#emc | Logs for 2008-02-05

[00:17:42] <SWPadnos> there's no way to mill the inner layers unless you can laminate multiple double-sided boards together (like the PCB shops do)
[00:19:06] <skunkworks> and somehow connect the pads
[00:19:34] <SWPadnos> sure - being able to plate through is a big help there ;)
[00:21:43] <maddash> jepler: http://emergent.unpy.net/01195490306 <-- why did you add "g"?
[00:22:04] <SWPadnos> he didn't write it, that was an example of terrible code
[00:22:24] <SWPadnos> click the "I am not making this up" link at the bottom for the source
[00:23:04] <skunkworks> the g was added - just in case
[00:23:20] <maddash> rofl
[00:24:12] <maddash> so what else is wrong with that function?
[00:24:51] <SWPadnos> other than the fact that it doesn't include '0'-'9' ?
[00:25:22] <SWPadnos> and that it writes two chars where one was before?
[00:25:42] <maddash> the argument is a char*
[00:25:53] <SWPadnos> yes, and?
[00:25:54] <maddash> and 0-9 is the same in dec as in hex
[00:26:01] <SWPadnos> no it isn't
[00:26:14] <SWPadnos> the ASCII character '0' is 0x30
[00:26:31] <maddash> for x in [0:9]
[00:26:41] <maddash> 0x30 is not inside [0:9]
[00:26:51] <SWPadnos> no, for x in 0-9, the ASCII character "x" is 0x30+x
[00:27:34] <SWPadnos> if I have a string "12", that does not contain the bytes 1 and 2, it contains the bytes 49 and 50 (0x31 and 0x32)
[00:28:03] <maddash> I think it's a (poorly done) hack for single chars...
[00:28:23] <SWPadnos> except that it writes two chars back ;)
[00:28:23] <maddash> obviously, getasdec('66') will fail
[00:28:37] <maddash> hence the "hack" and the "poorly done"
[00:28:41] <SWPadnos> sure
[00:28:43] <skunkworks> boy - you would think there is already a function for that.. ;)
[00:28:49] <SWPadnos> though that would be a compiler error
[00:29:06] <SWPadnos> you'd need getAsDec("66") ;)
[00:29:48] <maddash> I was using pseudo-C, and i also love my shift key very much, hence single quotes
[00:29:55] <SWPadnos> heh
[00:30:10] <maddash> btw, how's the arm-board?
[00:30:21] <SWPadnos> haven't worked on it, I was in Las Vegas
[00:31:06] <maddash> cool. i'd ask how you liked it, but what happens in vegas...
[00:31:25] <eric_U> SWPadnos, you do a lot of circuit design?
[00:31:53] <SWPadnos> eric_U, not a ton - usually more software contracts come my way
[00:32:04] <SWPadnos> maddash, oh - I wasn't a naughty boy - no worries there :)
[00:33:13] <eric_U> whenever I do a non-trivial design, I end up making a library part for too many of the devices, am I doing something wrong?
[00:33:21] <SWPadnos> probably not
[00:33:40] <SWPadnos> whenever I consider doing a design, I see that I'll have to define half the parts
[00:33:55] <eric_U> I was afraid of that, is it a conspiracy with the board houses?
[00:34:01] <SWPadnos> so it seems normal (even with a very expensive program with a 78000+ part library)
[00:34:04] <SWPadnos> dunno
[00:34:48] <eric_U> it's not like the chip manufacturers don't have the libraries or something
[00:35:05] <SWPadnos> well, there are several formats, and they wouldn't want to support all of them
[00:35:14] <eric_U> true
[00:35:25] <SWPadnos> so it seems they support none of them - maybe so it doesn't seem like they're playing favorites
[00:35:30] <eric_U> cadence will convert from most formats though
[00:35:40] <SWPadnos> so will Altium
[00:35:54] <eric_U> they need a standard
[00:36:00] <SWPadnos> I think Altium is pushing for some standard actually
[00:36:22] <SWPadnos> it can already access Orcad databases and convert a lot of things from the likes of PADS, P-Cad, and others
[00:56:55] <Roguish> SWPadnos: how does one submit a feature request?
[00:57:09] <SWPadnos> I think the SourceForge tracker is the best way
[00:57:16] <eric_U> just keep whining about it
[00:57:25] <SWPadnos> and/or an email to the developer list
[00:57:50] <Roguish> ok.. i would like to add a semicolon as a comment line indicator. lots of controllers use it.
[00:58:10] <Roguish> and i hate modifying post processors.....
[00:58:18] <SWPadnos> that sounds reasonable. I don't think that semicolons are used in the language at the moment
[00:58:33] <Roguish> I will try through SF
[00:58:43] <SWPadnos> ok. you have the link?
[00:58:56] <Roguish> i'm sure i can find it.
[00:59:19] <SWPadnos> heh - yep, I'm sure it's soemwhere on linuxcnc.org
[00:59:29] <eric_U> Nothing would make me save for retirement more than seeing Ed McMahon in person and realizing that no matter what you do to your body, you might live into your mid-80s.
[01:00:07] <SWPadnos> I think George Burns is a good example of that
[01:00:11] <SWPadnos> or was anyway
[01:00:19] <eric_U> that cigar was never lit
[01:00:25] <SWPadnos> heh
[01:00:40] <eric_U> McMahon was obviously drunk, every night
[01:00:52] <SWPadnos> lived to be 100 though, even with the cigars, wine, and young women
[01:01:10] <eric_U> he was still working
[01:01:50] <SWPadnos> there was a far side with flying cars whizzing around, and a comedy club with the sign "Tonight, George Burns!" outside ;)
[01:02:19] <Jymmmm> Ug
[01:02:30] <eric_U> too bad Gary Larkin succumbed to his psychosis
[01:02:50] <eric_U> I mean retired
[01:03:23] <skunkworks> I was listening to old time radio - the radio game show with groucho marx(sp) there was a performer that was 75 years old.. I didn't think people lived that long back then ;)
[01:04:25] <Jymmmm> skunkworks: no processed foods bake then, just diseases
[01:04:31] <Jymmmm> s/bake/back/
[01:04:37] <skunkworks> I suppose
[01:04:45] <eric_U> people have been living to 100 for quite some time
[01:05:00] <eric_U> the median sucked though
[01:17:19] <Roguish> SWPadnos: put it in. also commented on the request for 'support for gantry axis'. hope it makes it through.
[01:17:35] <SWPadnos> heh
[01:17:35] <maddash> brb
[01:17:35] <skunkworks> gantry axis?
[01:18:19] <Roguish> check the 'feature requests' at sourceforge for emc.
[01:19:33] <Roguish> need 'gantry axis' and '5axiskins' in the same machine.
[01:21:16] <SWPadnos> needy, aren't you?
[01:21:20] <SWPadnos> :)
[01:22:45] <Roguish> ya, that's me.
[01:26:24] <eric_U> gantry axis is two motors on one axis?
[01:27:15] <skunkworks> yes
[01:27:25] <SWPadnos> roght - imagine a dual-X gantry plus B/C axes
[01:27:29] <SWPadnos> s/roght/right/
[01:27:34] <eric_U> does that count as 2 of the 5axis?
[01:28:02] <Roguish> no
[01:28:12] <SWPadnos> it shouldn't
[01:28:24] <Roguish> 3 translation axes with 4 motor, plus 2 rotational axes.
[01:28:25] <skunkworks> you can have up to 9 in axis. it would take up 2 of those though
[01:28:35] <eric_U> the only problem with dual axis drive is the homing, correct?
[01:28:39] <SWPadnos> so you have XXYZAB or whatever, but at the moment I think you need gantrykins for the XXYZ part and 5axis for the BC
[01:28:53] <toastydeath> eric_u: dual axis drive needs to be compensated
[01:28:54] <SWPadnos> and you can't load two kins simultaneously, so - oops
[01:28:57] <toastydeath> for errors in distance
[01:29:19] <eric_U> what BC?
[01:29:30] <toastydeath> rotab in Y and Z
[01:29:42] <eric_U> think he meant ab
[01:29:54] <SWPadnos> AB, BC, AC - whatever
[01:29:58] <SWPadnos> two rotaries
[01:29:59] <toastydeath> BC is correct
[01:29:59] <Roguish> correct
[01:30:02] <toastydeath> usually
[01:30:03] <eric_U> whatever
[01:30:24] <eric_U> A is rotation in X?
