#emc | Logs for 2008-11-07

[01:05:08] <toastydeath> the only thing that i am a little upset about so far re: obama is that he is all for mandatory volunteer service for hs/college
[01:05:13] <toastydeath> i am going to throw a fit if that becomes a reality
[01:06:08] <JymmmEMC> toastydeath: it'll do ya some good
[01:06:20] <toastydeath> i hate volunteer work, i had to do it before with homeless people
[01:06:37] <toastydeath> now instead of a general feeling of ambevelence, i want them to starve
[01:06:55] <JymmmEMC> toastydeath: And if you throw a fit, we'll just let you volunter to work with the kids that have motor contorl impairments, you'll fit right it
[01:07:13] <toastydeath> exactly.
[01:07:18] <toastydeath> nobody will notice if i push them over.
[01:07:30] <toastydeath> it'll be the perfect outlet for my rage that I'm there in the first place.
[01:07:32] <JymmmEMC> oe they push YOU over
[01:07:45] <toastydeath> I'm not actually motor impared, I assume I can dodge.
[01:08:03] <toastydeath> though i guess there might be some motor-impaired savant who has superhuman reflexes
[01:08:07] <toastydeath> against whom i stand no chance.
[01:08:29] <JymmmEMC> It'll be a toastydeath gang bang
[01:08:55] <toastydeath> i wonder if there's an actual requirement to do anything
[01:08:55] <JymmmEMC> Then you'll get on You Tube
[01:09:08] <toastydeath> or if i can just hold a broom and stand there for a couple hours and go home
[01:09:22] <JymmmEMC> you'll have to clean up the puke
[01:09:27] <SWPLinux> "Set goals for middle-school and high-school students to serve 50 hours a year of public service, and for college students to serve 100 hours a year."
[01:09:31] <SWPLinux> that's the actual text
[01:09:33] <toastydeath> i'll stand there and contemplate cleaning up the puke
[01:09:40] <SWPLinux> so it says nothing about it being mandatory
[01:10:03] <toastydeath> swplinux: that's not the only article about it, and he's stated that he wants to make federal funding dependent on attendence
[01:10:13] <toastydeath> so if the students don't show, the college/hs doesn't get funding.
[01:10:14] <SWPLinux> where did you read that?
[01:10:21] <toastydeath> on other sites when I googled it
[01:10:32] <toastydeath> he's commented several times, not just once.
[01:12:43] <SWPLinux> you need to check your sources. try using something like http://www.sourcewatch.org/
[01:12:58] <SWPLinux> (we all need to check our sources - that's not meant to sound accusatory)
[01:13:14] <toastydeath> okay, except I've read the quotes, from him talking, on news websites
[01:13:26] <SWPLinux> which news websites?
[01:13:26] <toastydeath> i didn't think i needed to check direct quotes, but i guess i do
[01:13:37] <toastydeath> i don't have a list in front of me or i'd have linked it
[01:13:45] <SWPLinux> it's amazing how little truth and how much bullshit there is on the web
[01:13:52] <SWPLinux> and in print or on TV for that matter
[01:14:43] <toastydeath> i understand your criticism, but having seen it in several places that were not someone's random blog, i'm going to believe it
[01:14:53] <toastydeath> i'm googling it to try and find the one that stuck out
[01:15:41] <SWPLinux> well, I know there are a number of things I looked up before the election (Obama is a treeorits, Obama is a Muslim, Obama is a ...), and every one of them was false and/or totally blown out of proportion
[01:15:58] <toastydeath> i understand that
[01:16:15] <SWPLinux> even John McCain was repeating the bullshit (for instance, that Fannie Mae and Freddie Mac were among the highest contributors to Obamas campaign)
[01:16:36] <SWPLinux> (which I checked today, using the source he cited, and found his claims to be false)
[01:16:39] <toastydeath> i understand that
[01:16:52] <toastydeath> the same quote, attributed to obama, in several places, was compelling enough for me
[01:17:08] <SWPLinux> ok, I'm just giving you another data point on that situation ;)
[01:17:25] <fenn> geez it's not like they're sending you off to vietnam
[01:17:31] <SWPLinux> again!
[01:17:33] <SWPLinux> err
[01:17:42] <fenn> SWPLinux: did you really work on tanks?
[01:18:35] <SWPLinux> eys
[01:18:37] <SWPLinux> yes
[01:18:47] <fenn> ok. :)
[01:18:51] <SWPLinux> not designing them or anything, just repairing some parts
[01:20:41] <toastydeath> http://www.latimes.com/news/opinion/la-oe-goldberg8-2008jul08,0,368008.column
[01:22:28] <SWPLinux> the quote in that article is exactly as I pasted
[01:23:10] <SWPLinux> http://www.sourcewatch.org/index.php?title=Jonah_Goldberg
[01:23:27] <SWPLinux> I'm not positive it's the same guy, but he has the same name :)
[01:24:46] <SWPLinux> the actual policy statement (but not the actual legalese) is here: http://www.barackobama.com/issues/service/
[01:24:57] <toastydeath> yes, i read that too
[01:25:04] <SWPLinux> and I didn't see anything about withholding funds in my scan of that
[01:25:11] <toastydeath> and it doesn't include the withholding funds part
[01:25:35] <SWPLinux> they may be spinning the "25% of work study money to be used for service jobs instead of cafeteria work" as withholding funds
[01:25:38] <toastydeath> because he said it at a campaign speech
[01:26:27] <toastydeath> and regardless, I think it's true, but at the same time I hope you are right
[01:26:51] <toastydeath> it would be very nice to be wrong on this issue.
[01:27:11] <SWPLinux> well, I just get annoyed when I see people (apparently) go off the deep end because of something some spinmeister (on either side) said somewhere
[01:27:27] <SWPLinux> not that you're off the deep end, but it sure seemed like it got your blood pressure up :)
[01:27:33] <toastydeath> my blood pressure is up
[01:27:52] <SWPLinux> see! I was right!!!
[01:28:14] <toastydeath> but i've seen it enough from obama supporters that it concerns me.
[01:28:26] <toastydeath> that one link is hardly all, but it's what I found at the moment
[01:28:41] <SWPLinux> sure. I'll look into it more, even if you don't (I'm like that ;) )
[01:28:54] <toastydeath> excellent
[01:40:28] <JymmmEMC> I think it was 1000 hours hours for college
[01:41:07] <SWPLinux> this may be a more balanced analysis of the policy (but still no source cited): http://www.politico.com/news/stories/0708/11961.html
[01:41:43] <SWPLinux> we needed 140 credits or so, with each credit being 1 hour of class per week for 15 weeks
[01:41:50] <SWPLinux> so 2100 hours or thereabouts
[01:41:55] <SWPLinux> (over 4 years)
[01:52:26] <toastydeath> cool
[01:53:31] <boulabiar> hi
[01:53:40] <toastydeath> hi
[01:53:50] <toastydeath> swplinux: that would be awesome if there's just going to be money around for that kind of program
[01:53:58] <toastydeath> i just want to make sure I don't have to be part of it.
[01:53:59] <boulabiar> can I find help for Xenomai here ?
[01:54:48] <boulabiar> hmm ?
[01:56:41] <fenn> boulabiar: probably not. we only use RTAI
[01:57:47] <JymmmEMC> boulabiar: Try theri mailing list http://www.xenomai.org/index.php/Main_Page
[01:58:22] <boulabiar> isn't Xenomai more up-to-date ?
[01:58:28] <boulabiar> why do you use RTAI ???
[01:59:57] <SWPLinux> because it was there when it was needed
[02:00:16] <SWPLinux> xenomai is newer, and EMC2 could use it, but nobody has written the interface layer for it
[02:00:22] <SWPLinux> (patches gratefully accepted :) )
[02:01:05] <SWPLinux> toastydeath: here's the source document I think - let's see what actually happens: http://www.barackobama.com/pdf/NationalServicePlanFactSheet.pdf
[02:01:27] <toastydeath> opening
[02:01:40] <boulabiar> Obama ?
