Back
[02:59:19] <KimK> Referring to Chapter 6.2.7 of the Integrator's Manual,
http://www.pastebin.ca/1294776 , Is the ini file command HAL = command, now replaced by HALCMD = command? (It shows as HALCMD in the ini example I'm looking at, not mentioned in the Integrator's Manual) Are the SHUTDOWN and POSTGUI_HALFILE commands still valid?
[03:00:50] <jmkasunich> KimK: to be honest, I'm not sure
[03:00:56] <jmkasunich> simplest way would be to try it
[03:01:25] <jmkasunich> use a "show" command - it's output will be visible in the shell, but if it doesn't work, emc will start
[03:01:44] <jmkasunich> try HAL = "show", then HALCMD= "show" and see which works
[03:01:50] <jmkasunich> the "" may not be needed...
[03:02:18] <jmkasunich> as far as I know, SHUTDOWN and POSTGUI_HALFILE are correct
[03:02:51] <KimK> OK, thanks. I'll check a little later, still in reading mode
[03:03:50] <jmkasunich> I just checked the latest EMC script in CVS, HALCMD is correct
[03:04:11] <jmkasunich> I bet the HAL in the manual is a mistake
[03:04:50] <KimK> OK. Manual bug hereby reported?
[03:05:01] <jmkasunich> yeah, I'll dig into it
[03:05:44] <Skullworks> JMK - did you mount home or limit switches on your shoptask?
[03:06:04] <jmkasunich> I have a single switch on the quill, and another on the table (Y)
[03:06:06] <jmkasunich> nothing on X
[03:06:19] <jmkasunich> home only, not limit - I rely on soft limits once homed
[03:08:33] <Skullworks> hmm ok - I'm trying to get one set up for a friend - I'm also leaning towards using Axis for lathe config and Tkemc for mill so he won't forget which config he is running.
[03:09:10] <jmkasunich> eww
[03:09:13] <jmkasunich> I use axis for both
[03:09:27] <jmkasunich> the preview in mill mode is 3d, in lathe mode it is 2D
[03:09:47] <jmkasunich> the DRO for lathe mode is X, Dia, Z, the DRO for mill mode is X, Y, Z
[03:10:18] <jmkasunich> you could also use pyvcp to put a big-ole label in the right hand pane
[03:10:32] <Skullworks> I bet he will too - once past the initial learning curve
[03:10:35] <jmkasunich> in lathe mode I have a spindle speed bar-graph there, in mill mode nothing
[03:13:10] <Skullworks> I'm going to try using the 7i43 to make enough step pulses for the Geckos
[03:15:09] <jmkasunich> KimK: found and fixed, thanks for spotting that
[03:16:34] <KimK> No problem. I might have another one for you, give me a few minutes
[03:16:45] <jmkasunich> * jmkasunich hides
[03:16:51] <KimK> heh
[03:35:07] <cradek> the guis are so different - you wouldn't want to make anyone learn more than one
[03:35:49] <Skullworks> point taken
[03:43:14] <KimK> Referring to this example ini file, and Integrator's Manual chapters 6.2.9 and 6.2.9.1 ,
http://www.pastebin.ca/1294797 , there are HOME settings in the AXIS_n sub-sections that are not mentioned in the manual. There is a (different?) HOME setting (with multiple arguments) earlier in the TRAJ section. Is this a manual bug?
[03:45:21] <KimK> (I'm guessing that the example ini file is right again)
[04:00:46] <cradek> this is a messy thing. the AXIS_n HOMEs are joint homes. the TRAJ home is the world coordinate corresponding to the homed joint positions for nontrivkins machines.
[04:00:49] <cradek> I think.
[04:10:17] <KimK> OK. It's not giving me any problem, I just thought I should point out the discrepancy in case it helps. I'm just sitting and reading and comparing my example ini against the manual
[04:10:44] <cradek> http://www.linuxcnc.org/docs/devel/html/config_ini_config.html#sub:%5BTRAJ%5D-section
[04:11:08] <cradek> I think this matches my memory of [TRAJ]AXES - that's a good sign
[04:14:47] <KimK> OK. The html docs don't show the AXIS_n HOME either. How does the AXIS_n HOME differ from the AXIS_n HOME_OFFSET ?
[04:15:26] <cradek> http://www.linuxcnc.org/docs/devel/html/config_ini_homing.html
[04:15:50] <cradek> homing is so complex it gets its own document :-/
[04:17:45] <KimK> heh
[04:19:52] <KimK> I see the html docs have several new entries that are not in the manual. Not only the AXIS_n HOME we've been talking about, but HOME_IS_SHARED and HOME_SEQUENCE are likewise not in the (PDF) manual.