[01:30:28] <toastydeath> correct
[01:31:01] <eric_U> I'm too stupid to think how the xyzab even works
[01:31:24] <eric_U> is there a "machine tool kinematics for dummies" ?
[01:31:27] <SWPadnos> that's OK, there are videos :)
[01:31:34] <toastydeath> XYZ - ABC
[01:31:54] <toastydeath> linear X, rotary A; linear Y, rotary B; linear Z, rotary C.
[01:32:09] <toastydeath> and uvw are parallel axes to XYZ.
[01:32:17] <eric_U> I don't think that was in my kinematics book covered machine tools
[01:33:06] <toastydeath> probably wasn't
[01:35:10] <SWPadnos> http://www.youtube.com/watch?v=0jQy-zc5P6w&feature=related
[01:36:16] <eric_U> I need to see the g-code for that
[01:36:44] <Jymmmm> Baby SunConure... http://www.youtube.com/watch?v=RR2T0hhK2Y0
[01:38:31] <eric_U> that 5 axis machining is impressive and all, but can it whistle dixie while it's doing that?
[01:50:20] <maddash> SWPadnos: do you know of any arm-/avr-based counterparts to the rabbitcore boards?
[01:51:08] <SWPadnos> there are tons of little boards, but none that I know of that are meant to plug into something like the Rabbit does
[01:51:49] <maddash> actually, I'm looking for one of the "little boards" kind
[01:52:24] <SWPadnos> there are a bunch of hobby boards
[01:52:26] <maddash> the rabbitcores are great because I have to do minimal soldering work, but the rabbit architecture itself is horrible
[01:52:46] <SWPadnos> do you have size limitations?
[01:52:51] <maddash> size?
[01:53:01] <maddash> physical size, you mean?
[01:53:01] <SWPadnos> yes, how large a board can you use
[01:53:04] <SWPadnos> yes
[01:53:35] <SWPadnos> I think avrfreaks has a hardware section, but I'm not sure how easy it is to use
[01:53:47] <maddash> i guess anything roughly the size of the rabbits
[01:54:00] <maddash> definitely not as large as the avr stk500 board
[01:54:02] <SWPadnos> those are very small
[01:54:07] <SWPadnos> ok
[01:54:18] <maddash> is that www.avrfreaks.net?
[01:54:22] <SWPadnos> the STK500 was what I was thinking of suggesting, since it has the programmer and a reasonable power supply built in
[01:54:32] <SWPadnos> yes
[01:56:23] <maddash> the avrfreaks "Devices" section features only the MCU by itself, unless I'm missing something
[01:57:16] <SWPadnos> yeah, it may be the tools section you want
[01:57:26] <SWPadnos> but looking through that, it's not all that great a list
[01:59:17] <maddash> all of the "Development board" category seems quite large and purposed...
[02:01:35] <maddash> i know nothing about pcb design, but I so desperately want to rid myself of the rabbit
[02:02:18] <GNieport1> Hi guys. Is it possible to "home" an axis using a limit switch if the switch acts as both max and min limit? Right now the machine seems to be ignoring the switch and jogging to the hard stop. I did verify the pin axis.0.pos-lim-sw-in responds correctly.
[02:02:45] <maddash> GNieport1: yes
[02:03:03] <maddash> GNieport1: 'HOME_IGNORE_LIMITS_'
[02:03:27] <GNieport1> I should say, EMC registers the limit on the DRO but does not seem to do anything during the home routine. Yes, I did have to set ignore, or else it faulted out.
[02:08:37] <SWPadnos> did you connect the limit signal to the axis home input?
[02:12:10] <GNieport1> mayhaps this could be the problem "#linksp Xhome => axis.0.home-sw-in
[02:12:10] <GNieport1> "
[02:12:19] <SWPadnos> yep, that could be it :)
[02:12:46] <GNieport1> hey thanks SPW, we were in parallel on that one :)
[02:12:54] <SWPadnos> oh no ;)
[02:13:01] <SWPadnos> err, I mean, great minds and all ...
[02:13:07] <GNieport1> LOL
[02:13:30] <maddash> rofl
[02:15:38] <GNieport1> anyone know if the m5i20 encoder-index is working correctly? I seem to remember some time ago the driver didn't work as expected...
[02:15:51] <SWPadnos> I believe it's working now
[02:15:58] <GNieport1> sweet! :-D
[02:16:15] <SWPadnos> I'm not sure if there's still an index pin and an index-enable pin though
[02:16:20] <SWPadnos> the one to use is index-enable
[02:16:25] <SWPadnos> (the RW pin)
[02:17:02] <GNieport1> something like: homing_true => index-enable to arm the index pin?
[02:17:30] <GNieport1> It is masked iir
[02:17:47] <SWPadnos> no, more like - um - axis.whatever.<something with index in it> <-> index-enable
[02:18:30] <GNieport1> oh, I'll have to read up on bidirectional HAL pins
[02:19:38] <SWPadnos> they connect the same way
[02:20:16] <SWPadnos> I don't remember the exact rules, but it's pretty similar to other pins (one or more RW pins, as many R as you want, I think no W if there's an RW connected, etc.)
[02:24:26] <GNieport1> okay, it is working without index, now to fix the index pulse
[02:24:44] <GNieport1> first, dinner.
[03:25:55] <toastydeath> ff
[07:24:46] <micges> hi all
[07:26:07] <toastydeath> o rly
[07:26:22] <micges> logger_emc: bookmark
[07:26:22] <micges> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-02-05.txt
[07:27:47] <micges> hehe
[07:28:25] <micges> toastydeath: what time is it there ?
[07:32:52] <fenn> 2:30 am
[07:35:30] <micges> here 8:30 am
[07:35:43] <micges> begin o boring work..
[07:35:45] <toastydeath> 2:30
[07:36:48] <toastydeath> i begin fascinating work at 9!
[07:38:09] <micges> fascinating ?
[07:40:03] <fenn> only if you're toasty
[07:40:40] <fenn> it's so nice outside: 18C and blustery winds
[07:41:14] <toastydeath> lol
[07:41:31] <toastydeath> tommorow sometime i have a ton of bar stock to saw up then there's some copper to be machined
[07:41:52] <fenn> fascinating
[07:42:26] <jeetendra_g10> howdy!!
[07:42:31] <fenn> heloooo!!!
[07:42:35] <toastydeath> hai
[07:43:06] <jeetendra_g10> any "easy" way to remake a custom emc interface?
[07:43:23] <fenn> pyvcp
[07:43:35] <jeetendra_g10> ok...
[07:43:46] <jeetendra_g10> where is the dev forum?
[07:43:59] <fenn> this is where the devs hang out
[07:44:04] <jeetendra_g10> situation description:
[07:44:20] <jeetendra_g10> island wide client server for emc..
[07:44:29] <jeetendra_g10> no file/open/close
[07:44:33] <jeetendra_g10> no g-code
[07:44:56] <fenn> eh, first of all, what are you doing?
[07:44:56] <jeetendra_g10> re-engineering artwork viewing,purchase,download
[07:45:12] <jeetendra_g10> basically i got a grant from my min of industry to make emc
[07:45:19] <jeetendra_g10> accessible to all layman cncists
[07:45:32] <jeetendra_g10> i make the cnc routers for 3 applications:
[07:45:43] <jeetendra_g10> jewellwry,furniture,handicraft,metal work
[07:45:51] <fenn> so you want to put a skin on emc's GUI and sell it?
[07:46:10] <jeetendra_g10> i need a platform where users browse buy download encrypted artwork.
[07:46:23] <fenn> i dont see what that has to do with emc
[07:46:46] <jeetendra_g10> no sell. emc will be supplied free..but the user should not bother about g-code.
[07:46:54] <jeetendra_g10> more like printing from adobe photoshop
[07:47:23] <jeetendra_g10> they download artwork,pay for it,the router cuts it.
[07:47:40] <fenn> ok, and opening a file is too hard?
[07:47:42] <jeetendra_g10> artists have another version where they upload artwork
[07:48:05] <jeetendra_g10> its more that they should not be making unauthrised cuts,unless they paid for it
[07:48:23] <fenn> then dont let them download the .ngc file if they havent paid for it
[07:48:26] <jeetendra_g10> the files should not be interchangeable between machininsts
[07:48:32] <jeetendra_g10> exactly..
[07:48:48] <fenn> are you trying to make it impossible to cut the same file more than once?
