#emc-devel | Logs for 2007-06-03

[01:11:36] <SWPadnos> jmkasunich, still here?
[01:11:42] <jmkasunich> who me?
[01:11:42] <SWPadnos> yep
[01:11:50] <SWPadnos> planning on hitting HGR this week?
[01:12:04] <jmkasunich> maybe
[01:12:15] <jmkasunich> got something you want me to check out?
[01:12:18] <SWPadnos> yep :)
[01:12:22] <SWPadnos> that servo cable
[01:12:27] <SWPadnos> http://www.hgrindustrialsurplus.com/search-products/product-detail.aspx?id=30-285-142&searchtable=2&sortExpression=&SortASC=&pageSize=50&currentPageIndex=0
[01:12:29] <jmkasunich> URL?
[01:13:12] <SWPadnos> also: http://www.hgrindustrialsurplus.com/search-products/product-detail.aspx?id=30-285-149&searchtable=2&sortExpression=&SortASC=&pageSize=50&currentPageIndex=0
[01:13:39] <jmkasunich> bookmarked
[01:13:50] <SWPadnos> this digikey page has the pin pattern I'm looking for: http://dkc3.digikey.com/PDF/T072/P0342.pdf
[01:13:53] <jmkasunich> my plan is to take off either friday afternoon or all day friday
[01:13:54] <SWPadnos> it's 24-7
[01:14:01] <jmkasunich> if the latter, I'll probably visit HGR
[01:14:06] <SWPadnos> ok. you're heading out on Saturday, right?
[01:14:15] <jmkasunich> yep
[01:14:27] <jmkasunich> packing the truck friday afternoon/evening
[01:14:33] <SWPadnos> right
[01:14:56] <jmkasunich> can you send me an email with those links
[01:15:03] <SWPadnos> yep
[01:15:13] <jmkasunich> and other specifics - if your motor has a male conn and you need female on the cable, etc
[01:15:13] <SWPadnos> bummer - it looks like they aren't open on the weekend
[01:15:24] <jmkasunich> only one saturday a monty
[01:15:28] <jmkasunich> month
[01:15:39] <SWPadnos> and probably not late enough for me to arrive before closing
[01:15:56] <jmkasunich> you mean on weekdays?
[01:16:03] <SWPadnos> no, on the one Sat per month
[01:16:13] <SWPadnos> (if it's this Sat)
[01:16:13] <jmkasunich> they'll be open sat 6/16, 8 till 2
[01:16:17] <jmkasunich> its not, its next sat
[01:16:28] <SWPadnos> yep. closing waaaay too early ;)
[01:16:48] <jmkasunich> didn't you say something about heading home on monday?
[01:17:08] <SWPadnos> I'm not sure yet whether I'll leave on Sunday or Monday
[01:17:23] <jmkasunich> you'd have to leave at the crack of dawn to get to cleveland by 5pm when they close on monday
[01:17:24] <SWPadnos> wither way, I'll end up in the Cleveland area on Monday or Tuesday
[01:17:33] <jmkasunich> are you doing the trip in two days?
[01:17:37] <SWPadnos> and I sure don't want to do that
[01:17:41] <SWPadnos> yep
[01:18:02] <jmkasunich> ok, then all you gotta do is take your overnight stop in Cle
[01:18:03] <SWPadnos> Galesburg -> Cleveland area (Ashtabula, I think), then CLE -> Essex
[01:18:35] <SWPadnos> You're pretty close to Independence, aren't you?
[01:18:41] <jmkasunich> 20ish miles
[01:18:55] <jmkasunich> independence is south, I'm east (of downtown)
[01:19:17] <SWPadnos> ok. that's where I'm staying on the way out. it looked like it's not too far from I-80 on the maps
[01:19:30] <jmkasunich> when are you heading out?
[01:19:43] <SWPadnos> Saturday
[01:20:04] <jmkasunich> so arriving in cle sat night, and galesburg sun night
[01:20:13] <SWPadnos> yep
[01:21:12] <jmkasunich> coming from essex, are you on 80 or 90?
[01:21:18] <jmkasunich> I'd have thought 90, farther north
[01:21:36] <SWPadnos> oh right - they aren't the same road yet in Cleveland
[01:21:46] <jmkasunich> ;-)
[01:21:48] <SWPadnos> I'll be on 90 through BUffalo then Erie
[01:21:50] <SWPadnos> then on through
[01:22:04] <SWPadnos> I guess it's in Indiana that they merge
[01:22:04] <jmkasunich> then independence is kinda out of the way
[01:22:39] <SWPadnos> hmmm. maybe I'll look at that again. I stayed in Westlake before, which is nice because It's on the west side of town for the second day
[01:22:47] <jmkasunich> google maps "cleveland"
[01:22:55] <SWPadnos> heh - been there, done that ;)
[01:23:00] <jmkasunich> independence is south of 480
[01:23:11] <jmkasunich> 90 comes in along the lake, and leaves along the lake
[01:23:42] <jmkasunich> they join just west of avon point (W of cle)
[01:24:09] <jmkasunich> you really want to stay closer to 90 I think
[01:24:28] <SWPadnos> hmmm. is 480 usually packed on a Sunday morning?
[01:24:36] <jmkasunich> probably not
[01:25:12] <jmkasunich> 90->271->480->hotel->480->80 ?
[01:25:16] <SWPadnos> I had considered coming down 271 then 480 to get to this hotel
[01:25:16] <SWPadnos> right
[01:25:42] <SWPadnos> the hotel is 1/2 mile south of 480
[01:25:47] <SWPadnos> along 77
[01:25:51] <jmkasunich> probably on rockside?
[01:26:07] <SWPadnos> Quarry Lane, but it's very close to Rockside
[01:26:33] <SWPadnos> exit 156
[01:27:05] <jmkasunich> if you don't want to go out of your way, there are two hotels right off 90 in eastlake
[01:27:15] <jmkasunich> http://maps.google.com/maps?f=l&hl=en&q=hotel&near=eastlake&sll=41.588743,-81.548767&sspn=0.718985,0.587769&ie=UTF8&latlng=41638595,-81372393,10524538747516336078&ei=shhiRoHJKo-crQKw7vmADg&cd=10
[01:28:29] <SWPadnos> though it will be a Sunday morning, I think if I stay east of the city, it's best to not have to drive straight through it
[01:28:34] <SWPadnos> or stay west of the city
[01:30:28] <jmkasunich> heh, I found a hotel a couple blocks from HGR - perfect for the stop on the way home
[01:30:35] <SWPadnos> heh
[01:31:44] <jmkasunich> might have to dodge winos tho
[01:31:46] <SWPadnos> heh
[01:32:18] <jmkasunich> regarding driving thru on Sunday morning...
[01:32:18] <SWPadnos> ah - I should look at those exit numbers better - they go up as you head EAST!