[04:20:22] <cradek> oh really - I though they were generated from the same lyx docs
[04:20:40] <cradek> I keep giving you URLs pointing to trunk's documentation - which version are the pdfs?
[04:21:03] <cradek> the corresponding trunk pdfs should be here:
http://www.linuxcnc.org/docs/devel/
[04:21:51] <KimK> V2.2 Integrators Manual, December 14, 2008.
[04:23:46] <KimK> Oh, now this is V2.3 Integrators Manual, December 27, 2008. Doesn't the EMC2 update system update it's own manuals?
[04:23:48] <cradek> that homing stuff is in the 2.3 version at the above url
[04:23:59] <cradek> yes it does
[04:24:18] <cradek> you're probably running 2.2 unless you built 2.3 (trunk) yourself
[04:24:43] <cradek> does look like 2.2 has some out of date documentation
[04:25:03] <cradek> but I lose track of what is and isn't in 2.2, since I've been running trunk for so long
[04:25:04] <KimK> I'll check for updates right now. My systems has been on for a few days and it never said "Updates Waiting".
[04:25:30] <cradek> 2.3 (2.3.0) is not released yet. 2.2.8 is the latest release.
[04:26:43] <KimK> Oh, OK. Yes, that's me, 2.2.8
[04:28:54] <KimK> My fault, didn't know I was using an old manual, even if the latest released. Sorry to be bothering everyone about stuff that's probably fixed already.
[04:29:17] <jmkasunich> well, you spotted at least one that wasn't fixed
[04:31:22] <KimK> Can't complain about outdated docs if the latest ones are only two weeks old, lol. So when is 2.3.0 scheduled to appear?
[04:31:50] <cradek> pretty surely before summer (cnc workshop)
[04:32:52] <KimK> OK. Any news on the workshop, BTW?
[04:34:41] <cradek> hmm, I thought there were 2009 dates on
http://cnc-workshop.com/
[04:34:55] <jmkasunich> swp noticed that they went away
[04:35:07] <jmkasunich> he noticed a week or so ago, no idea when it happened
[04:35:47] <KimK> I wondered if our host was OK.
[04:35:57] <cradek> yeah I wonder too.
[04:40:15] <KimK> I have another partition, would it be useful to help test release candidates from CVS periodically?
[04:41:08] <KimK> (I was saving it for Windows, but, pfffft. Haven't needed it so far. )
[04:42:05] <cradek> you wouldn't need another partition - you could just get the cvs and build trunk as run-in-place
[04:43:19] <KimK> Might it not occasionally screw things up? I don't want ever to be without a runnable current release.
[04:43:34] <cradek> it's perfectly safe
[04:43:44] <KimK> OK
[04:43:59] <cradek> just configure with --enable-run-in-place
[04:44:25] <jmkasunich> cradek: we don't have an equivalent to #include for g-code do we?
[04:44:40] <jmkasunich> I guess I should RTFM
[04:44:42] <SWPadnos> the closest would be O<name>
[04:44:55] <SWPadnos> where <name> is the name of a gcode file in your nc_files dir
[04:44:58] <cradek> O-call can search for another file I think
[04:45:10] <cradek> yeah that
[04:45:10] <jmkasunich> I'm writing a "usefull subroutines" for lathe
[04:45:14] <cradek> I've never tried it
[04:45:19] <cradek> ooh, neat
[04:45:20] <jmkasunich> roughing, facing, boring
[04:45:35] <jmkasunich> s/roughing/turning
[04:45:37] <SWPadnos> I don't think you can have multiple named subs in the same file
[04:45:43] <jmkasunich> all of the cycles rough and then finish
[04:45:56] <jmkasunich> I'm not doing names subs (yet)
[04:46:01] <SWPadnos> but if you made something called roughing.ngc, and then called O<roughing> call in another file, that would act like an include
[04:46:04] <jmkasunich> are they in 2.2, or is that a trunk feature
[04:46:12] <SWPadnos> trunk I believe
[04:46:17] <cradek> I don't know - I've never used them either
[04:46:28] <jmkasunich> I'll stick with numbers then, I'm running 2.2.8 on the shoptask
[04:47:11] <SWPadnos> I don't know that numbers work. they might
[04:47:47] <SWPadnos> well, within the same file they do, but I don't know that you'll get the include-like behavior with numbered subroutines
[04:47:56] <jmkasunich> I suspect I won't
[04:48:04] <jmkasunich> I guess that can wait till 2.3
[04:48:14] <SWPadnos> heh
[04:48:19] <SWPadnos> or TRUNK :)
[04:48:33] <jmkasunich> I'll just copy the file with my routines in it, rename to whatever the part is, and then add the calls at the end
[04:48:59] <jmkasunich> somebody should write canned cycles for lathe
[04:49:07] <cradek> I do a lot of turning/boring with G91 G17 G85 L...
