Back
[00:44:30] <tjb1> Oh big surprise…USPS has changed the expected delivery date 3 times on this item...
[01:52:05] <Loetmichel> mornin'
[01:55:36] <Jymmm> Hi Loetmichel
[01:57:37] <Loetmichel> ... it seems i am getting old and brittle... somwhere in the 21hr shift wenedsday i must have dislocated a bone in the spine... have the feeling of someone putting a knife in my back every time i make wrong move or cough/laugh :-(
[01:58:59] <Jymmm> where in your spine?
[01:59:53] <Jymmm> mid back? Waist level? Below waist? between shoulder blades? Near the base of your neck?
[02:00:34] <Jymmm> just above waist level?
[02:01:17] <Loetmichel> mid back, about the "short riub"
[02:01:19] <Loetmichel> -u
[02:02:13] <Jymmm> Kind of a slow but strong sensation/throbe?
[02:02:41] <Jymmm> or a sharp jab?
[02:02:48] <DJ9DJ> moin
[02:02:54] <Jymmm> hi DJ9DJ
[02:03:00] <DJ9DJ> hi Jymmm :)
[02:07:38] <Jymmm> Loetmichel: make sense?
[02:11:43] <DJ9DJ> morning Loetmichel :)
[02:22:46] <DJ9DJ> re
[02:25:55] <Loetmichel> Jymmm: the usual sharp jab, dislocated rib/spine bone feeling.... i know that, bus usually on the other side of the spine... seems my vertebra are a bit wak there
[02:26:50] <Loetmichel> nothing dangerous, just annoying... especially beccause i have to pack 6 oeuro pallets with cvomputers today ;-)
[02:27:04] <Loetmichel> i know it woll fade over the next few days
[02:27:07] <Loetmichel> wqill
[02:27:10] <Loetmichel> grr
[02:27:11] <Loetmichel> will
[02:31:43] <Jymmm> Loetmichel: Well, if you start to feel any tingling or numbness in your fingers/toes, get it checked out asap.
[02:32:10] <Loetmichel> Jymmm: i know, will do ;-)
[02:32:18] <Jymmm> Loetmichel: You might try acupuncture too,
[02:32:54] <Loetmichel> or i might just ignore it until it snaps back, like every time
[02:33:48] * DJ9DJ kicks Loetmichel in his back. better now? ;)
[02:34:04] <Loetmichel> DJ9DJ: wrong side :-(/
[02:34:11] <DJ9DJ> oh, sorry ;)
[02:34:22] <Loetmichel> :-)
[04:03:31] <Tugge> I tried install the latest cnclinux for my computer. for some reason ubuntu installer doesn't find any hard drives from my computer.
[04:04:08] <Tugge> I have windows 7 intalled that computer already and I removed 30 GB from windows partition to ubuntu.
[04:04:41] <Tugge> But still installer doesn't even see any hard drives. So I need to install second hard drive just for linux?
[05:29:18] <psha> Tugge: to be brief - no
[05:37:50] <Tugge> So is there a possibility, that in this ubuntu installer there is no drivers for my sata drive?
[05:38:29] <Tugge> If so, how do I install the drivers. My motherboard is quite new and there is no IDE-connectors at all.
[05:50:27] <jthornton> Tugge, most of my machines have sata drives
[06:17:55] <Mamou-Mani> Hello
[06:21:01] <micges> hi
[06:22:45] <Mamou-Mani> I am trying to send a gcode to a reprap printer, I ve been looking at the pronterface code but cannot find the line that does that, was hoping you can help:
https://github.com/kliment/Printrun/blob/master/pronterface.py
[06:25:36] <dhoovie> Mamou-Mani: probably better to ask in #reprap
[06:26:06] <Mamou-Mani> dhoovie: they've redirected me here :)
[06:26:30] <dhoovie> ohhh.... is your reprap controlled via linuxcnc?
[06:26:33] <jthornton> does LinuxCNC control your machine
[06:27:53] <jthornton> seems you want to find these folks This file is part of the Printrun suite.
[06:35:37] <Mamou-Mani> Jthornton: Thx I will write to Kliment, I am not using LinuxCNC although I am quite curious about it now!
[06:55:23] <dhoovie> Mamou-Mani: kliment is usually on the #reprap channel :)
[07:55:01] <skunkworks> oktoberfest starts today...
[07:57:44] <r00t4rd3d> it should be called "hot girls in skirts with large teets chugging giant beers day"
[07:59:34] <skunkworks> not from what I remember...
[07:59:49] <r00t4rd3d> well you were at the wrong one then
[08:00:45] <jdh> they don't have to be large, just well displayed.
[08:01:34] <jdh> go measure your salinity, set it to that.
[08:02:46] <r00t4rd3d> what does my salinity have to do with it?
[08:03:54] <r00t4rd3d> or was that some misplaced fish tank fodder?