[01:32:48] <jmkasunich> 90 is a ghost town on sunday mornings
[01:33:05] <jmkasunich> I used to race on a sailboat out of edgewater, east of downtown
[01:33:28] <jmkasunich> 9:00 on a sunday morning, there was just about nobody on the road
[01:34:10] <SWPadnos> ok. that's a good thing :)
[01:35:30] <SWPadnos> hmmm. with the jeep, the few extra miles may outweigh the slight savings on the hotel
[01:35:48] <jmkasunich> the independence one is cheap?
[01:35:55] <SWPadnos> relative to other Hiltons ;)
[01:36:04] <SWPadnos> $100-ish
[01:36:12] <jmkasunich> one of the two in eastlake is a motel 6, if you want cheap! (the other one is a days inn)
[01:36:23] <SWPadnos> well, there's cheap, and then there's inexpensive ;)
[01:36:39] <jmkasunich> thats why I gave you the link for the days inn
[01:36:41] <SWPadnos> I get free breakfast (real breakfast) and internet at Hiltons, plus points!
[01:37:12] <jmkasunich> dang, I didn't realize how big HGR really is
[01:37:16] <SWPadnos> heh
[01:37:19] <jmkasunich> (looking at aerial photo)
[01:38:11] <jmkasunich> http://maps.google.com/maps?f=l&hl=en&q=hotel&near=eastlake&ie=UTF8&t=h&om=1&ll=41.566447,-81.541107&spn=0.010805,0.009184&z=16
[01:38:38] <SWPadnos> interesting. the Days Inn is only about $25 less than the Doubletree (or Hampton Inn in Westlake)
[01:39:11] <SWPadnos> holy crap - is HGR that entire huge building?
[01:39:31] <jmkasunich> I'm not entirely sure - I think they may be the full width but only about half the depth
[01:39:34] <SWPadnos> north / west of Euclid Ave
[01:39:51] <jmkasunich> yeah, the really big one in the middle of the pic
[01:40:54] <jmkasunich> its possible they only go from the northeast wall to the line where the roof turns blue
[01:41:12] <SWPadnos> ok. that's still pretty big
[01:41:22] <jmkasunich> 10 acres according to the website
[01:41:36] <SWPadnos> hmmm
[01:42:50] <SWPadnos> it would have to be the whole building then, it seems
[01:42:58] <jmkasunich> an acre is 200ft x 200ft, approximately, isn't it?
[01:43:09] <SWPadnos> yep, something like 208 square
[01:43:21] <SWPadnos> I think it's (1/25 mile)^2
[01:44:02] <SWPadnos> oops - not wuite. it's 1/640 of a square mile (thank you google calculator)
[01:46:08] <jmkasunich> the whole building is 16 acres if I scaled it right
[01:46:15] <jmkasunich> 840x840 feet
[01:47:10] <jmkasunich> they open at 8am on weekdays
[01:47:11] <SWPadnos> I measured closer to 750x750 :)
[01:47:23] <SWPadnos> well, that may work for me on the trip back
[01:47:27] <jmkasunich> you really should locate your return trip hotel such that you can stop there before you head for essex
[01:49:06] <SWPadnos> I'm looking into it right now
[01:49:20] <SWPadnos> I wonder if I could buy a trailer and a machine to put on it there ;)
[01:49:32] <jmkasunich> probably not the trailer
[01:51:16] <jmkasunich> there's a U-haul place not too terribly far from HGR, could do a one way rental (unless you want to have a trailer)
[01:51:45] <SWPadnos> I wouldn't mind a trailer, but it seems they're either too light duty for machinery, or they're very expensive
[01:51:53] <alex_joni> hi guys
[01:51:58] <jmkasunich> hi alex
[01:51:59] <SWPadnos> hi Alex
[01:52:04] <alex_joni> * alex_joni just got home ;)
[01:52:27] <SWPadnos> hmmm. late night^Wmorning for you :)
[01:52:40] <alex_joni> yeah, pretty late (close to 5am)
[01:53:02] <jmkasunich> SWP: http://maps.google.com/maps?f=l&hl=en&q=trailer&near=euclid&sll=41.39626,-81.53708&sspn=0.173327,0.146942&ie=UTF8&ei=WB5iRq3hLY3crQLov72ADg&cd=2&cid=41583350,-81531948,15529451904756150821&li=lmd&z=14&t=m
[01:53:18] <jmkasunich> print out the phone number in case you see a machine you really like at HGR
[01:53:30] <jmkasunich> (you can see the HGR building in the lower left of the pic)
[01:53:33] <SWPadnos> heh
[01:53:38] <SWPadnos> thanks
[01:54:09] <jmkasunich> something tells me the one way rental price for a suitable trailer will approach the buying one
[01:54:27] <SWPadnos> ot may be close. it's only ~700 miles though
[01:54:34] <SWPadnos> s/ot/it/
[01:57:28] <SWPadnos> wimpy trailers they have. the open-air ramp trailer only has a 1650 pound capacity
[01:57:42] <jmkasunich> no car-haulers?
[01:58:21] <SWPadnos> only the two-wheel type
[02:00:32] <jmkasunich> I think U-haul doesn't trust the general public with heavy stuff - folks are likely to try hauling an Expedition on a trailer behind a VW beetle
[02:00:39] <SWPadnos> heh
[02:01:03] <SWPadnos> the jeep actually has a 6500 pound tow capacity, so it's plenty capable
[02:01:12] <SWPadnos> but it still can't tow a hummer
[02:01:13] <jmkasunich> ISTR when I wanted to move my 13" lathe (from HGR btw) Uhaul wouldn't help me
[02:01:33] <SWPadnos> you usually need Ryder or another more pro outfit
[02:01:34] <jmkasunich> you already have a bport and a CNC lathe, right?
