#emc | Logs for 2006-06-08

[00:22:31] <giacus> guys, tryng a different approach to introduce the videoclip: www.giacus.org/files/emc2_preview.avi
[00:23:31] <giacus> starting from the head ..
[00:24:27] <giacus> but I would get some more good photos
[00:25:06] <giacus> some special effect will be used going on
[00:25:38] <giacus> resources are few and poor :(
[00:25:57] <fenn> need more, um, video
[00:26:15] <giacus> photos also, and at good resolution
[00:26:25] <giacus> I can't do miracles :/
[00:26:30] <fenn> there are movies of the mazak somewhere
[00:26:33] <fenn> ask swpadnos
[00:26:45] <Jymmm> giacus Yes, you can. you just are nto trying hard enough!
[00:26:46] <giacus> I got a bit of videos
[00:26:48] <SWPadnos> I have a 68M AVI from my digital camera
[00:26:55] <SWPadnos> 640x480 resolution, I think
[00:27:06] <giacus> SWPadnos: would be good
[00:27:29] <SWPadnos> a lot of it is crap, because I was moving around and there wasn't a lot of light. the last 15 seconds or so are good though
[00:27:42] <fenn> and there's cradek's lathe movie
[00:27:44] <giacus> I just cut the best scenes
[00:27:47] <SWPadnos> I'll stick it on cncgear, and let you know when it's done uploading
[00:27:52] <Jymmm> SWPadnos: Strange, on a 90 minute tape I get 40GB of broadcast quality video.
[00:27:55] <giacus> or maybe you could cut them before..
[00:28:15] <SWPadnos> not really - I don't have the software (or the time, really)
[00:28:29] <giacus> is it avi ?
[00:28:33] <giacus> avidemux
[00:28:34] <SWPadnos> yep
[00:28:38] <giacus> very simple to use
[00:28:38] <SWPadnos> not sure what codec though
[00:29:27] <giacus> or, if you can upload it somewhere, as it
[00:29:38] <SWPadnos> I had loaded it up on my big box, and cut out the earlier parts (with software I can't remember the name of)
[00:29:45] <SWPadnos> that's what I'll do
[00:31:21] <fenn> http://timeguy.com/cradek-files/emc/dscn6185.mov
[00:31:57] <giacus> fenn: already have it
[00:32:18] <giacus> I'll use it going on ..
[00:32:33] <fenn> also, the text is very small and slow and boring
[00:32:42] <giacus> I know
[00:33:04] <giacus> that's bad codec and low resolution (to be small right now)
[00:33:14] <giacus> actual size id biggest here
[00:33:17] <giacus> and good
[00:33:29] <giacus> at the end , if I done it ..
[00:33:39] <giacus> a good codec will be used
[00:34:26] <giacus> for fast streaming on the web I want to try in flash format
[00:34:40] <giacus> mencoder does it too
[00:34:43] <SWPadnos> ok. it'll be a while before it's uploaded. ADSL is very A here ;)
[00:35:06] <giacus> ty
[00:35:12] <fenn> yeah mencoder with mp4 is wonderful.. i dont know why someone would use any of this other crap
[00:35:28] <fenn> like asx or quicktime.. yuck
[00:35:35] <giacus> fenn: some codecs won't work at all sometime
[00:35:58] <giacus> I'm using the new repository has replaced marillat
[00:36:05] <giacus> having some issue ..
[00:36:28] <giacus> best way would be compile anything from tarball I guess
[00:36:44] <giacus> mplayer, ffmpeg, mencoder & co .
[01:14:40] <asdfq-shop> Hm...I wonder how well emc2 will do on a 500MHz machine
[01:15:41] <skunkworks> I am running it on a 450mhz with axis
[01:15:57] <asdfq-shop> What kind of video?
[01:16:28] <skunkworks> video card? I don't know - diamond stealth or something - old
[01:16:54] <skunkworks> been meaning to look - others have asked
[01:17:08] <asdfq-shop> I'm using a Voodoo3 PCI card - and getting accelleration with realtime :)
[01:17:48] <skunkworks> nice - have you run the rt latency test?
[01:17:59] <asdfq-shop> My install is debian/testing with paul's bdi .debs
[01:18:01] <CIA-8> 03jepler 07HEAD * 10emc2/src/hal/classicladder/Submakefile: make more gtk files depend on Makefile.inc
[01:18:02] <CIA-8> 03jepler 07HEAD * 10emc2/src/hal/user_comps/vcp/Submakefile: make more gtk files depend on Makefile.inc
[01:18:12] <asdfq-shop> No I haven't, on this machine
[01:18:56] <asdfq-shop> what's the command line for that, or should I dig out my RTAI testsuite CD
[01:19:14] <jepler> asdfq-shop: I don't think paul includes the test suite on his CD
[01:19:23] <asdfq-shop> Ah
[01:19:34] <asdfq-shop> [sighs]
[01:19:53] <asdfq-shop> Some things Paul does right, and some things not
[01:20:30] <jepler> *shrugs*
[01:21:04] <Jymmm> Paul has fought for a couple years to keep it to a single CD
[01:21:15] <SWPadnos> giacus, still around?
[01:21:17] <cradek> jepler: I installed the cd you gave me and it's breezy!
[01:21:24] <jepler> cradek: oh well
[01:21:28] <jepler> cradek: isn't the installer slick?
[01:21:28] <giacus> SWPadnos: yeah
[01:21:33] <cradek> jepler: yes, very nice
[01:21:41] <SWPadnos> ok - http://www.cncgear.com/Files/mazak.avi
[01:21:48] <giacus> thanks a lot
[01:22:06] <jepler> cradek: I'll give you another unlabeled CD and we can see if that one is dapper.
[01:22:10] <SWPadnos> giacus, no problem
[01:22:15] <cradek> jepler: ok
[01:22:21] <skunkworks> cradek: first install of dapper drake?
[01:22:24] <Jymmm> * Jymmm hands jepler a box of sharpies
[01:22:28] <cradek> skunkworks: zeroth
[01:22:41] <giacus> 225.36K/s not bad ;P
[01:22:59] <asdfq-shop> Ugh, I hate slick installers - doesn't give me that cold, clunky, impersonal computer feel
[01:23:02] <cradek> getting 600K/s here
[01:23:07] <cradek> what is it?
[01:23:16] <SWPadnos> the video of the mazak changing tools
[01:23:21] <cradek> cool
[01:23:38] <SWPadnos> the first half or so isn't that great due to the low light and me moving around a lot
[01:23:58] <SWPadnos> but the last couple of changes, when Roland turned on the shop light, are pretty good
[01:25:48] <cradek> it's taking longer to install mplayer than to download the video...
[01:25:54] <SWPadnos> heh
[01:27:37] <giacus> SWPadnos: download completed ;)
[01:28:03] <SWPadnos> yeah - much faster to download from DreamHost than to upload (with 256kbit DSL upload speed)
[01:28:59] <giacus> watching it, excellent video quality
[01:29:05] <SWPadnos> good deal
[01:29:12] <cradek> very nice, thanks for sharing that
[01:29:14] <giacus> cool
[01:29:40] <SWPadnos> sorry it took so long. I was going to upload a chopped version during fest, but didn't get to it
[01:29:43] <fenn> someone wanna re-encode it with a lower bitrate?
[01:30:11] <fenn> i just did it the other day so i think i can still remember how to do it
[01:31:07] <SWPadnos> I used some program that I think began with L (lives?) to cut out the shaky beginning, but the result was almost as large a file, for half the duration
[01:31:43] <fenn> mencoder -ovc frameno -o frameno.avi -oac mp3lame -lameopts preset=voice mazak.avi ; mencoder -o mazak-small.avi -oac copy -ovc lavc -lavcopts vcodec=mpeg4:vbitrate=50 mazak.avi
[01:32:03] <fenn> thats a "two pass" encoding
[01:32:28] <SWPadnos> I wonder if they have mencoder installed at DreamHost ;)
[01:32:29] <fenn> 50 is a pretty low bitrate
[01:32:44] <fenn> but generally movies are pretty wasteful of bandwidth i think
[01:33:11] <SWPadnos> nope - no mencoder on the server, and I don't have a Linux machine up at the moment
[01:34:03] <skunkworks> 62kbs here. in the boonies
[01:34:51] <CIA-8> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10emc2head/: build FAILED ; see http://linuxcnc.org/compile_farm/emc2head_slot4_log.txt
[01:36:36] <giacus> -rw-r--r-- 1 jack jack 72234000 2006-06-08 03:22 mazak.avi
[01:36:37] <giacus> -rw-r--r-- 1 jack jack 4725494 2006-06-08 03:39 mazak-small.avi
[01:37:03] <giacus> fenn: good
[01:37:48] <fenn> is it worth watching still?
[01:37:59] <fenn> * fenn wanna see
[01:38:28] <giacus> a second ..
[01:40:51] <giacus> check it, http://www.giacus.org/files/mazak-small.avi
[01:41:05] <fenn> thanks
[01:41:10] <giacus> lost a bit of quality ..