[07:48:48] <jeetendra_g10> a bit like wmv porn vids...you have to pay for it to view it :)
[07:49:12] <jeetendra_g10> no...more precisely: 3 purchase plans should be available:
[07:49:16] <jeetendra_g10> single use,
[07:49:21] <jeetendra_g10> exclusive use
[07:49:30] <jeetendra_g10> multiple use non exclusive
[07:49:34] <jeetendra_g10> anf free for all...
[07:49:43] <fenn> well.. good luck
[07:50:01] <fenn> you can see how well this worked for the music industry with billions of dollars of research
[07:50:12] <fenn> (total and complete failure)
[07:50:22] <jeetendra_g10> i got the grant already...on condition that athe artists have a protected artwork
[07:50:39] <jeetendra_g10> the issue is to make sure the artists can become professional designers.
[07:51:01] <jeetendra_g10> the only way is for a licensing their artwork
[07:51:14] <fenn> emc is free software, if emc is able to read your file, anyone can modify it to copy the file
[07:51:33] <jeetendra_g10> well...its just to satisfy ministry bean counters...
[07:51:45] <fenn> you can say that's illegal to do, and that may be satisfactory
[07:51:46] <jeetendra_g10> those who really want it ...should be able to do it
[07:52:19] <jeetendra_g10> the core of the cnc business plan is that everyone can make a living out of it
[07:52:24] <fenn> but you will have to make a new program to do the decryption (i.e. it cant be part of emc because it would conflict with the GPL)
[07:52:37] <jeetendra_g10> i make the routers..the artists make the vector/polygon files
[07:53:24] <fenn> this is more of a cultural issue than a technical one
[07:54:00] <jeetendra_g10> correct...its more that the cnc island wide plan is to increase and encourage design creativity.
[07:54:24] <jeetendra_g10> te only way to do that is that the rich contractors pay the poor artists some money
[07:54:30] <fenn> i mean trying to prevent unauthorized use
[07:54:48] <fenn> can only be solved by creating a culture that discourages it
[07:55:30] <jeetendra_g10> its just that
[07:55:59] <toastydeath> uh, doesn't it have to go to g-code at some point for EMC to make it
[07:56:02] <jeetendra_g10> contractors are very rich
[07:56:14] <jeetendra_g10> artista are very poor
[07:56:17] <toastydeath> contractors are rich because they don't pay poor artists for crappy art
[07:56:23] <fenn> toastydeath: not necessarily (g-code)
[07:56:27] <jeetendra_g10> lol
[07:56:36] <toastydeath> they pay artists whose art has a return value
[07:56:54] <fenn> toastydeath: but it will end up passing through some GPL'ed code eventually
[07:57:09] <toastydeath> they can always cash out with 'real' art; if you're making crap on a CNC, that's not worth the material its made with
[07:57:18] <toastydeath> it's like trying to resell a bumper sticker
[07:57:20] <micges> alex_joni: around ?
[07:57:22] <jeetendra_g10> no...they make local residences for people like beckham and saudi princes.
[07:57:31] <alex_jon1> micges: yeah
[07:57:40] <jeetendra_g10> the least they can do is pay the artist 20* Rs 2000
[07:57:45] <jeetendra_g10> one dollar=Rs 30
[07:57:52] <toastydeath> why would they want what the artist has
[07:57:53] <jeetendra_g10> if they cut it 20 times
[07:58:05] <fenn> i can see how artists would get screwed, if it is a buyer's market
[07:58:18] <jeetendra_g10> to somplify the issue:
[07:58:49] <jeetendra_g10> once again: the contractors make RS illions of profit on a bungalow for a prince on our resort island.
[07:59:04] <jeetendra_g10> someone made the artwork for the wood cnc carved panels.
[07:59:08] <jeetendra_g10> he should be paid for it
[07:59:26] <toastydeath> so you're asking the contractor to pay more money for things he already buys?
[07:59:41] <jeetendra_g10> no: the contractor buys cnc routers from me
[07:59:54] <toastydeath> i'm really lost as to why the contractor would WANT what you are selling
[08:00:13] <jeetendra_g10> the contractor in addition gets access to a hundred or so didgital artista who bid to get his decoration contract artwork
[08:00:45] <toastydeath> i really just don't see how every contractor is going to go to plain wood paneling art
[08:01:04] <jeetendra_g10> from there the contractor only has to specify the kind of artwork he needs,from there the artists then compete aong themselves to give the best artwork they can
[08:01:29] <toastydeath> also you're aware contractors hire design firms to do that crap, right?
[08:01:35] <toastydeath> they don't actually bother doing the art themselves?
[08:01:51] <jeetendra_g10> he then just browses the online catalogue...chooses the designs he wants...pays the artists...download and cuts the artwork on his routers
[08:01:52] <fenn> i think that's what he just said
[08:02:10] <toastydeath> he's talking as though the contractor is doing all the layout and design
[08:02:14] <jeetendra_g10> big contractors don't bother with all this.
[08:02:23] <jeetendra_g10> up to now they ordered everything from china.
[08:02:32] <toastydeath> and how are you cheaper than mass produced crap from china?
[08:02:41] <fenn> hah - saudi prince orders bungalow from china
[08:02:52] <jeetendra_g10> the last estimate was a house for beckham was sold for Rs 60 million
[08:02:52] <fenn> what's the world coming to
[08:02:57] <jeetendra_g10> 1 dollar=Rs 30
[08:03:28] <toastydeath> right you keep laying these dollar value amounts as if they're there for the taking
[08:03:37] <jeetendra_g10> the contractor complained to the ministry that cnc artwork ar any kind of artwork could not be provided locally becasue of the lack of cnc machines..
[08:03:51] <jeetendra_g10> and people trained to provide designs usable on cnc wood carving machines.
[08:04:10] <jeetendra_g10> they bulk imported cnc woodwork from china and malaysia.
[08:04:21] <jeetendra_g10> to build bungalows in mauritius.
[08:04:27] <toastydeath> right, which i imagine is way cheaper than trying to build it locally
[08:04:49] <jeetendra_g10> itsa fair that mauritian artisans and artists get a life out of their own country,and not watch china and malaysia take all the juicy contracts
[08:05:15] <toastydeath> okay just because you don't think it's fair doesn't mean it isn't economical and in the homeowner's and contractor's best interest
[08:05:30] <toastydeath> if nobody else's
[08:05:32] <jeetendra_g10> perhaps,,,but artisans of my country starve,while i can build routers,,and emc is available..
[08:05:43] <toastydeath> right, artists have been starving since cave paintings
[08:05:49] <jeetendra_g10> and the contractors are willing to aid local artisans..
[08:06:04] <eric_U> contractor probably doesn't even want to cnc things, he wants end product
[08:06:14] <jeetendra_g10> exactly eric
[08:06:29] <jeetendra_g10> he doesn't mind a little more expense...but being socially responsible
[08:06:38] <jeetendra_g10> towards those who host his bungalow
[08:06:42] <eric_U> but then the local cnc producers can pay the artists what they want
[08:07:00] <toastydeath> i just don't see how any artists are going to live off this
[08:07:12] <toastydeath> the market for art is much smaller than everyone who wants to stay home and paint all day
[08:07:13] <eric_U> they could export
[08:07:16] <toastydeath> smoking pot
[08:07:19] <jeetendra_g10> at most..we estimate only a sale of 25 cnc machines over the lifetime of my company
[08:07:39] <fenn> eh toastydeath i think your extrapolation of american culture probably doesn't apply to mauritius
[08:07:58] <jeetendra_g10> erm...you should try googlin "vacation in mauritius"
[08:08:16] <toastydeath> i think that "there is more art than demand for art" is a valid assumption, cross-culturally
[08:08:27] <jeetendra_g10> from there you would see that a good 40% of the population make living out of handicraft for tourists
[08:08:44] <toastydeath> okay, so why aren't you selling your CNC stuff to THAT market
[08:09:05] <jeetendra_g10> precisley: (1)they are poor people,
[08:09:13] <jeetendra_g10> (2)technically illiterate
[08:09:24] <toastydeath> start a company and employ some.
[08:09:28] <jeetendra_g10> (3)no way to them learn g-code
[08:09:30] <toastydeath> teach them how to hit go.
[08:09:44] <micges> alex_joni: pm to you
[08:10:09] <jeetendra_g10> so what min of ind proposed me after seeing firts emc based oruter..was to build a dozen or so routers.