[02:01:48] <SWPadnos> well, the lathe isn't here, and now they want more money for it, so who knows
[02:01:53] <jmkasunich> so the odds of finding a "can't pass it up" machine at HGR aren;t high
[02:02:00] <jmkasunich> oh, I thought that was a done deal except for delivery
[02:02:07] <SWPadnos> so did I
[02:02:11] <jmkasunich> bummer
[02:02:20] <SWPadnos> they originally wanted $500, now they want $2500
[02:02:22] <jmkasunich> heh
[02:02:26] <SWPadnos> I may offer them $1500
[02:02:41] <SWPadnos> it's amazingly clean, and it's still turning out aerospace-tolerance parts
[02:03:13] <jmkasunich> sounds good - even at $1500, if you don't have to spend hundreds on trailer or other haulage, its not bad at all
[02:03:27] <SWPadnos> yep. I can get it here for $100 or thereabouts
[02:03:35] <SWPadnos> (friend of a friend has a flatbed tow truck)
[02:03:41] <SWPadnos> maybe $50 and a case of beer
[02:04:52] <jmkasunich> A quick scan of HGR finds the cheapest CNC lathe is $2K, most are $3K
[02:05:19] <SWPadnos> yeah. and I'm sure they aren't capable of turning out parts with a tenth accuracy
[02:05:59] <jmkasunich> bet this one isn't bad, unless its wrecked:http://www.hgrindustrialsurplus.com/search-products/product-detail.aspx?id=42-138-002&searchtable=1&sortExpression=&SortASC=&pageSize=50&currentPageIndex=0
[02:06:28] <SWPadnos> I'm suspicious that I couldn't fit that through my garage door
[02:06:35] <SWPadnos> but it does look pretty nice
[02:06:38] <jmkasunich> http://www.hgrindustrialsurplus.com/search-products/product-detail.aspx?id=42-146-007&searchtable=1&sortExpression=&SortASC=&pageSize=50&currentPageIndex=0
[02:06:43] <jmkasunich> theres the 2K one
[02:06:59] <SWPadnos> that one's probably older than me
[02:07:03] <jmkasunich> lol
[02:07:09] <SWPadnos> well, older than skunkworks anyway
[02:08:11] <jmkasunich> sometime near the end of the fest when your plans firm up, we gotta take notes - if you are going to HGR monday or tuesday morning, I might play hooky and meet you there
[02:08:16] <SWPadnos> it may be older than me :_)
[02:08:42] <SWPadnos> okie dokie. I need to decide when I'm heading home so I can cancel one of my hotel reservations :)
[02:08:50] <jmkasunich> ;-)
[02:09:17] <alex_joni> * alex_joni heads to bed
[02:09:26] <SWPadnos> good - um - night
[02:09:31] <jepler> see you alex
[02:09:37] <alex_joni> yeah, thanks
[02:12:29] <alex_joni> sun is coming up.. gotta hide
[02:12:29] <SWPadnos> heh
[02:12:29] <jmkasunich> hey jepler - you want some junk^H^H^H^Hstuff from HGR, right?
[02:12:30] <SWPadnos> actually, that mori seiki looks very nice
[02:12:30] <jmkasunich> bport series 1 for $1300
[02:12:30] <jmkasunich> seems low, wonder whats wrong with it
[02:12:30] <jmkasunich> wow, series 2 for $1000
[02:13:24] <SWPadnos> and only $500 fora surfac egrinder
[02:13:28] <jmkasunich> moog - I think that is the one with hydraulic controls... $1K
[02:13:31] <SWPadnos> argh - what happened to my space bar
[02:13:57] <SWPadnos> that series 2 CNC is a beast
[02:15:10] <jmkasunich> need a toolbox? http://www.hgrindustrialsurplus.com/search-products/product-detail.aspx?id=01-967-004&searchtable=1&sortExpression=&SortASC=&pageSize=50&currentPageIndex=0
[02:15:29] <SWPadnos> ick
[02:15:42] <SWPadnos> that's not much less than the stainless one I got (new) at Costco
[02:15:50] <jmkasunich> lol
[02:15:56] <jmkasunich> you comparing costco to vidmar?
[02:16:07] <SWPadnos> sure - never heard of vidmar ;)
[02:16:26] <jmkasunich> that toolbox probably weighs 300 lbs empty, and will hold several thousand pounts
[02:16:29] <jmkasunich> pounds
[02:16:58] <SWPadnos> ok. mine only weighs ~175 pounds and can only hold ~2000 or so
[02:17:09] <jmkasunich> they are the industry standard for holding all sorts of tooling, stuff that gets heavy
[02:17:11] <SWPadnos> though the wheels may not be good for all that
[02:17:21] <SWPadnos> that one is bigger than mine as well
[03:59:37] <petev> cradek, u there?
[14:26:19] <a-l-p-h-a> quiet in here today
[14:53:53] <rayh> tried to respond to Claude regarding his CD issue but got...
[14:53:57] <rayh> Hi. This is the qmail-send program at trueband.net.
[14:53:58] <rayh> I'm afraid I wasn't able to deliver your message to the following addresses.
[14:53:58] <rayh> This is a permanent error; I've given up. Sorry it didn't work out.
[14:53:58] <rayh> <cem@aznex.net>:
[14:53:58] <rayh> Sorry, the subscriber you are emailing has opted not to receive email from your address.
[14:57:14] <SWPadnos> it must be all the high volume spammers on up.net :)
[14:57:30] <rayh> caught me.
[15:02:16] <SWPadnos> hey Ray - got a CHNC question for you
[15:02:34] <rayh> k
[15:02:53] <SWPadnos> I was all set to buy one for $500 locally, but now they want a bit more money for it (the bums)
[15:03:05] <rayh> Okay.
[15:03:10] <rayh> how much is a bit?
[15:03:16] <SWPadnos> what do you think an operational CHNC with GE control is worth?
[15:03:35] <SWPadnos> it's currently churning out aerospace parts and holding a tenth accuracy
[15:03:37] <rayh> operational
[15:03:47] <SWPadnos> yes, operational :)
[15:03:58] <rayh> Yea right!
[15:04:13] <SWPadnos> they're asking for $2500 now, which I think it's worth, but is more than its value for me
[15:04:43] <rayh> Some questions to ask. Is this their only one? Do they have spares for it?
[15:05:00] <SWPadnos> it's the last of 4 that they're getting rid of for updated machinery
[15:05:09] <SWPadnos> but I know where to get spares ;)
[15:05:40] <SWPadnos> one big advantage is that it's ~6 miles from my house
[15:05:51] <rayh> Sure.
[15:06:27] <rayh> What GE control is this?
[15:06:39] <SWPadnos> the 7 foot tall blue one :)
[15:06:49] <rayh> Cause all the ones I've seen had AB controls.
[15:07:00] <SWPadnos> hmmm. I think it was a GE
[15:07:06] <SWPadnos> it's only an NC right now, not CNC
[15:07:20] <rayh> Ah. A GE 550
[15:07:26] <SWPadnos> could well be
[15:07:32] <rayh> Matt has some spares for those.
[15:07:44] <rayh> I don't have anything.
[15:07:46] <SWPadnos> well, I'd probably junk the control pretty quickly
[15:08:02] <SWPadnos> I'll have lots of parts for sale at next year's swap meet
[15:08:09] <rayh> And the spindle speed is a snowmobile type dual belt clutch.
[15:08:28] <SWPadnos> hmmm. what does that mean in english? :)
[15:08:34] <rayh> Nothing wrong with it.
[15:08:37] <SWPadnos> do you have to manually change gear ranges?
[15:08:58] <rayh> The spindle motor is a single speed 3 phase motor.
[15:09:02] <SWPadnos> ah, ok
[15:09:19] <SWPadnos> hmmm. the ones you have use DC motors for variable speed, right?
[15:09:24] <rayh> a belt from it goes to an idler
[15:09:27] <rayh> Yep.
[15:09:59] <rayh> The idler shaft is driven toward or away from the motor using a little stepper and ball screw.
[15:10:46] <rayh> That motion forces the belts closer to or further away from the shaft.
[15:11:09] <rayh> It changes the size of the pulleys.
[15:11:08] <SWPadnos> ok - so it's a dual-belt sheave kind of thing?