[01:41:23] <giacus> but not bad at all
[01:42:46] <fenn> that is so cool
[01:43:35] <fenn> i never actually saw the whole cycle
[01:44:12] <SWPadnos> strange. I only get audio (using RealPlayer 10.5 on Windows)
[01:45:10] <fenn> whats a more standard codec than mp4?
[01:45:18] <SWPadnos> damfino
[01:45:34] <Jymmm> dont use a codec at all
[01:45:49] <fenn> what, raw bitmap array? :)
[01:46:01] <jepler> fenn: a carton full of polaroids, sent via us mail
[01:46:32] <fenn> i'll pipe it through a fiberoptic cable to your house, how's that?
[01:50:01] <jepler> it's Jymmm and SWPadnos you have to satisfy here, not me.
[01:50:56] <fenn> well, swpadnos has already seen it, and jymmm can go get a real OS and stop messin around
[01:51:35] <Jymmm> seen what?
[01:51:55] <SWPadnos> the large one works OK for me, what codec did it use?
[01:52:15] <fenn> dunno i didnt download it
[01:52:18] <SWPadnos> Jymmm, http://www.cncgear.com/Files/ - look at the mazak toolchange video
[01:52:31] <fenn> jymmm http://www.giacus.org/files/mazak-small.avi
[01:52:40] <SWPadnos> I put the small one on cncgear as well
[01:52:53] <giacus> Opening video decoder: [ffmpeg] FFmpeg's libavcodec codec family
[01:52:54] <giacus> Selected video codec: [ffodivx] vfm:ffmpeg (FFmpeg MPEG-4)
[01:53:21] <SWPadnos> ok. mpeg in, divx out
[01:53:26] <SWPadnos> ?
[01:53:44] <fenn> * fenn shrugs helplessly
[01:54:02] <giacus> I just use the command feen gave above
[01:54:11] <SWPadnos> ok, then it's fenn's fault
[01:54:15] <giacus> MPEG-4
[01:54:40] <skunkworks> SWPadnos: nice video
[01:54:46] <SWPadnos> thanks
[01:54:57] <SWPadnos> if only I had brought a tripod in that van full of stuff ;)
[01:54:59] <jepler> hm. what to do about deliberately infinite files in axis. for instance, a test program that repeats forever...
[01:55:33] <SWPadnos> automatically detect an infinite loop, and stop generating the preview ;)
[01:55:37] <fenn> giacus: try changing mpeg4 to mpeg2video
[01:55:46] <jepler> SWPadnos: yeah, but I was thinking of something easier to implement.
[01:55:54] <SWPadnos> oh, like something possible?
[01:56:02] <jepler> SWPadnos: I think I fell asleep in the class where they gave the solution to the halting problem
[01:56:17] <SWPadnos> hmmm -I haven't heard of that class
[01:56:45] <SWPadnos> the one with the solution anyway (or was that one of those "the solution is left as an exercise for the reader")
[01:56:58] <giacus> fenn: yeah, give me few minuts
[01:56:58] <jepler> it was between "we'll try again to teach you students recursion" and "recursion for comp sci seniors"
[01:57:07] <SWPadnos> heh
[01:58:04] <skunkworks> interesting - could you check for the posiblilty of an infanant loop? axis would have to be intelegent - not that it isn't already
[01:58:15] <cradek> both videos play fine for me
[01:58:34] <SWPadnos> the ability to detect any arbitrary infinite loop doesn't exist
[01:58:57] <SWPadnos> in the constrained world of a G-code interpreter, it may be possible
[01:59:08] <asdfq-shop> Sure there is - on a machine with finite resources - wait until it runs out of memory
[01:59:17] <skunkworks> but if you can detect it - I would think a program could.
[01:59:20] <SWPadnos> that would be a finite loop then ;)
[01:59:23] <giacus> Cannot find codec 'mpeg2' in libavcodec..
[01:59:34] <giacus> sorry .. right
[02:00:07] <skunkworks> I really have much of a clue - just thinking out loud again.
[02:00:26] <fenn> jepler: A machine with finite memory has a finite number of states, and thus any deterministic program on it must eventually either halt or repeat a previous state. Repetition of a previous state indicates a loop, so a program that repeats a previous state is thus known to not halt.
[02:00:27] <cradek> jepler: the 2.4.25-adeos is using gtk2 but still doesn't build...?
[02:00:48] <SWPadnos> some infinite loops can be easily detected, and it's of course easy to detect a loop that has finished, but you can't prove that a loop is not infinite without running it to completion
[02:01:25] <SWPadnos> fenn, right - that's the thing that could make it work - there are a finite number of variables and commands in G-code
[02:01:49] <fenn> it might take a really long time to overflow your variables and come back to zero
[02:01:51] <cradek> I'm not sure whether you guys are being serious, but it seems to me that in axis the time you want to abort is when the user gives up and tells it to
[02:02:05] <SWPadnos> but you still have to save all states (ie, all vars and all modes) for every line executed
[02:02:31] <SWPadnos> cradek - I agree. it's much better to keep the UI responsive, and put a "stop preview" button on it
[02:02:32] <skunkworks> btw my psyco spiro program - takes about 2 minutes to load on my 450mhz computer
[02:02:38] <skunkworks> that way it is
[02:02:55] <cradek> and you know the program is psycho and will take a while
[02:03:05] <jepler> axis won't show you how much has been loaded .. that would be more difficult than what is currently implemented
[02:03:09] <skunkworks> does the escape abort the program now? or does it stop previewing?
[02:03:14] <cradek> on another program you would know that two minutes of loading means it's doing something you don't intend
[02:03:22] <jepler> the progress bar seems to freeze, and when you finally hit escape it chugs for awhile longer
[02:03:40] <jepler> (that's the general way it behaves on any program that uses O-word looping)
[02:04:04] <cradek> the place to detect an insane program (when possible) is in the interpreter, not the gui. the interp should throw an error like it does for any other problem.
[02:04:29] <jepler> o1 while [1]
[02:04:29] <jepler> g2 f100 x0 y0 i[sin[#1]*.5] j0
[02:04:29] <jepler> #1=[#1+4]
[02:04:29] <jepler> o1 endwhile
[02:04:40] <jepler> here's my useful infinite program .. it does circles of varying radii, makes nice plots in halscope
[02:05:16] <jepler> until you abort it (which works fine)
[02:05:57] <skunkworks> will it let you run it after aborting?
[02:06:02] <cradek> oh so you're saying the program is fine (it's meant to run until the user aborts it)
[02:06:05] <cradek> I didn't understand
[02:06:10] <cradek> so it IS just a gui problem.
[02:06:15] <fenn> jepler: so you run that with tkemc or what?
[02:06:23] <jepler> fenn: no, I hit escape while axis is trying to load it
[02:06:30] <jepler> it shows a varying number of spirals depending on how long I waited
[02:07:08] <jepler> the only problem with hitting escape is that it's not self-documenting
[02:07:09] <skunkworks> but then can you cycle start the program? If you can - I say fine.
[02:07:43] <giacus> the mpeg2 is on the way
[02:07:51] <jepler> yes, I can
[02:09:07] <giacus> http://www.giacus.org/files/mazak-small-mpeg2.avi
[02:09:19] <skunkworks> it will have to be part of the axis directions - "if creating a gcode program that will never end - the escape key will have to be press to get out of the pre-draw loading" or something like that.
[02:10:16] <cradek> to make it self-documenting you would pop up a window with a cancel button if it took "a long time"
[02:10:49] <skunkworks> * skunkworks thought jepler had already done something like that - or atleast thought about it
[02:11:04] <jepler> skunkworks: I wrote that at first, but I became convinced it was a bad idea
[02:11:08] <cradek> oh
[02:11:11] <jepler> because there's no acceptable definition of "a long time"
[02:11:17] <jepler> like you with your 4-minute program
[02:11:32] <cradek> but no harm is done if it pops up.
[02:11:37] <jepler> sure there is -- the user panics
[02:12:39] <skunkworks> users don't panic - do they?
[02:12:41] <jepler> or, worse, asks me for a configuration option for how long is "too long"
[02:12:51] <cradek> arg
[02:12:55] <cradek> I retract my suggestion
[02:13:04] <cradek> I'm back in the "don't care, hit escape" camp now
[02:13:16] <skunkworks> which already works :)]
[02:13:31] <cradek> all the better
[02:13:40] <fenn> what's wrong with just a "cancel" button again?
[02:13:47] <fenn> cancel/escape
[02:13:51] <cradek> there's no obvious place to put it
[02:14:13] <jepler> there is also a technical problem with making only one button clickable -- it takes more than a few lines of code.
[02:14:16] <skunkworks> or - you allow people to disable the preview - which I guess is called tkemc with backplot :)
[02:14:39] <jepler> I did think about that: it still loads the preview even when that's turned off in the menu
[02:14:44] <cradek> even if the program preview is turned off ... yeah that
[02:14:55] <cradek> that's bordering on bogus
[02:15:05] <jepler> on the other hand, that means giving up a lot more than seeing it. Like knowing if it exceeds bounds when you hit 'run', or if it contains an outright syntax error
[02:15:11] <cradek> but if you load with it turned off, and turn it on, then what
[02:15:12] <skunkworks> I didn't know that was an option already - sorry
[02:15:31] <skunkworks> escape for the win
[02:16:08] <jepler> (the same problems you have if you abort the load early)
[02:16:54] <asdfq-shop> Question about AXIS: Is <F1> the only E-stop key?
