Back
[14:41:27] <alex_joni> hello steve
[14:41:35] <SteveStallings> hi alex
[14:42:19] <alex_joni> busy week?
[14:42:33] <SteveStallings> Thursday, Friday were a mess....
[14:42:59] <SteveStallings> spent Saturday chasing parts for my old Hardinge DV59
[14:44:07] <alex_joni> lathe?
[14:44:47] <SteveStallings> yes, a manual turret lathe, mine was missing the bed turret, picked one up Saturday
[14:45:35] <SteveStallings> actually bought it on eBay, but shipping was expensive so I drove 3 hours each way to pick it up
[14:46:12] <alex_joni> coo
[14:47:09] <SteveStallings> nice day for a drive, took some back roads on the return trip, stopped to see friends and their new toys
[14:47:30] <alex_joni> it's raining here :(
[14:47:37] <alex_joni> so it's a nice day to stay in
[14:48:14] <SteveStallings> did the question of how to connect CL get setteled?
[14:48:32] <alex_joni> it's pretty clear ...
[14:48:42] <alex_joni> both ways should be available
[14:49:11] <alex_joni> the hal-connection pretty much works (no rt-stuff yet)
[14:49:18] <alex_joni> for NML it's not that easy
[14:49:37] <alex_joni> hey rayh
[14:50:03] <rayh> Hi Alex
[14:50:31] <rayh> Saw a BBC report on the cold weather there.
[14:50:40] <alex_joni> no that cold now
[14:50:45] <alex_joni> it's +5 and raining
[14:50:52] <rayh> Good.
[14:51:05] <alex_joni> the raining part is not that good ;)
[14:51:13] <alex_joni> but it's better than -20
[14:51:36] <alex_joni> I started designing hal-components for gEDA
[14:51:50] <alex_joni> but I'm not sure if it's the best way to do stuff
[14:53:22] <rayh> Okay.
[14:53:35] <alex_joni> hmm.. maybe eagle would be a better choice
[14:53:57] <rayh> The idea is to use the electronic graphical capabilities to make connections in HAL?
[14:54:09] <alex_joni> yes
[14:54:30] <alex_joni> halgui (not existing) would need all the capabilities a schematic editor already has
[14:54:39] <alex_joni> and we wouldn't wanna reinvent on eof those
[14:54:46] <alex_joni> one of those
[14:55:06] <rayh> So each hal component would be an IC with pins.
[14:55:22] <alex_joni> yes
[14:55:27] <alex_joni> input/output pins
[14:55:39] <alex_joni> and the postprocessor (custom written) produces a .hal file
[14:55:47] <alex_joni> that can easily be loaded with halcmd
[14:57:28] <SteveStallings> does Eagle offer open source licenses and allow user written net listers?
[14:57:56] <rayh> I wonder what a HAL configuration would look like after the autorouter?
[14:58:14] <alex_joni> autorouter?
[14:58:43] <rayh> If you asked eagle to do a circuit board.
[14:58:47] <SteveStallings> autorouter is a PCB layout component, not used in schematic capture
[14:58:49] <alex_joni> lol
[14:59:01] <alex_joni> don't think you'd wanna go there ;)
[14:59:10] <alex_joni> we only need schematic
[15:00:53] <rayh> For an old dog like me, the new generation of schematic is so darned disconnected.
[15:01:47] <alex_joni> * alex_joni agrees that current schematic editors are a bit too much for halgui
[15:01:55] <alex_joni> maybe an older one will do ;)
[15:02:22] <rayh> I do like the idea of a visual aid during configuration.
[15:02:29] <alex_joni> steve: I know that cradek wrote his own post processor
[15:02:37] <alex_joni> for eagle
[15:04:09] <rayh> Does each pin, HAL wide, have a unique name?
[15:04:30] <alex_joni> might be the same name on more components
[15:04:42] <alex_joni> don't know if any are right now
[15:06:23] <rayh> I'd think the same naming on multiple components would be a drawback to current schematic stuff.
[15:06:52] <alex_joni> how do you mean that?
[15:07:05] <rayh> Abdul and I did a schematic for the "dirt simple" Rutex motherboard for Smithy and I got confused.
[15:07:30] <alex_joni> how so?
[15:08:02] <rayh> Six chips, a parport, and a couple dozen connections.
[15:08:34] <rayh> Even after all the looking, I wound up with two pins on the parport inverted.
[15:09:18] <alex_joni> hmmm...
[15:09:34] <alex_joni> anyways.. I still think hal stuff would be best if done schematic-like
[15:09:37] <rayh> Now add in an installer who is not visually oriented and I'm guessing you've got a real mess.
[15:10:00] <rayh> I don't entirely disagree...
[15:10:20] <alex_joni> rayh: not visuall oriented _might_ work if you only have to chose from existing setups
[15:10:34] <robin_sz> hmmm
[15:10:53] <rayh> Yes. There will be a lot of that.
[15:11:24] <rayh> And one these are configured, they will cover a lot of instances.
[15:11:44] <rayh> Look how long minimill and bridgeport have been applied.
[15:11:52] <robin_sz> so .. to get going with EMC, all the prosepectuve user needs to do is learn a bit of basic linux, configure the ini, get to grips with classic ladder and a short course on schematic layout .. and hes done. should be a snap.
[15:11:59] <alex_joni> great.. just the guy missing
[15:12:02] <alex_joni> hello john
[15:12:09] <robin_sz> * robin_sz plays devils advocate
[15:12:10] <jmkasunich> hi
[15:12:15] <rayh> but think that it would be easier to write and use a wizard.
[15:12:17] <alex_joni> robin: forgot about rtai
[15:12:23] <jmkasunich> robin_sz: you are correct except for one detail:
[15:12:48] <robin_sz> that being?
[15:12:48] <jmkasunich> all the prospective _INTEGRATOR_ needs to do
[15:13:07] <robin_sz> but 99% of emc installs the integrator is the user
[15:13:12] <jmkasunich> _users_ could/would use existing ini files
[15:13:31] <alex_joni> and no cl
[15:13:49] <rayh> Oh I disagree big time with the 99%
[15:14:14] <alex_joni> doesn't really matter
[15:14:20] <alex_joni> there are 2 kind of users
[15:14:20] <robin_sz> anyway
[15:14:26] <alex_joni> 1. is ok with standard configs
[15:14:27] <jmkasunich> tell me, how does a Mach2 _USER_ change the spindle control logic, for instance? ISTR, he can't... CL would be used for that kind of customization
[15:14:33] <rayh> It might appear that way from the user lists but we are not seeing anything outside a very narrow view.
[15:15:28] <rayh> I'm in agreement with robin's list and I don't think that path is all bad.
[15:15:35] <anonimasu> hey
[15:15:36] <anonimasu> iab
[15:15:41] <A-L-P-H-A> welcome back.
[15:15:47] <anonimasu> robin_sz: I hvae the xilinx webpack now
[15:15:49] <anonimasu> :)
[15:16:00] <alex_joni> rayh: what list?
[15:16:24] <rayh> learn a bit of basic linux, configure the ini, get to grips with classic ladder and a short course on schematic layout
[15:16:30] <alex_joni> ahhh that list ;)
[15:16:43] <alex_joni> the last two are optional
[15:17:06] <robin_sz> anonimasu: cool, wait one I'll post my G2002 FPGA somewhere
[15:17:09] <alex_joni> not everybody needs cl
[15:17:13] <anonimasu> yep
[15:17:21] <anonimasu> robin_sz: how do you program the fpga's ?
[15:17:22] <alex_joni> nor custom hal-scripts
[15:17:36] <A-L-P-H-A> robin_sz, what's this g2002 fpga you keep talking about?
[15:17:37] <alex_joni> an0n: try wireless :))
[15:17:45] <anonimasu> robin_sz: with a cable like that one i posted or do you need a programmer for that?
[15:17:50] <anonimasu> hehe
[15:18:06] <robin_sz> anonimasu: via a JTAG lead I believ .. not got THAT far yet
[15:18:30] <anonimasu> ah once you get that far drop me a message with what you need to program them.. I am really curious..
[15:18:36] <alex_joni> ok, to get back to the discussion
[15:18:38] <robin_sz> me too
[15:18:45] <anonimasu> robin_sz: I think you can use a paralell port interface for that...
[15:18:47] <anonimasu> ok
[15:18:54] <alex_joni> who agrees there is need for a visual setup for emc?
[15:19:07] <anonimasu> * anonimasu agrees
[15:19:16] <jmkasunich> * jmkasunich agrees, somewhat
[15:19:24] <robin_sz> if its simple and intuitive, I agree
[15:19:32] <alex_joni> I think we all should say what we expect from one
[15:19:51] <A-L-P-H-A> I would like a visual setup for emc!
[15:20:22] <robin_sz> I would expect it to adress all the setup, not just part of it
[15:20:39] <alex_joni> even if writing different config files, right
[15:20:49] <robin_sz> yeah, thats fine,
[15:20:51] <anonimasu> robin_sz: how much of it, from writing all ini's and redefining pins?
[15:21:02] <alex_joni> everything
[15:21:06] <anonimasu> and creating the links for classicladder?
[15:21:08] <A-L-P-H-A> gearing ratio calculations... show a gear on the motion system, the gear on the motor... show sizes and/or teeth. Ask pitch, step angle on motor, microstepping of driver, rotary encoders, etc.
[15:21:10] <robin_sz> reckon so
[15:21:27] <anonimasu> yeah, that's pretty much what I'd want too
[15:21:28] <jmkasunich> I dunno who's gonna write all that
[15:21:36] <A-L-P-H-A> anyways... I gotta pack, and I think I'm outta my league here anywaysh. :)
[15:21:39] <robin_sz> A-L-P-H-A that might be tricky with aschematic editor
[15:21:41] <alex_joni> jmk: did you look at emcconfig?
[15:21:52] <jmkasunich> call me a tech snob, but IMO if you need pics to figure out gear ratios, you gots problems
[15:21:54] <anonimasu> I could write the program but I have a large problem with writing graphical apps on linux..
[15:21:57] <anonimasu> I dont know GTK
[15:22:11] <robin_sz> use Qt then
[15:22:25] <anonimasu> jmkasunich: I dont know how to :)
[15:22:32] <A-L-P-H-A> jmkasunich, snob. :) I have issues following gear ratios like 3:1... or 1:3... what does that mean? 3 on the drive shaft, 1 on the motor, or the other way around.
[15:22:32] <anonimasu> err robin_sz
[15:22:34] <anonimasu> heh
[15:22:40] <A-L-P-H-A> TurboCNC has that issue, which bugged me a lot.
[15:22:49] <A-L-P-H-A> not that it was hard to reciprocate the value, but still.