[08:10:12] <fenn> jeetendra_g10: it's very difficult to make artistic carvings with plain g-code. usually this is done with a program like 'ArtCAM'
[08:10:21] <jeetendra_g10> min buys them...makes them available to artisans
[08:10:43] <jeetendra_g10> exactly: i have already done my own freeware artcam version
[08:10:59] <jeetendra_g10> waht lacks is a large catalogue
[08:11:05] <jeetendra_g10> of cnc artwork
[08:11:25] <jeetendra_g10> so who would make cnc design?
[08:11:30] <jeetendra_g10> artsts
[08:11:32] <jeetendra_g10> why?
[08:11:35] <jeetendra_g10> for money!!
[08:11:44] <jeetendra_g10> how to guarantee they will get paid?
[08:11:53] <jeetendra_g10> a secure mysql server
[08:11:55] <fenn> dont give away the art until you get paid
[08:12:04] <eric_U> when you solve this problem, you can solve the problem that authors have with their books
[08:12:05] <jeetendra_g10> where they upload artwork,and get paid
[08:12:26] <jeetendra_g10> when soemeone buys the artwork online.
[08:12:46] <fenn> jeetendra_g10: why would the contractors ever get a hold of the file? if they are just buying a piece of carved wood?
[08:13:17] <toastydeath> this is why i am so damn confused, why do you insist on being a hippie when there are literally five or six different manufacturing opertunities here
[08:13:23] <jeetendra_g10> some contractors prefer to buy routers from me...i am 28 times cheaper than techno isel wood routers
[08:13:23] <toastydeath> that will accomplish the same end goal
[08:13:49] <jeetendra_g10> but the contractor would still need artwork on a continuous basis
[08:14:20] <jeetendra_g10> local artists provide the artwork in theme to the country lifestyle
[08:14:43] <jeetendra_g10> so when a router is bought from me...
[08:14:50] <alex_jon1> micges: really?
[08:14:58] <fenn> the artists cant afford a cnc router, right? so that's why the contractors have to 'print' the design on their routers
[08:15:12] <jeetendra_g10> they also get a design studio thrown in...as artisans are working and uploading artwork continuously.
[08:15:16] <alex_jon1> micges: didn't get anything, maybe you are not registered
[08:15:29] <toastydeath> where are they working on computer designs?
[08:15:31] <jeetendra_g10> the model is identical to 3dexport.com
[08:15:44] <jeetendra_g10> they work from home,we have good adsl locally
[08:15:48] <kimron_> Hi, I am looing for a CNC table plan (want to build my own )
[08:16:11] <jeetendra_g10> the business model is identical to 3dexport.com
[08:16:29] <micges> yes im not :)
[08:16:42] <micges> new bad issues
[08:16:50] <micges> when Im run program and go to step mode
[08:16:57] <fenn> alex_jon1: you can set mode e+6 to allow unregistered users to pm you
[08:17:13] <alex_jon1> fenn: I know.. but why should I :P
[08:17:23] <micges> and step to M66 code program hang and no move or nothing is happening
[08:17:24] <alex_jon1> micges: run program and pause?
[08:17:27] <fenn> shrug. i've never gotten any spam
[08:17:45] <jeetendra_g10> ok...so...
[08:17:49] <alex_jon1> or simply start with step?
[08:17:59] <alex_jon1> alex_jon1 is now known as alex_joni
[08:18:05] <micges> run and pause
[08:18:35] <alex_joni> fenn: I do have e6.. but only for my registered nick :D
[08:18:36] <micges> step step step
[08:18:59] <alex_joni> micges: can you submit abugreport?
[08:19:18] <micges> ok I try
[08:19:24] <alex_joni> include a program which shows the problem
[08:19:41] <alex_joni> micges: I'm not trying to be difficult, but I am busy atm and can't look at it
[08:19:42] <fenn> jeetendra_g10: you can make it hard to 'intercept' the g-code but not impossible
[08:19:55] <jeetendra_g10> that will do...
[08:19:58] <alex_joni> micges: you can assigne the bug report to me
[08:20:54] <fenn> jeetendra_g10: just pipe the output of your decryption program to jdi.py
[08:20:59] <micges> understand
[08:21:25] <fenn> i think most contractors are not linux savvy enough to figure that out
[08:21:26] <jeetendra_g10> what is jdi?
[08:21:38] <fenn> its a script that opens and runs a g-code file
[08:21:39] <jeetendra_g10> exactly: they won't bother abt all that..
[08:21:43] <fenn> 'just do it'
[08:21:48] <jeetendra_g10> ok will looka at jdi
[08:22:16] <fenn> http://axis.unpythonic.net/01167419757
[08:22:22] <jeetendra_g10> also i want to make a direct interface between blender and emc
[08:22:28] <jeetendra_g10> blender3d
[08:22:58] <fenn> me too
[08:23:03] <jeetendra_g10> draw in blender...save in blender,,,and cut
[08:23:05] <fenn> call me when it's done
[08:23:19] <jeetendra_g10> so thats why you have to help me here...where to start?
[08:23:23] <jeetendra_g10> jdi.py
[08:23:28] <jeetendra_g10> pyvpc
[08:23:43] <jeetendra_g10> is that it?
[08:23:46] <fenn> just get something working first
[08:23:47] <jeetendra_g10> :0
[08:24:10] <jeetendra_g10> i already made a cm program on matlab.
[08:24:20] <jeetendra_g10> i want to port it to python or replace it with blender
[08:24:52] <jeetendra_g10> right now i have to switch winxp and ubuntu between designs
[08:25:13] <jeetendra_g10> i want to make the emc interface what lazycam is to mach
[08:25:42] <fenn> btw pyvcp is a little ugly
[08:26:15] <fenn> there is no 3d preview
[08:26:15] <jeetendra_g10> its not hard...its working already in fragmented mode: design on winxp(matlab)...savae as textgcode...reboot ubuntu...cut on emc
[08:26:38] <jeetendra_g10> i want a single emc platform...including artwork download
[08:26:42] <fenn> it's not intended to be a GUI designer, it's more for things like displaying sensor values and test buttons
[08:26:45] <jeetendra_g10> from mysql
[08:27:09] <jeetendra_g10> the other option is to buy mach in bulk..
[08:27:17] <jeetendra_g10> as mach gives a vbscript
[08:27:23] <jeetendra_g10> and interface apis
[08:27:27] <fenn> jeetendra_g10: do you think your code could run on a free clone of matlab? like octave or scilab
[08:27:55] <jeetendra_g10> dunno...scilab fares poorly in terms of equivalence
[08:28:02] <jeetendra_g10> there is no imread...
[08:28:07] <jeetendra_g10> function
[08:28:13] <fenn> i've never used matlab so i dont know really what it does better than any other programming language
[08:28:14] <jeetendra_g10> no neural networks
[08:28:36] <fenn> um, you use neural networks in a cam app?
[08:28:49] <jeetendra_g10> i don't mean to boast matlab ...but i made a cam program in 30/40 lines dude.
[08:28:56] <jeetendra_g10> its doesn't get more efficient than this
[08:29:02] <jeetendra_g10> yes i use NNs..
[08:29:14] <jeetendra_g10> as a clever form of raster to vector separator
[08:29:32] <fenn> this is interesting
[08:29:34] <jeetendra_g10> convertor
[08:29:53] <jeetendra_g10> and thats by knowing only a little bit of NN and genetic algorithm
[08:30:00] <jeetendra_g10> think what a real pro could do...
[08:30:26] <jeetendra_g10> my advice : get a demo copy of matlab..
[08:30:35] <jeetendra_g10> stunning this beast
[08:31:38] <jeetendra_g10> so is there a way to program my own interface to emc with python?
[08:31:52] <jeetendra_g10> inlcudin file download?
[08:31:58] <eric_U> of course
[08:32:09] <eric_U> multiple interfaces available already
[08:32:20] <jeetendra_g10> non...my appllication is weird..
[08:32:25] <eric_U> full source for each
[08:32:37] <jeetendra_g10> i want to rengineer the whole part just before emctask planner
[08:32:47] <eric_U> you have all the code available
[08:34:14] <jeetendra_g10> yes///but eher is the documentation of the api?
[08:34:21] <eric_U> yes
[08:34:23] <jeetendra_g10> is there and api?
[08:34:31] <jeetendra_g10> its documentation?