[04:49:25] <jmkasunich> O10020 sub
[04:49:25] <jmkasunich> (TURN DIAMETER TO SHOULDER ON LEFT)
[04:49:25] <jmkasunich> (O10020 call [start-dia] [finish-dia] [left] [right] [feed-per-rev] [depth-per-pass] [finish-feed] [finish-depth])
[04:49:28] <jmkasunich> (assumes that you have a tool suitable for turning from right to left and facing outward)
[04:49:30] <jmkasunich> (assumes that 'right' is clear of the work)
[04:49:40] <cradek> err G85 for boring, G81 for turning
[04:49:59] <cradek> it works but isn't very ideal
[04:50:11] <cradek> I agree 'someone' should write some cycles
[04:51:40] <SWPadnos> I think Lerman should do it
[04:52:06] <jmkasunich> lathe cycles will bring up the dia/radius issue big-time
[04:57:23] <jmkasunich> cycles done
[04:57:34] <jmkasunich> now I just have to figure out tool offsets
[04:58:03] <jmkasunich> (for the switch from turning tool to boring)
[04:58:10] <jmkasunich> crap. midnight
[04:58:20] <cradek> does your spindle reverse?
[04:58:24] <jmkasunich> I think I'll figure it out in the morning
[04:58:25] <jmkasunich> no
[04:58:28] <cradek> darn
[04:58:45] <jmkasunich> thinking of using a backwards boring tool for OD and ID?
[04:58:46] <cradek> sometimes I use the boring bar for turning
[04:59:20] <jmkasunich> regular bar? bore in front and turn in back?
[04:59:34] <cradek> upside-down, bore in back
[04:59:38] <jmkasunich> ah
[04:59:51] <ds3> finally sane discussions on lathe
[05:00:09] <jmkasunich> boring spindle fwd, turning reverse, right?
[05:00:26] <cradek> yes
[05:00:53] <cradek> I tried to face with it the other day, which made the CSS+M4 bug apparent
[05:01:03] <cradek> (fixed now)
[05:01:40] <ds3> CSS made it in to EMC? I been away too long
[05:02:11] <cradek> yeah it's been there two? years
[05:02:16] <cradek> maybe only one
[05:02:42] <ds3> definitely not 2... one and a half maybe
[05:03:20] <jmkasunich> that's about right - 2007 EMC workshop
[05:03:25] <jmkasunich> or thereabouts
[05:03:27] <ds3> looks like EMC will have full lathe support before i can finish my lathe
[05:03:27] <cradek> yeah maybe summer '07
[05:03:59] <ds3> got distracted around Spring '07
[05:04:18] <ds3> is there a GUI to show the profile being currently cut?
[05:04:23] <ds3> (for lathe)
[05:04:28] <cradek> sure, AXIS does that
[05:05:09] <ds3> so axis has a mode that will show either a 3D round or a half round with the profile?
[05:05:33] <cradek> no, it shows the tool tip motion, not the part
[05:06:12] <ds3> but for mill, the results are a lot like the part
[05:06:32] <cradek> maybe I don't understand what you're asking then
[05:06:51] <cradek> maybe you should check it out yourself
[05:07:26] <ds3> I intend to but holiday activities is preventing me from putting in time to do the fun stuff
[05:08:23] <ds3> I am hiding at the moment ;)
[05:09:03] <ds3> http://www.cncezpro.com/Images/canning.gif
[05:09:15] <ds3> the equivalent of the dotted lines; that is what I am askingg
[05:09:35] <cradek> yes the lines
[05:10:33] <ds3> nice
[18:02:58] <ChanServ> [#emc] "This is the #emc channel - talk related to the Enhanced Machine Controller and general machining. Website:
http://www.linuxcnc.org/, wiki at
http://wiki.linuxcnc.org/"
[18:04:34] <jymm> Skullworks: you must have it blocked
[18:04:36] <jmkasunich> Skullworks: do you have the renco part number? is the info on their site?