[02:01:50] <boulabiar> sorry but what is EMC ?
[02:01:53] <SWPLinux> it's longer than I wanted to read tonight, but I didn't see anything in a scan
[02:02:00] <SWPLinux> EMC is a machine controller
[02:02:11] <SWPLinux> we were havign a political discussion earlier though :)
[02:02:21] <fenn> boulabiar: RTAI/Xenomai was recently discussed on the mailing list here: http://thread.gmane.org/gmane.linux.distributions.emc.devel/1342
[02:02:48] <toastydeath> swplinux: part 2 has the stuff
[02:02:52] <toastydeath> depending on how you want to read it
[02:03:12] <toastydeath> there's a 4000 dollar tax credit for college, rather than compulsory
[02:03:18] <boulabiar> Enhanced Machine Controller = "is a machine controller" ???????
[02:03:21] <boulabiar> huh ?
[02:03:42] <SWPLinux> have you looked at http://www.linuxcnc.org/
[02:03:48] <SWPLinux> ?
[02:04:28] <boulabiar> I am there now
[02:05:01] <SWPLinux> oh, you're looking for xenomai only help, not related to EMC (I get it now ;) )
[02:05:36] <SWPLinux> toastydeath: yep, I saw that part, but forgot to point it out
[02:05:44] <fenn> well since we cant help you, perhaps you'd like to stir up the political discussion :)
[02:05:50] <toastydeath> it's not as bad as it was made out to be, thank you swplinux
[02:05:58] <SWPLinux> any time :)
[02:06:17] <fenn> american culture tends to over-exaggerate things
[02:06:36] <SWPLinux> specifically far left and far right commentators in American culture
[02:06:41] <toastydeath> well, from what I read in this, it will be mandatory for highschool
[02:06:54] <toastydeath> which does not make me happy at all, but obviously doesn't affect me directly so the impact is less
[02:07:15] <fenn> i dont see why it's a big deal when high school is practically mandatory anyway
[02:07:26] <JymmmEMC> fenn: so does toastydeath when he has to get off his butt =)
[02:07:46] <toastydeath> because i had to volunteer in highschool and it was awful
[02:07:58] <SWPLinux> I didn't see anything about a requirement, but I do see the phrase "should be expected to do 50 hours ..."
[02:08:00] <toastydeath> that's why i care about it
[02:08:05] <SWPLinux> so I can see how people would overreact
[02:08:20] <SWPLinux> have you read "Starship Troopers"?
[02:08:22] <fenn> well i thought high school was awful in general so i'm not surprised that extra high school is more awful
[02:08:34] <fenn> SWPLinux: hah that's what i was thinking
[02:08:39] <toastydeath> i have not
[02:08:45] <SWPLinux> good book
[02:08:52] <SWPLinux> stupid movie, different story with the same name
[02:08:55] <toastydeath> what's the reference
[02:08:59] <toastydeath> that you are making to it?
[02:09:12] <SWPLinux> in the book, people have to do 2 years of "federal service" before they get voting rights
[02:09:28] <SWPLinux> "federal service" could be being a nurse or in the military - whatever
[02:09:33] <JymmmEMC> SWPLinux: But did they have a fscked up voting system?
[02:09:53] <SWPLinux> the idea was that until you put aside your own greed and self-interest, you're probably unqualified to make decisions that affect everyone
[02:10:05] <SWPLinux> so you have to do it for 2 years if you want to vote
[02:10:13] <toastydeath> and i think that's terrible.
[02:10:20] <JymmmEMC> SWPLinux: Hey, isn't that an oxymoron (politions)
[02:10:29] <SWPLinux> you can choose not to, but then you don't get to vote and you just have to live with whatever the full citizens decide
[02:10:37] <toastydeath> yeah, and i think that's not the right way to go.
[02:10:55] <SWPLinux> yeah. I don't like it, but it does have its merits
[02:11:02] <fenn> i think there should be free beer for everyone
[02:11:16] <JymmmEMC> PARTY AT FENN'S HOUSE!
[02:11:29] <SWPLinux> note that volunteering to do IT work or machine repair for some community organization would also be considered "community service"
[02:11:29] <fenn> BYOFB
[02:11:39] <SWPLinux> it doesn't have to be "some politician telling you what to do"
[02:11:54] <toastydeath> the compulsory part is what bothers me.
[02:12:27] <toastydeath> the assumption that people don't want to help unless they're told
[02:12:39] <toastydeath> i think speaks volumes about the people who support it
[02:12:51] <toastydeath> and what they think about other people.
[02:12:59] <SWPLinux> no, actually I think the idea is that a lot of people want to help in some way, but the groups that need the help are underfunded
[02:13:05] <toastydeath> fund them!
[02:13:10] <SWPLinux> so the idea is to fund them
[02:13:11] <toastydeath> fund the crap out of them.
[02:13:12] <SWPLinux> see! :)
[02:13:20] <toastydeath> but don't make it mandatory, in any way.
[02:13:38] <toastydeath> give people the oppertunity to show that they're good people, don't tell them "help your fellow man or else."
[02:13:46] <SWPLinux> we'll see how mandatory it ends up being, I'm not too worried about it at the moment
[02:13:58] <toastydeath> that's the only part i'm worried about.
[02:14:01] <SWPLinux> heh
[02:14:07] <toastydeath> if he wants to pump funding into volunteer programs, do it.
[02:14:18] <SWPLinux> try to remember this conversation in a couple of years, we'll see what happens :)
[02:14:39] <toastydeath> haha, i will, and it will be nice to be wrong
[02:14:49] <SWPLinux> I disagree ;)
[02:15:00] <toastydeath> ?
[02:15:14] <toastydeath> I'm saying I'll be happy to be wrong
[02:15:18] <SWPLinux> (from my point of view, it wouldn't be nice to be wrong ;) )
[02:15:26] <SWPLinux> I agree, it would be nice if you're wrong
[02:15:30] <toastydeath> yeah, it will be bad if you are wrong
[02:15:30] <SWPLinux> :)
[02:15:32] <SWPLinux> heh
[03:08:14] <skunkworks> mounted one of the 2540 line encoders on the back of the monster servos.
[03:08:20] <skunkworks> Pretty painless
[03:10:47] <skunkworks> tm
[04:41:28] <steve_stallings> steve_stallings is now known as steves_logging
[04:59:47] <ewilhelm> question: "Several letters (I, J, K, L, P, R) may have different meanings ..." -- there is no L though?
[04:59:57] <ewilhelm> http://www.linuxcnc.org/docview/html//gcode_main.html
[05:02:56] <cradek> hmm, it is missing from the table
[05:03:14] <cradek> I believe L is used with G10, all the canned cycles, and G76
[05:04:50] <ewilhelm> ah, indeed: http://www.isd.mel.nist.gov/personnel/kramer/pubs/RS274NGC_3.web/RS274NGC_33a.html
[05:04:54] <cradek> also, E and O are used but not in that table
[05:05:06] <cradek> all letters are now used
[05:05:50] <ewilhelm> oh? is that being defined somewhere that I'm missing?
[05:06:30] <cradek> http://www.linuxcnc.org/docs/2.2/html/gcode_main.html#cha:O-Codes
[05:07:19] <cradek> this is my favorite docs page. it is a great summary and the links go to the detailed help for each gcode: http://www.linuxcnc.org/docs/2.2/html/gcode.html
[05:07:52] <ewilhelm> nice. thanks
[08:57:35] <tomp-road> to get good cad for linux, i used vmware on last laptop.
[08:57:45] <tomp-road> i just got new laptop and installed from hardy-emc live cd.
[08:57:53] <tomp-road> now have error installing vmware server 2 for linux on ubuntu8.04.
[08:58:05] <tomp-road>  err was gcc4.2.4 was installed but gcc4.2.3 was used to build system.