[15:22:53] <jmkasunich> that can be described in words, pictures is a lot of work, a lot of code bloat
[15:23:28] <A-L-P-H-A> jmkasunich, well... we wanted visual. :) So I mentioned it. it'd be a 15 minute work in a RAD programming
[15:23:52] <alex_joni> I think we could point the guy who started emcconfig in the right direction
[15:23:55] <robin_sz> one ting I will do on my plasmas is "mottion test .. make start mark .. itgoing to move 1000mm ... ok, measure it .. enter the actual value it moved ...finish"
[15:24:13] <anonimasu> robin_sz: did you think about using a editor, or writing the code by hand?
[15:24:30] <anonimasu> that's a good feature..
[15:24:33] <jmkasunich> better hope they're close, it might try to move 20meters if they are screwed up
[15:24:56] <robin_sz> anonimasu: for GUI? use the Qt 'develop' grapicla tool for layout, code the hard bits by hand later
[15:25:37] <rayh> * rayh reads back
[15:27:56] <rayh> The graphical wizard that i think you are speaking of would be rather easier in tickle than a compiled.
[15:28:29] <alex_joni> rayh: would it be possible to do the hal-schematic stuff in tickle?
[15:28:52] <rayh> Yes. The canvas is very flexible.
[15:29:05] <rayh> We use it now for tool path.
[15:29:09] <alex_joni> * alex_joni regrets not knowing tickle
[15:29:37] <jmkasunich> ray - having drawn a "schematic" in tck, can you generate a netlist?
[15:30:04] <alex_joni> I think the work started already emcconfig (written in python) could be extended
[15:31:01] <SteveStallings> learn Linux, tKL, CL, Python.... wheeeeeee!
[15:31:02] <alex_joni> but tcl sounds fine too
[15:31:22] <alex_joni> you don't need to learn tcl, python...
[15:31:33] <jmkasunich> users and integtators wouldn't need to know tck or pt
[15:31:36] <jmkasunich> users and integtators wouldn't need to know tck or py
[15:31:42] <SteveStallings> sorry, having an Aunt Tillie moment 8-)
[15:31:46] <alex_joni> do you learn C to use sirc?
[15:31:54] <rayh> I'm not certain what a netlist is but with a proper rule set you should be able to make a hal config.
[15:32:19] <alex_joni> jmk: I started writing components for gEDA,
[15:32:22] <jmkasunich> netlist says "on the schematic, pin 1 of part A is connected to pin 12 of part C"
[15:32:25] <alex_joni> and I must say I don't like it
[15:32:51] <rayh> aunt tillie may have to smart'n up.<g>
[15:33:17] <jmkasunich> a custom HAL "schematic capture" tool would be nicer than gEDA, but since I don't know how to do GUIs, I figured gEDA was the way to go
[15:33:23] <robin_sz> anonimasu: wanna see my attemeot a a 3 axis G2002 FPGA? email addy??
[15:33:40] <alex_joni> I use schematic editors for a while now, but it took me some time to feel not bothered by gEDA
[15:33:43] <rayh> I know that I'm a tille rule fan but.
[15:34:01] <jmkasunich> robin_sz: this is replacement "code" for Mariss's FPGA?
[15:34:18] <alex_joni> rayh: think you're amongst the few who could do the tcl stuff ;)
[15:34:19] <rayh> oh no "and" we must have tillie able configurations that she can grab.\
[15:34:38] <rayh> Like the old bridgeportio.
[15:35:03] <rayh> I think that John was on the right track with a xylotex drive config.
[15:35:17] <jmkasunich> right ray... no matter what the tools are, there should be a "library" of configs that can be used "as is" or as the starting point for a custom config... nobody should have to start with a blank page
[15:35:21] <rayh> in spite of my criticism for using the xylotex name.
[15:35:44] <jmkasunich> using the name is tillie friendly
[15:36:00] <rayh> What we would need is a wizard that can take these standard configs and allow the user to make small or big changes.
[15:36:10] <robin_sz> jmkasunich: this is my own attempt to replace the large number of chips on a G2002 board with a single FPGA, not seen Mariss' FPGA desing, just seen the original schematic
[15:36:30] <jmkasunich> I see
[15:36:32] <rayh> The name thing was a throwback to STG which wound up begin a generic for a while.
[15:37:04] <jmkasunich> robin_sz: so to use your design, you need a board with an FPGA on it
[15:37:18] <robin_sz> jmkasunich: I might use it with Steve Hardys EMC port
[15:37:22] <rayh> alex_joni: You are correct that it will take Paul and I to build the tickle framework.
[15:37:33] <robin_sz> jmkasunich: yeah and a rabbit, obviously
[15:37:49] <anonimasu> robin_sz: you could send me your design so Ican have a look at it
[15:37:51] <rayh> But I think that all of use can add in the specific branching and such.
[15:37:59] <alex_joni> robin: still not tired of wabbits?
[15:38:10] <anonimasu> * anonimasu is looking at the program now
[15:38:14] <robin_sz> anonimasu: i could if I had your email addy :)
[15:38:14] <alex_joni> rayh: I'd be glad to try to help
[15:38:19] <rayh> I committed a youpperism with mispelling on that one.
[15:38:33] <rayh> use should be youse guys 'eh.
[15:38:47] <alex_joni> heh
[15:39:05] <alex_joni> ok.. can we decide on stuff needed for the configurator?
[15:39:14] <alex_joni> jmk: btw, did you get my email?
[15:40:07] <rayh> Some of my thinking on this was leaning toward a config database
[15:40:16] <jmkasunich> alex_joni: not yet
[15:40:29] <rayh> so I can hear robin add database to his devils list.
[15:41:12] <rayh> I wrote that config years ago and used a single file for the data
[15:41:17] <alex_joni> jmk: I did send it last week
[15:41:44] <rayh> it got obsolete just about the time I put it in the dropbox.
[15:42:27] <jmkasunich> oh, I thought you meant a recent mail... I got's craploads of mail, what was it a bout
[15:42:36] <alex_joni> cl with hal support
[15:42:50] <alex_joni> I have a newer version though
[15:43:13] <jmkasunich> * jmkasunich looks
[15:43:30] <alex_joni> * alex_joni sent the new one
[15:46:25] <alex_joni> rayh: how would this database look like?
[15:47:01] <rayh> * rayh has a call from family.
[15:47:36] <alex_joni> ok.. later
[15:53:12] <alex_joni> jmk: say when you got the mail I just sent
[15:53:31] <jmkasunich> just arrived now
[15:53:39] <alex_joni> coo
[15:54:15] <alex_joni> btw, I just remembered
[15:55:27] <alex_joni> I noticed a little problem inside HAL_Introduction.pdf
[15:55:50] <alex_joni> parport configured as input has some bit numbers wrong
[15:56:01] <alex_joni> might be my version of Hal_Introduction though
[15:56:21] <alex_joni> didn't check the latest on www.linuxcnc.org
[15:57:24] <jmkasunich> I haven't changed it lately
[15:57:48] <jmkasunich> which bits are wrong?
[15:57:54] <jmkasunich> * jmkasunich opens lyx
[15:58:32] <alex_joni> 10, 11, 12, 13, 15 are mixed up
[15:59:57] <robin_sz> anonimasu: if you ghive me an email addy, you can have a copy of this FPGA to stare at ..
[16:01:06] <jmkasunich> oh, I think I see... you mean the drawings?
[16:01:11] <alex_joni> yup
[16:01:20] <jmkasunich> ok, I'll fix 'em
[16:01:27] <jmkasunich> the originals are in easycad, on winblows
[16:01:34] <alex_joni> heh ;)
[16:02:01] <jmkasunich> the easycad folks have made talk about a Linux version, but it hasn't happened yet
[16:02:13] <alex_joni> is easycad free?
[16:02:36] <jmkasunich> nope, about $200, and IMHO worth every penny... I'd buy the Linux version in a heartbeat
[16:03:52] <alex_joni> nice
[16:04:28] <jmkasunich> I think easycad is what autocad should have been, before it got bloated and multi-kilodollar pricing
[16:04:57] <alex_joni> yeah..
[16:05:14] <alex_joni> I know.. I wanted to buy a license (but the cheapest is around 1200 $)
[16:05:31] <jmkasunich> that's not too surprising, since the author of Easycad is one of the original Autocad developers (Mike Riddle)
[16:05:32] <gezr> morning, and I use turbocad have had it since version 3, I dont have to spend much to upgrade and it has a cam package out now, I havent tried it, I dont have 800 bucks
[16:05:33] <alex_joni> so I gladly skipped that
[16:06:50] <alex_joni> jmk: btw, weren't some of the parport pins inverted by default?
[16:06:53] <rayh> * rayh is back
[16:07:06] <alex_joni> rayh: you said smthg about a database
[16:07:12] <alex_joni> how would that look like?
[16:07:17] <jmkasunich> the hardware does invert one or two pins, but my parport driver corrects for that, so the user doesn't have to deal with it
[16:07:20] <rayh> Yes and I can't seem to find my notes on it.
[16:07:29] <alex_joni> ok ;)
[16:07:44] <alex_joni> files? or actual db?
[16:07:52] <rayh> It could be either.
[16:08:17] <rayh> A flat file, organized like a db
[16:08:33] <rayh> At least that would get us started.
[16:08:47] <rayh> That file could reside on the WiKi for now.
[16:09:01] <robin_sz> Mighty Wiki to the rescue!!
[16:09:09] <alex_joni> yay
[16:09:17] <rayh> What is essential is that everyone have access to the same data
[16:09:26] <alex_joni> CVS?
[16:09:29] <alex_joni> wiki?
[16:09:40] <robin_sz> badger!
[16:09:54] <rayh> That would work as well except only dev could grow it.
[16:10:08] <robin_sz> [OT]
http://www.badgerbadgerbadger.com/
[16:10:21] <rayh> robin_sz: could you serve a db from there?
[16:10:33] <robin_sz> sure .. I have MySQL
[16:10:46] <robin_sz> you have a frontend? or do you mean direct connections?
[16:10:57] <alex_joni> php?
[16:11:10] <rayh> How 'bout a html front end?
[16:11:28] <alex_joni> you'd need php for the mySql functions
[16:11:37] <alex_joni> it would deliver html pages though
[16:11:55] <alex_joni> but I don't think that's the way to go...
[16:12:03] <alex_joni> file as a db sounds more ok to me
[16:12:54] <danfalck> good morning
[16:13:00] <alex_joni> morning
[16:13:05] <robin_sz> * robin_sz waves
[16:13:25] <robin_sz> right time for tea ... and buns
[16:14:44] <alex_joni> jmk: say if you want to give CL a go
[16:15:18] <alex_joni> hello paul_c
[16:15:32] <rayh> If we make a set of categories, I could write a tickle front end.
[16:15:46] <paul_c> Morning all.
[16:16:04] <robin_sz> ideally, you would have some sort of flat file
[16:16:12] <robin_sz> that gave the categories
[16:16:24] <robin_sz> an the vars in each categroy
[16:16:26] <robin_sz> and the limits
[16:16:37] <robin_sz> (possibly with refernece to other vars)
[16:16:44] <rayh> Okay....
[16:16:48] <robin_sz> eg ThisVar cant be bigger than ThatVar
[16:17:06] <robin_sz> then just write a generic tickle front end
[16:17:08] <rayh> Ruleset.
