:22] <SWPadnos> the original development of all 3 (plus Linux itself, let's not forget) was non-commercial. there are commercial entities that contribute code for their own reasons
[22:17:30] <SWPadnos> actually, I'm not positive about MSQL
[22:17:47] <jepler> SWPadnos: by "original development" of firefox, you mean mosaic?
[22:18:01] <SWPadnos> mosaic, and Mozilla (before it was corporatized)
[22:18:01] <toastydeath> swpadnos: and they would have remained small if it wasn't for people able to devote full-time work to coding them
[22:18:10] <jepler> toastydeath: I understand that their business plan would fail if they were not able to do this anticompetitive thing
[22:18:14] <SWPadnos> could be true
[22:18:33] <anonimasu> SWPadnos: what about them?
[22:18:44] <toastydeath> jepler: they either have to sell the console at a price that nobody can afford, or exclude small developers who won't make them much money anyway
[22:18:54] <toastydeath> and focus on the best return for investment
[22:18:57] <SWPadnos> no, that's a false dichotomy
[22:19:07] <toastydeath> how so?
[22:19:07] <anonimasu> no it's not..
[22:19:11] <anonimasu> it's indeed true.
[22:19:35] <SWPadnos> they can sell the console at a loss, and then reqiore licensing for all games sold so they can make a buck
[22:19:35] <toastydeath> you CAN be a small development shop and get a contract, but, like most things "real business"
[22:19:40] <toastydeath> you need money up front to get started
[22:19:56] <SWPadnos> or, they can sell the console at a profit, and allow anyone to write games/other apps for the console
[22:20:13] <anonimasu> SWPadnos: it's easier to enforce who gets access to tools..
[22:20:15] <jepler> (reportedly, nintendo makes a profit on each Wii and DS sold -- the hardware is not sold at a loss)
[22:20:30] <toastydeath> and the wii is not exactly a powerhouse
[22:20:33] <toastydeath> nor is the DS
[22:20:41] <toastydeath> i love mine anyway, though
[22:20:41] <jepler> * jepler gives toastydeath yet another blank stare
[22:20:42] <anonimasu> then come up with a sane way to handle licencing..
[22:21:01] <anonimasu> I guess that will happen sooner or later though.
[22:21:11] <SWPadnos> consider the PC paradigm - you don't need a license to develop software for the platform
[22:21:12] <toastydeath> they came up with a sane way to handle licencing, they just happened to exclude the weekend warrior faction when they did o
[22:21:19] <toastydeath> *so
[22:21:25] <anonimasu> lol
[22:21:43] <jepler> I understand that the console makers have come up with this very effective way to limit the developer pool, in order to make their business model work
[22:22:03] <jepler> but this model is unjust and if a console maker had a monopoly I suspect it would be illegal.
[22:22:18] <toastydeath> thankfully anyone with the interest can start a company with venture capital
[22:22:19] <SWPadnos> they're probably not doing it to limit the developer pool, that's just a consequence of the "subscription" mentality
[22:22:21] <toastydeath> and create their own console
[22:22:28] <jepler> SWPadnos: well yes
[22:22:53] <jlmjvm> SWPadnos:which 2 pci express video cards did you say you got to work with emc?
[22:22:55] <anonimasu> hm, anyone can make windows programs if they buy a compiler..
[22:23:13] <anonimasu> but why should microsoft provide one?
[22:23:23] <jepler> anonimasu: I am not saying they should
[22:23:26] <SWPadnos> jepler, I wouldn't count on anything I've said as definitive - I've never used either with a machine attached
[22:23:37] <anonimasu> I dont think you can say that it's anticompetitive..
[22:23:42] <SWPadnos> however, both a 7800GT and 7300LE seemed to have good latency nunmbers
[22:23:48] <anonimasu> err monopoly.. or whatever..
[22:23:54] <jepler> I'm not saying they should provide a development kit
[22:24:11] <jlmjvm> thanks
[22:24:25] <SWPadnos> oops - that was meant for jlmjvm, not jepler
[22:24:50] <jepler> I'm saying that it is anticompetitive to work to exclude developers who don't enter into contracts with them from developing software for these devices.
[22:25:09] <toastydeath> i don't see how that's true
[22:25:13] <anonimasu> me neither
[22:25:25] <jepler> * jepler gives one last blank stare to the room, and leaves
[22:25:35] <SWPadnos> heh
[22:25:45] <anonimasu> it dosent mean I like it the least.
[22:26:05] <anonimasu> anticompetitive would be to prevent developers from developing on their console.
[22:26:08] <toastydeath> just because a person wants to do something, doesn't mean the person who owns the copyrights and patents on a device needs to allow public access to said property
[22:26:23] <SWPadnos> well, that's not always true, you see
[22:26:40] <SWPadnos> and aside from that, it's not a matter of copyright or patent
[22:26:51] <toastydeath> They own the software, do they not?