[08:07:52] <skunkworks> http://www.youtube.com/watch?v=CBHdNUQnArY
[08:16:38] <skunkworks> http://www.youtube.com/watch?v=CVlad7l9HvI
[08:16:48] <skunkworks> pretty neat. I guess I never ran across that.
[08:19:06] <jdh> misplaced diving fodder. computer has an option for setting water density via salinity.
[10:46:18] <Loetmichel> re @ home
[10:49:50] <pcw_home> Must be fall, there are turkeys in the backyard...
[10:51:18] <um_> hi, i ve been using linuxCNC for a couple of months without any issues. However today, after working flawlessly all this time, my cnc machine started cutting deeper than expected. On testing... it seems to be travelling .05 mm more for every 0.1 mm
[10:51:27] <um_> i havent changed any stepconf settings
[10:51:45] <um_> can anyone point towards where i can start debugging this
[10:52:12] <um_> x and y axis travels havent changed... only z axis travel has changed
[10:52:12] <mazafaka> Hello again :) Let me ask here. I am usin end mills on our 2.5D milling machine, and end mills are made of HSS. To mill construction steel with 0.05-0.40 % of carbon? what cutting speed, and therefore RPM (for 25 mm in diameter end mill) can I apply&
[10:54:16] <mazafaka> It works fine with 60-90 RPM for 25 mm end mill, but I want to use 250 RPM for 25 mm end mill made of HSS, and use pretty fast feedrade. The question is, does such cutting speed rounds the cutting edges of mill too quickly?
[11:07:29] <pcw_home> mazafaka: isn't that what feed&speed charts are for?
[11:08:21] <mazafaka> i don't have one, and we have a coordinate-drilling machine, it stops with drive errors if RPM is low like 60 RPM for 25 mm end mill
[11:09:14] <mazafaka> I want to know from your experience if higher RPMs wear off the mill bit too soon or not
[11:12:30] <mazafaka> pcw_home: ?
[11:18:37] <pcw_home> I dont have any experience but there are charts available that suggest 50 -120 SFM which is more like 200-400 RPM
[11:18:39] <pcw_home> (but Im just guessing, you need someone with some experience)
[11:21:24] <pcw_home> I just googled end mill feed and speed and there a re quite a few charts available if you weed through them
[11:27:36] <pcw_home> if you find the SFM and depth of cut for your steel and HSS endmill you can use
[11:27:38] <pcw_home> http://www.custompartnet.com/calculator/milling-speed-and-feed to get RPM and feed rate
[11:27:39] <pcw_home> (I'm sure there are many more or you can use your own calculator/spreadsheet)
[11:44:31] <mazafaka> pcw_home: thanks, it's something with what to start? i am not quite familiar with finding such sort of information in seconds. Although, our machine are made of electric schemes, and it simply stops. I had to use relatively high feedrate to work continously, but for HSS, high values of RPM are a bit wrong idea...
[11:45:03] <mazafaka> I meant is does is something with what to start! :)
[12:17:43] <IchGuckLive> hi all B)
[12:17:49] <Tugge> I did find the solution from my ubuntu problem. I needed to change my SATA hard drive to IDE mode from BIOS. It was in AHCI mode.
[12:18:28] <Tugge> So if someone is strugling with cnc ubuntu installation with newer hard ware, that might be one of the issues.
[12:22:00] <IchGuckLive> Tugge: why new hardware ?
[12:22:36] <L84Supper> Tugge, which maniboard and what BIOS?
[12:22:51] <L84Supper> might be a buggy BIOS
[12:29:49] <IchGuckLive> PC 3.2GB 120HDD 2GB Ram 40Euro with 17" 60Eur and good
[12:48:46] <Tugge> Motherboard was asrocks z77 Extreme 4
[12:49:24] <Tugge> I don't want to use my really old laptop for controlling my mill.
[12:49:35] <IchGuckLive> http://www.asrock.com/mb/Intel/Z77%20Extreme4/
[12:49:48] <Tugge> Yeah that's the one.
[12:49:53] <IchGuckLive> why not a D525 insted
[12:50:47] <Tugge> I bought this computer last spring.
[12:50:56] <Tugge> This seemd nice :D
[12:51:23] <IchGuckLive> i guess you winn not be amused on this system with the ubuntu stuff
[12:52:07] <IchGuckLive> a game PC is not good for Realtime stuff
[12:52:34] <IchGuckLive> Tugge: where are you from Country ?
[12:55:18] <Loetmichel> IchGuckLive: why do you think?
[12:55:55] <IchGuckLive> so many unsuportet fetchers at the mainboard
[12:57:01] <pcw_home> so the real question is how does the latency test look on that MB?
[12:59:53] <Tugge> IchGuckLive: Iäm from Finland
[13:00:01] <IchGuckLive> im off to the planetarium ,C14 in place,By clear sky but full moon O.O
[13:00:15] <IchGuckLive> Tugge: im from Germany
[13:00:35] <IchGuckLive> By for today
[13:00:38] <Tugge> Bye
[13:01:40] <pcw_home> Tugge: have you tried running the latency test?