[16:17:12] <robin_sz> exactly
[16:17:19] <robin_sz> perhaps ...
[16:17:37] <robin_sz> put < tags > around stuff
[16:17:58] <robin_sz> <category name="Axis">
[16:18:16] <rayh> On the original config I used the ; and [] in the ini file
[16:18:23] <robin_sz> <var name="someSetting">
[16:18:38] <rayh> That way I could read one back in.
[16:18:42] <robin_sz> I mimagin you need to include some notion of hierarchy
[16:19:47] <rayh> Certainly. If this then all of these.
[16:19:56] <robin_sz> eg the motion has 4 axes, each axis has 2 switch inputs ..each input has a pin and active_low/high setting
[16:20:22] <robin_sz> its all very hierarchical .. lends itself nicely to a GUI thing
[16:20:57] <alex_joni> how about the HAL stuff?
[16:21:23] <robin_sz> hmm ..
[16:21:31] <rayh> Think we could write a single configuration wizard
[16:21:46] <alex_joni> rayh: we agreed we should
[16:21:48] <rayh> HAL would be one of the up front choices?
[16:21:53] <robin_sz> that adresses the several config areas of EMC, through a common fornt end
[16:22:00] <rayh> CL another?
[16:22:17] <alex_joni> I think we should leave CL out for now
[16:22:21] <robin_sz> in EMC2, you'll be having HAL as standard, CL as an option
[16:22:27] <alex_joni> not sure how many will use that
[16:22:52] <robin_sz> * robin_sz really goes for a cup of tea
[16:22:56] <robin_sz> with badgrs!
[16:22:59] <rayh> k we can add things as we go if I can write the tickle that robin imagines.
[16:23:32] <rayh> My guess is this flat file will not be very man readable.
[16:23:42] <alex_joni> how about XML?
[16:23:53] <paul_c> * paul_c slaps alex_joni
[16:23:57] <alex_joni> use an XML-viewer to view it
[16:24:05] <alex_joni> it already looks like XML
[16:24:13] <rayh> and tickle to parse it?
[16:24:23] <alex_joni> dunno ;)
[16:24:43] <rayh> It certainly has some of the characteristics that folk want to use xml for.
[16:24:59] <alex_joni> * alex_joni is afraid to get slapped again
[16:25:24] <rayh> I tried a bit of that during my db experiment but I don't know nearly enought to pull it off.
[16:25:35] <alex_joni> about xml?
[16:25:56] <rayh> that whole structure leaves me in the dust.
[16:26:27] <alex_joni> you got stack support in tcl?
[16:26:35] <alex_joni> push/pop things?
[16:27:01] <rayh> I'd be much more comfortable with hard coded sections in the flat file.
[16:28:07] <rayh> I think my config program is still in the dropbox.
[16:28:12] <alex_joni> that would work too
[16:28:48] <rayh> We could certainly use a markup in the flat file.
[16:29:02] <rayh> <HAL> </HAL>
[16:30:09] <rayh> I'm not certain that such a thing would live in the WiKi though.
[16:30:21] <paul_c> parsing XML in C is a pain even with the aid of an xml-lib
[16:30:22] <alex_joni> rayh: join #flood
[16:30:33] <jmkasunich> I will be disapointed if the file is no longer readable without some kind of tool
[16:30:35] <alex_joni> paul_c: depends on what you need
[16:30:46] <alex_joni> jmk: it's still a textfile
[16:30:56] <jmkasunich> man readable and editable with a text editor is very important to me
[16:31:11] <rayh> Yes I very much agree.
[16:31:28] <jmkasunich> still a textfile, but when you get to XML like syntax, it is very unfriendly
[16:31:35] <rayh> I think that it would be desirable to have some input screens in ticke/tk as well.
[16:32:02] <robin_sz> why is xml "unfriendly"
[16:32:13] <rayh> I imagine each HAL module having a section.
[16:32:27] <alex_joni> <emcconfig>
[16:32:28] <alex_joni> <hardware>
[16:32:28] <alex_joni> <emc>
[16:32:28] <alex_joni> <axis0></axis0>
[16:32:28] <alex_joni> <axis1></axis1>
[16:32:28] <alex_joni> <axis2></axis2>
[16:32:29] <alex_joni> <axis3></axis3>
[16:32:31] <alex_joni> </emc>
[16:32:36] <alex_joni> <hal></hal>
[16:32:37] <alex_joni> </hardware>
[16:32:37] <alex_joni> </emcconfig>
[16:32:49] <robin_sz> eaxcatly ..
[16:33:02] <robin_sz> <axis0>
[16:33:11] <robin_sz> <P>1</p>
[16:33:13] <alex_joni> * alex_joni starts a wiki page
[16:33:20] <robin_sz> <i>2</i>
[16:33:27] <robin_sz> </axis0>
[16:33:53] <rayh> The wiki page could live if <nowiki> first and </nowiki> at the end.
[16:34:11] <paul_c> Miss a closing tag and things get messy
[16:34:23] <rayh> That would make everything plain text.
[16:34:48] <rayh> Paul is correct. The tickle parser would have to allow for messy code.
[16:35:05] <robin_sz> paul_c: true, but at least there are pre-made tings to check for errors like that and show you where you made them
[16:35:52] <rayh> Or it would have to be able to see two opening codes in a row. <axis_0> ... <axis_1>
[16:36:38] <alex_joni> robin: how do you do <br> between nowiki's ?
[16:36:49] <paul_c> XML is fine for some tasks, but it is not the solution for everything
[16:36:58] <robin_sz> alex_joni: just use a cr/lf
[16:37:03] <rayh> You don't it should honor returns
[16:37:05] <SteveStallings> </nowiki><br><nowiki>
[16:37:27] <alex_joni> http://www.redpoint.org.uk/cgi-bin/emcinfo.pl?EmcConfig
[16:38:19] <SteveStallings> just joking, really...
[16:39:16] <alex_joni> * alex_joni wonders who has actually seen/used emcconfig
[16:39:44] <rayh> Someone posted a file made with it years ago. I was surprised.
[16:39:56] <alex_joni> years ago?
[16:40:08] <alex_joni> I mean the new one released last month by some german folks
[16:40:31] <robin_sz> perhaps badgers made it?
[16:40:33] <alex_joni> http://5128.rapidforum.com/topic=115679365955
[16:41:48] <robin_sz> my germnan is a bit weak, does it say anything about nadger in there?
[16:41:58] <robin_sz> s/n/b/
[16:42:24] <rayh> oh. That is a copy of Martin's stuff.
[16:42:41] <alex_joni> my germbab is a bit weak, does it say abythibg about badger ib there?
[16:42:46] <rayh> He's got some good ideas in there.
[16:43:07] <alex_joni> that's what I was talking about all along
[16:43:12] <alex_joni> we could just extend that
[16:43:20] <alex_joni> although .. it's python ;)
[16:44:34] <danfalck> hey guys, trying to go through this morning's log. Why make things complicated with XML?
[16:44:51] <danfalck> doesn't that add a lot more complexity to things?
[16:45:15] <danfalck> man readable files w/ a document should do
[16:45:56] <paul_c> certainly using a 3rd party lib to parse an xml file adds a degree of bloat
[16:46:28] <rayh> And the parsing rules can get to be a mess.
[16:46:33] <paul_c> the smallest xml-lib is around 150K as I recall.
[16:46:37] <alex_joni> ok.. skip parsing
[16:46:48] <alex_joni> I did a program a few years ago
[16:46:53] <alex_joni> with an xml-like config
[16:47:00] <alex_joni> did the parsing in about 1 hour
[16:47:03] <alex_joni> no libs used
[16:47:09] <rayh> I had to write some simple rules for the original config but it was not that bad.
[16:47:16] <anonimasu> parsing somthing is pretty easy if you have a ruleset for parsing it..
[16:47:17] <anonimasu> :)
[16:47:19] <alex_joni> but it's certainly not error-tolerant or anything
[16:47:26] <rayh> And I believe that we could do that with a heading in the flat file.
[16:47:33] <CIA-4> 03jmkasunich * 10documents/images/hal-parport-block-diag.eps: fixed some mis-labeled pins on parport block diagram
[16:47:39] <alex_joni> that's ok too
[16:47:56] <rayh> foreach tag $tags {...}
[16:48:22] <paul_c> Agreed, you could write your own xml parser in less than 150K, but why re-invent the wheel..
[16:48:39] <rayh> I'm not thinking xml tags.
[16:48:40] <anonimasu> for portability..
[16:48:44] <alex_joni> in order not to bloat stuff
[16:48:54] <anonimasu> and not to depend on too many libs
[16:48:55] <alex_joni> but I agree that XML might not be the best thing
[16:49:16] <anonimasu> maybe XML-ish but not XML
[16:49:45] <alex_joni> rayh: feel free to edit EmcConfig on the wiki to explain the layout you'd like
[16:49:55] <rayh> The <nowiki> first and </nowiki> end doesn't work worth a darn on WiKi
[16:50:08] <alex_joni> I know
[16:50:10] <alex_joni> I removed them
[16:51:25] <rayh> That's better.
[16:51:29] <jmkasunich> how about <pre> and </pre>
[16:51:40] <alex_joni> pre should work
[16:52:13] <jmkasunich> revised Hal_Introduction.pdf is up on linuxcnc.org
[16:52:33] <alex_joni> jmk: quick fix ;)
[16:52:33] <jmkasunich> I wish the lyx->pdf and upload to linuxcnc.org could be automated
[16:53:06] <rayh> Can we live with something like the wiki EmcConf page now
[16:53:20] <alex_joni> I think so
[16:53:34] <rayh> The tickle parser would take very little work.
[16:53:38] <jmkasunich> seems like that's the only thing I've been able to do lately... if it takes much concentration or long term focus, I can't seem to make the time
[16:53:56] <alex_joni> jmk: care for another short term stuff?
[16:54:10] <rayh> We would leave it to the author/editor to keep stuff in the correct relationship.
[16:55:01] <jmkasunich> alex_joni: sure
[16:55:32] <alex_joni> try the CL stuff ;)
[16:55:41] <alex_joni> * alex_joni wonders if it works for jmk
[16:56:02] <rayh> I'd also like to see <compute> PERIOD= #+vv </compute>
[16:56:06] <jmkasunich> ok. I'll give it a try
[16:57:02] <alex_joni> rayh: shall we define the config file?
[16:57:07] <alex_joni> what it contains & such?
[16:57:12] <rayh> One thing that bothered me about my own config and martin's as well is that you can't set all axes the same with a single button.
[16:57:48] <rayh> Let's use the EmcConfig page to do it.
[16:57:58] <alex_joni> that's what I was thinking about
[16:58:31] <alex_joni> Martin's software has a emcconf.help file
[16:58:34] <alex_joni> do you have that?
[16:58:59] <rayh> Let's add a page EmcConfTickle for the code itself.
[16:59:33] <alex_joni> does the structure stay the way it's now?