[02:17:04] <cradek> F1 is estop, ESC is abort
[02:17:21] <asdfq-shop> Okay
[02:17:49] <asdfq-shop> Heh, finally loaded the AXIS page, now I have the keylist
[02:18:39] <asdfq-shop> I have an older gentleman for a friend, and he's used to using DESKNC in DOS
[02:18:52] <giacus> G'night all
[02:19:28] <asdfq-shop> The ESC key is good, but SPACEBAR would be good too
[02:19:30] <fenn> where did this "for the win" thing come from?
[02:19:51] <skunkworks> I don't know - normally I see if as FTW
[02:20:20] <skunkworks> didn't know if that would be reconized
[02:20:35] <skunkworks> twinkies FTW
[02:20:36] <fenn> Bangalore, apparently: http://www.google.com/trends?q=for+the+win
[02:21:57] <skunkworks> :) have not been there
[02:22:00] <jepler> huh. I never knew about google trends. that's neat.
[02:22:01] <jepler> http://www.google.com/trends?q=all+your+base%2C+hampster+dance%2C+badger+badger&ctab=0&geo=all&date=all
[02:22:19] <jepler> oops, don't know how that extra 'p' got in there
[02:23:04] <jepler> (I'm sad to see that 'badger badger' has beat out AYB by such a wide margin)
[02:23:37] <skunkworks> http://www.google.com/trends?q=bork+bork+bork&ctab=0&geo=all&date=all
[02:23:44] <fenn> badger badger??
[02:24:04] <fenn> http://www.google.com/trends?q=badger%2C+badger+badger%2C+badger+badger+badger&ctab=0&geo=all&date=all
[02:24:30] <jepler> http://www.google.com/trends?q=bush+sucks&ctab=0&geo=all&date=all
[02:24:58] <fenn> wow there's something statistically significant there
[02:25:12] <cradek> gee you think it's in november?
[02:25:47] <fenn> http://www.google.com/trends?q=bush+sucks&ctab=0&geo=all&date=2004-11
[02:25:50] <Jymmm> LOL http://www.google.com/trends?q=bush+sucks%2C+stupid+bush&ctab=0&geo=all&date=all
[02:27:41] <jepler> http://www.google.com/trends?q=gullible
[02:28:31] <jepler> I wish there was more old data. It looks like "epidermis" roughly repeated its 2004 performance in 2005, but maybe it's my imagination
[02:30:18] <jepler> http://www.google.com/trends?q=2003%2C+2004%2C+2005%2C+2006 http://www.google.com/trends?q=january%2C+april%2C+august
[02:31:44] <Jymmm> U.S. not even in the running.... http://www.google.com/trends?q=cnc&ctab=0&geo=all&date=all
[02:32:04] <fenn> weird
[02:33:23] <Jymmm> http://www.google.com/trends?q=cnc+mill%2C+cnc+lathe&ctab=0&geo=all&date=all
[02:34:52] <cradek> http://www.google.com/trends?q=gay+marriage&ctab=0&geo=all&date=all
[02:35:03] <cradek> (nice try, but nobody cares anymore)
[02:35:11] <fenn> good, because it's a non-issue
[02:35:24] <cradek> especially after today's vote
[02:36:43] <Jymmm> I think the question has just been answered.... http://www.google.com/trends?q=to+be%2C+not+to+be&ctab=0&geo=all&date=all
[02:36:59] <cradek> ha
[02:37:06] <jepler> http://www.google.com/trends?q=gay+marriage%2C+divorce
[02:37:33] <cradek> this is quite fascinating
[02:37:38] <cradek> google is still cool
[02:38:10] <Jymmm> http://www.google.com/trends?q=gay+marriage%2C+gay+divorce&ctab=0&geo=all&date=all
[02:39:30] <fenn> i hope they stay in the AI/database business and leave all the spreadsheets and webmail crap to yahoo
[02:40:00] <jepler> and what about the targeted ads?
[02:40:05] <cradek> and censorship and cooperation with the snooping govt agencies
[02:40:14] <cradek> spreadsheets??
[02:40:32] <jepler> cradek: yeah, didn't you hear? there's a google spreadsheet
[02:40:35] <jepler> beta, invitation-only
[02:40:37] <Jymmm> lil scarry... http://www.google.com/trends?q=rape%2Ccampus&ctab=0&geo=all&date=all
[02:40:48] <Jymmm> cradek http://spreadsheets.google.com/
[02:40:58] <jepler> gay marriage, polygamy,
[02:40:58] <Jymmm> It's a lil slow, but not bad
[02:40:59] <jepler> oops
[02:41:07] <jepler> I meant, http://www.theinquirer.net/?article=32259
[02:41:48] <cradek> I want google-draw-stuff-with-your-mouse-cooperatively
[02:42:11] <cradek> a shared sketch pad would be great, but nobody seems to have done it
[02:42:28] <fenn> yah or a TWIT interface
[02:42:39] <fenn> (sorry i had to)
[02:42:44] <cradek> I interface with twits on irc sometimes
[02:42:59] <cradek> (what's TWIT?)
[02:43:06] <fenn> its a play on the DWIT interface
[02:43:10] <fenn> "do what i'm thinking"
[02:43:31] <fenn> "type what i'm thinking"
[02:43:33] <cradek> ah I've always heard "do what I mean"
[02:44:36] <jepler> http://www.google.com/trends?q=dos%2C+linux
[02:44:44] <jepler> look out guys, DOS is catching up!
[02:45:19] <fenn> yeah DoS gets lots of nooze coverage
[02:45:31] <Jymmm> MS-DOS
[02:45:50] <skunkworks> whats wrong with ms-dos?
[02:46:04] <Jymmm> oh hell, they require you to have/get a google acnt to share the spredsheet with others.
[02:46:34] <Jymmm> skunkworks nothing at all. But what we used to call DOS (ms-dos), is now DoS (Denial of service)
[02:47:15] <Jymmm> well, not share, but someone else to view the shared spreadsheet.
[02:51:21] <skunkworks> oh well - time for bed
[02:53:00] <jepler> http://www.google.com/trends?q=0%2C+1%2C+2%2C+3%2C+4
[02:53:04] <jepler> I wonder what it means
[02:53:36] <cradek> maybe nothing?
[02:53:41] <jepler> clearly 2 is not just as bad as 1
[02:53:57] <cradek> well 1 is the lonliest number
[02:54:11] <cradek> lonelyest?
[02:54:23] <cradek> bah
[02:54:44] <jepler> yes, goodnight.
[02:54:48] <cradek> goodnight
[02:56:53] <jepler> ooh, glad we can settle this debate for once and for all: http://www.google.com/trends?q=emacs%2C+vi%2C+edlin%2C+wordstar%2C+notepad
[02:57:12] <jepler> (yes, that's right, roman numeral six is more popular than any text editor)
[02:57:42] <cradek> http://www.google.com/trends?q=emacs%2C+vim%2C+edlin%2C+wordstar%2C+notepad&ctab=0&geo=all&date=all
[02:59:18] <jepler> maybe this is the search you were looking for: http://www.google.com/trends?q=xemacs%2C+vim
[02:59:35] <cradek> heh cool
[03:00:01] <cradek> is it fair if I dislike both editors somewhat?
[03:08:18] <fenn> http://www.google.com/trends?q=xemacs%2C+vim%2C+kate&ctab=0&geo=all&date=all
[03:10:25] <fenn> huh weird that there actually seems to be a relationship between emacs/vim/notepad
[03:13:16] <fenn> http://www.google.com/trends?q=capacitor%2C+coil%2C+resistor%2C+diode%2C+transistor
[03:18:05] <Jymmm> http://www.jadeintegration.com/jackpc.php
[03:19:01] <fenn> that's nothin
[03:19:20] <Jymmm> if it'll stop all these trend crap, thats all I care about
[03:19:56] <fenn> http://www.lantronix.com/device-networking/embedded-device-servers/xport.html
[03:20:54] <fenn> actually that is pretty cool jymmm
[03:21:23] <Jymmm> fenn: Jsut remembr it's a THIN CLIENT.
[03:22:22] <fenn> only if you want it to be
[03:22:40] <Jymmm> it only has IE
[03:23:11] <fenn> why cant you boot any arbitrary ram disk?
[03:24:02] <fenn> god patents are so stupid
[03:24:08] <fenn> how can you patent "a computer in a wall"
[03:26:01] <Jymmm> fenn so how much is this xport?
[03:26:15] <fenn> i dunno probably like $300
[03:27:15] <fenn> oh wow its $50
[03:27:28] <Jymmm> where do you see that?