[22:26:56] <anonimasu> it's a matter of the right to profit from your own IP..
[22:26:57] <toastydeath> the development kit?
[22:27:05] <toastydeath> that's property, that right there
[22:27:15] <toastydeath> i don't have to let someone use my lathe if i don't want to.
[22:27:28] <toastydeath> and if I decide to rent time on it, it's going to be on my terms.
[22:27:39] <toastydeath> if you want to turn crap, build your own lathe.
[22:27:48] <toastydeath> or find someone who has one with terms you approve of.
[22:28:28] <SWPadnos> if Nintendo wants to make a game for the PS3, Sony can tell them to go screw themselves, since they are the single source for development tools, and they control all licensing of software for the PS3
[22:28:36] <toastydeath> yes
[22:29:00] <SWPadnos> so this is a monopoly on software available for a given platform
[22:29:07] <toastydeath> no, that's not how monopolies work
[22:29:19] <toastydeath> there are other consoles
[22:29:22] <SWPadnos> since there are multiple platforms available, it's not an absolute monopily
[22:29:26] <SWPadnos> right
[22:29:35] <toastydeath> and they are not crushing other, smaller consoles
[22:29:55] <SWPadnos> however, though it isn't an absolute market monopoly, it is a monopoly on a given market (ie, the PS3 game market)
[22:30:05] <SWPadnos> vs. the "console game market"
[22:30:07] <toastydeath> so i fail to see what the problem is
[22:30:22] <toastydeath> there are other consoles, and independant consoles
[22:30:27] <toastydeath> sounds like a pretty competative market
[22:30:46] <SWPadnos> well, it's the same old "hood welded shut" thing people get into for OSes
[22:30:54] <toastydeath> which is entirely legal
[22:31:01] <toastydeath> even without patents on software.
[22:31:04] <SWPadnos> if I buy the hardware, then I decide I'd rather use it for something other than gaming, I can't do that
[22:31:09] <toastydeath> yes, you can
[22:31:16] <toastydeath> just don't expect the manufacturer to support you
[22:31:22] <anonimasu> and how is that their problem
[22:31:23] <anonimasu> ?
[22:31:24] <SWPadnos> because I can't afford (or qualify for, probably) a development kit
[22:31:28] <toastydeath> so?
[22:31:33] <toastydeath> you own the hardware.
[22:32:25] <toastydeath> You buy parts made of plastic all the time
[22:32:36] <toastydeath> that doesn't mean you get to have the dimensions and engineering drawings for the mold that made them
[22:32:40] <SWPadnos> eh - it's a moot point, especially if you're an ultra-capitalist
[22:33:17] <SWPadnos> sure, but if I want to use my screwdriver as a pry-bar, I can do that. if there's a license that says it's not allowed, I can't (legally)
[22:33:21] <toastydeath> i just don't think i should be entitled to tell other people what to do with the things they've made.
[22:33:46] <toastydeath> So go buy a screwdriver without a license.
[22:33:48] <SWPadnos> it's a question of whether you believe corporations should have more protection than individuals
[22:33:53] <toastydeath> Or start a company that provides screwdrivers without licenses.
[22:34:27] <SWPadnos> personally, I believe that individuals are entitled to more protections than corporations
[22:34:31] <toastydeath> sure, that is true
[22:34:48] <toastydeath> i don't, however, agree that individuals should be able to dictate to corporations what to do with the corporation's property.
[22:35:04] <toastydeath> When you buy a PS3, that's yours. You can hit it with a hammer, or crack it, or whatever you want.
[22:35:18] <toastydeath> You are not entitled to the blueprints on how it's made, either in hardware or software
[22:35:28] <SWPadnos> wll, we'll have to leave it there, I've got to go move cars around so we can get some snow tires on my wife's vehicle
[22:35:32] <SWPadnos> *well
[22:35:34] <toastydeath> good luck
[22:35:38] <SWPadnos> heh - thanks
[22:36:08] <jlmjvm> wish we would get some snow
[22:46:57] <The_B> The_B is now known as The_Ball
[22:54:29] <Dogfishguzzler> So what are you guys gonna teach me tonight?
[22:55:19] <Dogfishguzzler> I hope it's not 'how to recognize a salad fork'
[22:58:10] <dmess> its usually the one above the bowl...
[22:58:30] <Dogfishguzzler> I call that a lighter but to each his own.
[22:59:19] <dmess> where you from?
[23:00:31] <jepler> toastydeath: so you would side with a modchip maker every time, right?