[16:59:37] <rayh> I think I'd use a <help> </help> within each.
[16:59:37] <alex_joni> or do we change that?
[17:00:10] <alex_joni> are you happy with <axis0></axis0> ?
[17:01:14] <rayh> Is one axis the same as all the others.
[17:01:32] <rayh> In a configurator that is.
[17:01:40] <alex_joni> should be
[17:02:02] <rayh> So <axis> ought to be enough </axis>
[17:02:17] <alex_joni> ahh.. I see
[17:02:28] <alex_joni> yes, only the output needs to have more than one
[17:02:49] <rayh> When we get to writing or posting the info created by the configuration we will need to sort the data into each.
[17:02:54] <alex_joni> I added EmcConfTickle to the text,
[17:03:09] <rayh> Good.
[17:03:29] <alex_joni> #comments in the emcconfig file?
[17:03:42] <alex_joni> or <!-- comments -->
[17:05:07] <alex_joni> rayh: remind me... this file is only to configure the tcl frontend, or does it contain sample data too?
[17:05:54] <rayh> Robin suggested that it contain the whole data. Structure as well as EMC specific config data.
[17:06:10] <alex_joni> well then we need more than <axis>
[17:06:23] <alex_joni> we need <axis0>... <axis5>
[17:06:41] <rayh> Only for output though.
[17:06:55] <alex_joni> if the settings are the same, yes
[17:07:10] <alex_joni> but if you have different PID's on axis0 and axis1...
[17:07:27] <rayh> As we work our way through the wizard, it builds axis0 -> n.
[17:07:39] <alex_joni> with the same values at first..
[17:07:42] <rayh> But that is not saved in this flat file.
[17:07:46] <alex_joni> ok.. got it
[17:07:49] <alex_joni> sounds ok
[17:08:07] <alex_joni> do <!-- such comments --> sound ok ?
[17:12:11] <rayh> * rayh starts the Tickle page.
[17:14:11] <rayh> oops. WiKi assumes numbering for # marks
[17:14:57] <alex_joni> use <pre>
[17:16:02] <rayh> Now the problem is \ for newline as extension of last line.
[17:17:03] <alex_joni> save the page (I wanna see it)
[17:18:07] <narnia> rayh, hello, how goes it?
[17:18:41] <rayh> Good Terry. Glad to see you here.
[17:19:36] <alex_joni> rayh: should we include also possible values?
[17:19:50] <rayh> yes.
[17:20:00] <alex_joni> e.g. <type>linear</type> (should contain linear/rotatory, etc?)
[17:20:17] <alex_joni> how should that sound like?
[17:20:45] <alex_joni> <type>linear<alt>rotatory</alt></type> ?
[17:21:00] <anonimasu> hmm, that sounds right
[17:21:06] <alex_joni> not sure
[17:21:31] <rayh> <type>linear rotatory</type>
[17:21:39] <rayh> Would be esier
[17:21:41] <alex_joni> <type>linear<alt></alt></type> - should state that any other value is ok ?
[17:21:51] <alex_joni> rayh: that doesn't include a default value
[17:21:52] <narnia> rayh, i am here. not sure how talkative i will be. on 11 feb 2004, my friend and companion of 12 yrs daily companion for 8 of those yrs, huckleberry finn, huck for short, cross the rainbow bridge. my dog is gone. it is difficult to be by the computers. huck was always here. either asleep on the floor by my side or at my feet.
[17:22:16] <narnia> rayh, 11 feb 2005
[17:22:38] <alex_joni> * alex_joni is sorry to hear that...
[17:23:14] <rayh> I know a bit about what that loss means. Vern left a bit ago.
[17:26:42] <rayh> What character do you use in the wiki for a literal next character.
[17:26:47] <alex_joni> what's default velocity ?
[17:27:13] <alex_joni> rayh: use <pre> and paste text only
[17:27:14] <rayh> Sets the low end of the scale display
[17:27:41] <alex_joni> that way you can copy paste the text to a .tcl script and run it
[17:28:17] <paul_c> * paul_c & Dan are going for a walk around beaverton
[17:28:35] <narnia> rayh, i have talking with the brlcad people. i suggested to them that they should consider attending n.a.m.e.s. in april. we have also talked about emc. what i plan to work on is the following:
[17:29:14] <narnia> rayh, add bi-drectional step file support to brlcad.
[17:29:52] <rayh> Really. Great.
[17:30:03] <narnia> rayh, take my tool-path code i have written for my project and integrate it in brlcad.
[17:30:19] <rayh> I've not worked brlcad at all.
[17:30:34] <rayh> Just seen the posts to the cad lists.
[17:30:49] <narnia> rayh, ultimately the goal is to provide an interface from brlcad to emc.
[17:31:01] <rayh> That should be fun.
[17:31:29] <rayh> alex. The \ is stripped from the copy paste.
[17:31:38] <alex_joni> how so?
[17:31:39] <narnia> rayh, brlcad just became opensource in january. it is an excellent cad program. price is right. free
[17:32:02] <rayh> Yes I saw the open announce.
[17:35:48] <rayh> Is the cam portion of brlcad open as well.
[17:35:50] <narnia> rayh, the step bi-directional interface will be written in c.
[17:36:24] <rayh> Oh. You are way ahead of me on getting info out of it.
[17:37:07] <narnia> rayh,
http://brlcad.org/
[17:37:26] <rayh> k
[17:38:28] <narnia> rayh, the nice thing about step files their specifications are iso standards. the bad thing about that is iso standards are not inexpensive.
[17:38:54] <narnia> rayh, are fred proctor and william still at nist?
[17:39:19] <rayh> Right. NIST promised an interface lib. Yep they are both there and Tom is on some project for them.
[17:39:45] <rayh> alex_joni: I'm gonna ignore the \ issue for the moment.
[17:39:52] <alex_joni> best thing
[17:40:00] <narnia> rayh, i was thinking that perhaps fred or william would have copies of the step standards that i could borrow.
[17:40:03] <alex_joni> rayh: cycle_time can be computed?
[17:40:14] <alex_joni> along with traj and servo ?
[17:40:19] <alex_joni> and PERIOD?
[17:40:28] <rayh> They would have them and it won't hurt to ask.
[17:40:55] <rayh> alex_joni: you bet. calculation will be valuable in lots of places.
[17:41:06] <alex_joni> ok, then I won't include those
[17:41:09] <rayh> Even if only to suggest a default for folk.
[17:41:23] <alex_joni> yup
[17:41:55] <narnia> rayh, is matt shaver still involved with emc?
[17:42:25] <rayh> He is still "Pres" -- spends a lot of time selling real estate these days.
[17:42:44] <rayh> EMC is still right up there in his thinking.
[17:42:57] <narnia> rayh, matt selling real estate? that is a switch.
[17:43:09] <rayh> Retrofitting is taking a back seat to sales.
[17:43:19] <narnia> rayh, what prompted that 'career' change?
[17:43:23] <rayh> Yes it is.
[17:43:43] <rayh> Money. His machine shop stuff trailed off.
[17:43:53] <narnia> rayh, umm.
[17:44:18] <rayh> One does what one imagines.
[17:44:24] <rayh> or tries to anyway.
[17:44:35] <narnia> rayh, i am still home bound. still in the legal nightmare 8+yrs now.
[17:45:08] <rayh> Wow. That is hard. Glad you are back here with development though.
[17:45:15] <rayh> These are a good bunch of guys.
[17:45:30] <rayh> * rayh needs to make lunch for the fam.
[17:45:52] <alex_joni> yet.. as rayh pointed it out... people with deep knowledge on emc are always needed ;)
[17:45:57] <rayh> I'll watch the discussion and be back in a couple hours.
[17:46:01] <rayh> * rayh is away: I'm busy
[17:46:13] <narnia> rayh, i miss huck. the house seems empty without him here.
[17:46:13] <alex_joni> rayh: before you go
[17:46:32] <alex_joni> I added some more stuff.. is this the right way?
[17:46:58] <robin_sz> whenever one of our cats passes on, we go straight out and get another .. its not the same, but well, its a start
[17:47:36] <alex_joni> robin: how does
http://www.redpoint.org.uk/cgi-bin/emcinfo.pl?EmcConfig look ?
[17:48:06] <alex_joni> * alex_joni grabs a bite
[17:48:26] <narnia> robin_sz, i have thought of that. the current adoption fees for a puppy are beyond my ability.
[17:48:31] <robin_sz> could do with a few more carriage returns
[17:49:02] <robin_sz> narnia: adoprion fees? coo. in the UK they are just glad for you to take one.
[17:49:06] <alex_joni> feel free
[17:49:18] <SWPadnos> Is there god documentation on the FFx PID parameters. (ie, not "look at the code" :) )
[17:49:49] <SWPadnos> good, not god :)
[17:50:22] <robin_sz> SWPadnos: there is documentatio somewhere, but whenever ive done it it always been corrected by people going .. no no, tru 1000 in there amd 1 in there etc, nothing like the docs
[17:52:20] <robin_sz> PID config is 2 teaspoons of luck and a pinch of experience
[18:01:09] <narnia> bbiab need to get away for a bit. :-(
[18:04:55] <narnia> brlcad, hello, thnak you for coming.
[18:05:25] <brlcad> howdy
[18:05:57] <narnia> brlcad, these is the enhanced machine controller people.
[18:06:30] <brlcad> * brlcad figured as much
[18:07:47] <brlcad> so what's the current development efforts under way with emc?
[18:08:06] <narnia> brlcad, i have been explain about writing the bi-directional step interface for brlcad and attempting to merge my tool-path code in brlcad so that eventually brlcad and emc would 'talk' to one another.
[18:09:13] <brlcad> yes, that would be nice
[18:10:13] <narnia> brlcad, fred proctor, at nist has the step iso standards. i am going to be sending fred an e-mail later and ask if i may borrow them.
[18:10:46] <narnia> brlcad, i have been able to do some work on the step interface.
[18:11:36] <SWPadnos> narnia: what features does your toolpath code have? (is there info on a website somewhere?)
[18:11:54] <narnia> brlcad, since friday, 11 feb 2005, i have had a difficult time being at the computers. my dog huckleberry finn, huck, for short, crossed to the rainbow bridge. huck is gone.
[18:12:22] <brlcad> aww, sorry to hear that
[18:12:24] <narnia> SWPadnos, the web site is severely out of date.
[18:12:52] <brlcad> i put a purchase order in for another copy of the spec myself last week
[18:12:57] <robin_sz> alex_joni: see the wiki,
[18:13:22] <narnia> brlcad, huck would always be alseep on the floor either at my side or at my feet. the house is empty without him.
[18:13:24] <brlcad> it'll take a couple weeks at least for the purchase to go through, but once it arrives can get started
[18:13:26] <robin_sz> how long before CAM programs begin to output in STEP?
[18:13:49] <robin_sz> mainstream I mean
[18:13:52] <narnia> robin_sz, some all ready output in step-nc.
[18:14:11] <robin_sz> really? coo. didnt know that
[18:14:22] <narnia> robin_sz, most are european.