[03:27:42] <fenn> http://www.lantronix.com/news/PR_03-08-06.html "Pricing and Availability"
[03:28:44] <fenn> search.ebay.com/7625271630
[03:30:02] <fenn> i kinda wonder where the i/o pins are
[03:35:35] <Jymmm> http://www.lantronix.com/pdf/XPort_IG.pdf
[03:38:42] <Jymmm> 8 pins on the bottom
[03:39:48] <Jymmm> Now, PoE would be very cool
[03:39:55] <Jymmm> (Power over Ethernet)
[03:41:16] <SWPadnos> the XPort has 8 or 10 pins, and the serial/I/O pins are among them
[03:41:24] <SWPadnos> + power and ground
[03:41:28] <SWPadnos> (of course)
[03:42:10] <SWPadnos> 8 pins + 2 for the shield can (on my dev board)
[03:42:25] <Jymmm> SWPadnos for your cameras?
[03:42:43] <SWPadnos> nope - remote controls for power supplies (defunct project)
[03:43:16] <Jymmm> I think it be cool for http://www.faqs.org/rfcs/rfc2324.html
[03:43:30] <SWPadnos> heh
[03:43:55] <Jymmm> I've always wanted to do that, just once.
[03:44:14] <SWPadnos> write an April fools RFC, or ssh into your coffee pot? ;)
[03:44:48] <Jymmm> in response to the rfc
[03:44:56] <SWPadnos> ah
[03:45:08] <Jymmm> make a working version of it
[03:45:33] <SWPadnos> hmmm - I was thinking that there was one at MIT, but I could be thinking of the coke machine
[03:45:46] <Jymmm> the guy who wrote it, is around the corner at HP in Palo Alto
[03:46:07] <Jymmm> err Xerox I mean
[03:46:15] <Jymmm> Xerox PARC
[03:47:02] <SWPadnos> great place (historically)
[03:47:21] <Jymmm> yeah, a tour would be cool.
[03:47:25] <A-L-P-H-A> ohla folks.
[03:47:58] <SWPadnos> howdy
[03:48:07] <A-L-P-H-A> so what's happening?
[03:48:34] <SWPadnos> well - I just watched the Chris Rock Saturday night live DVD
[03:48:43] <SWPadnos> he's a funny guy
[03:49:26] <A-L-P-H-A> want this. http://www.t3.co.uk/news/247/entertainment/mp3_player/impatient_japanese_make_own_video_ipod
[03:49:32] <A-L-P-H-A> Chris Rock is awesome.
[03:49:49] <A-L-P-H-A> SWPadnos, figured out the ATI driver issue yet?? :)
[03:49:54] <A-L-P-H-A> or how to build Xorg 7.1?
[03:50:02] <SWPadnos> haven't done anything with it today
[04:13:54] <A-L-P-H-A> I can't find any info on how to build Xorg 7.1
[04:14:02] <A-L-P-H-A> no binaries are out... so that's no help
[04:14:35] <SWPadnos> download / tar xvf (or jvf) / cd <new dir> / make / sudo make install
[04:14:48] <A-L-P-H-A> there's a ton of tar files.
[04:14:59] <A-L-P-H-A> do I need them ALL??
[04:15:31] <SWPadnos> I wouldn't think so, but it's probably easiest to just get them all and compile everything (overnight, probably)
[04:15:39] <A-L-P-H-A> http://ftp.x.org/pub/X11R7.1/src/everything/ and http://ftp.x.org/pub/X11R7.1/src
[04:16:45] <SWPadnos> I'd go for src/xerver/<big name>.bz2
[04:17:27] <A-L-P-H-A> that's only 6 megs...
[04:17:37] <SWPadnos> after bunzip, it
[04:17:37] <A-L-P-H-A> wonder how long it'd take...
[04:17:49] <SWPadnos> after bunzip, it'll probably be closer to 75M
[04:17:55] <A-L-P-H-A> for real?!?!
[04:17:58] <A-L-P-H-A> whoa.
[04:18:03] <SWPadnos> yep. don't know, jus testimating
[04:18:07] <SWPadnos> maybe only 50M
[04:18:16] <A-L-P-H-A> SWPadnos, http://ftp.x.org/pub/X11R7.1/src/xserver/ paste that back to me after I reboot?
[04:18:26] <SWPadnos> ok
[04:21:18] <A-L-P-H-A> hi
[04:21:21] <A-L-P-H-A> back
[04:21:23] <A-L-P-H-A> what was that url?
[04:21:33] <SWPadnos> http://ftp.x.org/pub/X11R7.1/src/xserver/
[04:21:59] <SWPadnos> there are some build instructions here: http://wiki.x.org/wiki/ModularDevelopersGuide#head-ef8aae011f5ba67c1884a28bb2f998fa87a46f95
[04:22:12] <SWPadnos> it looks like it may be a pain in the ass, actually
[04:23:24] <A-L-P-H-A> :((
[04:23:52] <SWPadnos> well - download all the tarballs (or the needed ones, if you know that subset), then run the build-from-tarballs script
[04:25:25] <A-L-P-H-A> I have the whole thing.
[04:25:29] <A-L-P-H-A> it's 50megs of scripts
[04:25:31] <A-L-P-H-A> source
[04:25:37] <SWPadnos> ok - not bad
[04:25:37] <A-L-P-H-A> 42.8 actually
[04:26:03] <SWPadnos> so that's the big tarball, not everything from the everything/ dir?
[04:28:03] <A-L-P-H-A> it's the same thing I think
[04:28:46] <SWPadnos> I was thinking it might be. the modular build system is kinda new
[04:33:21] <A-L-P-H-A> too much for my little brain
[04:33:41] <SWPadnos> heh. you may just have to wait for new packages or something :)
[04:41:18] <A-L-P-H-A> that maybe for forever.
[04:45:49] <A-L-P-H-A> so SWPadnos ... what's new today?
[04:46:03] <SWPadnos> not much. I was just thinking of going to bed, actually ;)
[04:46:18] <A-L-P-H-A> wimp. it's only 0050hrs
[04:46:35] <SWPadnos> I've been up since 0730 or so
[04:47:52] <A-L-P-H-A> I'm just bugging.
[04:47:59] <A-L-P-H-A> I've been up since oh .... 1600hrs.
[04:48:04] <SWPadnos> wimp
[04:50:43] <SWPadnos> SWPadnos is now known as SWP_Away
[08:32:23] <Bo^Dick> i wonder what's happened to my keyboard
[08:32:43] <Bo^Dick> the key "Alt Gr" doesn't work anymore
[08:33:20] <Bo^Dick> so it's a challenge typing "@" and "~"
[08:39:50] <robin_sz> morning
[08:43:11] <Bo^Dick> hi
[08:43:44] <Bo^Dick> do you know why my "Alt Gr" key on my kyeboard doesn't work?
[08:44:20] <Bo^Dick> could i've accidently configured my keyboard for a unix-style one?
[08:46:29] <robin_sz> shrug ... you dont configure keyboards
[08:46:44] <robin_sz> you configure the thing that rx the bytes and translates them to chars
[08:47:12] <robin_sz> the alt-gr on my linux box here works fine
[08:47:52] <robin_sz> with a US/UK keymap. alt-gr doesnt do much
[08:48:01] <robin_sz> try a french keymap and it might be different
[08:48:44] <Bo^Dick> it used to work fine for me too
[08:48:54] <Bo^Dick> but now i've screwed up something
[08:48:58] <Bo^Dick> apparently
[08:49:33] <Bo^Dick> maybe i should'nt have written "xfree86" when i installed it
[08:49:45] <Bo^Dick> am i a moron?
[08:52:13] <robin_sz> * robin_sz flips a coin
[08:53:07] <Bo^Dick> could you show me your config file
[08:55:06] <robin_sz> just set it like you want it ... from applications->desktop preferences->keyboard->layout
[09:02:11] <Bo^Dick> what distribution are you running?
[09:05:40] <robin_sz> debian Sarge
[09:06:07] <Bo^Dick> i'm using debian too
[09:06:17] <robin_sz> gnome desktop?
[09:06:24] <robin_sz> or that poxy KDE crap?
[09:06:27] <Bo^Dick> i'm running kde now
[09:07:09] <Bo^Dick> kde ca be configured for performance vs. look
[09:22:35] <Bo^Dick> i've discovered that there is a problem with the xfree software for the keyboard. the message i got was that this was likely due to an internal x-server problem
[09:23:08] <Bo^Dick> i was recomended to use a less complex configuration whatever that meant
[09:23:26] <Bo^Dick> any suggestions?
[11:07:59] <robin_sz> * robin_sz waves at giacus
[11:08:43] <robin_sz> I'm in Geneva today, so a little closer :)
[11:09:36] <giacus> hi robin_sz
[11:09:55] <robin_sz> nice weather in .it?
[11:10:04] <robin_sz> perfect here in Geneve
[11:10:22] <giacus> sstrange weather, clouds and sun ..
[11:10:31] <giacus> no wind
[11:10:47] <giacus> temperature is nice
[11:11:21] <robin_sz> we are sitting on the bacony in the sun
[11:11:28] <robin_sz> a nice way to work :)
[11:11:37] <giacus> cool
[11:14:10] <robin_sz> running another stamp auction with the Swiss guys ... going smoothly too
[11:14:36] <giacus> have a fun and shot some photo! :)
[11:14:37] <robin_sz> and I can watch my factory in the UK on the webcam
[11:15:40] <giacus> nice country over there
[11:15:58] <robin_sz> UK? or .ch?