[08:58:07] <tomp-road>  re-installing gcc 4.2.3-1ubuntu3 still had gcc symlinked to gcc4.2.4. any ideas? ignore warning?
[13:42:38] <archivist> cradek, dunno if this would be a simple mod but with your truetype prog, how about a version to use an engraving bit that goes down the center line and varies depth to set width, would require some path finding foo adding, I can see an immediate application with clock dials, will have a browse for algorithms as soon as I can
[14:36:52] <Fisia> :)
[14:41:56] <cradek> archivist: sounds neat, but very hard compared to what it currently does
[14:44:25] <archivist> cradek, I had a quick browse, seems the odd commercial one around (mastercam) will look in my graphic gems books as well later for inspiration
[14:46:03] <cradek> I think you have to inscribe circles and construct paths connecting the centers where the depth is proportional to the diameters
[14:47:54] <archivist> starting from a round shape something like that
[14:48:54] <cradek> I can picture the required path in my head, but I don't know how to calculate it
[14:49:28] <archivist> but from square/points then its tip length going deeper till it meets another path
[14:49:34] <SWPLinux> copy/paste the cutter comp code from EMC ;)
[14:50:07] <cradek> archivist: yes each sharp corner would have a path endpoint
[14:50:19] <archivist> I can see the paths in my head easily
[14:50:29] <cradek> yep, me too
[14:50:38] <cradek> SWPLinux: it's a completely different problem from that
[14:50:56] <SWPLinux> oh. bummer
[14:51:55] <cradek> imagine wanting to cut out a square. you have a V shaped cutter. you can make an X shape starting at the corners with depth zero, and at the center you have a depth where the diameter equals a side of the square
[14:52:20] <cradek> so with two "V" cuts you've cut out all of the square - the center will be the deepest
[14:52:32] <cradek> you can do this for arbitrary shapes
[14:52:56] <SWPLinux> hmmm
[14:52:57] <SWPLinux> mmmmm
[14:53:03] <archivist> really fancy engraving becomes possible
[14:53:08] <cradek> do you see it? it's easy to picture/draw, but hard to describe
[14:53:21] <SWPLinux> yes, I see the idea for simple shapes
[14:53:28] <SWPLinux> trying to think about more complex ones
[14:53:49] <SWPLinux> a circle becomes a Z plunge only (assuming the cutter is wide enough)
[14:53:50] <archivist> thin to thick back to thin as in caligraphy
[14:54:01] <cradek> in general you'll have a path starting at each point, going towards the "center" of the shape, getting deeper until it reaches some inscribed maximum diameter
[14:54:16] <SWPLinux> you need something closer to offs
[14:54:31] <SWPLinux> since it has infinite lookahead
[14:54:45] <cradek> yes you need to consider the whole shape at once
[14:54:46] <SWPLinux> (I know it's a path generating tool, but I think the concept is similar)
[14:54:49] <SWPLinux> yep
[14:54:55] <cradek> fortunately the shapes are closed finite-length paths
[14:55:01] <SWPLinux> heh
[14:55:31] <SWPLinux> so the simplest case is just assuming an infinite cutter size and varying depth to set path width
[14:55:54] <SWPLinux> then you need to deal with the reality of having a cutter that's finite width and a maximum desired depth
[14:56:37] <archivist> yes but first attempt make plate thick enough
[14:56:41] <SWPLinux> heh
[14:56:42] <fenn> the CGAL 'skeleton' algorithm does something like this
[14:56:44] <jepler> I have a suspicion that it's these red paths from the black polygons: http://www.cgal.org/Manual/3.3/doc_html/cgal_manual/Straight_skeleton_2/Chapter_main.html#ComplexSLS
[14:56:46] <SWPLinux> and use a big cutter
[14:56:47] <jepler> fenn: yeah exactly
[14:56:50] <jepler> too bad about CGAL's license
[14:57:25] <cradek> jepler: I agree those are the paths
[14:57:48] <fenn> also i remember reading that it was easier to do this with pixel based methods
[14:58:34] <jepler> yeah, there's also a relationship to gimp's shaped gradients
[14:58:53] <jepler> the color in the gradient giving depth, and the paths being something related to locally maximal pixels
[14:58:58] <SWPLinux> I think offs will generate those paths, if you add lines from the original cusps to the offset cusps
[14:59:15] <jepler> SWPLinux: except for how offs is unfixably buggy
[14:59:19] <jepler> * jepler has given up on that project
[14:59:29] <SWPLinux> bummer :(
[14:59:51] <fenn> the skeleton should be the same as Held's voronoi diagram
[14:59:58] <cradek> what remains wrong with offs?
[15:00:16] <jepler> cradek: it does not get all self-intersecting offset paths right, it only gets simple cases right.
[15:00:35] <jepler> and I never even got to islands
[15:00:36] <cradek> oh, heck.
[15:01:05] <SWPLinux> yeah - seeing that the offset path goes the "wrong direction" is a lot easier than writing a program to detect it
[15:03:24] <jepler> in this example, the upper middle area doesn't fill. the reason has to do with how the offset path has fallen apart, and offs doesn't find that there are parts of the fallen-apart path that it should still work on: http://emergent.unpy.net/files/sandbox/offs-one-more-bug-fixed.png
[15:03:53] <jepler> this is exactly the kind of "it gets very nonlocal" problem that all the papers cite when they say you should use these skeleton algorithms instead of directly offsetting.. but I wouldn't listen to them
[15:04:28] <cradek> but that looks SO close
[15:04:35] <jepler> yes, I know
[15:04:41] <jepler> feel free to fix it
[15:05:19] <jepler> source code is online: http://git.unpy.net/view?p=offs.git;a=summary
[15:06:14] <archivist> bah work getting in the way
[15:06:52] <jepler> "For a similar problem, which (I believe) formalizes what makes constructing the straight skeleton hard, see "Crashing Motorcycles Efficiently". "
[15:07:04] <SWPLinux> heh
[15:08:03] <SWPLinux> it's even hard for a square
[15:08:12] <SWPLinux> consider a 1" square, but offsetting by 2"
[15:08:34] <SWPLinux> (or anything more than 1/2" really)
[15:09:01] <jepler> that's "easy": the signed area of the offset polygon becomes negative
[15:09:09] <jepler> (what, you didn't know that polygon area is signed?)
[15:09:13] <SWPLinux> heh
[15:09:26] <SWPLinux> I was thinking of how to determine the starting position
[15:09:32] <SWPLinux> of the offset path
[15:10:12] <jepler> oh, I thought you were talking about when the inset path becomes degenerate
[15:10:15] <SWPLinux> assuming we start at (0,0) and move to (1,0), the offset path would start at 0,2 and go to 1,2
[15:10:20] <SWPLinux> sure, that's a problem too
[15:10:31] <jepler> offs assumes you can start anywhere on the offset contour
[15:11:05] <SWPLinux> you need to take into account the other connecting segment (in the case of a square, the last segment)
[15:11:15] <jepler> to put it another way, offs takes a closed path and turns it into a set of closed paths
[15:11:21] <jepler> it doesn't deal in starting points
[15:11:43] <SWPLinux> right, the assumption (requirement) is that you always have two segments at each point
[15:12:18] <SWPLinux> which isn't true for the actual source path, until you look at the entire thing
[15:15:44] <SWPLinux> damn, still available: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=290272892167
[15:15:51] <SWPLinux> and only 450 miles away
[15:16:21] <archivist> go fetch
[15:16:25] <SWPLinux> heh
[15:17:18] <cradek> I bet I know the ending price
[15:17:33] <SWPLinux> $0, $1000, or $2500 :)
[15:17:38] <cradek> $1000
[15:17:43] <cradek> well, if you buy it
[15:17:45] <cradek> otherwise, $0
[15:17:48] <SWPLinux> heh
[15:18:00] <cradek> looks like a 4 position turret, but it's long enough to put several tools on each side
[15:18:14] <SWPLinux> yeah, a lot of screws on that plate
[15:18:38] <cradek> but drills and boring bars, I dunno.