[18:14:34] <robin_sz> such as?
[18:14:57] <brlcad> iges was the standard (and still is widely used) but step is slowly becoming more popular
[18:15:01] <narnia> robin_sz, i will paste the links later.
[18:15:09] <robin_sz> iges is not a machine control standard though
[18:15:12] <alex_joni> * alex_joni is back
[18:15:26] <robin_sz> IGES is a physical model definition
[18:15:42] <robin_sz> you then have to create machining ops based on that
[18:15:54] <brlcad> * brlcad nods
[18:16:03] <anonimasu> iges are neat..
[18:16:06] <anonimasu> it works with everything
[18:16:07] <anonimasu> :)
[18:16:27] <alex_joni> robin: PID is used for freqmod too
[18:16:33] <alex_joni> not only segmodqueue
[18:16:40] <robin_sz> where as step-nc is presumably the output after processing a IGES model with some sort of processor
[18:16:47] <robin_sz> alex_joni: yes so?
[18:16:50] <anonimasu> * anonimasu nods
[18:17:15] <alex_joni> driver type="segmotqueue" should be driver type="segmotqueue freqmod"
[18:17:18] <robin_sz> alex_joni: if you had a frqmod type entry, you could still have that PID
[18:17:50] <robin_sz> alex_joni: no, here I am specifying the driver I am using in this *particular* instance will be segmotqueue
[18:18:04] <alex_joni> ok
[18:18:17] <robin_sz> we could write a DTD, like they do for html
[18:18:26] <alex_joni> DTD ?
[18:18:43] <SWPadnos> Document Type Definition
[18:18:43] <robin_sz> if you have a <foo> entry it must contain Bar or baz, and if bazm then bill as well
[18:18:44] <narnia> brlcad, did you order from ansi.org or iso.ch?
[18:19:22] <robin_sz> alex_joni: there are already standard tools that wioll compare a doc agaisnt a DTD, point out errrors from the stadard, excessive values etc
[18:19:50] <alex_joni> robin: sounds ok
[18:20:05] <robin_sz> I'll settle for a sane config first though
[18:20:08] <robin_sz> :)
[18:21:01] <narnia> robin_sz, from what i have been to figure out the iges/step files are converted to step-nc. step-nc looks almost identical to step but there are subtle differences.
[18:21:29] <alex_joni> robin: nice description on the bottom (didnt see that the first time)
[18:22:40] <SWPadnos> (I think it wasn't there the first time :) )
[18:23:12] <alex_joni> robin: I saved that file as an XML, and opened it with opera
[18:23:14] <narnia> brlcad, what are brlcad's abilities to do auto-dimensioning?
[18:23:38] <alex_joni> it said it contains an error on line 41, char:16 (you closed min instead of max;)
[18:24:34] <alex_joni> firefox does the same
[18:24:48] <alex_joni> so I think error checking for now, can be done with a current browser
[18:26:34] <SWPadnos> indeed. There is aklso a problem with there being two <driver> lines, and no </driver>
[18:26:44] <alex_joni> yup
[18:26:48] <alex_joni> seen that too
[18:27:23] <SWPadnos> Too bad you can't edit with Mozilla / FireFox
[18:27:50] <alex_joni> well.. you can (the wiki at least)
[18:28:20] <SWPadnos> Unfortunately, that page is read-only - I can't edit it.
[18:28:35] <alex_joni> you need to register to access the wiki
[18:28:41] <brlcad> narnia, i just put the request in, someone else orders
[18:28:44] <alex_joni> follow the info on the first page
[18:28:45] <SWPadnos> Well - that would help :)
[18:28:51] <brlcad> it'll be a week or two before I hear anything
[18:30:01] <narnia> brlcad, ah okay.
[18:30:03] <brlcad> there's not much of anything that is automated, much less the dimensioning in brl-cad
[18:30:14] <brlcad> you specify what you want
[18:30:34] <brlcad> constraints are on the todo list, but also not presently in either
[18:30:53] <narnia> brlcad, okay, understand.
[18:31:50] <narnia> brlcad, may take awhile but getting brlcad and emc to work together is a 'good' thing.
[18:32:02] <SteveStallings> I need to run, but wanted to toss out one issue before I go.... do we have any better details on plans for EMC meeting at NIST? Will leave chat running to log hopefully....
[18:33:04] <narnia> SteveStallings, i have not seen any. of course not being to drive or attend i have not been looking that hard.
[18:33:37] <SteveStallings> jmkasunich as making some sort of arrangements...
[18:33:43] <brlcad> narnia: agreed, though that path is not trivial -- brl-cad is a solid modeling system, not a cam package -- so bridges will need to be built
[18:34:45] <narnia> brlcad, yes. part of that bridge is all ready built. i am hoping to recycle code from my project for that bridge.
[18:36:44] <narnia> brlcad, perhaps the bridge should be named 'brl-cam' to avoid confusion. any thoughts?
[18:37:05] <brlcad> heh
[18:37:34] <jmkasunich> damn... X locked up
[18:37:48] <SWPadnos> jmk: Impossible! :)
[18:37:53] <alex_joni> welcome back ;)
[18:38:00] <narnia> jmkasunich, really? what distribution you running?
[18:38:20] <SWPadnos> * SWPadnos thinks it's a broken one :)
[18:38:23] <jmkasunich> BDI-TNG
[18:38:32] <jmkasunich> I don't know what really locked up
[18:38:36] <alex_joni> jmkasunich: hope it's not the CL stuff that broke X
[18:38:54] <jmkasunich> the last thing I did was save a bookmark in Konqueror...
[18:38:57] <SWPadnos> Actually - I had an odd thing happen on my BDI-4.08 install
[18:39:10] <jmkasunich> then it became unresponsive to the mouse
[18:39:16] <narnia> brlcad, we at least know were we do not want to go and that is down the path the opencascade.org has gone. :-)
[18:39:18] <jmkasunich> "it" meaning everything
[18:39:22] <SWPadnos> I clicked on the X axis display, and the machine froze. (don't remember if it was a right-click or a left-click)
[18:39:34] <jmkasunich> I had to ctrl-alt-F2 and start killing processes
[18:40:05] <jmkasunich> I expected that killing konqueror would do the trick, but it didnt, wound up killing everything and restarting X
[18:40:07] <alex_joni> SWPadnos: BDI-4.08 had a homing bug
[18:40:31] <alex_joni> that would freeze the pc
[18:41:23] <brlcad> there's a new modeling interface under development that wraps up all of brl-cad's functionality -- that environment will make cam integration pretty easy
[18:41:25] <narnia> okay here is a thought. would the 'keepers' of the bdi consider putting brl-cad on the disk?
[18:41:26] <robin_sz> narnia: I stil have great need of a 2d cam package for plsma cutting ..
[18:41:29] <robin_sz> plasma
[18:41:47] <robin_sz> keep us 2d guys in mind when you do the cam thing
[18:41:52] <alex_joni> narnia: don't think there's a reason not to
[18:42:15] <alex_joni> as long as there are debian packages... absolutely no problems
[18:42:37] <narnia> robin_sz, i understand. i am mainly a 2-d person. oxy-fuel, plasma, router, etc.
[18:42:58] <alex_joni> * alex_joni is also a 2-d guy
[18:43:00] <robin_sz> narnia; right. I have a commercial angle on this btw,
[18:43:16] <robin_sz> narnia: i know only of one and a half decent packages for 2d
[18:43:23] <SWPadnos> If you're willing to use Windows, then Solid Master is an excellent program (
http://www.cadmax.com)
[18:43:45] <robin_sz> SWPadnos for 2d?
[18:43:50] <SWPadnos> It does solids, as well as 2-D, and it's fully parametric
[18:44:16] <robin_sz> im prepared to look, but ive yet to see ANY solid package that is the slightest use for 2d, but .. I'll look
[18:44:47] <SWPadnos> There are several drawing modes, one or two of which are 2D only.
[18:44:54] <robin_sz> usually they cant even import half a dozen 2 d shapes into the same thing
[18:45:15] <robin_sz> Drawing.. I thought this was CAM?
[18:45:31] <SWPadnos> I found CadMax easy enough to use, that I would just do the job in 3D
[18:45:34] <alex_joni> robin: tried sheetcam lately?
[18:45:51] <robin_sz> alex_joni: yeah thats the 1/2 in my one and a half :)
[18:46:04] <alex_joni> it's pretty ok
[18:46:10] <alex_joni> what's the 1?
[18:46:16] <robin_sz> alex_joni: it has severla bad points .. but since its free :)
[18:46:28] <alex_joni> won't be long free
[18:46:39] <robin_sz> 1 .. well relaly 1 is one and a half .. because its absolutley excellent ..
[18:46:44] <SWPadnos> I think one of the best values around for 3D CAD/CAM is CadMax + STLWorks
[18:47:04] <SWPadnos> Full 3-D contouring CAD and toolpath generation for $370
[18:47:13] <robin_sz> Mecsoft Turbonest is my all world favourite 2d package
[18:47:27] <robin_sz> about 2200 USD
[18:47:35] <robin_sz> thats with MANUAL nesting ..
[18:47:41] <alex_joni> ouch
[18:47:42] <robin_sz> but, worth it.
[18:47:42] <SWPadnos> 2200>370 :)
[18:48:06] <robin_sz> yeah, but I bet cadmax is a pain in the arse for 2d
[18:48:55] <robin_sz> I'll try it and see.
[18:48:57] <SWPadnos> It seemed easy enough - you can just use a single sketch in 3D mode, then extrude it for the thicknes of your part.
[18:49:28] <SWPadnos> There is also a 2D drafting mode (as well as detailing modes, in case you actually have to print something for someone else :) )
[18:49:36] <robin_sz> ahh thats a minus point then
[18:50:17] <SWPadnos> I actually modeled the bearing brackets for a Bridgeport (both X and Y) in about 8-12 hours of fiddling the program.
[18:50:29] <robin_sz> thicknes for my part? .. theis is 2d ... things have no thickness in 2d
[18:50:35] <SWPadnos> With no mechanical CAD experience since CadKey v1.42e
[18:50:48] <brlcad> so what're other people using in the way of open source 2d and/or 3d cad/cadd/cam?
[18:51:55] <narnia> brlcad, my original 2d was to use xfig and perl scripts i had written to output the necessary control codes for the router table.
[18:52:13] <robin_sz> well, sheetcam is OK
[18:52:40] <robin_sz> mecsoft Turbonest is the best there is.
[18:52:41] <narnia> brlcad, there was no good opensource 3d cad/cadd/cam which is why i started my project 5 yrs ago.
[18:53:14] <alex_joni> there are a few cad packages, but no opensource-cam
[18:53:17] <alex_joni> or very few
[18:53:39] <robin_sz> i have no real interest in cad to be honest, you can get cad for �49 off the shelf, so thats a solved problem ..
[18:53:47] <robin_sz> CAM is VERY intersting though
[18:53:54] <alex_joni> there's gCAD3D, sagcad...