[11:16:15] <giacus> Swiss
[11:16:19] <robin_sz> yeah
[11:16:24] <giacus> UK too ;)
[11:16:28] <robin_sz> we are 1hr from Chamonix
[11:16:43] <robin_sz> I can see Mt Blanc from the hotel
[11:52:13] <skunkworks> logger_aj: bookmark
[11:52:13] <skunkworks> See
[12:06:04] <CIA-8> 03jepler 07HEAD * 10emc2/src/hal/utils/miscgtk.c: gtk_window_set_urgency_hit and gtk_window_is_active do not seem to be available in gtk2.6 (bdi4)
[12:18:54] <CIA-8> 03jepler 07HEAD * 10axis/tcl/axis.tcl:
[12:18:54] <CIA-8> rename "ESTOP RESET" to "OFF"
[12:18:54] <CIA-8> Make button "Touch off" instead of "Offset"
[12:18:54] <CIA-8> rename "Code Entry" to "MDI"
[12:23:12] <CIA-8> 03jepler 07HEAD * 10axis/scripts/axis.py:
[12:23:12] <CIA-8> New feature: AXIS will accept certain comments that begin 'AXIS,'
[12:23:12] <CIA-8> the first such comment is (AXIS,stop) which makes AXIS stop generating the
[12:23:12] <CIA-8> preview plot for a file at or soon after this line is encountered.
[12:23:12] <CIA-8> Improve the help text to note that ESC can abort the loading of the preview
[12:23:13] <CIA-8> plot.
[12:23:15] <CIA-8> Improve the text on the Touch Off window
[12:31:16] <CIA-8> 03compile-farm 07BDI-Live rc46 (2.4.25-adeos) * 10emc2head/: build PASSED
[12:40:27] <skunkworks> jepler: cool idea on AXIS, stop
[12:43:18] <jepler> yeah, it gives control to the user (the g-code author)
[12:43:31] <jepler> heck, you could put it at the very top of your program and disable the preview altogether
[12:43:54] <skunkworks> right
[12:48:52] <skunkworks> * skunkworks wonders what would happen if AXIS, stop was put inside a infanant loop
[12:52:26] <cradek> sounds like the preview would stop the first time through the loop
[12:52:37] <jepler> right, that's the intent
[12:52:54] <jepler> (by infanant do you mean a very young loop, or one that goes on a long time?)
[12:52:58] <jepler> bbl
[12:53:34] <cradek> skunkworks: I was just noticing that I also read your spelling of 'infinite' as 'infant' the first time because my brain's pattern matcher says it's closer
[12:53:59] <cradek> skunkworks: I try not to point out spelling errors unless they're funny...
[12:54:05] <SWP_Away> it's a really big infant ;)
[12:54:09] <SWP_Away> SWP_Away is now known as SWPadnos
[12:54:31] <cradek> g'morning
[12:54:37] <SWPadnos> morning
[12:57:34] <skunkworks> stepped away. I know it was spelled wrong - but was too lazy to look it up.
[12:58:06] <SWPadnos> you only have to remember Michael Nesmith's album: "Infinite Rider on the Big Dogma"
[12:58:13] <SWPadnos> lots of good spellings there ;)
[12:58:29] <skunkworks> google does most of my spell checking :)
[12:59:33] <Bo^Dick> could a software intended for programming a pic16f84 be used to a pic18f452?
[13:00:04] <SWPadnos> I'm not sure if the 18xx series uses the same algorithm as the 16xx series
[13:00:18] <SWPadnos> it should be similar, but there may be subtle differences
[13:00:38] <SWPadnos> just look at the datasheets, make the changes in the code, and you're done
[13:01:21] <Bo^Dick> i'm not that hardcore
[13:01:30] <SWPadnos> it's actually pretty easy
[13:01:37] <SWPadnos> (if you have the source)
[13:02:11] <SWPadnos> the low-level communications scheme should be the same, so it's just a matter of deciding how many words to send, and how many times to do it (and changing some command codes, possibly)
[13:04:11] <Bo^Dick> i was hoping of finding an existing software of that kind assuming one exists
[13:04:35] <SWPadnos> there are many systems for doing this, but your zero budget eliminates many of them
[13:05:23] <Bo^Dick> there is another problem as well. i don't even know if my pic is supposed to work or not
[13:05:29] <Bo^Dick> i've abused it
[13:05:46] <SWPadnos> a small budget increase would make that problem go away as well ;)
[13:05:50] <Bo^Dick> so the situation is quite confused right now
[13:06:26] <Bo^Dick> if i was to be really serious i'd need a power supply with adjustable current limiter
[13:06:38] <Bo^Dick> this would save my parts and my breadboard
[13:06:47] <SWPadnos> are you still working on a motor drive?
[13:06:52] <Bo^Dick> but they are really expensive
[13:06:57] <Bo^Dick> yepp
[13:07:19] <SWPadnos> then I'd say a good quality, adjustable voltage/current benchtop power supply is an absolute must
[13:07:45] <Bo^Dick> theyre incredibly overpriced though
[13:08:03] <SWPadnos> it's hard wnough to design something in a field in which you have no experience, but to attemp it without the proper equipment is folly
[13:08:09] <SWPadnos> *enough
[13:08:23] <SWPadnos> garrr -attempt, too
[13:08:33] <SWPadnos> I need caffeine
[13:09:48] <Bo^Dick> chinamill has managed to get the tait-programmer to work with his pic
[13:10:06] <Bo^Dick> so i don't really understand why it doesn't work for me
[13:10:54] <Bo^Dick> it didn't work even before i abused the pic
[13:12:14] <SWPadnos> programming PICs through the parallel port is a finicky process, which is usually very dependent on CPU speed, the port itself, cable lengths, etc.
[13:13:12] <Bo^Dick> highly unreliable in other words
[13:13:50] <SWPadnos> well - finicky. it can be a bitch to get working, but usually goes OK for some time, as long as you don't change anything
[13:15:49] <Bo^Dick> this sucks
[13:16:45] <SWPadnos> I could point out that the AVR series is quite comparable to the PIC in cost, and there's a $30 programmer that works via the serial port ($35 forthe USB version), and is utterly reliable
[13:17:11] <SWPadnos> unless there's some peripheral on the PIC you chose that isn't on the AVR (like a quadrature decoder or something), you may want to switch
[13:17:15] <Bo^Dick> ok, it's good that this budget alternatives exist such as the tait-programmer but it's so incomplete
[13:17:49] <Bo^Dick> the software is incomplete, the compability is incomplete, the documentation is incomplete........
[13:18:10] <SWPadnos> "the typical open-source project" as some would say
[13:18:25] <SWPadnos> (I don't agree, but sometimes you do get what you pay for)
[13:19:32] <SWPadnos> incidentally, this is the hopco programmer page: http://www.hopco.com/
[13:20:00] <SWPadnos> not really excellent stuff, but it worked for us (and we programmed probably 20,000 PICs)
[13:23:04] <Bo^Dick> it looks like the only thing to do now is to really buy a programmer hardware
[13:23:17] <SWPadnos> probably.
[13:24:03] <Bo^Dick> thank god i only hooked the tait programmer on a breadbord
[13:24:09] <SWPadnos> you have two choices: 1) find something that works, or is close enough to working that you can figure out how to make it work or 2) buy something that works, and if it doesn't, you can call someone and ask how to make it work
[13:25:11] <Bo^Dick> it's a disaster for those who have really built the tait-design in a layout
[13:25:54] <SWPadnos> there shouldn't need to be any circuitry on your end product... (or are you talking about building a separate board?)
[13:26:35] <Bo^Dick> some people build programmer boards
[13:26:54] <SWPadnos> yes - that's necessary ;)
[13:27:09] <Bo^Dick> but it shouldn't really be needed since this is considered "in circuit"-programming
[13:28:04] <Bo^Dick> it is starting do dawn on me that programming a microcontroller is not gonna be easy for me
[13:28:18] <Bo^Dick> i won't get away cheap or without a serious effort
[13:28:21] <cradek> I've had decent luck programming AVRs with just wires from the parport
[13:28:41] <SWPadnos> "in-circuit" means that the microcontroller can be programmed without needing to remove it from the circuit
[13:28:51] <Bo^Dick> i haven't had luck myself that's for sure
[13:28:54] <SWPadnos> it doesn't mean that there's no circuit needed for programming
[13:29:44] <Bo^Dick> the software i finally found that claimed it really could program a pic18f452 tait-style launched an error that said that after all this feature wasn't supported
[13:30:00] <Bo^Dick> i might sound like i'm joking but this is actually what really happened
[13:30:51] <SWPadnos> in general, a family of parts will have one or two methods for programming (usually word mode or block mode, at least)
[13:31:26] <SWPadnos> each chip will have a different amount of memory to be programmed, and the configuration registers (fuses) will have different layouts and features
[13:31:55] <SWPadnos> they also have different IDs, so the softwre can verify that it's programming the right kind of chip
[13:32:30] <SWPadnos> if the sottware doesn't have the ID in its list, it won't recognize the chip, so it may not be able to choose the correct programming method, and it won't know where the fuses are
[13:32:45] <Bo^Dick> well the error message had nothing with the actual hardware connected to it
[13:33:05] <Bo^Dick> it did the exact same regardless if the pic was connected to it or not
[13:33:21] <SWPadnos> unless the 18Fxx series uses a totally different programming algorithm (possible, since the 16F algo is really stupid), then the software may just not have that chip in its database
[13:33:39] <SWPadnos> understood. you tell it to program an 18Fxx, and it says "I don't know how"
[13:34:07] <SWPadnos> but it may be a very simple fix in the code (or not, if the 18F is fundamentally different)
[13:34:37] <Bo^Dick> as a matter of fact i haven't tried to use the tait config with another target chip
[13:34:55] <Bo^Dick> maybe the software would do differently when choosing for example the pic16f84
[13:35:01] <Bo^Dick> lets give it a try...