[15:11:11] <SWPadnos> right
[15:11:18] <rayh> exactly.
[15:11:38] <SWPadnos> kinda like my BPT vari-speed
[15:11:38] <rayh> That is the same machine that Matt is converting.
[15:11:41] <SWPadnos> cool
[15:11:46] <rayh> Yes.
[15:12:26] <SWPadnos> well, I was thinking of offering $1500 for it, though I'm a little annoyed that they decided to up the price
[15:13:16] <rayh> I can understand that. One I know of here was given away by the shop.
[15:13:27] <rayh> And I didn't even offer to take it.
[15:13:37] <SWPadnos> heh
[15:14:26] <rayh> If you could get it for that or 1k you'd do well.
[15:14:48] <SWPadnos> I'm not so sure that's the ideal lathe for me anyway - it's got the 5c collet, and I don't know what kind of chuck I could attach
[15:15:01] <rayh> It would be possible to continue to use the existing axis drives.
[15:15:17] <rayh> Matt has a spindle that would replace the resolvers with enocders.
[15:15:18] <SWPadnos> I was thinking that. it has resolvers and encoders, right? (or was it just resolvers)
[15:15:33] <rayh> resolvers and tachs
[15:15:35] <SWPadnos> ah, ok
[15:15:44] <rayh> The existing drives require the tach
[15:15:55] <SWPadnos> they're analog velocity input?
[15:16:09] <rayh> What they do is simply null the command against the tach return.
[15:16:35] <rayh> Yes. +-10 although you can change that with a resistor on the drive.
[15:16:39] <SWPadnos> ok
[15:16:59] <SWPadnos> do you know if the resolvers have an index output?
[15:17:21] <rayh> Not exactly
[15:17:41] <SWPadnos> of course, a zero position would indicate an index crossing ...
[15:18:08] <rayh> When the sine and cosine cross zero there is some sort of output from the counter board.
[15:18:33] <SWPadnos> hmmm. if they're analog resolvers (ie, they don't feed the sine/cosine out), then a microcontroller could be made to provide position output
[15:18:46] <SWPadnos> ok, so you ge tquadrant outputs
[15:18:54] <rayh> Sure. There are several folk who make converters.
[15:19:05] <rayh> vigilant among them.
[15:19:35] <SWPadnos> I know there are resolver -> quadrature devices. I was thinking of something that just tracks position directly
[15:19:40] <SWPadnos> since that's what EMC wants
[15:20:37] <rayh> I don't follow your thought.
[15:20:50] <SWPadnos> well, that could be because it's not a well-formed thought
[15:21:15] <SWPadnos> a simple analog input on a Mesa card could give a direct position readout
[15:21:39] <SWPadnos> the hardware can check the quadrant crossings and keep a turns counter directly
[15:21:59] <SWPadnos> so you'd have a direct position register in the mesa card (like the quadrature encoder readers)
[15:22:16] <rayh> Ah but the crossings only happen 4 times per rev.
[15:22:42] <SWPadnos> that's not problem
[15:22:47] <rayh> actually there is a gear increase in the feedback package
[15:22:48] <SWPadnos> ^a
[15:23:16] <rayh> I don't think you would get very good resolution.
[15:23:38] <SWPadnos> you get quadrature for the full cycle, so you know what quadrant you're in. then you use a relatively high resolution A/D for the position within the quadrant
[15:23:54] <SWPadnos> like a 16 bit one
[15:24:08] <rayh> Not quadrature
[15:24:12] <SWPadnos> (though you don't need that much)
[15:24:53] <SWPadnos> just using hte "index" outputs from the sin/cos zero crossings is fine
[15:25:16] <rayh> We are not understanding the nature of a resolver.
[15:25:18] <SWPadnos> heh
[15:25:36] <SWPadnos> I know how a resolver works, but I don't know the outputs you get on the ones they used
[15:25:37] <rayh> sine wave in,
[15:26:11] <rayh> Maybe I don't understand.
[15:26:43] <SWPadnos> ie, I don't know how much of the analog stuff is in the resolver, and how much is in the board it connects to
[15:26:49] <rayh> the angular position of the resolver is the difference between in signal and the two out signals.
[15:29:09] <SWPadnos> in any case, the effort to use the resolvers is probably a lot more than the cost of a USDigital encoder :)
[15:31:10] <rayh> and the cost of matt's adaptor.
[15:31:30] <SWPadnos> heh
[15:31:39] <rayh> The converters I've seen were in the $300+ cost range
[15:31:49] <SWPadnos> but with an NC lathe, I can make an adapter :)
[15:34:57] <rayh> Yes you can. And most likely the feedback packages you have use a split stepup gear to reduce backlash.
[15:35:00] <rayh> The one
[15:35:23] <rayh> s I've got here have an eccentric
[15:35:45] <rayh> turned into the tach and resolver to reduce backlash.
[15:36:22] <SWPadnos> hmmm - interesting
[15:36:24] <rayh> You would need to look at a feedback package before you design that adaptor.
[15:36:59] <rayh> I can help with getting a feedback package loose if you decide to get the machine.
[15:37:15] <rayh> There is a tiny bellow drive coupling.
[15:37:36] <rayh> That is easy to break and costs $80+ to replace.
[15:37:39] <SWPadnos> cool. thanks. I know you've always wanted to visit Vermont ;)
[15:37:46] <SWPadnos> bummer
[15:37:48] <rayh> Yes it do.
[15:38:04] <SWPadnos> well, I guess I'm trying to find out if it's worthwhile hardware. $1000-$1500 isn't a bad price for a lathe of that size, and I think this one is in very good shape, so the age isn't much of a factor
[15:38:24] <rayh> Good thinking.
[15:39:09] <SWPadnos> I even have a jeep, so I can move it myself (if I can find several hundred friends to get it on and off the trailer ;) )
[15:39:32] <rayh> There you go. I'll plan the visit after you get it into the shop.
[15:39:45] <SWPadnos> ok, so sometime in 2009 at the rate I move :)
[15:39:56] <rayh> anytime.
[15:40:08] <SWPadnos> sounds good to me
[15:40:33] <SWPadnos> our B&B doesn't have as nice a location as yours, but the area does have some nice features
[15:40:53] <rayh> I like varied views.
[15:41:04] <SWPadnos> (like a Thai restaurant closer than 150 miles :) )
[15:42:36] <rayh> There you go.
[15:43:09] <rayh> I had Thai food at one restaurant in Shanghi this time.
[15:43:19] <SWPadnos> I bet that was good
[15:43:22] <SWPadnos> or authentic
[15:43:30] <rayh> Not to bad.
[15:44:41] <SWPadnos> Thailand is one of those places I'd like to go one day
[15:45:03] <SWPadnos> since Shanghai is closer, I'd think the thai food there would be pretty good
[15:45:36] <SWPadnos> (closer than here, that is)
[15:47:21] <rayh> Most Oriental food here is different than in the country it purports to be from.