[15:18:40] <SWPLinux> I wonder if they're right that it repeats to 50 millionths
[15:18:51] <archivist> so many toys to get so little money and space
[15:19:13] <SWPLinux> yeah, there's the rub
[15:19:16] <cradek> that means a tenth on diameter? if it's the same class as mine, possibly
[15:19:46] <cradek> don't breathe on the part before you measure it though
[15:19:58] <SWPLinux> or breathe a lot while it's being machined ;)
[15:20:20] <SWPLinux> just keep the coolant at 98.6 degrees so you can measure easier
[15:20:33] <cradek> warmant
[15:20:40] <SWPLinux> lukewarmant
[15:21:22] <archivist> make it work it will warm itself up
[15:22:24] <SWPLinux> it looks pretty clean, and if it works as a CNC now, the retrofit should be easier
[15:24:33] <SWPLinux> oh hmm, I guess I should check out soon
[15:24:59] <cradek> it has a crt - the control might be quite usable as-is
[15:25:17] <SWPLinux> as long as I can get the "verbage" to come out right ;)
[15:26:03] <SWPLinux> maybe Stuart would want the screen for one of his machines ;)
[15:27:01] <archivist> looks like its done little work
[15:27:09] <BigJohnT> dang it is in New York
[15:27:14] <SWPLinux> Buffalo
[15:27:42] <SWPLinux> yeah, I don't know if they described everything, but skimming .007 off Aluminum for decades doesn't sound too hard
[15:30:22] <BigJohnT> If it was closer I'd buy it
[15:30:39] <SWPLinux> me too (still might :) )
[15:30:48] <SWPLinux> I got my mill in Buffalo
[15:31:24] <BigJohnT> it's about 18 hours to Buffalo from here :(
[15:31:34] <SWPLinux> bummer. only ~8 from here
[15:31:56] <SWPLinux> where are you? I thought you were in the northeast somewhere
[15:31:57] <archivist> ship from here :(((
[15:32:03] <cradek> driving a lathe for 8 hours is quite possible
[15:32:11] <BigJohnT> I'm in swamp east Missouri
[15:32:18] <cradek> get a few friends together and it's almost fun
[15:32:22] <SWPLinux> yeah, we picked up my mill then a Takisawa lathe in the same run
[15:32:26] <BigJohnT> I'd go 12 hours
[15:32:33] <SWPLinux> ah, ok then
[15:32:42] <cradek> BigJohnT: an 8 hour drive takes 12 hours at 50mph :-)
[15:33:07] <archivist> silly speed limits over there :)
[15:33:13] <BigJohnT> yep with 4000 lbs on your trailer...
[15:33:18] <SWPLinux> it's not the law, it's just a good idea
[15:33:26] <cradek> yeah
[15:33:28] <SWPLinux> when the trailer weighs 50% more than the tow vehicle
[15:33:34] <BigJohnT> I got my manual lathe from Dallas it was a long trip back
[15:33:37] <cradek> KE ~ m*v^2
[15:33:58] <SWPLinux> so V~= swrt(2) * normal_V
[15:34:01] <SWPLinux> sqrt
[15:34:06] <SWPLinux> for the same Ke
[15:34:06] <BigJohnT> google says it is only 14 hours
[15:39:45] <SWPLinux> I just can't seem to find the specs on the machine
[15:39:50] <SWPLinux> like spindle HP
[15:39:53] <SWPLinux> work envelope
[15:41:39] <cradek> looks like it's a 6 or 8" chuck
[15:41:50] <cradek> I bet it has 18" of Z travel or less
[15:42:02] <cradek> with no provision for a tailstock, more length is useless
[15:42:32] <cradek> would be nice to know the spindle bore size, and if it can take collets
[15:44:21] <cradek> http://img.photobucket.com/albums/v618/Chevy427z/toolturretandcolletnose.jpg
[15:44:23] <SWPLinux> oh yeah, forgot to ask that
[15:44:35] <cradek> might be 5C, but it looks suspiciously small
[15:44:50] <SWPLinux> others I've seen via google are 16c ...
[15:44:57] <SWPLinux> so it's hard to tell
[15:51:18] <BigJohnT> it's 16c with a tailstock and about 7" swing over the ways
[15:52:00] <cradek> oh so the turret goes in front of the work, not beside it?
[15:52:38] <archivist> do they have back gear, or just designed for light work
[15:52:51] <BigJohnT> can't tell from the Q&A's at the bottom of the page
[15:53:04] <BigJohnT> 1985
[15:54:43] <BigJohnT> I love paying bills with Iron Man playing in the background :)
[15:55:07] <SWPLinux> oh - I guess I could have read those Q/As :)
[15:55:20] <BigJohnT> :)
[15:55:49] <archivist> that many q's shows interest
[15:56:02] <SWPLinux> yep
[15:56:23] <SWPLinux> some from Europe though, which probably won't matter due to shipping/crating costs
[15:56:48] <archivist> last seconds bunfight coming up
[15:56:51] <SWPLinux> hmmm. 84" tall
[15:57:09] <SWPLinux> it ain't a BP where I can swivel the head down
[15:57:19] <BigJohnT> just buy it now and get it over with
[15:57:50] <SWPLinux> nah, I'm trying to save money ;)
[15:58:05] <SWPLinux> OK, checkout time is here. see you when I get home
[15:58:06] <archivist> we want to spend your money
[15:58:06] <BigJohnT> you going to bid on it?
[15:58:10] <SWPLinux> probably
[15:58:25] <SWPLinux> got a friend who I want to look at the auction, but he should be around during the day
[15:58:30] <SWPLinux> before it closes
[15:58:54] <micges> hi all
[15:59:19] <SWPLinux> see you
[16:00:50] <micges> emc will optimize this gcodes in G64P0.01 mode ?
[16:00:50] <micges> G1 X100
[16:00:50] <micges> M5
[16:00:50] <micges> M50 P0
[16:00:50] <micges> G1 X150
[16:00:50] <micges> M50 P1
[16:00:52] <micges> M4
[16:00:54] <micges> G1 X200
[16:00:58] <micges> ?
[16:05:54] <jepler> my recollection is that emc will exact-stop at an M3/M4/M5 regardless of the path control mode, but I don't know for certain.
[16:06:30] <cradek> yes, and possibly also M50
[16:07:00] <cradek> but definitely M4/M5
[16:09:07] <micges> so there is no way to change IO during move without stopping ?
[16:09:40] <cradek> I don't think so
[16:10:14] <micges> ok
[16:12:09] <BigJohnT> oh well off to the fab shop... talk to you guys later
[17:04:24] <jepler> this guy's cam software requires that there be a sound device to run !?
[17:04:45] <archivist> heh I saw that
[17:05:37] <archivist> cant see how he lost sound anyway
[17:35:36] <SkinnYPuPp> Just when you think you've heard it all,
[17:35:37] <SkinnYPuPp> http://www.13wmaz.com/article/20081106/NEWS05/81106011
[17:41:25] <archivist> didnt qualify for a Darwin award
[17:42:09] <SWPadnos> so close though, so close
[17:42:27] <SWPadnos> better luck next time I guess
[17:46:37] <Vq^> i also enjoy cleaning with fire
[18:10:58] <Fisia> hi
[18:14:11] <plattschnauze> hy everbody, i am planing to tune up my CNC to a fabber, i know someone has done it before but i have some differnet ideas for the extruder. There for i need information wether there is a pin which i can use to differ between G0 or G1/2 movements. does it exist ?
[18:14:57] <anonimasu> why not just program a Mxxx to turn off the extruder before you do rapids?