[13:35:43] <SWPadnos> well, there are 4 things to swap out when doing this: 1) the PIC chip, 2: the board, 3: the PC, 4: the software
[13:35:49] <SWPadnos> (and 5: the operator ;) )
[13:37:24] <Bo^Dick> this is what it says; "this device is not supported with the TAIT parallel programmer"
[13:37:37] <SWPadnos> ok - then it probably won't work ;)
[13:38:10] <Bo^Dick> and it appears to display this regardless of which chip i've chosen in the software
[13:38:59] <SWPadnos> what software is this?
[13:39:05] <Bo^Dick> that raises a very good question. why does the software allow the paralel tait programmer to be chosen when it actually doesn't support it
[13:39:19] <Bo^Dick> this software is "icprog v1.05"
[13:42:06] <jepler> maybe you should ask the author of "icprog v1.05" or at least someone who's ever used it or heard of it before this moment.
[13:42:12] <Bo^Dick> in my opinion this behaviour is equivalent with an application with the only function of displaying "this application is not available" on startup and then immediately shut down
[13:42:37] <Bo^Dick> do you agree with my conclusion
[13:43:20] <jepler> if there's a bad application out there on the internets, that fact fills me with disgust and rancor.
[13:45:03] <SWPadnos> Bo^Dick, have you fiddled with the hardware check menu?
[13:45:17] <Bo^Dick> yeah
[13:45:23] <SWPadnos> there's a note next to the 18F452 that says "pull PGM to GND!" - whatever that means
[13:45:28] <Bo^Dick> one can invert some of the pins and stuff
[13:45:43] <SWPadnos> also, if your machine is fairly fast, you amy need to increase the I/O delay
[13:45:48] <SWPadnos> s/amy/may/
[13:45:51] <Bo^Dick> i know that too
[13:46:26] <SWPadnos> well - I'd have to go with jepler then - you should ask these questions on the IC-Prog message boards
[13:47:10] <Bo^Dick> :D
[13:47:25] <Bo^Dick> "hardware check is not supported with the TAIT parallel programmer"
[13:47:30] <Bo^Dick> :D
[13:47:40] <SWPadnos> heh
[13:47:42] <Bo^Dick> this is just too much for me
[13:47:47] <Bo^Dick> holy crap
[13:48:15] <Bo^Dick> the software insults me
[13:48:59] <SWPadnos> does it say something like this: "This software doesn't do that, moron. stop trying and get a real programmer"
[13:49:01] <SWPadnos> ;)
[13:49:25] <Bo^Dick> well i can't come up with an alternative interpretation
[13:49:46] <Bo^Dick> but in my opinion this is equivalent to a software....
[13:50:06] <Bo^Dick> ...that does nothing but say "this app is not available"
[13:50:31] <cradek> is this free software or did you buy it?
[13:50:55] <Bo^Dick> well it's free but does that excuse this crazy philosophy
[13:51:18] <Bo^Dick> why does it have the TAIT parallel option when it's not supported
[13:51:30] <Bo^Dick> that's just ridicioulus
[13:51:36] <Bo^Dick> isn't it
[13:52:03] <jepler> cradek: it appears to be sourceless software for windows, gratis download.
[13:52:25] <SWPadnos> "sourceless" - I like that term
[13:52:39] <cradek> then your only remedy is probably to stop using it
[13:52:58] <SWPadnos> which should be easy, since it doesn't work
[13:53:01] <Bo^Dick> it's still a something to be remembered
[13:53:01] <cradek> ... and try real free software next time
[13:53:28] <Bo^Dick> if i only know another software for windows that supports the TAIT parallel interface
[13:53:33] <cradek> * cradek channels RMS
[13:53:51] <SWPadnos> I'd ditch the tait programmer as well, unless you know how to write software that'll use it
[13:54:07] <SWPadnos> since it seems that you can't find someone else's software that works with it
[13:54:09] <cradek> doesn't pic give away their own programming environment? I think I even had it once
[13:54:17] <SWPadnos> yes, but no hardware
[13:54:21] <SWPadnos> MPLAB
[13:54:24] <cradek> right
[13:54:36] <Bo^Dick> mplab supports only microchip programmers
[13:54:38] <cradek> I guess I had eprom programming hardware I used (somehow)
[13:54:39] <Bo^Dick> afaik
[13:54:50] <cradek> that was so long ago!
[13:55:05] <SWPadnos> most of the multi-device programmers program EPROMS or chips (but not usually in-circuit(
[13:55:07] <SWPadnos> )
[13:55:24] <SWPadnos> chips = microcontrollers (oops)
[13:55:27] <cradek> I suggest you ditch the pics and use avr, which has honest-to-god free software support under linux
[13:55:57] <cradek> and you can do in-circuit programming with no special hardware other than 5 wires and a db25
[13:56:14] <skunkworks> cradek: now are the servos coming?
[13:56:38] <SWPadnos> as long as you have a clock source
[13:56:39] <cradek> still working on the mountings (bearings etc)
[13:56:51] <cradek> SWPadnos: they'll program with the built-in oscillators
[13:56:56] <SWPadnos> hmmm - no VCC in those 5 wires, I guess
[13:57:03] <cradek> right
[13:57:07] <SWPadnos> right - as long as the fuses are set to use the internal osc
[13:57:17] <cradek> yes, and they come that way
[13:57:31] <SWPadnos> right.
[13:57:42] <SWPadnos> I'm thinking of the problem with the 1200 vs. 1200A
[13:57:45] <cradek> brb
[13:58:19] <SWPadnos> the serial programming enable fuse is set to disallow serial prog in one of them (which makes it hard when you need to use ISP)
[14:00:18] <Didier> Hello everybody
[14:00:45] <SWPadnos> hello
[14:03:32] <Didier> SWPadnos : are you french ?
[14:03:39] <etla> what does the scope_rt HAL component do ?
[14:03:48] <SWPadnos> non. je ne sui pas Francais
[14:04:10] <SWPadnos> etla, it's the thing that actually samples the HAL signals for the UI to display
[14:04:28] <etla> but it starts automatically with emc ?
[14:04:39] <SWPadnos> Didier, you can tell from my terrible french grammar ;)
[14:05:04] <SWPadnos> etla, it may in some configs. it's also loaded by halscope if necessary
[14:05:50] <Didier> SWPadnos : your french is good "Non, je ne suis pas français"
[14:06:09] <Didier> (:
[14:06:10] <SWPadnos> heh - I was too lazy to find the c-with-cedilla character ;)
[14:06:35] <Didier> lazy ?
[14:06:43] <SWPadnos> it's not on my US keyboard
[14:07:05] <SWPadnos> so I have to find it with a character set viewer (or remember, which is impossible before I've had enough coffee)
[14:07:25] <Didier> then you are american
[14:07:32] <SWPadnos> yep
[14:07:51] <Didier> a lot of americans, no french (,:
[14:08:14] <SWPadnos> there are several who speak french, though I think they're from Canada
[14:08:48] <Didier> :)
[14:09:05] <robin_sz> ah, quand meme, je ne suis pas Canadien!
[14:09:27] <Didier> tu es français robin_sz
[14:09:39] <SWPadnos> hmmm - no emc users in France, according to frappr
[14:09:46] <robin_sz> heh, no. English ... but in Geneve
[14:10:00] <SWPadnos> heh - I don't remember Robin typing in french before ;)
[14:10:14] <Didier> what is frappr
[14:10:21] <SWPadnos> some of the swedes also probably speak french
[14:10:36] <SWPadnos> an image / mapping host - there's a link in the forum topic here
[14:10:47] <Didier> i will try
[14:11:21] <robin_sz> the office here used to mainly english speaking, but these days, its mostly french, I manage .. just.
[14:12:51] <skunkworks> I had 2 years of french in high scholl - but I think I can only count to 20
[14:12:57] <skunkworks> school
[14:13:33] <Didier> to work in native language it's easy
[14:13:41] <Didier> it's not my case
[14:15:35] <Didier> i have a question
[14:15:50] <Didier> when i install rtai, all is ok
[14:16:13] <Didier> but after rebooting, i have "error opening /dev/rtf..."
[14:16:25] <SWPadnos> what distribution?
[14:16:30] <Didier> Mandrake
[14:16:41] <SWPadnos> hmmm. does it use udev?