[15:47:30] <SWPadnos> heh - too true
[15:47:40] <SWPadnos> hence the "good or authentic" comment :)
[15:47:43] <rayh> You running 6.06 ubuntu there?
[15:47:52] <SWPadnos> I have it on one machine
[15:47:57] <SWPadnos> hmmm. nope, two
[15:48:07] <rayh> I wonder if xvidtune is on 6.06
[15:48:14] <SWPadnos> well, 3 if you count the laptop. but none of them are on at the moment
[15:48:22] <SWPadnos> one sec
[15:49:05] <SWPadnos> it should be there
[15:49:10] <SWPadnos> or available
[15:57:32] <jepler> it's in a package called "xvidtune"
[16:01:15] <rayh> Thanks jeff.
[16:53:46] <jmkasunich> morning folks
[16:53:52] <jmkasunich> (I'm sure its morning somewhere)
[17:00:15] <jepler> hi jmk
[17:00:21] <jmkasunich> hi
[17:00:47] <jmkasunich> I'm making great progress - in another year or two I might be able to write a GUI program ;-)
[17:01:10] <jepler> one of the gimmicks I saw on another PC CNC software last year was using a jogwheel to advance through the program a short distance for each click. turns out you can do it in emc as well, using the feed-hold pin. http://emergent.unpy.net/index.cgi-files/sandbox/jogwar.comp http://emergent.unpy.net/index.cgi-files/sandbox/jogwar.hal
[17:01:45] <jmkasunich> I've heard some folks say they want to go backwards too... thats a bit harder
[17:01:50] <jepler> yeah
[17:01:56] <jepler> no matter which way you turn the knob, this goes forwards
[17:04:22] <jmkasunich> whats div for?
[17:04:45] <jepler> I want 'out' to be on for N periods for each 'count'
[17:04:51] <jmkasunich> ah
[17:05:45] <jmkasunich> delta = in - data.old_in
[17:05:52] <jmkasunich> runtime += delta * scale
[17:06:12] <jmkasunich> if (runtime > 0) { out =1 ; runtime-- } else { out = 0 }
[17:06:21] <jmkasunich> isn't that simpler?
[17:06:26] <jepler> yeah
[17:07:08] <jmkasunich> only works for one direction of rotation, you'd need an abs to work like yours, and an if(delta<0) delta = 0 to prevent reverse rotation from moving the machine
[17:07:45] <jmkasunich> I think the latter would be preferred - if someone has an ohcrap momemt because they turned the wheel too far, their instinct is gonna be to turn it back
[17:08:49] <jmkasunich> actually, I think I'd not ignore reverse rotation, but let it unwind runtime
[17:08:57] <jepler> yeah I was thinking that
[17:09:01] <jmkasunich> runtime += delta (delta might be negative)
[17:09:07] <jmkasunich> if runtime < 0 runtime = 0
[17:09:58] <jmkasunich> and to make it easier to config, make runtime a float, rename scale as "runtime-per-click" or something shorter with the same meaning
[17:10:09] <jmkasunich> runtime-- becomes runtime-=period
[17:11:50] <jmkasunich> have you tried jogwar on a real machine?
[17:12:04] <jepler> no
[17:12:29] <jmkasunich> I wonder if its gonna have the same "bumpiness" issues that the original non-RT implementation of jogwheels had
[17:12:41] <jmkasunich> when feedhold changes state, the machine accels/decels at the limit
[17:13:55] <jepler> it may, but I don't know what else you would do
[17:14:22] <jepler> (my jogwheel is usb so there's no telling when the report of a count will come in -- that's bound to make it worse)
[17:14:34] <jmkasunich> it might be fun to experiement with it at the fest, cradek will have his lathe (I think he has a jogwheel) and I have my stepper testbed with 3 motors and a wheel
[17:14:53] <jmkasunich> I'm thinking adaptive feed could be used
[17:15:16] <jmkasunich> runtime = the desired "amount" of motion that you have dialed in so far, assuming full speed
[17:15:37] <jmkasunich> but we could modify adaptive feed so if runtime is small, the machine goes slow
[17:15:50] <jmkasunich> as you get ahead of the machine by cranking the dial, feed would increase
[17:16:14] <jmkasunich> runtime -= period would become runtime -= period*current_adaptive_feed
[17:17:12] <jmkasunich> or something, I might be off in left field
[17:18:22] <jepler> I updated the files on my website to the "runtime-per-count" version
[17:19:10] <jmkasunich> is "variable" new?
[17:19:20] <jepler> yes it's recent in 2.2.
[17:19:20] <jmkasunich> I thought you needed to make a struct to have local data
[17:19:27] <jmkasunich> cool
[17:20:03] <jmkasunich> oops, you might want to remove the nofp ;-)
[17:20:50] <jepler> indeed
[17:33:34] <jmkasunich> the indexing in the python reference leaves a bit to be desired
[17:33:41] <jmkasunich> I'm trying to understand "global"
[17:34:23] <jmkasunich> I think I originally misunderstood it to mean I should use "global foo" at the beginning of any function where I want to access foo
[17:35:04] <jmkasunich> but now I think it means I should use "global foo" at the beginning of a function where I'm gonna define foo and want it to be global
[17:36:55] <jepler> if your function *uses* foo but does not define it, clearly it must be the global foo
[17:37:35] <jmkasunich> so the statement is not needed if I only read globals - but to write to them I need global
[17:37:46] <jepler> by contrast, things you assign are taken to be local names
[17:37:57] <jepler> .. except in the presence of a 'global' statement, which is put at the top of the function by convention
[17:39:07] <jmkasunich> that explains why "make_menus(window)" needs to declare the menus it makes global, but "disable_buttons" which grays out some buttons when no file is loaded doesn't need global to accss the menus and buttons
[17:40:48] <jmkasunich> so far, I have open, close, and save working - of course they don't actually display the file they've opened or anything, that would be too impressive
[17:48:38] <jmkasunich> it seems that data members of a class instance aren't "global variables", even if the class instance is a global
[17:48:52] <jmkasunich> report_menu.add_checkbutton(label='Pins', variable=fpga.report_pins)
[17:49:02] <jmkasunich> fpga is a global instance
[17:49:25] <jmkasunich> and report_pins is created by __init__, with a value of 1
[17:49:44] <jmkasunich> but the button doesn't change the class member
[17:58:04] <jmkasunich> report_menu.add_checkbutton(label='Pins', variable=report_pins, onvalue=1, offvalue=0, command=print_report_values)
[17:58:03] <jmkasunich> report_menu.add_checkbutton(label='Modules', variable=report_modules, onvalue=1, offvalue=0, command=print_report_values)
[17:58:34] <jmkasunich> these two buttons both change color when I click either one, and neither one actually modifies report_pins or report_modules
[17:59:05] <jmkasunich> report_pins and report_modules are assigned to (thus created) at the outer level, before the menus are created
[17:59:45] <jmkasunich> I've tried it with and without declaring report_pins and report_modules global in the make_menu function that contains those two lines
[18:00:23] <jmkasunich> its like the "variable=report_pins" option doesn't exist
[18:02:40] <jmkasunich> that option is being parsed - if I change "variable" to "vaariable" I get an error
[18:02:54] <jmkasunich> and if I change "report_pins" to "reeport_pins" I get an error
[18:03:04] <jmkasunich> so its finding the global variables
[18:37:42] <SWPadnos> hmmm. I would like to get fwd/rev jogging through g-code working
[18:37:55] <jmkasunich> jogging thru g-code?