[18:15:09] <plattschnauze> sorry for my english ! :-) out off training
[18:18:33] <plattschnauze> I want to use a Stepper for the Extruder in a Kinematic so that i can use ramps and other things. that is what im missing in repraps at this time.
[18:25:09] <plattschnauze> I think you can get much more precision an eventualy speed, if you are able to run the program with ramps, and the other thing is to leave the g-code unchanged as possible
[18:26:37] <plattschnauze> by the way, is alex online sometimes ?
[18:27:18] <skunkworks_> are you talking some sort of feed rate override?
[18:27:57] <archivist> seems to want to program acceleration rate
[18:30:17] <plattschnauze> i want to drive a Extruder as a Stepper . running only when G1/G2 operations are executed , in a defind speedrate to the XY movements
[18:32:18] <plattschnauze> ( endless in one direktion )
[18:35:09] <fenn> what do you mean "ramps"?
[18:35:21] <plattschnauze> acceleration
[18:35:29] <fenn> emc does that automatically
[18:36:07] <fenn> i think you should just couple spindle speed to your stepper in HAL somehow
[18:37:10] <fenn> then, if it really matters, you could do a spindle-synchronized move when extruding
[18:38:09] <fenn> or maybe i'm thinking adaptive feed
[18:38:28] <plattschnauze> yah, i am searching for a pin that differs between G0 and G1 movements.
[18:38:40] <fenn> there is no pin
[18:39:12] <fenn> you can turn certain pins on and off with other g-codes though, and put those before/after any g0 move
[18:39:59] <skunkworks_> you could probably turn a pin on above a certain velocity..
[18:42:35] <fenn> hmm is there a way to make the spindle slow down in response to feed override?
[18:42:44] <fenn> is that just a scale block in hal?
[18:46:42] <Fisia> :)
[18:46:49] <plattschnauze_> sorry computer shutdown
[18:47:19] <Fisia> Can any one tell me, How fast is, Stepper Step can run on EMC2 ?
[18:47:53] <Fisia> ...The fastest Step Clock...
[18:48:27] <Fisia> is it 80k step / sec ?
[18:50:11] <Fisia> cos maybe different kind of stepper could result different strength...
[18:50:33] <fragalot> Fisia: you should be able to do 30k easilly
[18:50:38] <Fisia> big one and small one... or other...
[18:50:54] <fragalot> depending on your system, that is
[18:51:06] <Fisia> is it 30K is the fastest EMC2 able to?
[18:51:18] <fragalot> depends on your computer how fast EMC can do it reliably
[18:51:31] <fenn> Fisia: it depends on the computer and stepper driver timing requirements
[18:51:32] <fragalot> the 30k is for a computer that has 11uS latency, iirc
[18:52:07] <fenn> more than 30k is not necessary in my opinion
[18:52:11] <plattschnauze_> I want to run the extruder in a exact follow off the XY Axis (example : X moves 3 mm / Extruder has to make 12 steps ) only while G1 / G2 moves and with the same accelerations !
[18:52:13] <Fisia> can u specify kind of computer like xxGHz Processor
[18:53:01] <Fisia> fenn... explain why please..
[18:53:18] <fenn> plattschnauze_: emc has a "constant surface speed" mode which might work for that, typically used for facing cuts on lathes
[18:53:55] <fenn> plattschnauze_: but the "only while g1" isn't going to happen, sorry.
[18:53:58] <Fisia> <plattschnauze_> u can use 2 computer ?
[18:54:04] <fragalot> Fisia: http://wiki.linuxcnc.org/uploads/StepTimingCalculator.ods
[18:54:06] <Fisia> at a time..
[18:54:29] <fragalot> Fisia: or http://wiki.linuxcnc.org/uploads/StepTimingCalculatorPlus.ods
[18:55:03] <fenn> Fisia: if you need more speed, decrease the number of microsteps in your stepper driver. if you need high speed and high resolution, you should probably be using servo motors anyway
[18:55:12] <plattschnauze_> I can use 5 computers at a time , if i want, but that was an accident :-)
[18:55:18] <fragalot> lol
[18:55:26] <fenn> Fisia: basically it's because step/dir is a crappy interface
[18:55:41] <Fisia> ok... ill go there, Thxs ! :)
[18:56:06] <fenn> sometimes i ssh from one computer to another and back, does that count?
[18:56:21] <Fisia> fenn, step/dir is a crappy interface, what does it mean, explain please..
[18:56:57] <fragalot> Fisia: step/dir is one signal line is for taking one step on the stepper motor, and the dir is the line that tells the controller what direction to step in
[18:57:09] <archivist> Fisia, openloop
[18:57:33] <Fisia> :>
[18:57:43] <Fisia> :)
[18:57:50] <fenn> Fisia: it takes a long time to tell the motor to hurry up and move, because you have to send a pulse for each micron
[18:58:54] <fenn> but each pulse has to come at exactly the right time
[18:59:59] <fragalot> fenn: don't you have to send the information at the exact right time.. nomatter the interface? ;p
[19:00:15] <archivist> fragalot, yes
[19:00:19] <Fisia> :)
[19:00:30] <anonimasu> step and dir is a nice hack for nice steppers..
[19:00:43] <archivist> but with servos you can react to error
[19:00:43] <anonimasu> if you want a better/nicer/faster/more accurate machine use servos..
[19:00:59] <fragalot> I'm not arguing that
[19:02:00] <anonimasu> steppers have big tradeoff between accuracy/speed/torque
[19:02:16] <Fisia> yes, true... but i still have kind of interrest on stepper control ...since PC has a 'little' changes
[19:02:53] <anonimasu> Fisia: then you are stuck with the limitations of steppers
[19:03:26] <anonimasu> there's also hardware step generators like jon elsons boards
[19:04:00] <Fisia> :)
[19:04:51] <anonimasu> well, that wont cange that when you try to go fast steppers suck
[19:06:14] <Fisia> true... but still interresting for hurry machines
[19:07:45] <Fisia> stepper weak on precision, speed, ... all u name it cos its a true open loop archivist says
[19:08:29] <Fisia> but all in emc2 is important knowledge for me
[19:08:52] <Fisia> including stepper dinosaur
[19:08:58] <fenn> anonimasu: my issue with hardware step generators is that they're basically a servo interface
[19:09:05] <fragalot> price-wise, steppers aren't that bad, drivers can be made on the cheap too, and for most homemade stuff, they are more accurate than the machine itself :p
[19:09:27] <anonimasu> fenn: yeah, but the motor's still a stepper
[19:09:28] <anonimasu> :)
[19:09:41] <fenn> ideally you'd just run the stepper as an AC servo
[19:09:44] <Fisia> :) true... still enteresting, yeah
[19:09:59] <fenn> sensorless
[19:10:06] <Fisia> hehe
[19:10:13] <fragalot> lol
[19:10:26] <fragalot> fenn: isn't that basically what microstepping is?
[19:10:53] <fenn> fragalot: sorta, but you have this silly step/dir interface to deal with
[19:11:11] <fragalot> I can live with that
[19:11:12] <Fisia> can stepper (big one/strength one) do 50k steps/sec ...(i go check site u reffer to)
[19:11:13] <SWPadnos> and no feedback, and nothing to do with the feedback if you had it
[19:11:31] <SWPadnos> larger steppers have lower top useful speeds
[19:11:41] <fragalot> makes it easier to hook up alternative interfaces,.. like you could hook up a pulse generator and it'd work
[19:11:42] <anonimasu> -_-
[19:12:04] <fragalot> for me, .. I don't care about speed, all i'll use it for is PCB's anyways,.. :p
[19:12:14] <fenn> me, i dont really care about pulse generators
[19:12:24] <Fisia> :)
[19:14:14] <Fisia> btw. other day (at night day) i ask about RTAI that only active when G-Code being execute and CNC really goes running, how about it?
[19:14:22] <Fisia> any comment please...