[13:10:16] <Tugge> Not yet. I needed to install PCI-e to parallel port card and now I can't find the drivers for it.
[13:10:38] <Tugge> Also I'm getting a little drunk so I won't try it anymore :D
[13:16:25] <pcw_home> no drivers are needed for parallel ports for LinuxCNC
[13:21:33] <pcw_home> LInuxCNC has it own real time driver
[13:32:53] <Tugge> So LinuxCNC should find the parallel port which is connected via PCI-express card?
[13:36:01] <pcw_home> Yes, though you may need to tell LinuxCNC the base address of the card (obtained via lspci)
[14:34:14] <Tugge> Okay I just tested the jitter. It was around 8000
[14:34:30] <andypugh> That's very good.
[14:34:53] <andypugh> did you abuse the machine at the same time
[14:35:15] <Tugge> Yes. I used fire fox and downloaded putty to get here :)
[14:35:30] <andypugh> Sounds promising.
[14:36:13] <Tugge> Now I just need to config the motors and stuff and then it my mill should be ready to do some work.
[14:36:17] <pcw_home> running a bunch of glxgears and flash videos is standard abuse
[14:36:55] <Tugge> Okay. I set the jitter to 15 000 so there is quite much buffer.
[14:36:58] <andypugh> "Oi! Computer! You're ugly!"
[14:37:21] <pcw_home> launching big programs is something to try as well
[14:37:55] <andypugh> Tugge: You found Stepconf?
[14:38:56] <Tugge> Yeah. I also did find that the stepconf rewrites the .hal file.
[14:39:12] <Tugge> And I needed to modify the .hal file for parallel port configuration.
[14:39:33] <andypugh> You can re-modify in stepconf.
[14:40:15] <andypugh> You do need to hand-edit the HAL for some things, but parallel port configuratio is not one of them.
[14:41:34] <andypugh> Once you have started to hand-edit, though, it's best to stop using Stepconf, or use it on dummy configs then copy/paste the code sections across.
[14:42:29] <andypugh> Right, how much red wine do I need?
http://xkcd.com/323/
[14:42:49] <Tugge> I needed to do this: [#linuxcnc] I needed to do this:
[14:43:12] <Tugge> Hmm.. Wheres the link :D
[14:43:23] <Tugge> http://www.linuxcnc.org/index.php/english/component/kunena/?func=view&catid=18&id=18917
[14:44:00] <andypugh> I am pretty sure that is possible in Stepconf.
[14:44:38] <andypugh> <starts up the virtual machines>
[14:45:26] <andypugh> Yes, the very first page lets you choose up to three parports and set the direction to IN or OUT
[14:46:17] <andypugh> The only thing you can't do there is set the secret, slightly dangerous, "X" mode :-)
[14:52:33] <mazafaka> Can steel with 1% of Mn round the cutting edges of HSS end mill more than stainless steel would do?
[14:53:09] <andypugh> Possibly, Manganese steel is tough stuff.
[14:53:59] <andypugh> www.xkcd.com/1114
[14:56:47] <mazafaka> end mill bends off the material and adds about 0.5 mm to the desired size at one side which is being milled
[14:56:57] <andypugh> Has anyone seen this? EMCw2 running in a little box on the end of a USB cable:
http://www.ecklersoft.com
[14:58:11] <mazafaka> What does '-w2' mean?
[14:58:11] <andypugh> Oh my! A MacOS version!
[14:58:36] <andypugh> It means my fingers are too fat for the keyboard.
[14:58:52] <mazafaka> Oh///
[14:59:22] <L33TG33KG34R> so I am looking at limit switches
[14:59:24] <pcw_home> Yeah there are a number of things like that around (a bit of the gcode interpreter ported from LinuxCNC to the uProc)
[14:59:44] <L33TG33KG34R> people have begun using hall effect sensors as switches. is there any advantage to this?
[15:01:57] <andypugh> This one is fairly cheap ($79) and supports MacOS though. And I just got a 1-year backdated payrise + pension bribe. I see an Ultimaker in my future, jooked up to the Mac. :-)
[15:02:37] <andypugh> L33TG33KG34R: The main advantage is that the magnets look very pretty with swarf stuck to them.
[15:02:52] <L33TG33KG34R> swarf?
[15:02:56] <L33TG33KG34R> what that?
[15:03:19] <andypugh> Anti-components in small pieces?
[15:03:51] <L33TG33KG34R> what?
[15:04:29] <andypugh> http://en.wiktionary.org/wiki/swarf
[15:04:44] <L33TG33KG34R> oh...
[15:04:45] <L33TG33KG34R> nvm
[15:04:46] <L33TG33KG34R> k
[15:04:51] <L33TG33KG34R> physical switches it is
[15:05:28] <andypugh> L33TG33KG34R: Well, I am using optical sensors on my machine
[15:06:05] <L33TG33KG34R> and what advantage does that offer as opposed to mechanical variety?
[15:07:12] <andypugh> Possibly more repeatable. (probably not). No moving parts. On balance I might not use them again.