[14:16:48] <Didier> udev ?
[14:17:03] <SWPadnos> heh - what version of Mandrake?
[14:17:09] <robin_sz> mandrake? eeuw.
[14:17:15] <robin_sz> your ARE french! :)
[14:17:44] <SWPadnos> (udev is a system that creates device nodes when devices are present, it's very useful for USB and that kind of thing)
[14:17:53] <Didier> :-(
[14:18:06] <cradek> SWPadnos: multi-key c , => ç
[14:18:25] <SWPadnos> not on Windows, I think -> c, c ,
[14:18:47] <cradek> ok, I don't know about windows
[14:18:52] <SWPadnos> me either
[14:19:01] <cradek> but I have a multi-key (compose) that lets me type lots of things in iso8859-1
[14:19:02] <robin_sz> its shift 4 on my keyboard
[14:19:28] <cradek> but I rarely use it
[14:19:47] <etla> cradek: you develop AXIS ?
[14:19:49] <robin_sz> this is a qwertz kb
[14:20:01] <cradek> etla: yes jeff and I do
[14:20:22] <cradek> mine's regular american qwerty, I use Scroll Lock for compose
[14:20:27] <Didier> ç= Alt 0231
[14:20:27] <SWPadnos> it's alt-0231 on mine "ç"
[14:20:40] <SWPadnos> but I had to look it up ;)
[14:20:45] <cradek> ugh
[14:20:46] <robin_sz> we have US, swiss-french and swiss-german kbs here
[14:20:49] <etla> cradek: one suggestion: when switching between MDI and Manual control (jog), could pressing F3/F5 change the mode to manual / MDI ?
[14:20:51] <cradek> compose is better
[14:21:12] <etla> cradek: now I have to first press F3, then press a jog (+ or -) and then the EMC mode changes to manual
[14:21:18] <robin_sz> a few azerty french ones too ... and thats before we confuse ourselves with the pesky MACs!
[14:21:19] <SWPadnos> no way. Microsoft has been innovating using alt+keycode for years!
[14:21:48] <cradek> etla: why does it matter what mode it's in? axis changes to the necessary mode for you
[14:21:56] <jepler> cradek: I think etla is jogging with the jog wheel
[14:22:10] <Didier> so
[14:22:20] <etla> jepler: you got it ! I wan't to switch to manual mode and then jog with halui
[14:22:29] <etla> and not have to press +/- in between
[14:22:57] <cradek> ah I see
[14:23:07] <alex_joni> morning all
[14:23:08] <jepler> I believe halui should do the same thing as axis: issue a command to change the mode from MDI to manual before issuing a jog command
[14:23:37] <etla> jepler: well, but I don't plan on having an external mode button, I'd like to use the AXIS one
[14:24:02] <alex_joni> jepler: maybe an autoswitch param to halui?
[14:24:06] <SWPadnos> etla, you're mixing up two UIs then
[14:24:08] <cradek> there's a bigger problem - it's not obvious when it's safe to jog. You don't know when the machine might be doing something.
[14:24:10] <alex_joni> if set, it does like AXIS
[14:24:25] <alex_joni> but I'd rather see it work like tkemc, with mode selection
[14:24:25] <SWPadnos> that would be like hitting a button in tkemc to change something in axis
[14:24:43] <etla> SWP: No. I'm simply asking that in axis F3 would set emc to manual mode and F5 would set it to MDI mode
[14:25:15] <etla> now you have to do F3+jog, or F5 + write g0x0 or similar to get the mode to change
[14:25:17] <Didier> where can i find help on rtai install /dev/rtf
[14:25:27] <jepler> cradek: you can do the same as axis (manual_ok() and ensure_mode(emc.MODE_MANUAL))
[14:25:35] <alex_joni> Didier: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?TroubleShooting
[14:25:50] <SWPadnos> Didier, I think you need to set up something with udev, but I'm not sure how to do it. alternately, you can create the RTAI device nodes at startup, by making an rc script or something like that
[14:26:09] <jepler> Didier: http://www.captain.at/rtai-error-opening-dev-rtf.php (but on systems that use udev, which is most modern linux systems, you have to create the /dev/rtf/N device entries at each reboot)
[14:26:34] <jepler> Didier: however, emc2 does not use /dev/rtf during normal operation; it uses only /dev/rtai_shm. the latency test uses one of the /dev/rtf devices.
[14:27:22] <Didier> it's the same problem /dev/rtai_shm /dev/rt...
[14:28:09] <SWPadnos> etla, yes - axis is a little different from the other UIs in that it tries to be modeless (as far as possible), so it automatically switches for you when you jog or enter MDI commands
[14:28:50] <cradek> halui is going to have to do the switches
[14:28:53] <Didier> i will try, thanks
[14:28:58] <Didier> :)
[14:29:03] <cradek> imagine you have JUST halui with a cycle-start (run) button and a jog wheel
[14:29:04] <SWPadnos> good luck
[14:29:15] <cradek> it has to do the switch somehow.
[14:29:21] <SWPadnos> right - each UI has to control how emc should interpret its commands
[14:29:59] <etla> cradek: but in this case it's unneccessary, I wan't to do certain things with halui, and certain things with a normal UI (AXIS). Now you're saying that _just_ changing modes is not possible in AXIS...
[14:30:06] <cradek> I think the artificial modality stuff is silly. we talked about this at length at workshop.
[14:30:14] <Didier> well received, SWPadnos
[14:30:19] <SWPadnos> you have to look at halui as being equivalent to tkemc / mini / axis. it would be weird to have to change something in mini so that axis would do behave differently
[14:30:41] <cradek> brb
[14:31:29] <etla> SWP: still, I think when I press F3/F5 in AXIS, AXIS whould make emc change to the appropriate mode. I don't see why you could be against this ??
[14:32:05] <SWPadnos> I'm just pointing out that it's a strange scheme to use one UI (axis) to modify the behavior of another (halui)
[14:32:45] <etla> SWP: I'm not modifying halui in any way, my halui HAL script will check that the mode is correct before issuing jog commands
[14:33:15] <SWPadnos> can you not have a button on the panel that toggles jog mode (and a light to tell you if it's on)?
[14:33:35] <alex_joni> SWPadnos: 3 buttons, one for Manual, one for Auto, one for MDI
[14:33:40] <etla> yes, but I think it's unneccessary, I want to use the keyboard for changing modes
[14:33:41] <alex_joni> and one led for each
[14:33:56] <alex_joni> etla: ideally the halui panel, and the PC are far apart
[14:34:04] <alex_joni> at least on some systems
[14:34:21] <alex_joni> etla: if your GUI changes the mode, then halui should be happy
[14:34:44] <SWPadnos> theoretically, the entire machine should be operable from either halui *or* axis, independently of each other
[14:34:54] <SWPadnos> alex_joni, axis doesn't - that's the problem ;)
[14:34:57] <etla> alex: yes ! and I think AXIS is the GUI of choice, but the developers don't want to have AXIS change modes on F3/F5...
[14:35:05] <alex_joni> etla: patch it
[14:35:06] <SWPadnos> it only changes when you issue an MDI command or jog
[14:35:23] <alex_joni> SWPadnos: if they feel that's the way to do it, then I can't say anything ;)
[14:35:23] <etla> alex: yeah ! ;)
[14:35:36] <alex_joni> if I wouldn't like it, I would probably get the sources and modify it
[14:35:41] <alex_joni> but I don't really care about that
[14:35:47] <SWPadnos> right. so etls, you can change to tkemc or mini, and it'll work the way you want
[14:35:51] <alex_joni> otoh, AXIS works well as it is
[14:35:51] <SWPadnos> etla, that is ;)
[14:36:04] <alex_joni> etla: for what I care you could even run tkemc & AXIS & halui
[14:36:14] <alex_joni> but that would be actually way too crazy
[14:36:27] <etla> SWP: I really like the backplot/preview in AXIS
[14:36:33] <SWPadnos> me, too ;)
[14:36:36] <alex_joni> so.. I guess the actual answer is having buttons in halui
[14:37:32] <SWPadnos> that's my view
[14:38:22] <etla> I might just have to patch AXIS then...
[14:39:02] <SWPadnos> actually, you could patch halui instead: make it look for MDI mode, and switch to jog mode if necessary
[14:39:11] <SWPadnos> but don't do that if it's in auto mode
[14:40:26] <etla> I think I want the jogwheel active only in manual mode
[14:41:37] <etla> but AXIS does not give me a way to only change modes... something silly like coolant on/off does change to manual mode
[14:41:40] <SWPadnos> it makes sense that most things in halui be disabled when in auto mode, but operable in either manual mode (MDI or jog)
[14:42:09] <etla> yes.
[14:43:03] <SWPadnos> I think the way to look at this is "how can I make halui operate correctly, independent of other UIs" - not "how can I get axis to make halui work the way I want?"
[14:43:53] <etla> well I think it's illogical to have F3/F5 buttons with the modes as labels and then have pressing F3/F5 do nothing !