[18:54:02] <robin_sz> loads, i know
[18:54:14] <narnia> brlcad, now that brl-cad is opensource it would appear easier to intergrate various opensource parts to work together.
[18:54:15] <robin_sz> CAM though ... well
[18:54:56] <robin_sz> I'd almost consider getting hold of Les newell the sheetcam bloke and offering him cash to moidify sheetcam like I need it.
[18:54:57] <brlcad> question was specifically what're you using, though? sounds like nobody is
[18:55:45] <robin_sz> for 2d? ... I already said.. sheetcam and turbonest
[18:56:01] <brlcad> those don't seem to be open source
[18:56:05] <robin_sz> theres little call for 2d cad, most 2d stuff is auto-generated
[18:56:18] <robin_sz> oh what OPENSOURCE cam are people using for 2d
[18:56:24] <robin_sz> nothing. there aint none
[18:57:04] <robin_sz> I have very very simple requirements
[18:57:12] <robin_sz> 1) import a DXF
[18:57:13] <narnia> brlcad, there is qcad which is opensource but it is strictly 2d.
[18:57:20] <robin_sz> 2) and another
[18:57:32] <robin_sz> 3) move them about on a sheet
[18:57:42] <robin_sz> 4) press "cut" and it happens
[18:57:52] <robin_sz> thats it.
[18:58:02] <brlcad> narnia: heh, I know of a few -- it was more just inquisitive as to whether anyone was actually using anything open source right now
[18:58:38] <robin_sz> * robin_sz knows of no opensource cam packages. cad, yes plenty, cam .. none
[18:58:57] <alex_joni> robin: there's ace (but that's soo limited ;)
[18:59:05] <robin_sz> ace?
[18:59:14] <alex_joni> dxf to gcode convertor
[18:59:23] <robin_sz> oh yeah that
[18:59:29] <alex_joni> do the placing in the cad, and run it through ace
[18:59:36] <robin_sz> nah
[18:59:50] <robin_sz> Id forgtoeen about ace convertor
[19:00:24] <robin_sz> ace cant work out what bits are internal cuts what bits are external cuts.
[19:00:57] <robin_sz> shetcam you have the same problem plus you have to apply the tool to the part, set the path as external
[19:01:03] <robin_sz> all no good at all for my users
[19:01:40] <SWPadnos> Well - gotta run. (The wife wants to get out of the house - with me :) )
[19:01:55] <robin_sz> right, cya
[19:01:59] <alex_joni> bye
[19:02:24] <SWPadnos> * SWPadnos is logging, for posterity :)
[19:02:28] <robin_sz> brlcad: do feel free to enlighten us about the open source CAM packages, id be VERY interested to see them
[19:02:31] <alex_joni> brlcad: are you aware of any opensource CAM packages?
[19:02:45] <alex_joni> robin: you beat me to it ;)
[19:02:48] <robin_sz> :)
[19:02:49] <rayh> * rayh is back (gone 01:16:49)
[19:03:04] <alex_joni> welcome back rayh ;)
[19:03:04] <robin_sz> ooh, cadmnax has finished downloading
[19:03:40] <rayh> I confess to being completely lost by what's on EmcConfig
[19:03:41] <brlcad> heh, no I'm not aware of any
[19:04:07] <alex_joni> how so?
[19:04:08] <brlcad> brl-cad's the first open source solid modeler of any real use as far as I know
[19:04:55] <rayh> I copied it to a clean file and opened with konqueror. The result is unintelligible because the tags are removed.
[19:05:08] <alex_joni> bummer .. try it with firefox
[19:05:17] <alex_joni> it does xml parsing
[19:05:29] <robin_sz> it not xml
[19:05:37] <robin_sz> it just a web page with ome text on it
[19:05:57] <alex_joni> nah.. not the page
[19:06:02] <alex_joni> but the info inside it
[19:06:06] <brlcad> robin_sz | alex_joni: as narnia mentioned, a bridge from brl-cad to cam would be easy enough -- so maybe there will be a better answer next year
[19:06:14] <robin_sz> right
[19:06:23] <alex_joni> that would be very welcomed ;)
[19:06:57] <robin_sz> most/all the cam packages ive looked at (and ive looked at LOTS) have totlaly sucked at 2d, if we got a 2d solution that would be wonderful
[19:07:08] <brlcad> brl-cad has just had absolutely no need (nor funding) for cam in our area of interest
[19:07:16] <robin_sz> right
[19:07:24] <rayh> with mozilla I can see the tags.
[19:07:48] <robin_sz> I can see the tags in explorer and firfox ..
[19:08:02] <robin_sz> havent tried downloading the page and diecting it though
[19:08:06] <robin_sz> disecting
[19:08:23] <alex_joni> robin: I think this config we're writing should behave as a template
[19:08:29] <alex_joni> not with actual values in it
[19:08:51] <alex_joni> if you say for example DISPLAY: tkemc
[19:08:52] <rayh> Right.
[19:08:55] <alex_joni> and the user needs to change that
[19:09:05] <alex_joni> he'll have to know what the alternatives are
[19:09:09] <robin_sz> well .. are you going to use standard templeat e language or invent your own?
[19:09:32] <robin_sz> all the USER should see id a drop down box
[19:09:39] <alex_joni> yup
[19:09:43] <rayh> What are the advantages of each
[19:10:11] <robin_sz> well, I seen absolutely no value in putting up a template for discussion on the wiki :)
[19:10:12] <alex_joni> how do you use a standard template language?
[19:10:23] <alex_joni> it's not for discussion
[19:10:23] <brlcad> robin_sz: there's development under way in brl-cad for a new interface that intends to make it trivial to work with both 2d and/or 3d -- I'll drop by again when we've got something of interest to play with
[19:10:31] <robin_sz> alex_joni: see the XHTML DTD on W3C.org
[19:10:46] <brlcad> it's going to have a plug-in interface that'll make features like drafting and machining simple extensions
[19:11:26] <robin_sz> brlcad: kewl, that will be intersting .. id say 90% of the 2d user will be outputting shapes as DXF from some sort of auto-generatin package, so if you are going to do 2d, bear that in mind
[19:11:56] <brlcad> brl-cad already does dxf import/export
[19:12:01] <rayh> I could easily take something like <driver_types> freqmod steppermod stgmod newstgmod </driver_types>
[19:12:02] <robin_sz> excellent.
[19:12:17] <brlcad> among a variety of major formats
[19:12:25] <rayh> and build a set of selection buttons.
[19:12:28] <robin_sz> I saw the thng on the wiki as a example of how it could work
[19:12:52] <alex_joni> I saw the thing on the wiki as an template for the configurator
[19:13:00] <alex_joni> to build dropdowns based on it
[19:13:02] <robin_sz> brlcad: what every single 3d cam package i have seen has got wrong is the inablility to work with more than one 2d object at once
[19:13:11] <alex_joni> and select the first value by default
[19:13:29] <rayh> first value default is fine.
[19:13:46] <alex_joni> but have the others as alternatives
[19:14:07] <alex_joni> rayh: how about a character to specify that it's a form, not drop-down
[19:14:08] <rayh> Certainly.
[19:14:30] <alex_joni> so user can type in stuff (or make that default where only one value is?)
[19:14:37] <robin_sz> I suggst you do it on a different page .. it will confuse users totally
[19:14:45] <robin_sz> or hide that page
[19:14:51] <alex_joni> it's not shown from anywhere
[19:14:56] <robin_sz> fair enough
[19:15:00] <paul_c> Yo Dave.
[19:15:07] <dave-e> hwllo
[19:15:13] <robin_sz> let me copy that stuff out as an example ..
[19:15:17] <dave-e> oh well can't tuype
[19:15:21] <alex_joni> well.. it is from my page (emcwiki.pl?AlexJoni)
[19:16:40] <robin_sz> brlcad: all 2d user will want to be able to treat each imported DXF as an object and move it with respect to other 2d objects .. all the 3d packages seem to want to either fix them in space .. so you end up with no way to move one with restpest to an other
[19:17:05] <alex_joni> restpest?
[19:17:19] <robin_sz> brlcad; or .. worse, they forget that say, the hole in a middle of a sqaure is assoicited with that square, so you can move th square but leave the hole where it is
[19:17:27] <robin_sz> respect
[19:17:47] <alex_joni> restpest sounds better :-)
[19:17:55] <robin_sz> agreed :)
[19:18:28] <alex_joni> rayh: do you consider some hardcoded parts to the tcl stuff? or should everything be in this configfile?
[19:19:12] <alex_joni> robin: what I'd love from a 2D CAM package... (the ability to place the same part over and over again)
[19:19:22] <rayh> What the tickle needs to read are variable values, lists of things that can be expanded into the pages of the config.
[19:19:26] <alex_joni> like you do in a schematic editor when placing components
[19:19:54] <robin_sz> yeah
[19:20:04] <robin_sz> alex_joni: thats waht turbonest does
[19:20:05] <dave-e> like rubber stamp?
[19:20:09] <alex_joni> I was thinking about stuff you need to calculate (PERIOD, and such)
[19:20:29] <robin_sz> alex_joni: you inport a DXF, say how many you need, import some more with the number you need of each ..
[19:20:30] <rayh> I'd think that is where a formula would come in.
[19:20:50] <robin_sz> alex_joni; then you go to lay them out, you drag em from the left menu, onto the sheet
[19:20:55] <alex_joni> rayh: I know, but that formula would be hardcoded
[19:21:07] <alex_joni> robin: why complicate it?
[19:21:15] <robin_sz> alex_joni: arrow keys move em till they 'bump' onther object (or sheet edge)
[19:21:16] <rayh> No need to it could be read from the flat file and run with exec.
[19:21:16] <alex_joni> robin: open a part (type 1)
[19:21:29] <robin_sz> alex_joni; thats what it does
[19:21:29] <alex_joni> start clicking till you fill a sheet
[19:21:34] <robin_sz> nah
[19:21:42] <alex_joni> with placing
[19:21:50] <robin_sz> waht about when you need 12 of this and 5 of that?
[19:21:58] <alex_joni> click 12 times?
[19:22:00] <robin_sz> how do you keep track of the numbers?
[19:22:15] <alex_joni> I agree some find it difficult to count
[19:22:23] <robin_sz> yeha, but as I am fitting them in, I delte this one, move that one .. wait m, delete another twoi.. move that
[19:22:27] <dave-e> fingers and toes
[19:22:35] <alex_joni> dave: yay ;)
[19:22:51] <robin_sz> tubonest just shows a menu with <shape> 12 of 20 placed
[19:22:57] <alex_joni> rayh: nice ideea
[19:23:03] <robin_sz> so yo know you have more to drag in
[19:23:08] <alex_joni> that would be ok too
[19:23:11] <robin_sz> anyway
[19:23:18] <robin_sz> the best bit is ...