[18:38:01] <SWPadnos> yes
[18:38:02] <jmkasunich> thats rather twisted
[18:38:08] <SWPadnos> sort of
[18:38:18] <jmkasunich> why not just use G0
[18:38:42] <SWPadnos> not stepping fwd/back, but acutally controlling the effective time elapsed as discussed
[18:39:04] <SWPadnos> because I may want to check a program for workholder clearance or something
[18:39:28] <jmkasunich> oh, not "jogging thru g-code" as in "using g-code to control jogging", but "using a jogwheel to control the execution of the g-code"
[18:39:36] <SWPadnos> that would probably require additional features, such as the ability to edit upcoming blocks without reloading the program
[18:39:40] <SWPadnos> yes
[18:39:51] <SWPadnos> sorry for my english language ambiguity :)
[18:40:04] <jmkasunich> reversing g-code is a VERY non-trivial operation
[18:40:21] <SWPadnos> I know it is, but I don't understand conceptually why that should be so
[18:40:33] <jmkasunich> imagine an o-word loop
[18:40:35] <cradek> SWPadnos: to anyone else I'd just say `PGA'
[18:40:45] <SWPadnos> if it's due to the implementation, then it's understandable
[18:40:48] <jmkasunich> when running backwards, how many times to you "un-run" the loop
[18:40:56] <SWPadnos> PGA - no thanks, I don't play golf
[18:41:12] <jmkasunich> cradek: do you know any tkinter?
[18:41:14] <cradek> it seems possible to allow going backwards 'a bit'
[18:41:29] <SWPadnos> at least back to the start of this block
[18:41:38] <jmkasunich> you could probably unwind the current move, but not multiple lines of g-code
[18:41:54] <SWPadnos> as for unrolling loops, you'd have to do it on a line-by-line basis
[18:42:01] <SWPadnos> ie, step into the sub or loop
[18:42:10] <cradek> not block, maybe this basic move
[18:42:13] <SWPadnos> it does get complex for loops
[18:42:22] <jmkasunich> SWP: imagine the g-code equivalent of:
[18:42:26] <jmkasunich> n = 4
[18:42:28] <cradek> (a block can make an arbitrary number of moves)
[18:42:36] <SWPadnos> sure, I thought I was using the word "block" as it's used in the G-code manual (ie, one line)
[18:42:40] <jmkasunich> while ( n-- ) { do something }
[18:42:51] <jmkasunich> if you find yourself at the end of that loop, you have no clue what N was
[18:43:02] <cradek> you're only thinking of the simplest lines then
[18:43:07] <cradek> lines/blocks
[18:43:13] <SWPadnos> jmkasunich, true, but some checkpointing would help there
[18:43:30] <SWPadnos> cradek, right - not G76 or some other canned cycle
[18:44:00] <jmkasunich> cradek: do you know any tkinter?
[18:44:05] <jmkasunich> * jmkasunich is stumped
[18:44:06] <cradek> I think starting at the gcode level is the wrong approach - you should think about the canon/motion level, where the commands are fairly invertable
[18:44:22] <cradek> jmkasunich: sort of - it's just tk/tcl commands wrapped up
[18:44:48] <jmkasunich> you can make checkbuttons as part of menus
[18:45:08] <cradek> yes
[18:45:08] <jmkasunich> they are supposed to use a -variable to tell them what global var to toggle
[18:45:18] <cradek> ok
[18:45:29] <jmkasunich> it doesn't toggle the variable
[18:45:57] <cradek> hmm
[18:46:06] <jmkasunich> it is definitely referring to the proper variable - if I don't declare the global var before creating the menu entry I get an error
[18:46:09] <cradek> there's also -command I bet
[18:46:22] <cradek> sorry I'm not clueful enough
[18:46:27] <jmkasunich> yes, I have command invoking a function that prints out the value of the global
[18:46:31] <jmkasunich> and that works
[18:46:51] <jmkasunich> but there seems to be no way to get at the state of the checkbutton _except_ the variable option
[18:47:02] <jmkasunich> I've read the man page end-to-end a couple times
[18:47:36] <jmkasunich> used entryconfigure to look at the data, I get: 'variable': ('variable', '', '', '', <parsedVarName object at 0x81cde60>)
[18:47:54] <jmkasunich> but I don't know how to tell if that is the same as the global variable I want to modify
[18:48:50] <cradek> in axis, there are checkbuttons on menus, and they have only -command and -variable
[18:48:58] <cradek> it seems like it just works...?
[18:49:12] <jmkasunich> yeah, it should just work
[18:49:20] <jmkasunich> I think its some kind of scoping thing
[18:50:07] <cradek> sorry, clueless
[18:50:15] <jmkasunich> I just added a call to entrycget("variable") to my "print the value" function
[18:50:34] <jmkasunich> it prints the original value of the global, not the toggled value, and not the current value
[18:51:02] <SWPadnos> does this help at all? http://mail.python.org/pipermail/python-list/2000-August/049640.html
[18:51:34] <jmkasunich> its like there are three variables - the global (which I can change using other methods and verify that it changes), the one its supposed to toggle (which is buried somewhere, but must be changing because the indicator changes) and the one that entrycget(variable) reports, which never changes
[18:53:13] <jmkasunich> that guy isn't using -variable at all
[18:53:20] <SWPadnos> nope - oops :)
[18:55:26] <cradek> jmkasunich: http://pastebin.ca/534289
[18:55:31] <cradek> here is the basic (working) idea
[18:55:50] <cradek> you don't declare the variable first
[18:58:10] <jmkasunich> hmm
[18:58:18] <jmkasunich> if I don't declare it first, I get an error
[18:58:31] <jmkasunich> lemme see if I can reduce this to a trivial example
[18:58:46] <SWPadnos> is the menu created before any other access to the variable?
[18:59:09] <jmkasunich> I assign a value to the variable (to create it) before creating the menu
[18:59:27] <SWPadnos> and that's all
[18:59:29] <SWPadnos> ?
[19:01:56] <cradek> haha Top Donations: Donald Becker: $2.00
[19:02:02] <cradek> (pastebin.ca)
[19:09:01] <jmkasunich> http://pastebin.ca/534316
[19:09:12] <jmkasunich> can you run this and see if you get the same results I do?