[23:01:40] <jepler> there should be no law preventing a modchip maker from developing a modchip that allows backup and homebrew, and no civil liability for doing so either
[23:03:57] <Dogfishguzzler> I'm in NC now but from SC
[23:04:03] <Dogfishguzzler> I like to cut up when I can dmess, I try to not act like a fool when someone needs help though.
[23:04:15] <Dogfishguzzler> I work with a guy who has a long history with machine tools; service etc. He seems fairly handy when it comes to finding whats wrong but in the last week of coming here and reading what I've been given to read I think I know more about what makes the machines do what they do than he does.
[23:05:03] <BigJohnT> what's a modchip?
[23:05:17] <Dogfishguzzler> jepler: How about kids who unlock phones? Should they be crucified or glorified
[23:05:55] <Dogfishguzzler> Its the hardware equivilant of a hack BigJohnT
[23:06:07] <BigJohnT> ok
[23:06:10] <jepler> BigJohnT: broadly, it's a modification to a game console that makes it possible to use out-of-region, backup, and/or homebrew software on them
[23:06:45] <BigJohnT> thanks jepler
[23:07:17] <jepler> no relevance to emc except that I was pontificating earlier that I think it's unjust that mere mortals can't develop software for video game consoles
[23:07:47] <BigJohnT> ok I was just wondering out loud what it was
[23:08:08] <BigJohnT> so they keep you locked out as opposed to computer software
[23:08:12] <jepler> right
[23:10:15] <BigJohnT> Anyone get a chance to check out the facing software I wrote in python?
[23:10:57] <Dogfishguzzler> I would like to see it just so I can see how those scripts are written.
[23:11:09] <BigJohnT> it's on emcwiki
[23:11:34] <BigJohnT> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Simple_EMC_G-Code_Generators
[23:11:57] <BigJohnT> There is a couple at the moment one that generates the g code for facing a part
[23:12:15] <BigJohnT> and on that does boltcircles that another guy did
[23:12:43] <jepler> BigJohnT: the GUI looks fine to me.
[23:13:07] <BigJohnT> thanks, any thing I need to add to it?
[23:13:59] <jepler> BigJohnT: for "tool diameter" I tried typing "1/8" and of course it wasn't accepted. It would be nice if it was.
[23:14:07] <Dogfishguzzler> I don't have the desire to learn py that good lol.
[23:14:20] <BigJohnT> ok, I didn't think of that
[23:14:35] <BigJohnT> most cad programs will do conversion and math even
[23:15:02] <BigJohnT> Dogfish: it is fun
[23:15:03] <jepler> here's a simple way to accept a fraction (no whole-number part is accepted, though):
[23:15:07] <jepler> def my_float(s):
[23:15:09] <jepler> if "/" in s:
[23:15:12] <jepler> n, d = s.split("/", 1)
[23:15:14] <jepler> return float(n) / float(d)
[23:15:17] <jepler> return float(s)
[23:15:31] <Dogfishguzzler> thats cool
[23:15:47] <BigJohnT> ok let me play with that a bit here
[23:16:41] <JymmmEMC> skunkworks: thanks for the url.
[23:18:01] <jepler> BigJohnT: you might want to give the user a choice about what corner of the material (0,0) is -- I would be more likely to put the lower-left corner there than the upper-left corner
[23:18:45] <BigJohnT> the corner near to you and to the left ie. the face on the movable jaw?
[23:18:59] <JymmmEMC> and where Z0 is.... top of material -or - bottom of material
[23:19:44] <Dogfishguzzler> Not to interupt here, but if I write a macro for emc does it have to be in py?
[23:19:53] <jepler> Dogfishguzzler: no
[23:20:05] <jepler> Dogfishguzzler: it is simply any program that writes gcode that emc understands
[23:20:15] <Dogfishguzzler> Okay, good
[23:20:17] <BigJohnT> what situation do you use Z0 on the bottom of the material
[23:20:31] <Dogfishguzzler> I'm 1337 with macros but not so much with python
[23:20:53] <JymmmEMC> BigJohnT: When you use the table as the reference point and offset for material thickness.
[23:21:33] <BigJohnT> ok, I usually zero my tools off the top and set the tool offset from there in my CNC mill (Anilam)
[23:22:03] <JymmmEMC> So do I, but it's an option I've seen some use.
[23:22:07] <dmess> fixed jaw... is best
[23:22:45] <BigJohnT> jepler: I understand the conversion now from fractions to decimal, thanks I'll add that
[23:22:47] <JymmmEMC> BigJohnT: Take for example that you don't car how thick the material is, you just want the finished piece 2" thick.
[23:23:53] <Dogfishguzzler> +1 on the fraction converter japler that is sweet and simple even I can understand it....and that's sayin something.