[19:23:26] <robin_sz> you click on a piece on the sheet
[19:23:32] <robin_sz> left click moves it
[19:23:38] <robin_sz> mouse wheel rotates it
[19:24:22] <robin_sz> my customers happily paid 2000+ bucks a seat for that
[19:24:24] <alex_joni> sound user-friendly
[19:24:37] <alex_joni> how about autonesting in 2000$ ?
[19:24:42] <dave-e> once upon a time I had a program that distributed g-code in a circle...12 parts around a center
[19:24:44] <robin_sz> nah ]thats more
[19:24:55] <alex_joni> bummer
[19:25:00] <robin_sz> autonesting starts at 7000 bucks
[19:25:15] <alex_joni> really?
[19:25:18] <alex_joni> * alex_joni wonders why
[19:25:27] <robin_sz> and thats for "optimized rectanglualr"
[19:25:33] <robin_sz> (ie bounding box)
[19:25:46] <alex_joni> rectangular? bleah... that's not even nesting ;)
[19:25:48] <robin_sz> "true shape nesting" is LOTS More
[19:26:31] <robin_sz> why? thats easy
[19:26:46] <dave-e> ray ... g83 works like a charm...sherline site handbook is just fine but linuxcnc seems behind.
[19:26:56] <robin_sz> if you need autonesting it because you are doing sufficient volume not to have time to manually nest
[19:27:17] <alex_joni> or you want to waste less material
[19:27:20] <robin_sz> if you have that volume, you can affod 7000 bucks right?
[19:27:33] <alex_joni> nope ;)
[19:27:37] <robin_sz> hmm, generally, manual nesting wastyes less than auto nesting
[19:28:00] <alex_joni> well then autoesting is not done properly
[19:28:26] <robin_sz> cadmax ... looks complicated, crashed before I imported anything
[19:28:32] <alex_joni> autonesting even
[19:28:44] <alex_joni> before?
[19:29:14] <robin_sz> before anyting
[19:29:34] <alex_joni> and it costs money?
[19:29:38] <alex_joni> :D
[19:29:50] <robin_sz> right, looks unuseable anyway .. seems to be 3d package like solidworks on a shoestring
[19:31:40] <robin_sz> yeah, probably very handy for a 3d desinger ... no use at all for my purposes. :(
[19:31:48] <robin_sz> another one bites the dust.
[19:32:24] <narnia> sorry had to get away for a bit.
[19:33:14] <robin_sz> I cant even fin the CAM functions in CadMAX ...
[19:33:16] <robin_sz> find
[19:34:03] <alex_joni> I think he said smthg about STLWorks
[19:34:42] <robin_sz> hmm... I have to hand it to SWPadnos, thats probably the most unsuitable CAD package for 2d ive yet come across ;))
[19:35:07] <robin_sz> I can imagine STL works will take STL files ...
[19:35:15] <robin_sz> bet it wont take DXF's
[19:35:33] <robin_sz> I'll lay odds STLworks is a 3d machining packages
[19:36:13] <robin_sz> this is the trouble ... the 3d guys just don't understand the 2d guys
[19:37:59] <robin_sz> the 2d plasma/laser/flame guys just want to drag shpaes about on a sheet and press 'cut' ... no "tooling selection" no "export as STL", no hand edit gcode .. just press and go
[19:38:01] <alex_joni> jmk: still around?
[19:38:10] <robin_sz> * robin_sz stops ranting
[19:38:30] <alex_joni> robin: ever seen the Trumpf CAM package?
[19:38:34] <robin_sz> yeah
[19:38:39] <alex_joni> looks good
[19:38:44] <robin_sz> I have that with my laser
[19:38:50] <robin_sz> its *almost* good
[19:39:07] <alex_joni> almost?
[19:39:09] <robin_sz> first you have to covert the drawing from DXF to .geo
[19:39:22] <robin_sz> then ..
[19:39:27] <robin_sz> create a nest
[19:39:43] <alex_joni> the one I've seen used dx
[19:39:44] <alex_joni> dxf
[19:39:54] <danfalck> some of the guts for cam is alreay out there needing some tlc
[19:39:58] <robin_sz> (idiotic click and drool sytem) one of the worlds worst user interface
[19:40:14] <danfalck> sagcad and varkon come to mind
[19:40:15] <alex_joni> out there?
[19:40:41] <danfalck> yes, I have Sagcad on my system and it can generate G-code from 2D profiles
[19:40:46] <danfalck> albeit one at a time
[19:40:47] <robin_sz> alex_joni: nah .. really, it imports DXF, but you have to open it in the 'drawing' tool and save as .geo .. or use their (more money) bulk convertor
[19:41:13] <danfalck> it's very beta (segfaults)
[19:41:20] <danfalck> but has some good points
[19:41:31] <robin_sz> danfalck: does it recognise internal and externl contours?
[19:41:43] <alex_joni> tried sagcad a while ago (segfaulted ;)
[19:41:58] <danfalck> all it does right now, is put out profile data
[19:42:09] <danfalck> as in follow the line
[19:42:15] <danfalck> no pocketing
[19:42:18] <robin_sz> so it just follows the DXF curve?
[19:42:21] <danfalck> yes
[19:42:34] <robin_sz> oh. same as ace covertor then
[19:42:37] <danfalck> kind of like Ace with more of a gui
[19:42:42] <robin_sz> right
[19:42:56] <danfalck> it's suppose to be a CAD program, although it's very hard to use that way
[19:43:01] <robin_sz> right
[19:43:15] <robin_sz> this is the other thing these cam packages get wrong
[19:43:22] <danfalck> I have just put it through some tests
[19:43:23] <robin_sz> they all try and do cad ..
[19:43:30] <danfalck> not used for anything useful yet
[19:43:36] <robin_sz> and they manage it ususally very badly
[19:43:47] <danfalck> but, you can use Qcad to do the dxf and import it into Sagcad
[19:43:56] <robin_sz> quite
[19:44:01] <robin_sz> cad we have lots of
[19:44:05] <robin_sz> too much in fact
[19:44:14] <robin_sz> decent 2d cam we have almost none of
[19:44:15] <danfalck> you can make the path go the direction you want, which is a plus
[19:44:38] <danfalck> agreed
[19:44:44] <robin_sz> id want it to work it out itself ideally
[19:44:47] <alex_joni> robin: I remember trying DesKam
[19:45:03] <robin_sz> oh yeah, I vaguely remember that
[19:45:09] <robin_sz> and Dolphin CAD/CAM
[19:45:10] <alex_joni> and it worked pretty good
[19:45:33] <danfalck> I need to try out websys synergy soon
[19:46:17] <alex_joni> I think it's on the latest BDI
[19:46:22] <danfalck> yes
[19:46:49] <robin_sz> whats varkon?
[19:47:12] <danfalck> varkon is a framework for a cad system that is open source
[19:47:22] <danfalck> look for the link on the linuxcnc link page
[19:47:30] <danfalck> it's quite mature
[19:47:59] <danfalck> it's not so much a cad program as a way of building your own cad program
[19:48:34] <danfalck> the authors claim that they have helped people create g-code processors with it
[19:50:04] <danfalck> opencascade also comes to mind
[19:50:38] <danfalck> a large library of cad source code
[19:51:08] <robin_sz> hmm
[19:51:15] <robin_sz> looking at varkon now
[19:54:07] <danfalck> we have an irc channel called sagcad that gezr set up. There is absolutely nothing going over there right now
[19:54:21] <danfalck> should we bring the cad/cam discussion over there?
[19:54:27] <robin_sz> opencascade looks interesting
[19:54:37] <danfalck> #sagcad
[20:11:50] <alex_joni> * alex_joni leaves
[20:11:52] <alex_joni> bye guys
[20:13:30] <narnia> robin_sz, opencascade sucks
[20:13:49] <narnia> robin_sz, been there done that it sucks.
[20:13:49] <robin_sz> narnia: looks too big
[20:14:18] <robin_sz> all I need is to import DXF's and then be able to navigate the elements in some way
[20:14:32] <narnia> robin_sz, source code is 210+mbytes. then of all things they use java ( gag, spit) to unpack it.
[20:15:12] <robin_sz> if I could import them as dxfs and keep them as objects (ie keeping relationships between inner and outer parts of a profule) Id be well on the way
[20:16:01] <narnia> robin_sz, does not compile out-of-the-box. had to write several scripts to correct errors in their generated makefiles. two glaring errors in two c++ files that even a novice programmer would have caught.
[20:16:54] <narnia> robin_sz, opencascade running on a 1.4ghz amd athlon with 1gbyte ram takes 2+hrs to load a 40mbyte step file.
[20:17:15] <narnia> robin_sz, opencascade is bloatware.
[20:17:34] <anonimasu> argh
[20:17:37] <anonimasu> that's heavy
[20:18:32] <robin_sz> no its not
[20:18:43] <robin_sz> bloatware is tihngs up to a cuple of tonnes
[20:18:50] <robin_sz> this is way way more than that
[20:18:57] <anonimasu> hehe
[20:19:10] <robin_sz> right, thats gone then
[20:19:27] <robin_sz> ANdrew Mustun had / has some open source cad stuff
[20:19:39] <robin_sz> ribboncad was GPL at one point iirc
[20:19:53] <robin_sz> or at least the libs where lgpl
[20:21:31] <narnia> robin_sz, that is qcad.
[20:24:09] <narnia> robin_sz,
http://www.ribbonsoft.com/qcad_downloads.html
[20:24:15] <anonimasu> laters..
[20:26:53] <robin_sz> yeah, he has dxflib as open sores
[20:27:18] <robin_sz> fsck it ... I'll just have to buy in someting commercial and resell it.
[20:27:44] <robin_sz> maybe sheetcam is good enough anyway
[21:00:34] <Imperator_> les: are you arround ??
[21:01:02] <ottos> day gents
[21:01:15] <Imperator_> hi
[21:01:27] <ottos> did I miss anythign major...?
[21:03:00] <Imperator_> some discussion about CAD
[21:03:38] <ottos> hmm...ok
[21:17:18] <robin_z> wahey!
[21:17:40] <robin_z> so that was a interesting, but frustrating discussion and websearch
[21:23:07] <jmkasunich> robin can't decide whether he's here or not
[21:27:14] <paul_c> Yo John
[21:30:50] <Imperator_> hmmm
[21:31:04] <Imperator_> trying emc2 with two axis and real programms
[21:31:12] <robin_s> oh its my poxy proxy loggin me in and out again
[21:32:10] <robin_s> * robin_s boggles more at his inbox
[21:32:30] <Imperator_> emc2 reads at startup the whole program then it travels to the first position and stops there for a while
[21:32:43] <robin_s> perhaps it likes it there?
[21:32:45] <Imperator_> the programm has 500 lines
[21:33:00] <Imperator_> maybe
[21:33:31] <robin_s> could be many tings
[21:34:00] <robin_s> program header ? lots of preliminary codes?
[21:36:54] <Imperator_> nope
[21:37:19] <robin_s> plain old gcode?
[21:37:52] <robin_s> g0 g1 etc .. and feedrates have been set OK?