[19:14:43] <fragalot> rephrase that in a question please
[19:14:54] <anonimasu> I think you need to download and start looking at the sourcecode of emc
[19:15:35] <anonimasu> ^_^
[19:15:38] <Fisia> at that time all GUI-luxury and Multitasking being freeze to image HDD memmory swap,
[19:16:23] <archivist> * archivist frezes, cant digest that sentence
[19:16:47] <Fisia> can RTAI only be activated on 'only certain time' like when G-code being execute
[19:16:49] <Fisia> ?
[19:17:06] <Fisia> other time, RTAI can being turn off
[19:17:11] <anonimasu> did you try emc yet?
[19:17:22] <SWPadnos> it's like that now, except that RTAI is activated when you start EMC
[19:17:28] <Fisia> yes... it is fast already
[19:17:28] <SWPadnos> and deactivated when you close EMC
[19:18:06] <Fisia> i know that just now.
[19:18:13] <Fisia> :)
[19:19:31] <Fisia> (it is like EMC2 reads all my mind, all i ve thinkin just already happen)
[19:22:34] <fragalot> Yes, that is why the economy is shot.
[19:22:41] <Fisia> i opening hall as a home work, stepper is my first learning object... :)
[19:23:10] <Fisia> :)
[19:30:09] <partyzun> hi all
[19:30:24] <Fisia> cool john
[19:30:24] <jepler> hi partyzun
[19:30:37] <Fisia> welcome
[19:30:59] <partyzun> can anyone help ? got an cnc
[19:31:14] <partyzun> a chinese one
[19:32:04] <jepler> if you have some specific questions, we can try to answer them
[19:32:40] <partyzun> ok, sorry here it is
[19:34:08] <anonimasu> anyone knows anthing about scraping?
[19:34:26] <anonimasu> think making scrapers out of a old file would work?
[19:34:30] <anonimasu> err a scraper
[19:35:18] <cradek> anonimasu: I think that's what some people do, yes
[19:35:54] <Fisia> :) ? (i still not understand the question)
[19:36:03] <archivist> anonimasu, yes or braze lump of hss on something
[19:37:02] <partyzun> one of my ex bosses bought a cnc directly from china a cte1225s, couple months later had a car crash and is gone (: now i'm trying to discover the machine
[19:37:09] <cradek> I'd be tempted to try mounting a carbide insert to the end of something long
[19:37:27] <partyzun> i'm asking here because i'm a linux user
[19:37:44] <cradek> sorry to hear about your boss. what is the question?
[19:37:59] <Fisia> :)
[19:38:17] <partyzun> for the moment ain't got no docs and don't know where to start
[19:38:33] <anonimasu> got a name of the machine?
[19:38:45] <anonimasu> I'd call up the representative of that company in your country
[19:38:46] <jepler> is the mill controlled by emc2 now? or it's controlled by something else, and you want to make emc the controller?
[19:38:47] <Fisia> i think u might use ebay for a start.. :)
[19:38:51] <partyzun> yes ofc CTE1225S
[19:38:57] <anonimasu> then call the reseller.
[19:39:29] <archivist> google says its a cnc engraver
[19:40:21] <partyzun> they don't even knoew a few words in english u may check their site, even the manual i found (which is few pages) is in a poor english
[19:41:26] <Fisia> ...
[19:41:33] <archivist> are you thinking of a retrofit
[19:41:36] <jepler> I found this page which implies that it uses a proprietary PCI card to communicate with the machine: http://www.fuzing.com/vli/001995204216/CNC-cutting-and-engraving-machine-CTE1225
[19:41:58] <anonimasu> do ou know anything about cnc machines or g-code or anything?
[19:42:01] <partyzun> anonimasu: http://www.pcut-cn.com/index.php
[19:42:45] <jepler> whatever this pci card is, it's unlikely it works with emc
[19:43:13] <partyzun> i see
[19:43:30] <anonimasu> partyzun: do you know anything about cnc machines?
[19:43:38] <partyzun> haven't tested though on other pc
[19:43:55] <jepler> unless you can obtain programming information for this pci card, and are willing to write the emc driver or pay someone to do it, it won't work
[19:44:20] <anonimasu> partyzun: isnt there already software from the manufacturer to operate the machine?
[19:44:32] <partyzun> now is runin on a dell with win2k and the vga card is defective so the image is a little unclear
[19:45:10] <partyzun> anonimasu: there is just a soft delivered with the machine named creation
[19:45:16] <anonimasu> well, I'd suggest playing around with it, until you get familiar
[19:45:53] <jepler> it says the system uses stepper motors. if the stepper motors use standard "step & direction" signals, then you can throw away the PCI card and use one of emc's stepper interfaces such as the parallel port, pico-systems, or mesa
[19:46:13] <Fisia> u can use EMC2 with mesa
[19:47:50] <Fisia> jepler is right, that would do.
[19:47:55] <archivist> another states 700watts on the axis
[19:48:08] <archivist> could be servo on some models
[19:48:35] <Fisia> still the same dir/step inside
[19:48:51] <archivist> I might be tempted to retrofit
[19:49:09] <Fisia> true..
[19:50:14] <Fisia> congratulation, your fun emc2 adventure just to begin. :)
[19:50:41] <jepler> the first step is to gather much more information about the machine.
[19:51:10] <Fisia> low cost u get
[19:52:37] <Fisia> but it is deppent just like car-modify
[19:53:04] <Fisia> but it would run for sure, once u fix it
[19:58:56] <dmess> anon... what mat'l are you scrapping???
[20:02:55] <Fisia> :)
[20:04:19] <Fisia> I get alot of knowledge in just because 'iam' here... it is true,
[20:04:22] <Fisia> :)
[20:06:27] <Fisia> it is my honour
[20:06:29] <Fisia> :)
[20:07:38] <Fisia> i have to log off now
[20:09:37] <Fisia> 'The stepper controller requires a minimum step pulse width (typically 1µs).' i rellly got almost whole my answer here u see, Alot of thanks to you all
[20:09:50] <Fisia> :)
[20:16:01] <partyzun> anonimasu: cradek jepler thanks guys for your kindness hope to be back with good news
[20:16:29] <anonimasu> partyzun: I'd try the stock program first and see if I could decipher it, then go on as to how to make it run
[20:16:32] <anonimasu> :)
[20:16:41] <alex_joni> have fun partyzun
[20:17:07] <anonimasu> I'd think it's better to make parts ten spending time figuring out how to retrofit it with emc or something else
[20:17:13] <anonimasu> err ten
[20:17:17] <anonimasu> err damn keyb.
[20:17:20] <anonimasu> THEN
[20:17:22] <partyzun> good point i'll try
[20:17:49] <anonimasu> a machine that wont run is a money sink
[20:17:51] <anonimasu> :)
[20:18:50] <partyzun> i woke in the middle of this with no intention but at least want to continue
[20:18:59] <anonimasu> :)
[20:20:40] <partyzun> nice to have you around thanks again and keep in touch
[20:20:51] <anonimasu> good luck!
[20:36:11] <SWPadnos> well damn. lathes.co.uk says this about the HXL "In 1983 the company launched its first CNC machine, the Model HXL; although originally designed as a CNC toolroom lathe it was marketed as a production machine and, with a weak turret and cantankerous GE1050 control system, was destined not to be a success."
[20:36:31] <SWPadnos> weak turret doesn't sound good - I wonder what specifically they mean by that
[20:36:37] <cradek> good question
[20:37:01] <SWPadnos> it looks weak in that it only holds 4 tools, and it has 100 bolts you have to tighten (so repeatability/alignment may suck)
[20:37:22] <SWPadnos> if it's weak as in "you can't take heavy cuts", that's a totally different problem
[20:37:22] <archivist> perhaps written by a machine abuser
[20:37:40] <SWPadnos> lathes.co.uk is ueually pretty good (I think)
[20:37:40] <SWPadnos> usually
[20:37:43] <SWPadnos> http://www.lathes.co.uk/cataract/page23.html
[20:37:48] <cradek> surely the bottom of the turret will be a certain distance below the spindle centerline, so tools of a certain height are automatically aligned
[20:38:00] <cradek> "Page Title" heh
[20:38:09] <SWPadnos> yeah, a lot of them are liek that ;)
[20:38:11] <SWPadnos> like
[20:38:14] <archivist> he is not far from us
[20:38:15] <dmess> AHHHH you posted those letters "GE1050" FLASHBACK... :0
[20:38:32] <SWPadnos> dmess, do you know about the Hardinge HXL?