[08:34:53] <fenn> the API between GUI and emc proper is the NML message layer
[08:35:03] <jeetendra_g10> ok noted
[08:35:37] <fenn> i think the GUI just feeds emc a g-code filename though
[08:36:08] <jeetendra_g10> that would do:
[08:36:21] <jeetendra_g10> the gui downloads a file,decrypts it,fedds it to emc
[08:36:42] <jeetendra_g10> also,i get to interface it to other xmla based polygons
[08:36:46] <jeetendra_g10> xml
[08:37:02] <jeetendra_g10> <polygon/>
[08:37:07] <fenn> for displaying the preview?
[08:37:19] <jeetendra_g10> <x,y,z>
[08:37:22] <jeetendra_g10> ,1,2,3
[08:37:25] <jeetendra_g10> 456
[08:37:31] <jeetendra_g10> </x,yz>
[08:37:38] <jeetendra_g10> <polygon/>
[08:37:53] <jeetendra_g10> not for preview...a text based xml form of dxf
[08:37:53] <fenn> hmm.
[08:38:14] <jeetendra_g10> rather 3d surface
[08:38:22] <jeetendra_g10> in fact its stl in xml
[08:38:32] <jeetendra_g10> there ...thats accurate description
[08:38:57] <jeetendra_g10> my countries own 3d format..
[08:39:00] <fenn> what would emc use that information for?
[08:39:00] <jeetendra_g10> simple..
[08:39:26] <fenn> it's not a toolpath
[08:39:33] <jeetendra_g10> basically an artist makes an artwork ..which is a collection of polygons
[08:39:38] <jeetendra_g10> upload it to server
[08:39:47] <jeetendra_g10> as xml
[08:39:54] <jeetendra_g10> server encrypts it
[08:40:22] <jeetendra_g10> when purchase occurs,gui download the xml..decrypts it...converts it to g-code...the end user cuts
[08:40:26] <jeetendra_g10> no visible g-code
[08:41:02] <jeetendra_g10> no file open/close..
[08:41:07] <jeetendra_g10> menu
[08:41:27] <jeetendra_g10> the download process puts the decrypted g-code in atemp folder
[08:41:40] <jeetendra_g10> for emc to load automatically
[08:41:58] <jeetendra_g10> when cutting is done,the g-code is deleted...
[08:42:10] <jeetendra_g10> pending another purchase
[08:43:12] <jeetendra_g10> also,its scale invariant..
[08:43:29] <jeetendra_g10> all polygon files are normalised to 1 unit of max size
[08:43:52] <jeetendra_g10> the user's emc scales it to the the local cnc machine
[08:44:11] <jeetendra_g10> same polygons used for jewelly,cutting,routing etc..
[08:44:54] <fenn> there's more to CAM than just making it the right size
[08:44:59] <jeetendra_g10> there is no documentation as to where to start customising emc
[08:45:29] <jeetendra_g10> i know...but its a good start..
[08:45:34] <fenn> have you looked at jdi.py or axis.py?
[08:46:04] <jeetendra_g10> ok ...looking now
[08:49:52] <jeetendra_g10> there is a file...the mailing list archive is 404
[08:50:16] <fenn> this one? http://axis.unpy.net/files/01167419757/jdi.py
[08:52:20] <fenn> now i cant remember the script that provided a python API instead of g-code
[08:52:54] <jeetendra_g10> i think i can do with that:
[08:53:06] <jeetendra_g10> now how do i get emc to load that script?
[08:53:29] <fenn> ./jdi.py ?
[08:53:55] <fenn> oh it needs some arguments..
[08:54:48] <jeetendra_g10> so basically if i put it as a user interface...like mini etc...it works just like an interface
[08:55:28] <fenn> env DISPLAY=.../jdi.py emc .../foo.ini .../foo.ngc
[08:55:52] <jeetendra_g10> ok...i start to understand
[08:56:08] <jeetendra_g10> c.program_open(filename) is the part i have to put the g-code file?
[08:56:08] <fenn> oops that's wrong
[08:56:14] <fenn> DISPLAY is in the .ini file
[08:57:15] <fenn> filename comes from the argument list
[08:57:28] <fenn> or you could hard-code it if you wanted
[08:59:07] <fenn> jdi.py is a quick way to get started, but modifying axis might be better in the long run
[09:00:00] <jeetendra_g10> yes...thats the problrm...there is no documentation where to start tinkering...
[09:00:10] <jeetendra_g10> the source code is there...
[09:00:17] <jeetendra_g10> the docs are not :(
[09:00:28] <alex_joni> use the source luke
[09:01:06] <fenn> i'm not sure what docs you could write for the axis source
[09:02:00] <jeetendra_g10> catch you guys later...i have a theird p4 sleeping in a corner...i put emc...then i see waht i can do with jdi/axis
[09:03:38] <fenn> it could use some comments i guess
[09:07:06] <fenn> i think he should just make a filter
[09:07:14] <fenn> and then use that as the splash program
[09:25:04] <fenn> i wonder if i was thinking of emcmodule.cc
[09:27:31] <alex_joni> fenn: using a filter is the cleanest solution imo
[09:27:53] <alex_joni> (even if it's a filter which connects through a secure socket and grabs the program from somewhere, doesn't save it locally)
[09:27:53] <fenn> yep the rest of it was just my stupidity getting in the way
[09:28:38] <alex_joni> but otoh, this is just like security "features" they build into CD's and DVD's these days
[09:28:49] <alex_joni> (if someone really wants to overcome them, they will)
[09:30:00] <fenn> the 'emc' module could use some nice API reference documentation
[09:30:23] <fenn> or maybe i'm just spoiled
[09:31:07] <alex_joni> or maybe you can just write it :P
[09:31:12] <fenn> right
[09:31:31] <fenn> i was thinking something like the pygtk docs
[09:32:21] <fenn> though really just some docstrings would be sufficient
[09:33:46] <fenn> heh my cat ate a mole and now he's making funny noises
[10:34:08] <alex_joni> mole?
[10:35:48] <fenn> small furry animal, like a blind mouse that digs tunnels
[10:40:18] <alex_joni> ahh.. ok, didn't know that was the english name for it
[10:43:22] <fenn> http://fennetic.net/pub/camera/DCP_0795.JPG
[10:50:03] <alex_joni> bbl
[11:00:38] <archivist> one full cat?
[11:03:24] <archivist> I picked a mole up once, amazingly strong arms
[11:09:55] <fenn> did you have to do a cool-down stretch afterwards? you big burly man
[11:12:41] <alex_joni> archivist: yours or the moles?
[11:14:05] <archivist> hehe
[11:59:10] <archivist> * archivist starts cutting second production gear, still not finished the build!
[12:04:49] <BigJohnT> got pics?
[12:08:34] <archivist> www.archivist.info/cnc for upto yesterday
[12:12:20] <BigJohnT> your making gears?
[12:12:23] <archivist> yes
[12:13:08] <archivist> first was a ratchet, second is a 72 tooth .8 module cycloidal wheel
[12:14:20] <BigJohnT_> crap, lost my connection
[12:14:29] <BigJohnT_> how are you cutting the profile?
[12:14:46] <archivist> profiled cutters
[12:14:58] <BigJohnT_> gear on the rotary table?
[12:15:02] <archivist> yes
[12:15:10] <BigJohnT_> cool
[12:16:01] <BigJohnT_> profile cutter like a saw blade or end mill?
[12:17:39] <archivist> http://www.clock-works.clara.net/cata/wnpc.htm top left pic
[12:18:13] <BigJohnT_> Hey, my uncle must have made them LOL
[12:18:28] <BigJohnT_> even spelled the same as me
[12:18:35] <archivist> who is your uncle
[12:19:10] <BigJohnT_> uncle Thornton
[12:19:17] <BigJohnT_> just kidding same last name
[12:20:12] <BigJohnT_> kinda pricy cutters
[12:20:31] <archivist> the late Mr Thornton's company now further south in the UK under new ownership for many years
[12:21:01] <archivist> they are pricy but small market
[12:21:45] <BigJohnT_> my great grandfather was born int the UK somewhere
[12:22:35] <BigJohnT_> so you built this from scrounged bits and pieces?
[12:23:29] <archivist> yes
[12:23:38] <archivist> and a few new parts
[12:23:46] <BigJohnT_> super cool, I love doing that here
[12:24:09] <BigJohnT_> I'm building a plasma cutter from scratch at the moment
[12:26:51] <archivist> I will measure the tooth pitch on this gear, that will tell me how good the rotary table is/isnt
[12:27:11] <BigJohnT_> I know what you mean
[12:29:08] <BigJohnT_> is it a new rotary table?