[21:40:11] <Imperator_> think so
[21:40:42] <Imperator_> N0006 X+0055.220Y+0008.360G00 F1000
[21:40:44] <Imperator_> N0007 Z+0050.000G00
[21:40:46] <Imperator_> N0008 X+0055.220Y+0008.360G00
[21:40:47] <Imperator_> N0009 X+0055.220Y+0008.360G01 F500
[21:40:49] <Imperator_> N0010 X+0055.478Y+0009.416
[21:40:50] <Imperator_> N0011 X+0055.751Y+0010.421
[21:40:52] <Imperator_> N0012 X+0056.038Y+0011.377
[21:40:53] <Imperator_> N0013 X+0056.343Y+0012.286
[21:40:55] <Imperator_> N0014 X+0056.664Y+0013.148
[21:40:57] <Imperator_> N0015 X+0057.003Y+0013.966
[21:40:58] <Imperator_> N0016 X+0057.360Y+0014.740
[21:41:00] <Imperator_> and so on
[21:42:50] <gezr> move the first g00 to the beginning of the line, then where the first g01 is put it at the front of that line
[21:43:59] <Imperator_> i think emc reads the whole program at startup
[21:44:07] <Imperator_> is that right ??
[21:44:40] <gezr> I dont know how its lookahead works
[21:45:24] <paul_c> Yo Ho Steve
[21:50:38] <paul_c> Imperator_: Get rid of the F1000 with G0
[21:51:49] <Imperator_> ok
[21:52:06] <jmkasunich> hi paul
[21:52:41] <paul_c> Thought you might be ignoring me ;
[21:52:49] <paul_c> :}
[21:52:54] <jmkasunich> been here and there
[21:53:18] <jmkasunich> hard to find time to focus on any one thing lately
[21:53:39] <jmkasunich> I'm starting to think my mind is as cluttered as my basement
[21:53:49] <paul_c> Same here - Been travelling a lot in the last few days
[21:53:58] <jmkasunich> where you at now?
[21:54:05] <paul_c> Oregan
[21:54:11] <jmkasunich> Dan's place?
[21:54:15] <paul_c> yup
[21:54:36] <paul_c> Nice town - No snow.
[21:54:50] <jmkasunich> no snow is good
[21:54:58] <jmkasunich> ours is nearly melted
[21:55:02] <jmkasunich> but we'll get more
[21:55:33] <Imperator_> paul_c: you are flying to emc fest ?
[21:55:48] <paul_c> Haven decided yet
[21:55:58] <Imperator_> ok
[21:56:13] <jmkasunich> flying beats swimming
[21:56:23] <paul_c> or walking
[21:56:27] <Imperator_> have a price now, something around 500EUR from germany
[21:56:39] <jmkasunich> if you can walk on water, you don't need to come to the fest
[21:56:41] <Imperator_> are you shure John :-)
[21:57:41] <SteveStallings> Ok, hi Paul, now that I have mostly caught up reading....
[21:58:34] <SteveStallings> my logger dropped off, did the NIST meeting get discussed?
[21:59:57] <jmkasunich> not that I noticed
[22:00:10] <paul_c> 250 pounds to fly London<=>Washington
[22:00:38] <jmkasunich> that's about $350 or $400?
[22:00:52] <paul_c> around that...
[22:01:37] <jmkasunich> where'd you fly in for this trip? CA, WA?
[22:01:55] <paul_c> LAX
[22:02:15] <SteveStallings> The room at NIST is reserved for the Monday-Wednesday after NAMES?
[22:02:32] <jmkasunich> yes
[22:02:40] <jmkasunich> mon-thurs I think
[22:03:19] <SteveStallings> Just a conference room, or a lab-like room with Internet and space for a tabletop machine?
[22:04:20] <jmkasunich> conference room, normally seats about 30 for conference style events
[22:04:35] <jmkasunich> but will be configured with significant table space, and only about 12-15 seats
[22:05:02] <jmkasunich> so yes, room for sherline and similar machines
[22:05:27] <SteveStallings> Do we plan to announce publicly or just here on IRC?
[22:05:40] <SteveStallings> push... push..... 8-)
[22:05:58] <paul_c> or on the emc-dev list ?
[22:05:59] <jmkasunich> I need to get my a$$ in gear and add some content to the page at linuxcnc
[22:06:11] <jmkasunich> any announcment should be on the dev list, IMO
[22:07:04] <SteveStallings> I am pushing because people need time to arrange travel etc.
[22:07:18] <jmkasunich> I know
[22:08:00] <paul_c> Mid March would be the latest I would want to wait before organising flights
[22:08:33] <jmkasunich> the dates are firm now (and on the webpage - it is Mon-Thur)
[22:08:41] <jmkasunich> I need to fill in more info
[22:08:47] <jmkasunich> no time like the present
[22:08:52] <jmkasunich> * jmkasunich opens an editor window
[22:09:57] <paul_c> Do we have to submit names & addresses to NIST for security passes ?
[22:10:07] <jmkasunich> yes
[22:10:44] <jmkasunich> I have several mails from fred with various details - let me gather them together in one place and edit a little...
[22:10:47] <paul_c> That should filter out the casual users
[22:12:16] <jmkasunich> yuck - Microsoft tools make ugly HTML
[22:13:16] <SteveStallings> Frontpage is OK if you just use the basics.
[22:14:09] <gezr> I wont be able to make it to the meeting at NIST
[22:19:32] <paul_c> I need to look at the price of motels before deciding for certain.
[22:19:54] <jmkasunich> about $60 a night at the Motel 6 (other's in the area are higher, some much higher)
[22:22:38] <paul_c> * paul_c looks again @ airfairs....
[22:23:10] <paul_c> 741GBP from Dulles to London during the last week of April
[22:23:29] <paul_c> and that is just one way before taxes !!
[22:23:41] <jmkasunich> crap!
[22:24:07] <SteveStallings> forget Dulles, check on BWI
[22:24:54] <paul_c> BA and Virgin both use Dulles
[22:25:13] <jmkasunich> paul doesn't like 'mercian airlines
[22:25:15] <paul_c> Don´t know of anyone using BWI
[22:28:50] <jmkasunich> http://www.linuxcnc.org/EMC_news_history/Programmers-Fest2005.html
[22:29:00] <jmkasunich> very preliminary, needs a rewrite
[22:29:09] <SteveStallings> BA flight 229 London/BWI roundtrip $565
[22:31:17] <paul_c> 262GBP - Fewer flights to choose from.
[22:31:29] <SteveStallings> true, but much cheaper
[22:31:57] <SteveStallings> and BWI is closer to NIST
[22:32:16] <paul_c> ´bout the same price for Dulles or BWI
[22:33:02] <SteveStallings> so you found Dulles flight for much less than the 741GBP quoted earlier
[22:34:11] <paul_c> that was one price for a single flight from Dulles to London - Not on the right dates for me...
[22:41:13] <Imperator_> paul_c. are you only flying for emc ?� Or do you make also some weeks holidays there ?
[22:42:06] <paul_c> Take a few weeks holiday
[22:43:23] <Imperator_> ok
[22:44:55] <Imperator_> I realy also want to come, but only for a week that is also to expensive for me. But I don't want to travel alone around in the USA. At the moment I try to get one of my friends to come with me, but that is not that easy
[22:45:40] <Imperator_> they don't want to make holiday in the states
[22:45:53] <Imperator_> :-(
[22:47:08] <paul_c> ´pends on the sort of things that you like to do on holiday
[22:47:27] <paul_c> There are some good caves in Virginia....
[22:48:46] <Imperator_> I like mountains and caves also
[22:49:04] <paul_c> The Blue Ridge Mountains are nice that time of year if you enjoy hills & walking
[22:50:34] <Imperator_> have no exact plan at the moment
[22:50:43] <paul_c> same here...
[22:51:28] <Imperator_> like i said the problem is that at the moment i have nobody to travel with me, and traveling alone makes no fun
[22:51:43] <paul_c> I know the feeling.
[22:52:04] <paul_c> Catch me next weekend - I will be back home then.
[22:52:12] <Imperator_> ok
[22:52:30] <Imperator_> btw. my servos are runing fine on EMC2 now
[22:52:35] <paul_c> We can talk about the options open to us.
[22:52:51] <paul_c> * paul_c has a new card to write a driver for...
[22:53:15] <paul_c> Sensoray 526
[22:54:25] <Imperator_> ok Paul
[22:54:49] <Imperator_> i have some sampels of the LS7266 now for makeing also a new card :-)
[22:55:56] <paul_c> LS7266 is fairly easy to read - dro.c has all the code needed.
[22:56:55] <Imperator_> it is the LS7166 I think but they are the same I hope. the LS7266 has two counters the LS7166 ony one
[22:57:30] <paul_c> yes - Same core for both chips, so the registers are the same.
[22:57:49] <Imperator_> whow it is realy the LS7266
[22:58:53] <Imperator_> they have a 8bit bus. I want to conect both parallel to the 16bit isa bus
[22:59:02] <Imperator_> I hope that runs ????
[22:59:21] <Imperator_> and makes sense ?? It is faster I think
[23:01:56] <paul_c> chip one on the lower 8 bits, chip 2 on the upper
[23:02:06] <Imperator_> jep
[23:02:07] <paul_c> the STG card does that..
[23:02:20] <Imperator_> realy
[23:02:30] <Imperator_> so that works
[23:02:36] <paul_c> yup
[23:03:06] <Imperator_> then i have to wire that on my ISA prototype board the next days
[23:03:29] <Imperator_> but removing the old circuit is not that easy
[23:04:45] <Imperator_> I want to use the DAC7744 from Analog. If i make no mistakes on the analog design I will get a very precise 16bit output
[23:05:14] <Imperator_> it is a 4x16 bit DAC
[23:05:42] <Imperator_> normal price 50EUR
[23:06:03] <Imperator_> I got 12 peaces for 6$ each on ebay USA
[23:06:16] <Imperator_> enough for testing :-)
[23:08:50] <paul_c> thatś a good price - I have a few that I bought from Dan Mauch a while back...
[23:11:15] <paul_c> I´ve got to go - Driving down to San Francisco this evening..
[23:11:42] <paul_c> Won´t be back online till next weekend.
[23:13:30] <Imperator_> havent noticed that pal is in the USA
[23:24:06] <jmkasunich> OK, a preliminary version of the Fest webpage has been posted
[23:24:08] <jmkasunich> http://www.linuxcnc.org/EMC_news_history/Programmers-Fest2005.html
[23:30:29] <SteveStallings> Great, I will plan on attending. May have to bum a ride from NAMES in order to make it Monday. My travel companion will be overnighting in Columbus on Sunday night.
[23:32:18] <jmkasunich> if you can get to cleveland, you can ride with me
[23:33:23] <SteveStallings> OK, may also have others from my area there. Thalx is probably going.
[23:35:01] <jmkasunich> I'm probably gonna drive down Sunday night... start at 5pm or so, get there around 10-11
[23:35:27] <SteveStallings> k
[23:36:33] <Imperator_> ok late here
[23:36:40] <Imperator_> cu