[18:05:00] <JymmmEMC> http://www.renco.com/index.php?WCMSGroup_8_3=12&WCMSGroup_4_3=8&WCMSGroup_12_3=159&WCMSArticle_3_159=114
[18:05:25] <jymm> RCML15_646770.pdf
[18:05:58] <jmkasunich> jymm how do you know what encoder skullworks has?
[18:06:07] <Skullworks> thats it
[18:06:16] <Skullworks> from the pdf file name
[18:06:17] <JymmmEMC> Becasue I have the PN from the PDF he was trying to send me
[18:06:22] <jmkasunich> ah
[18:07:39] <jmkasunich> wtf is with the internets this morning - something must be stuck in the tubes
[18:08:05] <Skullworks> the actual P?N is a 90325-110
[18:08:23] <jmkasunich> that will help if the renco page ever loads
[18:08:57] <JymmmEMC> Heh, they have a 3/16 and a 3/16+
[18:10:15] <jtr> traceroutes from central NC to smallparts.com die in washington, to weather.com die in atlanta
[18:10:32] <Skullworks> yeah - but these don't seem well suited to disassembly like the US Digital
[18:10:40] <alex_joni> hi guys
[18:11:47] <JymmmEMC> Skullworks: Where did you get these encoders from?
[18:12:33] <skunkworks> I don't have any 1kohm resistors here..
[18:12:46] <JymmmEMC> Skullworks: As I can't get any 9xxxxx series PN to show up
[18:12:53] <JymmmEMC> might be an OEM PN
[18:12:59] <Skullworks> Renco - Direct they dumped a huge amount of overstock on ebay a few months back on certain part numbers
[18:13:06] <JymmmEMC> ah
[18:15:01] <alex_joni> level3 seems clogged from here
[18:19:07] <JymmmEMC> Skullworks: Are these the ones you got? I noticed on renco's websit ethey don't list index
http://cgi.cafr.ebay.ca/RENCO-Low-Profile-Encoder-RCML15-512-0_W0QQcmdZViewItemQQitemZ200055585379
[18:19:15] <Skullworks> here - They are still selling them -
http://cgi.ebay.com/RENCO-Low-Profile-Encoder-RCML15-512-0_W0QQitemZ200055585379QQcmdZViewItemQQptZBI_Automation_Components?hash=item200055585379&_trksid=p3911.c0.m14&_trkparms=72%3A1205%7C66%3A2%7C65%3A12%7C39%3A1%7C240%3A1318
[18:19:51] <JymmmEMC> Ah Non-RoHS
[18:20:08] <jmkasunich> don't you just love RoHS
[18:20:23] <jmkasunich> perfectly good stuff available cheap ;-)
[18:20:31] <Skullworks> yup
[18:20:54] <jmkasunich> and the europeans can't have any of it - muwahh ha ha
[18:21:50] <alex_joni> :P
[18:22:15] <alex_joni> crap, half the US pages are dead from here :/
[18:22:42] <JymmmEMC> And they didn't even bother to put up a PN conversaion chart
[18:22:47] <Skullworks> Hmm the Chinese must be attacking again.
[18:24:02] <JymmmEMC> alex_joni: what url?
[18:25:10] <alex_joni> http://isc.sans.org/
[18:25:38] <alex_joni> JymmmEMC: trying a traceroute on my domain (robcon.ro) somewhere in texas
[18:25:51] <alex_joni> it dies after the first US hop in washington
[18:26:08] <skunkworks> the revolution has started! ;)
[18:26:10] <alex_joni> 11 418 ms 419 ms 419 ms ae-44-44.ebr2.Washington1.Level3.net [4.69.137.6
[18:26:10] <alex_joni> 2]
[18:26:12] <alex_joni> 12 * * * Request timed out.
[18:26:26] <jtr> alex_joni: same as I got here
[18:26:32] <JymmmEMC> 11 65.99.253.235 (65.99.253.235) 43.502 ms 43.158 ms 43.429 ms
[18:26:43] <alex_joni> JymmmEMC: yeah, I know it works from there
[18:26:54] <alex_joni> I pinged it from dreamhost and it's up and running
[18:26:59] <Skullworks> I bought 8 back in August - then they dried up for awhile, now they have more on the ebay auction block.
[18:27:30] <alex_joni> http://www.internethealthreport.com/
[18:27:39] <alex_joni> level3 looks pretty bad
[18:28:03] <JymmmEMC> Heh L3 always looks bad
[18:28:37] <Skullworks> Next issue will be to find a stocking source for the required Berg connectors.