[12:29:58] <archivist> yes
[12:30:27] <archivist> but often the worm/wheel in them are rubish
[13:14:36] <skunkworks_> logger_emc: bookmark
[13:14:36] <skunkworks_> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-02-05.txt
[14:42:32] <skunkworks_> http://www.cnczone.com/forums/showthread.php?t=49968&highlight=emc2
[14:54:04] <alex_joni> "The computer controlling the machine is an old Compaq which used to be my server, running for weeks on end with no problems. I am using the free program EMC2 (Enhanced Machine Controller 2) to control the machine. Installation was a breeze using their Ubuntu live CD."
[14:54:15] <alex_joni> * alex_joni is always pleased to read things like that
[14:54:27] <alex_joni> http://www.guitar-list.com/node/23253
[14:56:53] <skunkworks_> cool huh
[15:07:51] <tomp> if emc could be drip-fed, then emc might get each gcode line from an internet 'stream', and no one would 'have' the file
[15:10:05] <tomp> uh, that was regarding jettendra_g10
[15:10:17] <SWPadnos> which board?
[15:11:00] <skunkworks_> earlier on irc #emc
[15:11:21] <SWPadnos> oh - I guess I missed him
[15:12:30] <jepler> SWPadnos: basically, the guy seemed to want to add DRM to gcode
[15:12:41] <jepler> so that it was necessary to purchase the right to mill a design
[15:12:57] <SWPadnos> ah - I'll have to read back a bit
[15:13:13] <tomp> jeetendra
[15:19:42] <skunkworks_> seems like that would not play well with open source software.. :)
[15:20:07] <alex_joni> bbl
[15:20:23] <cradek> on the contrary, you're perfectly free to write DRM into GPL software (but the next guy is just as free to take it out)
[15:22:02] <SWPadnos> and you have to include the source ;)
[15:22:11] <archivist> and the gcode if machine specific needs mods before use
[15:22:54] <SWPadnos> I think the idea is to have all the machines use EMC, and about the only thing that would change (as far as G-code is concerned) is the work area
[15:23:49] <archivist> and scale of object produced
[15:24:13] <SWPadnos> sure, but that's actually doable with ugly equations everywhere
[15:38:11] <cradek> jensor: I think you made a mistake in your email about MOD - there are no negative numbers, and the right result is 3, not 1 as you state
[15:38:50] <cradek> jensor: I tested it and did get 3 as expected
[15:42:11] <cradek> although, 15 MOD 7 gives me 1, which is pretty hard to understand
[15:42:50] <SWPadnos> 7*2=14, 15-14=1
[15:43:20] <skunkworks_> that seems right
[15:43:42] <cradek> oh, duh
[15:43:43] <skunkworks_> I used mod in the spirograph o-word program and it seemed right
[15:43:50] <skunkworks_> *work as expected
[15:44:07] <cradek> for some reason I confused MOD with integer division
[15:44:15] <SWPadnos> heh - or AND :)
[15:44:29] <cradek> well I was thinking DIV
[15:44:41] <cradek> are you all glad I didn't jump in and fix it?
[15:44:42] <maddash> it says, "MOD"
[15:44:50] <SWPadnos> you were close, it's REMAINDER(N div mod)
[15:45:07] <SWPadnos> yes, very ;)
[15:45:07] <maddash> there is no "close" in mathematics...
[15:45:24] <SWPadnos> sure there is, if you're dealing with the mathematics of hand grenades or horseshoes
[15:45:47] <maddash> uh, no.
[15:45:50] <cradek> I seem to recall that MOD for negative numbers is not well agreed-upon, and in fact in C it's implementation-dependent
[15:46:07] <cradek> * cradek hopes someone has a better reference than "I seem to recall"
[15:46:11] <SWPadnos> yes, that's a tough one
[15:46:20] <fenn> the class of predicates wherein 'close' is a satifying criterion
[15:46:20] <maddash> huh? modular arithmetic is well-defined across the integers
[15:46:29] <SWPadnos> maddash, statistics is the mathematical science of "close" ;)
[15:46:51] <maddash> SWPadnos: but statistics in itself is precise
[15:46:51] <cradek> http://mathforum.org/library/drmath/view/52343.html
[15:47:10] <maddash> even "close" needs to be precisely defined
[15:47:13] <fenn> cradek: he's mod'ing two positive numbers, why would it return a negative?
[15:47:33] <cradek> no, I think he just made a mistake in his email (posted the wrong code)
[15:47:51] <fenn> why do you think that?
[15:48:06] <cradek> because otherwise his report makes no sense
[15:48:36] <SWPadnos> when was this email?
[15:48:44] <cradek> "The mod function does not work for the negative numbers 15 and 4 as shown in my example. excel gives an answer of -3" makes no sense
[15:48:48] <cradek> recently
[15:49:00] <SWPadnos> is this a bug report or to one of the lists?
[15:49:23] <jensor> Your right I did make a mistake in the email. I meant to enter negative numbers
[15:49:27] <maddash> cradek: this is mod 60, right? so isn't 40 = -20 mod 60?
[15:49:44] <cradek> maddash: did you read my url?
[15:49:49] <maddash> cradek: there is no "dis-agreed-upon"-ness
[15:50:02] <cradek> maddash: you're wrong
[15:50:37] <fenn> i sense disagreement
[15:50:40] <maddash> if it returns 20, then that is incorrect. -360 mod 60 is -20, not 20
[15:50:49] <maddash> -340*
[15:50:55] <skunkworks_> I don't sense a disagreement. ;)
[15:50:55] <cradek> I'm sure python and gcc disagree, as do (apparently) excel and lotus
[15:52:06] <fenn> you can divide and get the remainder - that's consistent right?
[15:52:52] <maddash> no, to correct my earlier statement,
[15:53:04] <maddash> -340 mod 60 is 20, 340 mod 60 is 40 = -20
[15:53:26] <maddash> the negative sign only givees an illusion of difference
[15:53:37] <SWPadnos> as they point out in that URL, DIV and MOD are related, but if DIV goes toward zero you get one answer, and if DIV goes toward -INF you get the other answer
[15:54:13] <cradek> I agree emc's answer is the "more correct" of the two MOD answers you might expect
[15:54:23] <maddash> cradek: as for your "disagreements," there are none. if the result of a " x mod y" is negative, then adding y will give the same results
[15:54:36] <jepler> For the "%" operator on integers, ISO C doesn't guarantee much when an input is negative
[15:54:39] <jepler> K&R2, p. 205 says: [paraphrased] If both operands are non-negative, then the
[15:54:42] <jepler> remainder is non-negative and smaller than the divisor; if not, it is guaranteed only
[15:54:45] <jepler> that abs(remainder) < abs(divisor).
[15:55:05] <maddash> [sigh] this is pathetic
[15:55:45] <jepler> however, I believe the sign of the fmod() result is defined for all signs of the inputs
[15:56:36] <skunkworks_> did I get so worked up about things like this when I was younger... * thinks to himself.
[15:57:08] <jepler> The fmod functions return the value x - ny, for some integer n such that, if y
[15:57:08] <jepler> is nonzero, the result has the same sign as x and magnitude less than the
[15:57:08] <jepler> magnitude of y.
[15:58:17] <jepler> (that's from ISO/IEC 8988:TC3, the ISO C standard -- the description in 'man fmod' is different; it describes that 'n is the quotient of x/y rounded towards zero to an integer'; I don't immediately see that those are identical values)
[16:00:13] <cradek> heh, ada has both :-)
[16:00:26] <SWPadnos> yeah, I'm just skipping over that part now ;)
[16:00:26] <jepler> >>> fmod(-15,4), (-15)%(4)
[16:00:26] <jepler> (-3.0, 1)
[16:00:32] <jepler> so does Python, apparently
[16:01:24] <cradek> I think python % disagrees with gcc % (and FORTRAN?)
[16:01:31] <jepler> I can see arguments for both rounding the quotient towards zero, or taking the floor of the quotient. I have a harder time seeing the justification for rs274ngc's "always positive".