[19:10:10] <jmkasunich> toggling the checkbuttons changes nothing
[19:10:39] <jmkasunich> clicking INC increments the globals, toggling the buttons still does nothing, AND the values returned by entrycget are still the initial values
[19:10:57] <cradek> yep
[19:12:15] <jmkasunich> not sure how much more simplification I can do - guess I could get rid of the function make_menus and do everything inline
[19:13:27] <jmkasunich> no surprizes there, same thing
[19:13:57] <jmkasunich> if I don't define cb1 and cb2 before creating the menu, I get an error
[19:14:06] <jmkasunich> so that at least doesn't work like it does in Tk
[19:14:57] <cradek> http://effbot.org/tkinterbook/checkbutton.htm
[19:15:27] <jmkasunich> http://pastebin.ca/534327 <-- simpler version
[19:16:35] <cradek> jmkasunich: see the Patterns section on that page
[19:16:56] <jmkasunich> what is IntVar()?
[19:17:12] <cradek> gives a tcl integer variable I guess
[19:17:23] <cradek> then you later use var.get()?
[19:18:06] <jmkasunich> so Tk has its own private stash of variables
[19:18:08] <jmkasunich> grrr
[19:18:17] <cradek> apparently
[19:18:32] <cradek> tkinter is just a 'layer of crap' smeared on top of tcl
[19:19:05] <SWPadnos> variable=
[19:19:07] <SWPadnos> Associates a Tkinter variable to the button. When the button is pressed, the variable is toggled between offvalue and onvalue. Explicit changes to the variable are automatically reflected by the buttons. (variable/Variable)
[19:19:21] <SWPadnos> note that they say "a tkinter variable"
[19:19:25] <jmkasunich> yep
[19:19:36] <SWPadnos> so I guess that means the typeVar() thing is mandatory
[19:19:45] <SWPadnos> don't you love spaghetti documentation? :)
[19:19:59] <jmkasunich> seems to work that way
[19:21:26] <cradek> btw that was the first google hit for 'tkinter checkbutton' :-)
[19:23:28] <jmkasunich> I read the tk manpage (jepler pointed out that tkinter docs aren't that good and recommended reading the tk ones), I also actually looked at that page or something like it
[19:24:09] <jmkasunich> the thing that escaped me is that tkinter treats tk variables as a completely different world, sort of
[19:24:41] <jmkasunich> sort of, because defining the variables in py was good enough to avoid the error message, but not good enough to transfer the data from tk to py
[19:24:49] <cradek> yeah unfortunately you have to learn the tcl level (tk man pages) and then also learn how to glue the layers together (Tkinter.IntVar())
[19:25:10] <jmkasunich> http://pastebin.ca/534356 <-- working version
[19:25:23] <cradek> I'm not a fan of tkinter as maybe you can tell
[19:25:33] <jmkasunich> neither is jepler
[19:25:33] <cradek> wish there was a good way to do gui stuff in python, but I don't knwo what it is
[19:25:47] <cradek> wish there was a good way to do gui stuff, but I don't know what it is
[19:25:49] <jmkasunich> but I guess its the best of the less than stellar alternatives
[19:26:20] <cradek> for jepler I'm sure it's the devil he knows - I'm not sure that makes it the best of anything
[19:27:38] <cradek> hmm, now that emc is done, I'm not sure what to work on
[19:27:44] <jmkasunich> lol
[19:28:17] <jmkasunich> you got a 5i20 card didn't you?
[19:28:24] <cradek> yes
[19:28:32] <jmkasunich> that means you have a vested interest in helping me with this stuff ;-)
[19:28:39] <cradek> hm I should stick it in this machine
[19:32:02] <cradek> I think the current driver would work fine for me - I won't have need for step/dir from it
[19:32:11] <jmkasunich> darn
[19:32:42] <jmkasunich> FYI, once the new is working, I'm going to do my best to deprecate the old one
[19:33:06] <jmkasunich> doesn't mean much if people want to keep using it tho
[19:33:37] <cradek> it'll be able to do a combination of step/dir, pwm, dio?
[19:33:42] <jmkasunich> yep
[19:33:50] <jmkasunich> and encoders, and etc
[19:34:02] <cradek> that sounds very nice
[19:35:17] <jmkasunich> yay! the fix works in my program
[19:35:25] <cradek> yay
[19:35:53] <jmkasunich> I only spend the last hour or two fighting with those fscking buttons
[19:35:55] <jmkasunich> nefs
[19:37:07] <jmkasunich> cradek: dunno if you saw the discussion with SWPadnos last night - anything at HGR you want me to look at for you?
[19:37:17] <jmkasunich> planning on going friday morning
[19:37:23] <cradek> hmmmm
[19:46:14] <cradek> http://www.hgrindustrialsurplus.com/search-products/product-detail.aspx?id=20-760-068
[19:46:21] <cradek> wonder what this funny little thing is
[19:47:33] <jmkasunich> expensive is what it is
[19:48:01] <cradek> yep
[19:48:06] <cradek> it says 5 axis
[19:48:21] <tomp2> http://www.barer.com/proteo5.html
[19:48:22] <jmkasunich> older version of http://www.barer.com/proteo5.html ?
[19:48:55] <jmkasunich> if they took a digit off the price, I'd be interested - its nice and small
[19:49:21] <jmkasunich> I'd prefer they take the last digit, but even the first might make it interesting
[19:49:50] <cradek> I agree
[19:50:08] <jmkasunich> did you see the drawings at the bottom of the page I posted
[19:50:17] <jmkasunich> that machine does look kinda like the drawing
[19:50:21] <cradek> yes
[19:50:23] <jmkasunich> interesting axis arrangement
[19:50:51] <jmkasunich> bet its pretty rigid
[19:51:12] <cradek> it would be fun to play with
[19:51:20] <cradek> not sure what I'd generate 5 axis gcode with though
[19:51:30] <alex_joni> synergy
[19:52:00] <SWPadnos> cnc-toolkit
[19:52:07] <cradek> python
[19:52:11] <SWPadnos> though you'd need gmax or 3dsmax for modeling
[20:24:10] <petev> cradek, u there?
[20:28:34] <alex_joni> he was a bit earlier
[20:28:46] <petev> alex, I have a question for you
[20:28:50] <alex_joni> shoot
[20:29:03] <petev> I tried to install the rtai source package, but it installed the wrong version
[20:29:12] <petev> I got version 3.2 instead of 3.3
[20:29:25] <petev> does the apt repo need updating?
[20:29:24] <alex_joni> hmm.. probably you got the one from the main ubuntu repo
[20:29:35] <alex_joni> no, you probably used the wrong name
[20:29:45] <petev> hmm, I only saw one in synaptic
[20:29:49] <petev> rtai-source
[20:30:04] <petev> did you build it with the LXRT scheduler enabled?