[15:08:38] <andypugh> These particular ones:
http://docs-europe.electrocomponents.com/webdocs/02af/0900766b802aff18.pdf
[15:09:31] <L33TG33KG34R> hmm
[15:09:42] <L33TG33KG34R> I think I'll stick with the mechanical variety
[15:09:54] <L33TG33KG34R> I am still debating whether to go servo or not with my machine
[15:09:58] <andypugh> I neither advocate nor disparage the option, I merely mention it.
[15:14:23] <pcw_home> I think the issue with moving part of LinuxCNC to an off board engine is that if you need any real time features they also need to be ported to the off board engine (probing, threading, rigid tapping etc) so you end up eventually porting all of LinuxCNC (and needed enough horsepower for all) so at the end you are pretty much where you started. (except the added need to keep LinuxCNC current...
[15:14:25] <pcw_home> ...on the embedded hardware
[15:14:27] <pcw_home> But if the real time limitations are OK theres a large subset of machines that would work fine with buffered controller and it would be nice if LinuxCNC had some built in support/ canonical API for such devices
[15:16:02] <andypugh> I sort of agree, but at the same time one advantage of FOSS is that separate projects can use parts of the main project, and get updates for free, while remaining separate.
[15:16:25] <andypugh> (and not distracting the main developer)
[15:17:41] <andypugh> I can't decide if my metaphor is inspired or stupid:
http://www.linuxcnc.org/index.php/english/component/kunena/?func=view&catid=10&id=23995&limit=6&start=12#24701
[15:20:20] <pcw_home> I think the different accelerations are pretty confusing...
[15:20:49] <andypugh> In the INI you mean?
[15:20:56] <pcw_home> And I wish the hardware stepgen work work with the 0 maxaccel option
[15:21:04] <pcw_home> would work
[15:21:07] <andypugh> I have no real idea why the stepgens need any physical limits.
[15:22:12] <andypugh> You mean the hm2_stepgen? I have never really paid much attention to what the problem is.
[15:22:25] <pcw_home> It has to do with imperfections in the stepgen driver
[15:22:36] <andypugh> (seb)
[15:23:21] <andypugh> Is that "imperfections" spelt B-U-G-S ?
[15:23:34] <pcw_home> well not even really Seb, the software stepgen was forklift ported to the hardware but its not exactly the same
[15:24:18] <andypugh> I would look at it, but I can't even figure out how to write a printf..
[15:25:17] <andypugh> (Currently mine breaks badly with negative numbers)
[15:25:24] <pcw_home> Ha
[15:32:44] <andypugh> Here's something. When did anyone last have a message from the Mailing Lists?
[15:33:11] <Tugge> Hmm. Intresting stuff. I did configure the machine and everything else seems to be working except X-direction.
[15:33:36] <andypugh> Tugge: It always moves in one direction?
[15:33:45] <Tugge> Yes.
[15:34:14] <Tugge> So I hooked up my multimeter to that corresponding pin and it only gives 0.0 V and 1.3 volts.
[15:34:31] <andypugh> You can get that if "step" and "dir" or "enable" are swapped
[15:34:35] <Tugge> And in Y direction pin the voltages are 0.0 and 3.3 volts.
[15:34:51] <cradek> andypugh: ~ 3h ago
[15:35:09] <andypugh> cradek: Hmm. Subject?
[15:35:27] <Tugge> So there is something starge going on. I think this can't be driver issue. Maybe that hardware is flawed.
[15:35:31] <cradek> From emc-users-bounces@lists.sourceforge.net Fri Sep 28 12:13:42 2012
[15:35:32] <cradek> Subject: Re: [Emc-users] Helical interpolation not quite round, WTF?
[15:36:06] <andypugh> Ah. Yes. I suspect my memory isn't what it was, but can't recall what it was.
[15:36:40] <andypugh> But that's the only one since yesterday.
[15:38:20] <cradek> there was one to -commit and two to -users in the last 24h
[15:38:46] <cradek> and a whole bunch in the 24h preceding
[15:39:29] <andypugh> Seems atypical, but may be true
[15:43:51] <andypugh> I wonder what the payback is for forum spammers? As far as I can see _all_ they do is waste our time when we reject them. Even if they slip through, all they get is a URL in their profile that no-one ever bothers to look at.
[15:44:47] <andypugh> Someone (who I would happily break all the fingers of) created 370 bogus identities in 10 minutes last night.
[15:45:06] <cradek> ugh.
[15:45:20] <cradek> I sure appreciate the work you guys do on the forum.
[15:46:18] <andypugh> They were all gmail, and we used to block gmail, but since I persuaded JT to let them in again we have had a dozen or so real folk. (and I am a gmailer myself)
[15:46:56] <cradek> isn't gmail the most or nearly the most common email provider now?
[15:47:44] <andypugh> So I think the balance is in favour of allowing gmail in. A warning in the signup page "gmail is blocked, please obfuscate thusly" would have sufficed, but neither of us can edit the web content.