[16:01:48] <jepler> cradek: But % isn't defined in ISO C, as I said above
[16:01:55] <cradek> right, I understand
[16:02:01] <jepler> gcc might give differing results on different architectures -- you should at least say "gcc/x86"
[16:02:11] <jepler> unless you found that gcc specifies it somewhere
[16:02:26] <cradek> 'a certain version of gcc on a certain machine that I used once'
[16:03:45] <cradek> case MODULO: /* always calculates a positive answer */
[16:03:45] <cradek> *left = fmod(*left, *right);
[16:03:45] <cradek> if (*left < 0.0) {
[16:03:45] <cradek> *left = (*left + fabs(*right));
[16:03:45] <cradek> }
[16:04:01] <cradek> it sure was on purpose
[16:04:05] <jepler> yep it sure was
[16:04:07] <jensor> where can I get a good explanation of the usage of if , else and other NGC functions?
[16:04:48] <jepler> jensor: depends what you mean by "good". If you mean "I've read the documentation and the wiki and I haven't found a good explanation" then the answer is "nowhere".
[16:04:53] <jepler> or maybe "in the source code"
[16:05:02] <cradek> jensor: http://linuxcnc.org/docs/devel/html/gcode_main.html#cha:O-Codes
[16:05:56] <jensor> Thanks,I'll take a look
[16:06:00] <skunkworks_> if all else fails - you could email lerman
[16:06:48] <jepler> I apologize -- my last comment was unnecessarily rude.
[16:08:40] <cradek> I see the info at the manual link I pasted is really pretty decent, but it is terse/dense
[16:10:27] <micges> logger_emc: bookmark
[16:10:27] <micges> Just this once .. here's the log: http://www.linuxcnc.org/irc/irc.freenode.net:6667/emc/2008-02-05.txt
[16:11:11] <skunkworks_> actually - I think I looked at the nist ngc pdf for a lot of the functions.. I think.
[16:13:45] <skunkworks_> page 85 in the pdf http://www.isd.mel.nist.gov/documents/kramer/RS274NGC_3.pdf
[16:14:11] <skunkworks_> gives a test program with results.
[16:14:15] <skunkworks_> c.2
[16:17:39] <cradek> as far as I can see, the ngc spec does NOT define how mod should work
[16:18:16] <skunkworks_> right - I ment what fuctions where available. atan,sqr,..
[17:24:36] <Jymmmmm> Hey folks, excuse me if I bounce in/out, trying out a new irc client.
[18:48:02] <skunkworks_> so if someone wanted a gui for python... what would one use?
[18:50:01] <jepler> skunkworks_: there are so many to choose from. for ubuntu systems, I'd sure recommend python-gtk, because it will look like other desktop applications.
[18:51:11] <skunkworks_> http://wiki.python.org/moin/IntegratedDevelopmentEnvironments
[18:53:34] <skunkworks_> jepler: thanks
[18:54:02] <jepler> maybe I don't understand what you meant by "gui for python"; I assumed you meant "python modules for creating GUIs", not "GUI for using/developing Python software"
[18:54:10] <jepler> I don't know the answer for that -- I use a terminal
[18:54:27] <jepler> vim ten_thousand_lines_and_counting.py
[18:54:33] <jepler> :%d
[18:54:33] <skunkworks_> actually - both
[18:54:46] <jepler> you going to become a programmer? or are you answering somebody on cncmumble.com?
[18:55:12] <skunkworks_> no - we are at a turning point here a work. Looking for options.
[18:55:30] <jepler> oh
[18:56:11] <skunkworks_> our infrustructure is a bit on the dated side..
[18:56:46] <skunkworks_> moving towardw a mysql back end and looking for front end options.
[18:57:26] <skunkworks_> non microsoft preferibly
[18:58:34] <skunkworks_> heh - I consider myself a programmer.. but not compared to you guys.
[18:59:44] <jepler> fwiw it appears there is a python-gtk for windows, same for python-mysql -- with adequate testing, I would expect the same software to run on linux and windows both.
[19:05:15] <archivist> I get multi platform by using php apache mysql, then the clients only need a browser
[19:07:10] <jepler> yeah you could do the same with python + httpd + sql
[19:08:29] <jepler> not all applications are a good match for dumb client software like a web browser, though.
[19:08:33] <jepler> POST /?estop=true HTTP/1.1
[19:08:54] <alex_joni> one could put AJAX in there
[19:08:59] <ds2> and get back a 500 response? ;)
[19:09:10] <alex_joni> ds2: 404 is worse
[19:09:13] <alex_joni> or 403 :D
[19:09:16] <jepler> 404 SORRY ARM ALREADY SEVERED
[19:09:26] <archivist> hehe
[19:09:49] <ds2> somehow I don't think it will pass EU's version of OSHA ;)
[19:10:47] <ds2> anyone tried using the antistatic stuff on monitors with plexiglass/acrylic to prevent dust and small bits of chips from sticking to it?
[19:11:07] <jepler> http://blog.vlad1.com/2007/11/26/canvas-3d-gl-power-web-style/
[19:12:56] <tomp2> need a small surface grinder? http://www.wttool.com/product-exec/product_id/40789/nm/Bench_Top_Surface_Grinder
[19:14:20] <skunkworks_> that is cute
[19:16:23] <alex_joni> 465 lbs?
[19:16:49] <archivist> bigger than cute
[19:17:04] <tomp2> yeh, liteweight, but table top heavy
[19:17:14] <ds2> hmm guard on the side, wheel spins clockwise I guess ;)
[19:17:22] <ds2> doesn't look setup for coolant though
[19:17:26] <tomp2> jepler: a browser based cnc gui? also http://my.opera.com/timjoh/blog/2007/11/13/taking-the-canvas-to-another-dimension
[19:22:21] <archivist_win> if you dont mind a download and are on windows http://www.archivist.info/herron/pictest.php was a 3d thing I started writing bottom pic
[19:24:38] <archivist_win> again windaz and a different library and solidworks www.archivist.info/chime/index.html
[19:24:44] <skunkworks_> http://www.electronicsam.com/images/KandT/hexapodworkarea.JPG
[19:25:09] <alex_joni> bbl
[19:26:49] <archivist> skunkworks, that looks like the same viewer as Im using above for the herron(a window company) pic
[19:27:39] <skunkworks_> It was one i found to view the extents of that hexipod calculator
[19:31:42] <tomp2> the canvas3d.xpi requires firefox 3.0 thru 3.0b1 !?!?! 2.0 is current 0b1? ( as in 0b1-kenobe? :)
[19:33:19] <jepler> yeah I guess you have to get a beta version of firefox to try it out .. too bad
[19:35:43] <archivist> vrml seems to be available for firefox now (except the old one from 2004 just crashed the winbox)
[20:28:50] <gezar> some sheet, Ive got a hum dinger of a cal problem and im not sure how to approach it yet?
[20:33:49] <cradek> cal?
[20:34:42] <gezar> calculus
[20:35:45] <cradek> in these circles saying something like that is called a "geektrap"
[20:36:21] <gezar> yeah, I know, ive made it to the current peak of my brains I belive, and I have stalled out on #16 of page 194
[20:36:40] <gezar> I dont want my homework done for me, but I cant see the method to solving this one
[20:36:58] <cradek> oh it's schoolwork? that's invalid for geektrap material
[20:37:06] <gezar> yep
[20:37:31] <gezar> however, having gezar knowing C and advanced math makes for better machine motion
[20:38:53] <jepler> hm my page 194 only goes to exercise 5; page 195 has additional exercises but only through #11, then it's on to chapter 9.
[20:39:06] <gezar> hahahaha
[20:39:10] <jepler> oh wait, what book is it?
[20:39:36] <gezar> early transendential functions of calculus 3rd edition
[20:39:38] <jepler> (I thought it was funny that the first book I opened to page 194 had exercises on it at all)
[20:40:34] <gezar> this one is just a simple solve for the derivitave of an interesting problem, I cant see how to look at it, my first try ended up with a string of numbers about 25 units long
[20:40:38] <cradek> all I can find are "fundamentals of differential equations" and "engineering electromagnetic fields and waves"
[20:40:55] <cradek> and "sendmail"
[20:41:14] <gezar> hahahaha, you guys are having too much fun with me rightnow
[20:41:15] <cradek> "applying SCCS and RCS"
[20:41:18] <gezar> thats okay
[20:41:21] <jepler> this was 'The C++ Programming Language'
[20:41:31] <jepler> I don' think we have what it takes to help you
[20:41:36] <cradek> nope
[20:41:48] <cradek> although I'm not down to my desk surface yet, you never know
[20:41:50] <gezar> maybe Im needing different conversation then math :)
[20:43:08] <gezar> and no, I havent been able to make my mini machine, too damn busy trying to remember algebra, and do calculus at the same time
[20:44:22] <gezar> im soo wanting to machine sugar cubes too
[20:45:37] <gezar> and as far as compsci so far this year, im sick of payroll programs
[20:48:37] <archivist> be adventurous cut metal
[20:49:21] <gezar> if my math professor would give me extra credit I would machine him a nice ocarina
[20:50:17] <archivist> I hated calculus and algebra
[20:55:41] <BigJohnT> archivist: how did your gear come out?