[20:38:53] <archivist> have met the person who runs the site
[20:39:53] <SWPadnos> the other practical problem is that I only have an 83" high garage door ;)
[20:40:20] <archivist> large hammer should sort the door
[20:41:13] <SWPadnos> or a couple of small ones, then some replacement nails
[20:41:56] <archivist> thinking about light cuts, was getting belt slip here today :) and only 30thou wide cut 4 thou feed
[20:42:14] <SWPadnos> in Ti?
[20:42:46] <archivist> no freecutting steel
[20:42:53] <anonimasu> haha
[20:43:37] <archivist> old machines just not built for a good meaty cut
[20:47:09] <anonimasu> lies
[20:48:13] <archivist> well depends on speeds and feeds and material, have had swarf over the tailstock on the same machine
[21:15:37] <alex_joni> SWPadnos: "A rocket engine is a jet engine[1] that uses only propellant mass for forming its high speed propulsive jet. " [wikipedia]
[21:15:48] <SWPadnos> heh, thanks
[21:18:56] <fragalot> * fragalot wants a solar hydrocar toy
[21:19:00] <fragalot> nearly Xmas :o
[21:21:54] <alex_joni> SWPadnos: it might also be a starting system..
[21:22:11] <SWPadnos> hmmm. could be
[21:22:15] <alex_joni> for jet engines, rockets you just light up
[21:22:48] <SWPadnos> turbine jet engines need a starter. I'm not sure about other types
[21:23:39] <alex_joni> I think they all do
[21:23:57] <skunkworks_> match?
[21:26:33] <SWPadnos> ah - an intake
[21:26:46] <SWPadnos> for the liquid (also includes gases) propellant
[21:27:34] <SWPadnos> so my answer was nearly correct - the propellant is carried by a rocket, but it's not for any other type of jet
[21:28:49] <fragalot> jet engines are lovely
[21:28:54] <fragalot> some types don't even have any moving parts
[21:36:34] <skunkworks_> SWPadnos: isn't this something your mom would know?
[21:39:02] <anonimasu> :D
[21:40:03] <fragalot> http://pixdaus.com/pics/1226001873yuF5iwJ.jpg photoshop CS5 Beta
[21:41:20] <anonimasu> lol
[21:53:51] <SWPadnos> skunkworks_, no, she wouldn't know anything that has to do wit (gasp) engineering!
[21:55:17] <jymm> SWPadnos: http://cuzyerdumb.com/
[21:55:26] <dmess> my grandma is a wiz in engineering... you should see her knit socks.... shes a machine
[21:55:39] <jymm> isn't that manufacturing
[21:55:53] <markdashabout> Got a question about EMC2
[21:56:03] <jymm> (or slave labor, one of the two)
[21:56:04] <dmess> she had to figure out the needle things
[21:56:16] <archivist> engineering is the sock coming out the right size
[21:56:19] <SWPadnos> markdashabout, woohoo! an EMC2 question. shoot :)
[21:56:35] <jymm> http://emc.com/2
[21:56:52] <jymm> *shrug*
[21:57:00] <markdashabout> I am trying to run it with my sherline mill and Ive even seen people suggest that there is a shereline option to select from the config screen
[21:57:06] <markdashabout> but I dont have it.
[21:57:15] <markdashabout> Is there? And if not, how can there not be?
[21:57:19] <dmess> it is her labour of love.. at 88 yrs old she still makes 4 pair of new wool socks every fall / thanksgiving
[21:57:35] <archivist> markdashabout, its in the stepconf program
[21:57:42] <cradek> in stepconf, there are sherline presets to make it easy to build a configuration for a sherline
[21:57:51] <cradek> I have heard that one or more of the timings is wrong, however
[21:57:53] <SWPadnos> well, the sherline button in stepconf (the config program) only sets the step pulse timing and which pins are step/direction
[21:58:10] <archivist> markdashabout, but I saw a comment about a bug with it anyway
[21:58:13] <cradek> for the last part of your question: there could not be, since sherline has not given us this information.
[21:58:24] <markdashabout> Yes, I heard that, too and that doesnt seem like a problem.
[21:58:29] <SWPadnos> it does not make a complete config for your sherline - you still need to know things like pulley ratios, screw pitch, travel limits, etc.
[21:59:20] <SWPadnos> markdashabout, first off, are you talking about stepconf when you say
[21:59:26] <SWPadnos> "the config screen" ?
[22:00:06] <markdashabout> No, I fired up emc2 with the emc command and have a list of (I think) .ini files ona gui.
[22:00:39] <markdashabout> There are thousands of these tiny table top mills around I suppose and it seems odd that I should need to configure?
[22:01:02] <archivist> all different
[22:01:18] <markdashabout> I am nervous that if I run stepconf it will not be as good as my pre-configured emc1 configuration.
[22:01:30] <cradek> markdashabout: I agree, it would be nice if sherline would contribute configurations for the machines they are currently selling; we would be happy to host and distribute them with emc2 releases.
[22:02:00] <cradek> if you recently bought a new sherline machine, you are in a position to let them know that it is important to you.
[22:02:18] <markdashabout> Would it be possible to parse info out of an emc1 config or is it very very different?
[22:02:39] <cradek> it is quite different
[22:03:07] <markdashabout> Yeah, thought it would be, wouldnt be much point in writing emc2 if it was basicically the same now would there !!!
[22:03:14] <SWPadnos> heh
[22:03:15] <cradek> we (like you) are disappointed that they ship such very old EMC software
[22:03:28] <cradek> but of course, that is up to them.
[22:04:09] <markdashabout> It must be said that the EMC1 machine works very well, what really hacks me off is the linux is so old that even getting a network to work is a challenge.
[22:05:05] <SWPadnos> I'm not sure if there's still any development going on for that version of EMC
[22:05:27] <cradek> I have not seen any for a few years? now
[22:05:30] <SWPadnos> there were a couple of developers who were maintaining it, but I don't know their current level of involvement
[22:05:33] <markdashabout> But I am sure I have seen people say just select it from the menu, I even got the name of the file Sherlinemill.ini, but of course anyone could make that up.
[22:05:34] <cradek> but EMC2 is very active
[22:05:43] <markdashabout> No , no development. I want emc2 anyway!
[22:06:02] <cradek> go ahead and work through the stepconf wizard, and let us know if you have any questions.
[22:06:14] <SWPadnos> huh, there it is: http://www.linuxcnc.org/docs/2.1/html/common/emc2_introduction/emc2_introduction.html
[22:06:23] <SWPadnos> (references to Sherlinemill.ini)
[22:06:24] <markdashabout> Alright I'll do that, if thats what people do.
[22:06:59] <jepler> the sherline timing preset is wrong in the current version of emc 2.2
[22:06:59] <SWPadnos> but it doesn't say they exist, it only says that they're more descriptive names than "generic"
[22:07:19] <jepler> in stepconf, that is
[22:07:20] <markdashabout> Ah yes.
[22:08:10] <markdashabout> But its really odd. SL ought to help us and the emc people here, after all they do make a lot of use of emc1
[22:08:39] <markdashabout> There are a lot of these inexpensive mills around after all, and at least some of them must be the same???!!