[15:48:17] <cradek> that kind of IQ test always seems to filter out a lot of people :-/
[15:48:29] <andypugh> Yes, it probably is. (it's really very good). But it is also very easy to create a wothless acciunt in it.
[15:48:46] <cradek> I had a hell of a time editing the web content last time I tried. I think if I turned off the special editor it worked.
[15:49:55] <andypugh> It's an admin rights thing, JT and I are not enabled.
[15:50:00] <cradek> ah
[15:50:30] <DJ9DJ> gn8
[15:50:51] <andypugh> I think I have my printf (floats in kernel space) working now. In fact I think it improves on the normal printf in one way that might even be worth trying to mainline.
[15:51:22] <cradek> writing printf is something I'm sure I'm not smart enough to do
[15:51:31] <andypugh> I can use &06,5f as a format, and get a comma as the decimal separator.
[15:51:49] * Jymmm lol @ cradek =)
[15:51:49] <andypugh> (sorry %06,5f )
[15:52:11] <cradek> it's not clear to me that that is a feature
[15:52:38] <andypugh> Because you are American. About 40% of the world write decimals that way
[15:53:13] <andypugh> (I made that number up :-)
[15:53:18] <cradek> yes I know this, and they shouldn't have to change their source code and recompile to get it
[15:54:09] <andypugh> I have never tried (we use . in the UK) but it might happen automatically with localisation?
[15:54:32] <cradek> yes it certainly does
[15:54:55] <andypugh> In plain old printf?
[15:54:57] <Jymmm> My HP 20S has the option to flip ./,
[15:55:17] <cradek> http://pubs.opengroup.org/onlinepubs/009695399/basedefs/xbd_chap08.html#tag_08
[15:55:21] <cradek> see LC_NUMERIC
[15:56:09] <andypugh> Aha!
[15:56:52] <andypugh> So, the programmer uses %3.3 and printf creates 123,456 ?
[15:56:53] <cradek> All forms of the fprintf() functions allow for the insertion of a language-dependent radix character in the output string. The radix character is defined in the program's locale (category LC_NUMERIC ). In the POSIX locale, or in a locale where the radix character is not defined, the radix character shall default to a period ( '.' ).
[15:57:26] <Jymmm> Hey, in a calculator, what is "10 superscript x" as opposed to y^x?
[15:57:48] <andypugh> Jymmm: It save you typing a 10
[15:58:09] <Jymmm> Hmmm
[15:58:24] <andypugh> No use to you and your antiquated units, but handy for decimalised countries
[15:58:48] <Jymmm> I thought it might have helped with 10^(x/10)
[15:59:33] <cradek> it's useful as the inverse of log base 10 which also might be a special key
[15:59:45] <andypugh> it would. Type x/10={magic button}
[16:00:14] <andypugh> Good point cradek. It's also antilog
[16:00:53] <andypugh> (bear in mind that until last week Jymmm had never heard of arctan)
[16:01:44] <cradek> meh, a lot of people are a bit short on math education, no shame
[16:01:52] <andypugh> cradek: So, can you confirm that comma-centric coders use a dot in the format specifier, and their users get a comma?
[16:02:32] <cradek> yes coders don't have to worry about it, and always use dots. users can get their output localized by setting their environment properly.
[16:02:53] <Jymmm> andypugh: No, I heard of it, just like I've heard about bigfoot =)
[16:03:02] <andypugh> And can I see LC_NUMERIC inside HAL module code?
[16:03:35] <andypugh> I was just going to let the integrator choose.
[16:03:39] <Jymmm> Ok, I'll jsut do the 10^(x/10) the long way then =)
[16:04:04] <cradek> I suppose you could read the environment, not sure
[16:04:18] <cradek> are you sure you need to reimplement printf? seems like that way lies madness
[16:04:26] <Jymmm> Afte tomorrow, hopefully I won't care =)
[16:04:33] <cradek> % LC_NUMERIC=, dc -e '4k1 3/p'
[16:04:34] <cradek> .3333
[16:04:37] <andypugh> Jymmm: That key will work fine. divide your number by 10, press "=", press the button.
[16:04:42] <cradek> ^ (this is surprising to me)
[16:05:07] <Jymmm> andypugh: trying it now
[16:06:01] <andypugh> cradek: Yes, I am pretty sure that if I want to parse float format specifiers in kernel code then I need to write my own sprintf.
[16:06:27] <cradek> what is your eventual goal?
[16:06:58] <andypugh> cradek: Try a minimal comp that tries to print a floating point value in HAL code
[16:07:17] <Jymmm> andypugh: Oh, that is a lil faster =) ty
[16:07:23] <cradek> print to what?
[16:07:31] <andypugh> My goal is to output numerical values byte-by-byte to an LCD
[16:07:51] <andypugh> But sending to stdout is equally hard
[16:07:53] <cradek> can't that be a userland hal component?