[23:23:58] <Dogfishguzzler> jepler*
[23:24:38] <BigJohnT> OK, Jymmm I with you now. It really helps to chat with other users...
[23:24:43] <jepler> Dogfishguzzler: good
[23:24:47] <dmess> fraction converter???? jepler did youteach them how to divide???
[23:25:07] <BigJohnT> no he taught us how to split
[23:25:12] <BigJohnT> not spit
[23:25:32] <Dogfishguzzler> He did dmess
[23:25:39] <Dogfishguzzler> via python
[23:25:46] <BigJohnT> yea it's cool
[23:25:49] <dmess> cool
[23:26:25] <jepler> BigJohnT: with the changes shown here, I can run face.py directly from the AXIS GUI .. when you press "generate" then "write to axis and quit", the code automatically loads in axis and you're read to go. http://pastebin.ca/807734
[23:26:36] <jepler> this requires that your configuration file be set up to understand that .py files can be run to make gcode: http://linuxcnc.org/docs/2.2/html/gui_axis.html#r1_11_1
[23:27:14] <Dogfishguzzler> Why would I use a script for a macro though? Is there a rreason I can't write them like I always do in regular Gcode? Does EMC not under stand G00 X#5 or G00 XV5; anything like that?
[23:27:22] <jepler> BigJohnT: I was a little baffled because I had to remember to click "generate gcode" first before I clicked "write" .. perhaps if the gcode hasn't been generated yet, the buttons that will save or copy the gcode can generate it first?
[23:27:52] <jepler> Dogfishguzzler: emc has parameters (G0X#5) and flow control, but personally I find it a very painful language to write in compared to Python..
[23:28:31] <Dogfishguzzler> That my friend is subjective. Looking @ John's script I had to take two asprin.
[23:28:34] <jepler> Dogfishguzzler: http://linuxcnc.org/docs/2.2/html/gcode_main.html#cha:O-Codes <-- for more information
[23:28:55] <BigJohnT> jepler: Ok that makes sense to have the save button generate it too.
[23:29:08] <Dogfishguzzler> Parametric code on a machine does'nt bother me one bit.
[23:29:35] <jepler> Dogfishguzzler: I am sure it depends on what you're familiar with
[23:29:50] <Dogfishguzzler> that's what I'm saying.
[23:29:56] <jepler> I programmed "real languages" for 10+ years before I became aware of gcode .. while it is almost certainly the opposite for many people who are using emc
[23:30:30] <Dogfishguzzler> LOL as if Gcode isn' a real language. Your just jealous of me aren't you?t
[23:31:16] <jepler> Dogfishguzzler: I put '"real language"' in quotes for a reason..
[23:31:20] <Dogfishguzzler> Acually I'm always jealous of you hardcore coders. I never stuck with it long enough to get good. But I can make a machine make sweet potatoe pie if I'm hungry.
[23:31:22] <BigJohnT> jepler: on the pastebin link is the top code the one I want or the bottom code?
[23:31:53] <BigJohnT> nevermind I see that the bottom is the same
[23:31:57] <jepler> BigJohnT: yeah
[23:31:57] <dmess> same here DogFish...
[23:32:15] <jepler> BigJohnT: do you understand what the format means? "-" is a line I removed, "+" is a line I added, and the other lines are shown to give context
[23:32:27] <dmess> APT 360 is a rather high level language for its day...
[23:32:29] <jepler> BigJohnT: there is actually a program ("patch") which can read a file like this and automatically make the changes in your file
[23:32:52] <jepler> patch face.py < patchfile
[23:32:53] <BigJohnT> no I didn't know that
[23:33:07] <jepler> (after downloading the "raw" link on that page to "patchfile")
[23:33:16] <BigJohnT> ok
[23:36:46] <dmess> its all fortran... and assembler modules.... the source is like a wad a spagetti...headin' here there and everywhere..
[23:37:03] <skunkworks> sounds like foxpro..
[23:37:35] <dmess> this was developed over 15 yrs startin in the '60s
[23:38:12] <dmess> and it seems the NEVER removed anything.... just comment it out
[23:39:03] <dmess> no comment saying why?? what was hacked or slashed..
[23:43:22] <BigJohnT> jepler: do I need to add face.py to axis or does IN_AXIS = os.environ.has_key("AXIS_PROGRESS_BAR")
[23:43:36] <BigJohnT> tell face.py that axis is running?
[23:45:51] <BigJohnT> jepler: when I run IN_AXIS = os.environ.has_key("AXIS_PROGRESS_BAR") from python then do IN_AXIS I get False
[23:46:08] <BigJohnT> yet i have axis up and running...
[23:48:15] <Dogfishguzzler> Good link jepler.
[23:48:22] <Dogfishguzzler> Thanks.