[18:29:19] <alex_joni> JymmmEMC: usually they don't have > 40% loss
[18:29:53] <JymmmEMC> alex_joni: I've seen them with all of west coast down, being rerouted thur tx then up to WA
[18:30:37] <JymmmEMC> let me see what I get from .nl
[18:31:44] <alex_joni> some sites I get by getting routed through ALTER.NET
[18:32:00] <alex_joni> but anything going through level3 is dead
[18:33:20] <JymmmEMC> 215ms via HK
[18:34:31] <JymmmEMC> alex_joni: 131ms from .nl
[18:35:53] <JymmmEMC> which really isn't bad for across the pond
[18:35:58] <alex_joni> yeah, seems back up now
[18:36:29] <alex_joni> yeah, I'm getting routed to vienna, frankfurt, amsterdam, london -> new york
[18:36:39] <alex_joni> (no more paris -> washington) ;)
[18:37:22] <alex_joni> I get about 108ms for NY
[18:39:33] <JymmmEMC> alex_joni: It's all over the place =)
[18:50:00] <robin_sz> meep?
[18:50:09] <eric_u1> 200055585379
[18:56:04] <alex_joni> hey robin_sz
[18:58:44] <robin_sz> dood!
[18:58:49] <robin_sz> are we well?
[18:59:05] <alex_joni> sure thing
[18:59:10] <alex_joni> yourself?
[18:59:44] <ds3> wheeeee xmas and boxing day over
[19:00:17] <robin_sz> yeah, not bad
[19:00:31] <robin_sz> cant say the economy is helping much though
[19:00:39] <robin_sz> but, still movign forward
[19:00:48] <robin_sz> just installed a powder coating plant :)
[19:00:53] <alex_joni> sounds ok then
[19:01:02] <alex_joni> cool, what mfr?
[19:01:14] <robin_sz> oh, seconhand ..
[19:01:23] <robin_sz> maywick oven and some E8 conveyor
[19:01:31] <alex_joni> I see
[19:01:44] <robin_sz> RDM cyclone and filter
[19:01:46] <alex_joni> I was wondering about the gun
[19:01:53] <robin_sz> eurotech guns
[19:02:08] <alex_joni> hmm, doesn't sound familiar ;)
[19:02:15] <alex_joni> uk local?
[19:02:22] <alex_joni> I see rdm is
[19:02:27] <robin_sz> 7m gas infrared oven, 300kw
[19:02:40] <robin_sz> umm, US i think ...
[19:03:04] <robin_sz> http://www.eurotecfinishing.com/systems.html
[19:04:00] <robin_sz> just two manual guns, fine for waht we do
[19:04:08] <alex_joni> coo
[19:04:50] <robin_sz> we dotn do shit loads of powder coat, but the bills were mounting up ...
[19:04:59] <robin_sz> about 4K to 5k a month
[19:05:13] <robin_sz> enought ot make it worthwhile putting in our own
[19:07:15] <alex_joni> * alex_joni discovers fortune
[19:07:24] <robin_sz> and you? still install ing evil robots with lasers?
[19:07:39] <alex_joni> yeah, pretty much
[19:07:56] <robin_sz> they turned evil yet?
[19:08:03] <alex_joni> not if I can help it
[19:08:26] <robin_sz> mmm ... they are probably keeping quiet
[19:08:35] <robin_sz> until they rise to strike agaisnt us ...
[19:09:51] <toastyde1th> toastyde1th is now known as toastydeath
[19:10:08] <robin_sz> apart from that, business as usual ... and I'm learing to play bass :)
[19:10:22] <ds3> robin_sz: aren't there a ton of gov't rules regulating powder coating there?
[19:10:34] <robin_sz> not as far as I know
[19:10:45] <robin_sz> theres a million regulating wet spray
[19:10:49] <robin_sz> solvents and all that
[19:10:58] <alex_joni> yeah, waterbased ftw :)
[19:11:09] <ds3> thought there might be something about out gassing when it gets baked
[19:11:15] <robin_sz> errr
[19:11:21] <robin_sz> no, it doesnt do that
[19:11:45] <robin_sz> virtually voc free
[19:13:13] <robin_sz> water based is good, but a right pain with electrostatic guns
[19:14:05] <alex_joni> well, at least it works
[19:14:13] <alex_joni> solvent based is nto allowed ;)
[19:16:01] <ds3> so you are reduced to a bunch of kids using water colors? ;)
[19:17:02] <robin_sz> I thought most cars where painted that way for a while now?
[19:17:06] <robin_sz> even in the US?
[19:17:23] <alex_joni> robin_sz: most usually cars are painted using dip techniques
[19:17:33] <robin_sz> really?