[16:08:01] <Jymmm> andypugh: sounds like a scrolling 2x16 lcd
[16:08:10] <andypugh> Jymmm: It could be
[16:08:43] <Jymmm> andypugh: I did that with a SINGLE 5x7 matrix LED once, pretty cool actually
[16:08:59] <andypugh> cradek: No, because the interface pin is a HAL realtime pin, and if it isn't serviced every thread you get doubled/missed characters
[16:09:11] <cradek> eww
[16:09:55] <andypugh> I _am_ trying to do strings in HAL, I deserve the pain
[16:10:02] <cradek> are you sure you can't restructure the task somehow?
[16:11:52] <andypugh> Oh, I could. There could be a userland function to create the buffer in a separate block of shared-memory, and a realtime function to stream the data to the LCD. But from where I am now (90% done) that isn't easier.
[16:12:28] <andypugh> If I had known that kernel string formatting can't do floats I probably wouldn't even have started.
[16:12:41] <cradek> :-(
[16:12:50] <Jymmm> andypugh: lcd monitor?
[16:13:01] <Jymmm> or like a 2x16
[16:13:35] <andypugh> And I challenge you to find any documentation in a language other than french that mentions the fact that there is no %f in printk (which is what the rt-code abstracts to)
[16:13:57] <andypugh> Jymmm: 2x16, 4x80, whatever
[16:14:09] <andypugh> But basic serial LCD display
[16:14:16] <cradek> http://www.kernelthread.com/publications/faq/335.html
[16:14:33] <cradek> sorry, first google hit for printk floating point formatting
[16:14:40] <Jymmm> andypugh: would a 5x7 LEd matix work for you? then you cna choose how big and bright it is?
[16:15:05] <Loetmichel> hmmm
[16:15:35] <Loetmichel> i once made a 640*32 dots text display out of 3mm green leds.
[16:15:41] <PCW> You dont actually have to send stuff every thread you just have to send nulls if you have nothing to change (since sserial has to send something)
[16:15:48] <andypugh> I have absolutely no use at all for this code. I don't want an LCD display. I bought a cheap one on eBay to write this driver because someone on the forum thought they might need it. Do I need to mention how much not-fun this has become?
[16:15:52] <Loetmichel> i wont do THAT again...
[16:16:14] <Loetmichel> tahts a work for someone with some brain damage.
[16:17:11] <Jymmm> andypugh: then I'd suggest you tell him about this and let him figure it out (it does 5x7 matrix with scrolling and daisychain) MAX7221
http://www.maximintegrated.com/datasheet/index.mvp/id/1339
[16:17:15] <PCW> so the rtai print functions dont deal with FP numbers at all?
[16:17:33] <Jymmm> andypugh: serial input to it
[16:17:38] <mhaberler> no, the varargs list is queued to userland
[16:17:52] <andypugh> PCW: Yes, but then you need to send the nulls too. You still need to synch the user thread to the realtime thread to send the nulls at the right time. (I guess you could clock them in from userland to realtime)
[16:17:53] <mhaberler> it's converted once its out of the kernel
[16:18:03] <Jymmm> andypugh: like $7/each but very nice
[16:18:27] <andypugh> PCW: No, no %f formats in kernal space.
[16:19:31] <andypugh> Anyway, it's nearly done. It was almost useful to me as I now almost undertand how C handles strings. (or, possibly, how C totally fails to handle strings)
[16:19:59] <PCW> maybe a userland formatting utility that gets sent numbers and returns strings
[16:20:49] <PCW> so you managed to make strings from a FP number without printf?
[16:22:40] <PCW> (formatted rounded padded etc)
[16:22:44] <andypugh> Yes. It's just arithmetic.
[16:23:32] <andypugh> divide by the base of your choice, take the remainder, make that into a char, repeat.
[16:23:43] <PCW> the LCD is fastest if you just send all four lines of stuff padded with spaces so you get automatic wrap
[16:24:34] <andypugh> PCW: That involves telling the component what the display size is. It's already hard enough getting the format spec in.
[16:24:59] <PCW> You can also use cursor addressing to the beginning of the number part of each line
[16:25:15] <andypugh> I will worry about speed when I have more than zero users. And he complains the third time.
[16:26:01] <andypugh> Sorry, but I am not totally motivated at the moment by a chunk of code that even I don't want :-)
[16:26:42] <PCW> :-)
[16:27:31] <PCW> Its also faster if you use the quad character mode (4000 CPS at 1Khz servo thread vs 1000 CPS)
[16:28:35] <andypugh> Ths started as a small and elegant comp. It first went wrong when it became clear that unknown-length argument lists into printf are tricksy. Then it became a lot of not-fun when %f turned out to be turned off.
[16:29:05] <andypugh> PCW: I might add that, though after I have written the utility to flash nvram
[16:29:43] <andypugh> Unless it is on by default?
[16:29:45] <PCW> sserial EEPROM parameters?
[16:29:50] <andypugh> Yes
[16:30:13] <PCW> Thats nice
[16:30:18] <andypugh> Actually, on by default would work, because the spare bytes would be null by default.