[20:56:02] <archivist> it looks like a gear
[20:56:09] <BigJohnT> cool...
[20:57:10] <archivist> made swarf twice now, must finish the machine
[20:58:27] <seb_kuzminsky> where do you guys buy small quantities of metal stock?
[20:58:28] <BigJohnT> were making bullets
[20:58:38] <BigJohnT> mcmaster carr
[20:59:06] <cradek> jmk says he likes 'metal express' but I've not tried them yet
[20:59:29] <seb_kuzminsky> anyone try OnlineMetals.com or SpeedyMetals.com?
[20:59:36] <archivist> scrap yard or local stockist offcuts
[21:00:18] <BigJohnT> I order so much from mcmaster carr that it is easy to add a bit of metal when I need some. If it is a big job I get it from a local metal supply
[21:01:27] <cradek> yeah I usually go to local places - depends what I need
[21:02:23] <seb_kuzminsky> scrap yards seem like a good idea if you know how to identify what you need. hardware store (at least around here) have poor selections of expensive metal...
[21:16:18] <BigJohnT> * BigJohnT going to shoot guns
[21:29:32] <gezar> hardcore techno music is helping me overcome the problem
[21:29:38] <gezar> ill show it to yall
[21:30:02] <gezar> would it better suit me to do all the inside math, then chain rule, or should I take a deeper look atf(x)= [(x^2+1)(x^(1/2)+1)^3)]^1/2 solve for f'(x)?
[21:30:57] <gezar> i just finished up buring up an entire sheet of paper, I know the answer, but im working it out
[21:37:29] <jepler> chain rule is what you need to find the derivative of sqrt(g(x))
[21:38:01] <jepler> that's a f(g(x)) which is pretty much the definition of when to apply the chain rul
[21:38:02] <jepler> e
[21:43:02] <gezar> yeah i know that part, its breaking it down to do it right
[21:48:22] <maddash> i'm using 2.2.2, and I just discovered that continuous jogging does not halt at the limit switches, even after overrides have been disabled
[21:49:45] <maddash> if an emcmot-jog-cont is allowed to occur without any emc_aborts, jog_ok() in command.c isn't called again after the first time
[21:49:50] <cradek> soft limits don't allow you to even hit the limit switches if you have them set right
[21:50:18] <maddash> my problem's not with the soft-limits, so I don't see their relevance here.
[21:50:21] <cradek> and hitting a limit turns machine off
[21:51:36] <cradek> I don't understand the behavior you're seeing - can you say again what happens at the gui and machine, not the source code?
[21:51:43] <gezar> all be damned, it was the product rule, then using chain to solve the individual derivitaves for each section
[21:51:50] <gezar> still had to use the calculator
[21:52:01] <maddash> tkemc's x-axis coords turn red, but the numbers keep decreasing( even after the neg-lim sw is triggered)
[21:52:28] <jepler> maddash: first, are you talking about jogging from a UI that sends NML messages, or jogging by putting values on the axis.N.jog-* pins?
[21:53:01] <maddash> jepler: holding the "<--" on my keyboard inside tkemc?
[21:53:59] <gezar> hmm, ran into a wall at looking how to do that
[21:55:12] <maddash> jepler: could the problem be with the limit switch?
[21:55:26] <maddash> i just looked at it in halscope,
[21:55:32] <maddash> (the lim sw)
[21:55:52] <maddash> and there's a lot of digital noise, even after i hold it down
[21:56:02] <cradek> it stops for me...
[21:56:16] <cradek> joint 0 on limit switch error
[21:56:20] <cradek> and then machine turns off
[21:56:41] <SWPadnos> maddash, just to be sure, are you looking at the port input pin, the signal, or the axis.n.lim-sw-xxx-in pin?
[21:57:02] <maddash> SWPadnos: in halscope? parport.0.pin-08-in
[21:57:12] <jepler> if you hold the limit switch closed and find that it's somtimes TRUE and sometimes FALSE in emc, then the problem is probably not a bug in emc's handling of limit switches.
[21:57:21] <jepler> er, sometimes TRUE and sometimes FALSE in hal
[21:57:22] <SWPadnos> ok, look at the motion controller input, in case the pin/signal isn't properly connected in HAL
[21:58:08] <cradek> I don't see any problem
[21:58:45] <cradek> I metered axis.0.joint-pos-cmd and it stops, tkemc numbers turn red, drops to ESTOP RESET state, and I get "joint 0 on limit switch" error popup
[21:58:45] <maddash> "net Xhome parport.0.pin-07-in-not axis.0.home-sw-in"
[21:59:02] <cradek> maddash: ...limit-sw?
[21:59:11] <SWPadnos> yep - that's the home switch, not the limit
[21:59:23] <maddash> home-sw-in isn't hooked up internally to lim-sw-in?
[21:59:25] <cradek> 32770 bit IN FALSE axis.0.neg-lim-sw-in
[21:59:25] <cradek> 32770 bit IN FALSE axis.0.pos-lim-sw-in
[21:59:28] <SWPadnos> no, it isn't
[21:59:44] <maddash> brb, then
[21:59:52] <SWPadnos> if you connect them together, you need to say that in the ini file, something ilke HOME_IS_SHARED = true
[22:00:09] <cradek> that's the wrong ini entry
[22:00:19] <cradek> HOME_IGNORE_LIMIT is what you want, I think
[22:00:33] <SWPadnos> uh - right, something like one or the other or both of those ;)
[22:00:42] <jepler> You mean to say HOME_IGNORE_LIMIT but that doesn't cause additional connections in HAL -- it just makes homing still work in case you make those connections in HAL
[22:00:46] <cradek> yeah, better read the docs :-)
[22:00:52] <jepler> yes this is in the documentation
[22:01:01] <cradek> yes definitely
[22:01:33] <SWPadnos> right - I was going to point out that the ini entry only tells EMC how HAL is supposedly connected, it doesn't make the connections
[22:01:35] <maddash> gosh, I thought I read somewhere in the docs that setting H-I-L = 1 causes the home to associate with the lim sw automagically
[22:01:54] <SWPadnos> no, it causes the limit switch to be disassociated while homing ;)
[22:01:58] <maddash> geez. sorry to bug you guys.
[22:48:19] <alex_joni> good night all
[23:15:35] <eric_U> at least maddash doesn't want to add DRM to EMC2
[23:16:57] <eric_U> are there any open source projects that implement DRM?
[23:19:01] <archivist> I hope not
[23:19:10] <eric_U> I'm thinking they may
[23:21:14] <eric_U> if you want to know why I'm talking about that, someone came on here about 3:30 us eastern time and started asking how to use EMC with copy protected files instead of g-code
[23:22:26] <eric_U> toasty was having his way with the guy
[23:22:36] <archivist> I was around
[23:23:11] <eric_U> that's a reasonable hour in the U.K. now that I think of it :)
[23:23:48] <archivist> I was supposed to be working!
[23:23:53] <eric_U> I was supposed to be sleeping
[23:26:38] <cradek> wow, I tried to read back, but I can't make it through
[23:27:23] <eric_U> the guy could solve his problem more readily with cnc machines owned by an artists collective
[23:27:38] <eric_U> DRM is nearly universally a bad idea
[23:28:09] <cradek> yes, selling "prints" that require this cnc equipment to make would be a reasonable business plan
[23:28:48] <cradek> he could of course use emc to run the "printer"
[23:29:19] <cradek> but the thing he seemed to want instead would have so many problems
[23:30:49] <eric_U> seemed like his market was local anyway
[23:31:14] <eric_U> and that there were no local suppliers
[23:31:42] <eric_U> so why come up with a split business model, just make the artist and producer vertically integrated
[23:32:07] <eric_U> problem is, he got a government grant to do the copy protection thing, at least I think that's what he said
[23:32:40] <ds2> copy protection does not work. it is a waste of everyone's time, past, present, and the future.
[23:35:40] <cradek> like war, it may push technology and innovation