[19:17:38] <robin_sz> well, there you go
[19:17:53] <robin_sz> I know VW/Audi went water based long time ago
[19:19:20] <robin_sz> and I remember that BMW were powder coated from ages ago
[19:21:48] <alex_joni> "It wasnt lost. It always knew exactly where it was. It was always HERE. It was just that everything else seemed to have been temporarily mislaid."
[19:37:57] <eric_u1> powder coating is great on relatively non-complex surfaces like a car
[19:48:22] <robin_sz> its fine on relatively complex surfaces too, with a skilled operator and careful adjustment
[19:48:51] <robin_sz> I think i remember the BMW spec called for +- 10u of variation, even in tight corners
[19:55:14] <alex_joni_> alex_joni_ is now known as alex_joni
[19:56:21] <robin_sz> you'll never beleive this either ...
[19:56:43] <robin_sz> if you have a 400w bass amp, your neighbours will hear it
[19:57:12] <BigJohnT> not enough power for my neighbors to hear it
[19:57:28] <robin_sz> mmm, i guess it depends how far away they are
[19:57:35] <BigJohnT> :)
[19:57:44] <robin_sz> might need a few more watts then :)
[19:58:00] <alex_joni> or neighbours
[19:58:15] <robin_sz> also a valid solution
[19:58:34] <robin_sz> personally, i think more wats and less neighbours is the answer :)
[19:59:33] <robin_sz> went with a freind to demo his new amps to a UK based PA maker, that was fun
[20:00:42] <robin_sz> the amp was putting out about 1650w into 4R, 18" bass cabs
[20:01:06] <robin_sz> on the kick drum they moved about an inch across the floor on each beat
[20:15:54] <alex_joni> heh, didn't know there is an electric drill chuck key fairy
[20:53:57] <skunkworks> http://imagebin.ca/img/gIZqi5S.jpg
[20:55:44] <Skullworks> this the rev with current limiting?
[20:55:51] <skunkworks> yes
[20:56:09] <skunkworks> 2nd rev
[20:56:34] <alex_joni> I could have sworn it's rev #17
[20:56:53] <skunkworks> heh ;)
[20:58:34] <alex_joni> skunkworks: looking good
[20:59:06] <Skullworks> He just wants to be able to crank those accell numbers up without worrying.
[20:59:29] <alex_joni> and motor sizes
[20:59:54] <alex_joni> skunkworks: if it works, will you order PCB's ?
[21:00:24] <skunkworks> I am really thinking about it. I have wet dreams about plated thru holes.
[21:01:06] <BigJohnT> yea, my class works
[21:01:06] <jmkasunich> TMI
[21:01:42] <Skullworks> BatchPCB.com
[21:03:52] <skunkworks> it is going very well.. These sockets that have a barrel give just enough hight that I can solder the top easy
[21:03:54] <alex_joni> or olimex
[21:04:24] <alex_joni> skunkworks: I think you can buy thingies for thru holes
[21:04:43] <skunkworks> I found out tining the top of the circuit board helps a lot. (where I need to make the connections)
[21:04:49] <alex_joni> they look like tiny tubes, you can solder both ends, and then insert whatever
[21:04:57] <skunkworks> I have seen rivets.. Never eally looked into it.;
[21:05:04] <alex_joni> need slightly bigger holes though
[21:05:22] <skunkworks> that is what turned me off. room is an issue.
[21:07:19] <alex_joni> http://olimex.com/pcb/pcb-price.html
[21:08:03] <alex_joni> skunkworks: what's your board size?
[21:09:35] <alex_joni> DSQ 320x200 mm (12.6 x 7.8")
[21:10:16] <skunkworks> 4.5"X3.75"
[21:11:30] <stustev> why would acos be defined and asin not be?
[21:12:27] <alex_joni> skunkworks: then you can fit 4 boards on one panel
[21:12:48] <alex_joni> a bit pricey
[21:12:53] <alex_joni> stustev: from where?
[21:13:06] <stustev> in the rtapi_math
[21:13:19] <stustev> I am trying to compile a kins file
[21:13:21] <skunkworks> cool
[21:13:22] <alex_joni> probably someone needed acos, and ddn't need asin
[21:13:34] <stustev> doesn't have tan either
[21:13:42] <stustev> has atan2 but not tan
[21:13:59] <alex_joni> stustev: chances are you can use the rtai ones directly
[21:14:00] <stustev> why would those not be standard?
[21:14:24] <alex_joni> iirc rtapi_math only defined some problematic stuff
[21:14:33] <alex_joni> which wasn't available from rtai's math
[21:14:34] <stustev> you are talking to a 'C' neophyte
[21:15:00] <stustev> if the rtapi is included then the functions should be there - correct?