[16:30:44] <PCW> Yes as long as you null the top 3 bytes its the same
[16:31:18] <andypugh> So, is the 7i73 in 4-byte or 1-byte mode as-supplied?
[16:31:33] <PCW> just software modes
[16:31:48] <andypugh> Hmm?
[16:32:08] <PCW> sserial mode
[16:32:55] <andypugh> Yes, just found that.
[16:33:01] <PCW> sserial.port.0=1230xxx or however that works
[16:33:28] <andypugh> OK, yes. Looks like a parameter to add then.
[16:33:59] <andypugh> For some reason I thought it was an eeprom option
[16:34:25] <PCW> we toss nulls efficiently so theres no real harm using the quad char mode (other than a few more usec data time)
[16:35:30] <andypugh> The data time would be the same, I guess, except that you get analogue in as a free bonus in that mode?
[16:39:19] <PCW> Yeah most of the higher modes add more data (the lower ones are for higher update rates 4 KHz and above)
[16:47:04] <PCW> Didn't think about the issue of formatting the FP numbers though, maybe we should have implemented that in the display but formatting is a PITA
[16:48:16] <andypugh> cradek: Any ideas?
http://pastebin.com/nJn9EQiS The only thing not working is the - sign in fixed-width formats (ie, fill = '0' rather than fill = ' '. It doesn't matter what index I use, the - only appears if I use [i]. No literals seem to work <puzzled>
[16:49:04] <andypugh> I suspect there is something I don't know about indexing in to *chars
[16:52:56] <andypugh> Ah, no, this isn't some reseverved-word issue is it? where out[int] is not the same as out[literal]?
[18:15:52] <andypugh> I might have to try this:
http://youtu.be/TLPWqCMb7DE
[18:18:50] <L84Supper> ok PMMA fuel, don't use polyester
[18:19:14] <andypugh> Because?
[18:19:44] <L84Supper> it releases nitrogen when burned so it tends to self extinguish
[18:19:47] <andypugh> I have a feeling that teflon would be bad.
[18:20:14] <andypugh> I think that nitrogen burns happlily in hot oxygen?
[18:21:13] <andypugh> (In fact I know it does, it is one of the main things we struggle with at work, trying not to form oxides of nitrogen in the engine)
[18:21:16] <L84Supper> there's a small ramjet engine vid as well
[18:21:52] <Jymmm> teflon (PTFE) release toxic fumes when it hits 550F
[18:21:52] <L84Supper> you get more bang out of the acrylic
[18:22:09] <Jymmm> 525-550F
[18:22:16] <andypugh> If that had been my rocket I would have had to run it to destruction :-)
[18:22:25] <L84Supper> there are lots of papers on the decomposition of many polymers as they burn
[18:22:49] <L84Supper> i went through this recently working with polymers for lost wax castings
[18:23:23] <L84Supper> whats the safest as it burns, what ends up burning out most completely, etc
[18:23:39] <andypugh> Bloodhound SSC uses some sort of rubber, I think.
[18:26:02] <L84Supper> polyethylene + oxygen hybrid
http://www.youtube.com/watch?v=YcmXxxD8H_w&feature=related
[18:26:11] <andypugh> AH, yes. rubber and peroxide:
http://www.bloodhoundssc.com/car/rocket_propulsion/update_-_march_2011.cfm
[18:27:36] <L84Supper> In order to accelerate the car to 1000mph it will provide an average thrust of 111 kN (25,000lbf) for 20 seconds. The peak thrust will be 122kN (27,500lbf).
[18:28:05] <andypugh> That ought to be quite exciting for the driver
[18:28:17] <Jymmm> andypugh: That was an EXCELLENT engine video; simple, straight forward, and just enough to make it interesting
[18:28:22] <L84Supper> 0-1000mhp in <20 sec
[18:30:51] <andypugh> L84Supper: The same team are the only ones to have broken the sound barrier on land, so if they say they can do it, they probably can.
[18:31:17] <L84Supper> I'll take 100 lbs rubber, 55 gal drum of peroxide, 2 gyros, some aluminum tubing ..... 3,2,1 police sirens
[18:32:23] <L84Supper> pretty much just time and money, I can't wait to see the video
[18:33:21] <L84Supper> Wheels rpm 10,000
[18:33:43] <L84Supper> Wheels diameter 0.9 metres
[18:38:40] <L84Supper> whats the best way to polish plastic without leaving any residue or melting the plastic?
[18:39:12] <andypugh> Melting is a good way.
[18:41:17] <L84Supper> trying to get a clear plastic film that has matte on one side to be clear
[18:41:24] <andypugh> There is a kit meant for aircraft windscreens that is well thought of.