[22:08:51] <SWPadnos> they should probably start using EMC2
[22:09:01] <SWPadnos> nope, they're almost all different in the little details
[22:09:04] <jepler> some user reported that by switching the "step length" and "step space" values shown he got a working configuration, but I'm not sure that matches the step timings emc1 gave for sherline
[22:09:11] <alex_joni> mine is red :D
[22:09:16] <SWPadnos> some use Xylotex drives, otehrs Geckos, others custom (like Sherline)
[22:09:34] <SWPadnos> the travel limits are different, some people have home switches and others don't ...
[22:09:34] <markdashabout> Well of course if nobody does then that would explain it.
[22:09:52] <markdashabout> I'm just going to look at the url you sent... (pause)...
[22:10:05] <SWPadnos> don't bother
[22:10:08] <cradek> no don't, that's extremely outdated
[22:10:10] <SWPadnos> it's an old manual for EMC2 2.1
[22:10:19] <cradek> just work through the stepconf wizard.
[22:10:21] <jepler> SWPadnos: is that doc reference still in the current docs? if so, I'll drop bigjohnt a message to fix it
[22:10:39] <SWPadnos> google didn't find it in later versions (at least not a tthe top of the list)
[22:10:59] <cradek> markdashabout: sherline used to be very active in the emc community; recently, they aren't.
[22:11:19] <archivist> redirect the web page to current docs and make old docs download only
[22:11:50] <SWPadnos> archivist, well, that's one way, but people using 2.1 should still have docs available for that version
[22:12:09] <archivist> thats why I say download
[22:12:23] <markdashabout> Trouble is, do you think I'll get a workable config out of stepconfig. I mean one that makes circles circular and the right size??
[22:12:40] <cradek> markdashabout: if you answer the questions correctly, yes
[22:12:41] <SWPadnos> yes, if you know what your machine is
[22:12:57] <cradek> markdashabout: if you get it wrong, then you fix it
[22:13:43] <SWPadnos> as long as you don't edit something manually, you can load the config back into stepconf and fix things
[22:13:56] <archivist> loop a couple or three times, save setups as different versions if you wish
[22:14:05] <markdashabout> So if I try step conf, do you think I'll get a machine that produces repeatable results,
[22:14:23] <SWPadnos> yes
[22:14:27] <alex_joni> sure
[22:14:31] <cradek> I already answered that: 'if you answer the questions correctly, yes'
[22:14:37] <alex_joni> you'll get repeatable results with either config :D
[22:14:38] <SWPadnos> again, that's only if you can answer the questions correctly
[22:14:40] <SWPadnos> heh
[22:14:43] <archivist> I used stepconf for my machine
[22:14:51] <alex_joni> if you answer the questions correctly you'll get good repeatable results
[22:15:31] <SWPadnos> but, there is a problem with the Sherline timings - hold on a sec
[22:15:45] <markdashabout> Sorry asked same question twice thought client had crashed!!!
[22:15:50] <markdashabout> Ok will go and try it!
[22:15:57] <markdashabout> Thanks for your help guys!
[22:16:14] <markdashabout> Ok will hold ok !!
[22:16:16] <cradek> welcome, hope you get it going.
[22:18:11] <alex_joni> markdashabout: you don't need to close the IRC chat just to configure emc2 :)
[22:18:21] <alex_joni> (unless they are on different PC's)
[22:18:34] <SWPadnos> hmmm - I don't see the fix I'm looking for
[22:18:46] <markdashabout> Ok, I was waiting to see what SWPadnos was going to say, he said hold on so I'm holding....
[22:18:48] <jepler> markdashabout: here's the person who says the stepconf timings are wrong and gives a fix: http://www.cnczone.com/forums/showpost.php?p=432441&postcount=10
[22:19:15] <markdashabout> It is on a different machine but one of the joys of modern X is I can have the display here and run it there...
[22:19:20] <jepler> but the bottom line is when you choose the "sherline" driver timings line, you get wrong numbers that may not work reliably
[22:19:33] <jepler> (this is my fault, I wrote stepconf and didn't find the right numbers to put in)
[22:20:02] <SWPadnos> oh hey - that's the information I was looking for (but I thought it was changed in stepconf)
[22:20:10] <jepler> SWPadnos: it's been changed but not in a released version
[22:20:20] <alex_joni> jepler: not your fault that the timing are obscure/hidden
[22:20:20] <SWPadnos> I was looking at commit messages
[22:20:20] <markdashabout> Oh, wow, well I am honoured to be talking to the king!
[22:20:34] <jepler> markdashabout: no kings here -- just a bunch of citizens
[22:20:38] <alex_joni> markdashabout: all kings around here, and a couple of queens like SWPadnos
[22:20:40] <markdashabout> Haha!
[22:20:47] <SWPadnos> smoooooch! :)
[22:21:01] <alex_joni> * alex_joni blushes
[22:21:02] <SWPadnos> I like Queen
[22:21:03] <jymm> well, that queen eats lie a pig!
[22:21:08] <jepler> alex_joni: uhhhhh maybe there's something you need to know about Steve...
[22:21:08] <SWPadnos> heh
[22:21:10] <jymm> like
[22:21:21] <alex_joni> jepler: I prefer not knowing :P
[22:21:28] <SWPadnos> it's the long hair - gets 'em every time
[22:21:40] <alex_joni> markdashabout: we also are a really happy/crazy bunch in here
[22:21:41] <jymm> proof that alex_joni is in de nile
[22:21:45] <alex_joni> don't take it all too serious
[22:21:55] <archivist> nothing wrong with being a hairy
[22:22:08] <alex_joni> jymm: your jokes would be lots more appreciated if you would spell them right :P
[22:22:24] <jymm> alex_joni: Learn to speak typo =)
[22:22:29] <SWPadnos> I think that was a double joke
[22:22:31] <archivist> tyop
[22:22:33] <SWPadnos> on jymmm
[22:23:22] <jymm> it's all good, I just want to get out of here and start the weekend
[22:23:43] <jymm> dont care about anything else
[22:23:56] <archivist> make something
[22:24:17] <jepler> ooh, time to get out of the office
[22:24:21] <jepler> see all you guys around
[22:24:27] <jymm> hasta jepler
[22:25:02] <alex_joni> see you jymm
[22:25:10] <alex_joni> and jeff ;)
[22:25:22] <jymm> =)
[22:25:26] <archivist> for me its time to stay at work and do MY jobs
[22:25:56] <alex_joni> archivist: should be 8pm there.. ?
[22:26:15] <archivist> 10:30 pm
[22:26:24] <alex_joni> err. sorry
[22:26:33] <alex_joni> my math is terrible.. 00:30 here
[22:27:42] <archivist> might get another 1/2" of steel converted to swarf
[23:08:17] <markdashabout> Haha!! I missed all that looking at stepconf. Seems like it knows quite a lot about sherline...
[23:08:42] <markdashabout> I have swaped the two parameters round in the file it generated...
[23:09:13] <markdashabout> Am trying to test but get messages like
[23:09:21] <markdashabout> joint 0 on limit switch error
[23:09:34] <markdashabout> Of course, there are no limit switches!!
[23:16:10] <alex_joni> markdashabout: rerun stepconf, and disable the limit switches
[23:16:12] <jepler> markdashabout: on the pinout page, change all the inputs to "not connected" or "unused" or whatever it says
[23:16:40] <alex_joni> * alex_joni is off to bed
[23:16:44] <alex_joni> good night all
[23:51:15] <markdashabout> Thanks jepler, yes thats what I figured and that overcame that problem. But current problem is to try to figure out what the pralell pin connections actuallya
[23:51:52] <markdashabout> actually are, because the two pre-config options (sherline, xysomething) neither seem to work.
[23:55:13] <markdashabout> (There is info for a hand-controller on the sherline page that gives connections same as non-sherline connections on stepconf)
[23:56:34] <JymmmEMC> alex_joni: Yep you're right, your math is really off, it's 2:30pm, not 00:30 =) 02:26:32 PM) alex_joni: my math is terrible.. 00:30 here