[21:16:18] <alex_joni> only the ones defined by rtapi_math
[21:16:43] <stustev> sooooo - then I don't have asin - ugh
[21:17:11] <alex_joni> you could define it though :)
[21:17:29] <stustev> yeah right - I will look and try
[21:17:34] <alex_joni> extern __inline double acos (double __x) { return atan2(sqrt(1.0 - __x * __x), __x);
[21:17:37] <alex_joni> }
[21:17:42] <alex_joni> that's for acos()
[21:17:59] <stustev> but if I do then when I next update or download I will have to do this again
[21:18:19] <alex_joni> stustev: cvs diff -u > foo.patch
[21:18:19] <stustev> that is in rtapi_math?
[21:18:32] <alex_joni> then send to cradek or to me, or to the dev list (preferably)
[21:18:47] <stustev> will work on it - thanks
[21:18:48] <alex_joni> stustev: in emc2/src/rtapi/rtapi_math_i386.h
[21:18:58] <alex_joni> but try rtai's math first
[21:20:11] <alex_joni> https://www.rtai.org/documentation/magma/html/api/dir_000034.html
[21:22:38] <stustev> looks as if I can change my calculations just as easily
[21:23:44] <alex_joni> in emc1 people used to include <math.h>
[21:23:54] <alex_joni> alhough I'm not sure that's a good idea ;)
[21:27:32] <stustev> I can do what I want to do with the current definitions - it will be a little more work but not any more than what would be done with a new function (asin) - thanks
[21:27:55] <stustev> I think :)
[21:29:12] <Skullworks> Stuart - you inspire me, I would just get a massive headache...
[21:34:04] <stustev> Skullworks: thanks for the compliment - I think :) - I am down to just one page of compile errors now
[21:38:08] <skunkworks> heh
[21:40:55] <skunkworks> damn it - where the heck do I have some 1kohm resistors?
[21:46:57] <BigJohnT> in the second drawer to the left
[21:47:19] <skunkworks> I looked
[21:47:22] <alex_joni> stustev: at least you're modeling something fairly simple, not like superstrings :)
[21:47:29] <alex_joni> skunkworks: no, you didn't
[21:47:58] <BigJohnT> I'll e-mail you a couple if you can't find the one in the drawer
[21:48:43] <skunkworks> what does it mean when you start assembling a board that you designed --- and you start putting componants on the wrong side?
[21:50:14] <alex_joni> fuse test
[21:50:17] <BigJohnT> you have fell into a worm hole and warped to Australia
[21:50:28] <BigJohnT> now everything is upside down
[21:51:47] <skunkworks> heh - I did catch it only after I soldered in one componant but still - duh
[21:52:10] <alex_joni> http://upload.wikimedia.org/math/c/3/8/c380d0267a35923731d9b1a06fd25128.png
[21:52:21] <alex_joni> that explains what BigJohnT just said
[21:52:43] <micges> ot question: anyone have programmed video4linux2 device? if so help is needed to program two cameras to be installed on machine
[21:53:13] <alex_joni> skunkworks: hich integrates (sums) over all possible embeddings of this 2D surface in 26 dimensions...
[21:53:39] <alex_joni> 2D as it's doublesided of course :)
[21:53:55] <eric_u1> micges, on one framegrabber?
[21:54:03] <BigJohnT> exactly
[21:55:04] <micges> eric_u1: two signals and two previews
[21:55:39] <eric_u1> micges, didn't really answer the question
[21:56:00] <alex_joni> analog cameras?
[21:56:20] <eric_u1> v4l is a driver standard
[21:56:34] <micges> what is framegrabber ?
[21:56:54] <micges> alex: yes
[21:56:54] <eric_u1> what kind of cameras do you have?
[21:57:11] <eric_u1> do you have a board in your computer?
[21:57:35] <micges> yes with 4 connectors
[21:58:08] <eric_u1> what brand, if it is a bt878, it has 4 channels
[21:58:12] <eric_u1> but one at a time
[21:58:30] <micges> problem is that I cannot find any howto to preview it
[21:59:08] <micges> one at time ?
[21:59:18] <alex_joni> micges: interpolated fps
[21:59:20] <micges> will be needed second card ?