[18:42:23] <Tom_itx> L84Supper i have one such kit
[18:42:27] <L84Supper> buffing with cotton by hand works pretty well for small areas
[18:42:41] <Tom_itx> comes with some mighty fine sandpaper and a couple different polishing compounds
[18:43:03] <L84Supper> sub micron aluminum oxide powder
[18:43:19] <Tom_itx> like 4800 and 6000 grit or some such thing
[18:43:25] <Tom_itx> then polish
[18:43:36] <L84Supper> but i don't want any powder remaining in the plastic
[18:43:49] <Tom_itx> wax when done
[18:44:06] <L84Supper> can't have any wax or polish on the surface when it's down
[18:44:16] <L84Supper> or it would be easy
[18:44:38] <L84Supper> it would be nice if they just had a smooth roller on both side when they extruded it
[18:45:28] <L84Supper> maybe the 6000 grit and di water
[18:46:07] <Tom_itx> still leaves it hazy
[18:46:08] <ink_> whoa
[18:46:10] <ink_> cool links
[18:46:10] <Tom_itx> without polish
[18:46:57] <L84Supper> contact lens media
[18:52:24] <L84Supper> might be easier to heat it and pass it through two polished rollers
[20:02:47] * JT-Shop can call off the search for a surface grinder now
[20:03:08] <jdh> wooho
[20:03:28] <jdh> now find me a Haskell AGT15/30 gas booster.
[20:03:50] <Jymmm> jdh: andypugh has a rocket engine you can use
[20:04:33] <jdh> I think I'll pass on that.
[20:04:48] <andypugh> Not yet I haven't, but I saw how to make one on that Youtube
[20:05:56] <jdh> I saw a video of how to...err, nevermind.
[20:07:43] <JT-Shop> I didn't even realize it had a micro feed on the Z axis before I went to look at it
[20:08:04] <JT-Shop> 500 miles round trip... makes John a tired puppy
[20:09:14] <JT-Shop> http://littlerock.craigslist.org/tls/3208959211.html
[20:10:16] <JT-Shop> time to chill out
[20:10:30] <andypugh> Do you have a wheel balancing jig?
[21:10:20] <ReadError> http://dl.dropbox.com/u/25091878/Photo%20Sep%2028%2C%209%2056%2041%20PM.jpg
[21:10:23] <ReadError> came out pretty good
[21:11:03] <andypugh> CF really is pretty isn't it?
[21:11:18] <ReadError> very ;)
[21:11:26] <ReadError> its like trippy to look at
[21:11:33] <ReadError> like hypnotic almost
[21:12:18] <Tom_itx> what feed/speed did you use?
[21:14:22] <ReadError> 10ipm
[21:14:37] <ReadError> 3flute 1/16th endmill
[21:16:14] <ReadError> came out pretty nice
[21:16:20] <ReadError> very nice edges
[21:16:22] <ReadError> no finish work
[21:21:33] <Jymmm> http://www.themarysue.com/gangnam-style-klingon/
[21:32:02] <r00t4rd3d> even i dont dare click that
[21:33:28] <Jymmm> They actually did a good job on it.
[21:33:48] * ReadError has never seen starwars
[21:34:01] <r00t4rd3d> hmm, you can drag and drop gcode on firefox icon and it will open and show the code
[21:34:08] <Jymmm> ReadError: It's StarTrek =)
[21:34:12] <ReadError> yea that
[21:34:15] <ReadError> ive seen starwars
[21:34:16] <ReadError> my bad
[22:05:41] <tjb1> Google is a PITA, I type in "toe" and all I get are results for "tow"
[22:15:30] <r00t4rd3d> actually use the quotes in google
[22:18:56] <tjb1> it is searching toe but its like saying the word and finding tow too...
[22:22:36] <tjb1> r00t4rd3d:
http://sphotos-a.xx.fbcdn.net/hphotos-ash3/s480x480/577016_4373987783839_52552144_n.jpg
[22:23:04] <FinboySlick> r00t4rd3d: Big green machine.
[22:24:11] <FinboySlick> r00t4rd3d: That yours?
[22:26:13] <tjb1> FinboySlick: Thats mine
[22:27:06] <FinboySlick> tjb1: Nice work.
[22:27:11] <tjb1> Thanks
[22:27:49] <FinboySlick> What do you plan to use for motion? Belts or screws?
[22:27:54] <tjb1> Rack
[22:30:35] <FinboySlick> I know it's not polite to ask you to show your rack... But I'm curious as to what approaches to minimize backlash would be with those.
[22:31:12] <tjb1> FinboySlick:
http://www.cncrouterparts.com/rack-and-pinion-drive-nema-23-p-50.html
[22:31:19] <tjb1> r00t4rd3d: Did you see CNCRP changed the site?
[22:31:47] <tjb1> Its pulled into the rack with a spring
[22:32:33] <FinboySlick> tjb1: OK, so the rack teeth are flat walled and the pinion just sinks 'till there's no play.
[22:35:46] <tjb1> Would take a while to wear that rack out ;)
[23:17:11] <r00t4rd3d> seems all they did was make it prettier
[23:19:51] <ReadError> huh
[23:22:24] <r00t4rd3d> http://i.imgur.com/PDDjR.jpg
[23:32:02] <toastyde1th> i'll allow it