[20:30:23] <alex_joni> http://linuxcnc.org/emc2/dists/dapper/emc2.1/source/
[20:30:27] <alex_joni> don't think so
[20:30:27] <petev> I wanted to try some user level stuff, but it didn't seem to work right
[20:30:57] <petev> I have the standard apt sources from the live CD
[20:31:32] <petev> definitely didn't get that package
[20:31:50] <SWPadnos> petev, try this: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#Realtime_on_64_bit_systems
[20:32:03] <SWPadnos> it has a good set of RTAI build options
[20:32:12] <SWPadnos> (which presumably also work on 32-bit systems)
[20:32:21] <alex_joni> SWPadnos: that's sim
[20:32:28] <alex_joni> err.. nope
[20:32:38] <petev> I don't see linuxcnc in the sources.list anywhere
[20:32:49] <alex_joni> petev: then you don't have the proper sources
[20:32:54] <petev> should that have been in the list from the live install?
[20:32:59] <alex_joni> yes
[20:33:06] <petev> hmm, that is strange
[20:33:12] <alex_joni> but most probably synaptic or whoever changed it
[20:33:16] <petev> I'll have to check my other machine
[20:33:31] <alex_joni> do you have emc2 installed on that machine?
[20:33:36] <petev> oops, I see it there
[20:33:47] <alex_joni> emc2 or emc2.1 ?
[20:33:53] <petev> wonder why the package deosn't show?
[20:33:56] <SWPadnos> it's -magma, not -rtai
[20:33:56] <petev> emc2.1
[20:34:06] <petev> magma is a version of rtai
[20:34:11] <alex_joni> petev: dpkg -l | grep rtai
[20:34:15] <petev> supposedly 3.3, I think
[20:34:23] <alex_joni> please run that command
[20:34:32] <SWPadnos> right, but that's the package name, so if you look for rtai, you may not find it
[20:34:50] <SWPadnos> linux-image-2.6.15-magma or some such
[20:35:02] <petev> petev@bridgeport:~$ dpkg -l | grep rtai
[20:35:03] <petev> ii rtai-modules-2.6.15-magma 3.3-1+aj4 rtai modules for Linux (kernel 2.6.15-magma)
[20:35:03] <petev> petev@bridgeport:~$
[20:35:03] <petev> only shows the rtai-modules-2.6.15-magma installed
[20:35:17] <alex_joni> try apt-get source rtai-modules-2.6.15-magma
[20:35:20] <SWPadnos> do the same with magma instead of rtai
[20:35:46] <petev> alex, that works, but doesn't install it as a package
[20:35:51] <petev> just gets the source file
[20:35:59] <alex_joni> apt-get source doesn't ever install a package
[20:36:01] <SWPadnos> apt-get source never installs a package
[20:36:04] <SWPadnos> :)
[20:36:08] <petev> I wanted to use module-assistant, but it needs a package
[20:36:08] <alex_joni> I thought you want to get the sources to recompile?
[20:36:15] <petev> yes, with m-a
[20:36:17] <alex_joni> petev: no it doesn't
[20:36:28] <alex_joni> eeek.. m-a?
[20:36:29] <petev> well I probably don't know how to use it very well
[20:36:30] <alex_joni> what for
[20:36:45] <petev> I like it, it's simple and saves time
[20:36:46] <alex_joni> you simply place the rtai source under /usr/src/
[20:37:10] <petev> m-a will do the whole build, just saves time
[20:37:20] <jepler> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?UbuntuBreezyPackages section Rebuilding linux-image-2.6.12-magma and rtai-modules-2.6.12-magma
[20:37:23] <petev> you can use a-i and it gets, builds, installs all at once
[20:38:05] <jepler> I don't know that chris's packages work with module-assistant, that's not what these instructions have you use
[20:38:30] <alex_joni> just fyi I used almost the same commands as cradek
[20:38:34] <petev> ok, Ill take a look at what chris did
[20:38:54] <petev> jepler, do you know the diff between motionLine and currentLine in task status?
[20:39:00] <jepler> petev: no
[20:39:20] <jepler> current may be the line most recently interpreted, while motion would be the one for ongoing motiojn
[20:39:22] <petev> hmm, what do you use to determine the current g-code line in axis?
[20:39:24] <jepler> but that's just a guess -- I suggest you read the source code
[20:39:41] <petev> I think readLine is for the interp
[20:40:06] <alex_joni> petev: the important part for rtai is to place the stuff in /usr/src/modules/rtai-3.3
[20:40:17] <alex_joni> then make-kpkg will pick them up and generate the deb
[20:40:29] <jepler> I use the field that is called 'stat.id' in Python
[20:40:57] <petev> alex_joni, ok, I can use make-kpkg also
[20:41:01] <jepler> that's apparently motion.traj.id in the NML structure
[20:41:16] <petev> hmm
[20:41:57] <petev> the comments say that is an id, do you use it as the line number?
[20:42:48] <alex_joni> * alex_joni heads to bed
[20:42:51] <jepler> yes
[20:42:58] <petev> interesting
[20:43:01] <alex_joni> going on a business trip tomorrow .. (again)
[20:43:02] <petev> alex_joni, gn
[20:43:08] <alex_joni> night all
[20:43:12] <jepler> alex_joni: have fun
[20:43:25] <alex_joni> jepler: yeah, hopefully it'll work out this time :/
[23:56:07] <cradek> does anyone remember if we decided to have CDs, gcode cheat sheets, or neither made?
[23:56:17] <cradek> if CDs, I better get to work soon
[23:56:33] <SWPadnos> if either, it's gotta get done soon
[23:56:47] <SWPadnos> I think we liked cheat sheets better, because they're more permanent
[23:56:51] <jmkasunich> yeah
[23:56:56] <cradek> ok
[23:57:08] <SWPadnos> but I also think they'll take longer to get done. I forgot to ask about lead time
[23:57:30] <cradek> I guess I'll want to make an updated CD anyway
[23:57:30] <jmkasunich> did we decide where we're gonna get them done?
[23:57:34] <jmkasunich> (and by implication, who)
[23:57:45] <SWPadnos> the local print shop here had good pricing
[23:58:16] <SWPadnos> @28/100 double sided prints, half sheet
[23:58:21] <SWPadnos> $28/100
[23:58:35] <cradek> that's laminated?
[23:58:39] <SWPadnos> lamination with a pre-drilled hole, $45/100
[23:58:41] <SWPadnos> for 3 mil
[23:58:46] <SWPadnos> $68 for 5 mil
[23:58:52] <cradek> ah
[23:59:01] <cradek> still that does seem like good prices
[23:59:04] <SWPadnos> still <$100 printed, laminated, and drilled
[23:59:06] <SWPadnos> yep
[23:59:08] <jmkasunich> so is the total $28 for printing plus the lam charge?
[23:59:18] <SWPadnos> yes
[23:59:26] <SWPadnos> $73 for 3 mil, $96 for 5 mil
[23:59:28] <jmkasunich> so what we need is the artwork
[23:59:37] <SWPadnos> yep, and I can email a PDF to her
[23:59:45] <cradek> how can I help?
[23:59:51] <SWPadnos> generate the PDF ;)