[21:59:27] <alex_joni> you get one frame from each channel in a loop
[21:59:36] <alex_joni> no, it'll work, but not very fast
[21:59:46] <alex_joni> although I'm sure 25 fps or so will be enough for you
[21:59:54] <micges> yes surely
[22:00:04] <micges> only to preview position of material
[22:00:05] <eric_u1> we have some cards that have 4 bt848s on there
[22:00:08] <alex_joni> eric_u1: in my experience bt878 boards have different ways of changing channels
[22:00:10] <micges> with 20x magnify
[22:00:23] <eric_u1> he hasn't told us what equipment he is using
[22:00:25] <alex_joni> and it depends on the manufacturer how they did it
[22:00:33] <micges> eric: Im checking..
[22:00:38] <alex_joni> they have some DIO which the driver needs to toggle
[22:01:22] <eric_u1> a bt878 is one chip on a pci board, not much to change
[22:01:31] <micges> bt878
[22:01:46] <alex_joni> eric_u1: yeah, but it has usually some mux outside
[22:01:55] <alex_joni> to change between 2..16 inputs
[22:02:03] <alex_joni> and the mux is driven by DIO
[22:02:26] <eric_u1> never seen one with that many inputs, 878 has 4 or 5
[22:02:40] <eric_u1> built in
[22:02:57] <alex_joni> I have some which have connectors for extenders
[22:03:09] <alex_joni> risers with another 4 BNC connectors on them
[22:03:17] <micges> I've initilaised input, video format, video standard but cannot initialize framebuffer
[22:03:21] <alex_joni> but at more than 4 inputs the 878 suck real time
[22:03:49] <eric_u1> micges, are you sure your input was accepted?
[22:04:07] <eric_u1> what resolution are you using?
[22:04:21] <eric_u1> 320x200 should always work
[22:04:45] <micges> ok
[22:05:18] <eric_u1> just as a test, you should be able to go to your camera's full resolution
[22:06:42] <micges> how to test if camera is previewing without actually preview it on screen ?
[22:07:38] <eric_u1> you should be able to tell when you got a frame
[22:07:43] <eric_u1> it's been a while
[22:10:14] <LawrenceG> micges, you could have a look at the source for tvtime.... I have a btty card that supports tv,composite1,s-video,and composite3 inputs
[22:10:31] <eric_u1> tvtime is too much code to be useful
[22:10:35] <LawrenceG> a wintv go card
[22:11:06] <micges> same xawtv
[22:11:07] <LawrenceG> all he needs is the capture subroutine I think
[22:11:13] <eric_u1> but I've always tested with tvtime or xawtv to see if it works
[22:11:57] <eric_u1> LawrenceG: I'm just not that smart I guess :)
[22:12:29] <micges> ok I'll look into source once again, thanks eric alex and lawrence
[22:12:38] <LawrenceG> I did a video capture app called gspy and the capture code reduced to less than 1 page of code
http://gspy.sourceforge.net/
[22:13:02] <eric_u1> that's about what mine does, I'm looking for it
[22:13:30] <LawrenceG> see src/v4lif.c
[22:13:45] <LawrenceG> but the code is the v4l 1 interface
[22:14:56] <LawrenceG> the code captures 1 image/call... not streaming video
[22:16:53] <stustev> it compiles!! - now if it will do what I want :)
[22:17:11] <fragalot> lol
[22:17:19] <alex_joni> sim first :)
[22:17:41] <stustev> straight to the machine - :)
[22:17:46] <stustev> not
[22:17:49] <alex_joni> oh-oh
[22:19:09] <stustev> alex_joni: if this is simple I don't want to try anything complicated
[22:19:59] <alex_joni> stustev: same here ;)
[22:20:16] <micges> LawrenceG: thanks for example, must sleep with it
[22:20:33] <micges> good night all
[22:20:49] <LawrenceG> micges, source always helps get ideas rolling! cheers
[22:25:57] <alex_joni> good night all
[22:26:07] <alex_joni> stustev: good luck :)
[22:31:05] <stustev> alex_joni: imagine that - something in sim didn't work as expected :) - thank you for your help today
[22:43:37] <stustev> how do I run through the kins file 'one line at a time' so I can troubleshoot the results?
[23:16:22] <JymmmEMC> I guess I should let you guys have a crack at these first... $120 for all three, $50/ea (124lbs each)
http://www.apcc.com/resource/include/techspec_index.cfm?base_sku=su2200rm3u
[23:23:39] <skunkworks> http://imagebin.ca/img/F8yx9V8.jpg
[23:42:13] <skunkworks> leds light up as expected.. :)
[23:42:30] <LawrenceG> yay
[23:48:10] <Skullworks> I guess you found your 1K R
[23:52:33] <skunkworks> yes - I scavenged them