[14:44:04] <SWPadnos> heh - that may be true ;)
[14:44:51] <etla> brb
[14:45:43] <cradek> it's a gui blooper if bringing a tab to the top performs an action
[14:46:14] <cradek> another problem is that there are two tabs but three modes, your halui gui needs to switch to auto mode before you can use the run/abort/pause buttons.
[14:46:49] <cradek> so you need to switch the modes with halui, either with buttons, or automatically when appropriate
[14:47:06] <cradek> did I say lately that I think the modes are bogus?
[14:47:10] <SWPadnos> thatwas my thought - make halui switch modes as necessary
[14:47:25] <SWPadnos> hmmm - I think you may have said that more recently than I mentioned that I hate LabView
[14:47:34] <cradek> ok your turn then
[14:47:42] <SWPadnos> thanks. I hate LabView
[14:47:48] <SWPadnos> phew - I feel better now
[14:47:59] <cradek> I think the emc modes are bogus
[14:48:34] <SWPadnos> yes and no - having status tell the UI (or multiple UIs) that it's executing a program, rather than in "free" mode is a good thing
[14:48:49] <SWPadnos> so a n auto/manual switch makes sense
[14:48:56] <cradek> I agree, that's the real modality, which is unrelated to these gui modes we are discussing
[14:49:15] <SWPadnos> I thought emc itself reflects those 3 modes ...
[14:49:22] <cradek> and as far as I can tell, that real modality is NOT represented in the stat buffer anwyhere
[14:49:40] <SWPadnos> ugh
[14:49:54] <cradek> emc has three modes (the gui ones) and when in one, it rejects commands from the others
[14:50:02] <cradek> I think it has no other effect
[14:50:23] <cradek> I don't see any reason for it. There's real modality that we need (like you said, in free mode or not)
[14:50:26] <SWPadnos> hmmm. then why would there be an issue with halui sending jog commands when axis is in MDI mode?
[14:50:34] <SWPadnos> (or is there one?)
[14:50:47] <cradek> because jogs are rejected when in MDI mode
[14:50:58] <SWPadnos> then it is an emc mode, not just the UI
[14:51:02] <cradek> yes
[14:51:21] <SWPadnos> ah - OK. more coffee for me :)
[14:51:29] <cradek> axis hides them from the user (takes care of a detail the user shouldn't have to worry about)
[14:51:54] <cradek> only now that the user has to worry about it (because halui doesn't) there's a problem
[14:53:05] <SWPadnos> right. so we keep coming back to "fix halui"
[14:53:20] <SWPadnos> or use it with sometihng other than axis
[14:53:39] <etla> maybe a third UI app that would have a mode selector switch
[14:54:06] <SWPadnos> I think it would be easier to modify halui to automatically switch modes as necessary (like axis)
[14:54:15] <Didier> bye everybody
[14:54:19] <SWPadnos> see you Didier
[14:54:30] <Didier> and thanks
[14:54:47] <etla> well, if I'm in auto or mdi mode and I accidentally brush the jogwheel I don't want it to autmatically switch to manual mode and jog
[14:54:50] <SWPadnos> you're welcome
[14:55:26] <cradek> etla: I envisioned the jogwheel's axis-selecting switch to have another setting "disabled"
[14:55:29] <SWPadnos> true. that begs for buttons and lights on the panel
[14:55:59] <alex_joni> SWPadnos: that's the whole purpose of halui
[14:56:12] <SWPadnos> heh - I know that ;)
[14:56:13] <etla> cradek: yes that's one possibility
[14:57:39] <etla> but when I have a whole keyboard full of buttons and about 1000x700 lights in front of me I thought I'd implement only what I really want on the jog-pendant. not a mode selector...
[14:58:16] <SWPadnos> etla, I think you and I are looking at this differently
[14:58:34] <SWPadnos> you're thinking of halui and the pendant as an extension of the on-screen GUI
[14:58:50] <SWPadnos> I'm thinking about it as a totally separate and independent UI
[14:59:15] <SWPadnos> and as it happens, axis isn't written to have an auxiliary pendant attached, and halui is written to be independent of other UIs
[14:59:18] <alex_joni> SWPadnos: ideally it should do both
[14:59:27] <alex_joni> halui that is
[14:59:36] <etla> swp, I'm not going to want to do _everything_ from the pendant anyway. not load files, issue g0x5f600, etc. etc. so why should I have to do mode selection ?
[14:59:41] <alex_joni> it should have enough pins & co to be usefull in any conditions
[14:59:54] <SWPadnos> only if there are hooks into the GUIs to let halui signal them that things are happening (like keystrokes or mouseclicks)
[15:00:16] <SWPadnos> because it's the operation of the pendant that requires you to change modes
[15:01:12] <SWPadnos> alex_joni, halui talks to emc via NML, and to HAL via pins - it has no way of directly communicating with a GUI, right?
[15:01:39] <SWPadnos> or is there an interface for that?
[15:01:40] <etla> I think so yes
[15:01:43] <alex_joni> SWPadnos: no
[15:02:05] <SWPadnos> ok, so halui needs to operate independently of any other UI that may be connected to a running emc
[15:02:10] <alex_joni> right
[15:02:21] <alex_joni> but it shouldn't be bothered if theres another one running
[15:02:37] <SWPadnos> right, but it shouldn't depend on another one running either
[15:02:41] <cradek> unfortunately that means it has to know how to switch modes.
[15:02:47] <SWPadnos> right
[15:03:00] <cradek> because otherwise its messages will be rejected
[15:03:26] <etla> I already got a few 'can't do that while in MDI mode'
[15:04:24] <alex_joni> cradek: it also needs a way to display error messages :)
[15:04:31] <alex_joni> error led / clear error pin
[15:04:58] <SWPadnos> heh - unfortunately, every UI attached will get the error, no?
[15:05:09] <alex_joni> luckily
[15:05:12] <alex_joni> yeah ;)
[15:05:13] <SWPadnos> so there could be dialogs on 6 computers ;)
[15:05:41] <robin_sz> wow, thats almost as good as that wiondows pop-up message thing ...
[15:06:20] <etla> ok, so AXIS does not like the idea of modes. maybe it's possible to make a simple mode selector UI that works on-screen
[15:06:43] <robin_sz> you know, the one when you spend a morning testing your net-enabled UPS and discover that ALL the users in your 1200 person site have been suffering them and the helldesk is going crazy ...
[15:06:54] <robin_sz> not that I'd know anything about that
[15:07:37] <SWPadnos> heh - oops
[15:07:48] <SWPadnos> etla, that would work
[15:08:25] <SWPadnos> another possibility that's a little more convoluted, but may be the "right" solution:
[15:08:46] <SWPadnos> halui should be able to take care of itself, so it should have the ability to check and change modes as needed
[15:09:13] <SWPadnos> but, you can also add a pin to halui that enables/disables this functionality (or that directly sets the mode)
[15:09:29] <alex_joni> SWPadnos: might be true, but unlike AXIS you can't display stuff to the user.. so it's not trivial to be done
[15:09:35] <SWPadnos> that pin can be connected to hardware via HAL, or to VCP via HAL, if you want an onscreen button
[15:09:55] <SWPadnos> VCP + HALUI can do a lot, I think
[15:10:01] <etla> yes, I want thinking about VCP
[15:10:12] <etla> probably need to learn that next... hope it's not too hard
[15:10:22] <alex_joni> etla: there's a sample config
[15:10:25] <alex_joni> halui + halvcp
[15:10:42] <SWPadnos> alex_joni, does halui have a mode input pin?
[15:10:47] <etla> how would I start halvcp
[15:10:53] <etla> swp: three inputs for the three modes
[15:11:05] <SWPadnos> halcmd loadusr halvcp <probably some file goes here>
[15:11:15] <alex_joni> etla: I put a shell script in there
[15:11:25] <alex_joni> run emc, select the halui-halvcp config
[15:11:29] <alex_joni> and then run that script
[15:11:37] <SWPadnos> ok, then if halvcp has radio buttons yet, then your GUI is very simple to make
[15:12:03] <alex_joni> SWPadnos: no radio buttons
[15:12:08] <alex_joni> but pushbuttons
[15:12:08] <SWPadnos> bummer
[15:12:20] <SWPadnos> I was thinking of the mutual exclusivity aspect
[15:12:32] <alex_joni> that's ok.. you can't push more than one at once :D
[15:12:40] <alex_joni> your mouse doesn't move that quickly
[15:12:52] <alex_joni> and they do act mutual exclusive, because the feedback comes from emc
[15:12:57] <SWPadnos> ok, are the halui pins one-shots, or "level sensitive"?
[15:13:02] <alex_joni> one shot
[15:13:05] <SWPadnos> cool then ;)
[15:13:45] <alex_joni> etla: please allow halvcp a few seconds to finish loading before pushing buttons (they won't do anything at first)
[15:13:54] <alex_joni> 2-3 seconds or so
[15:14:08] <alex_joni> err.. 5 seconds
[15:14:24] <etla> alex: I now started the halui halvcp config, what do I doo next ?
[15:14:31] <alex_joni> open a new terminal
[15:14:39] <alex_joni> go to emc2/configs/halui_halvcp
[15:14:48] <alex_joni> ./load_halvcp.sh
[15:15:03] <etla> yep, got it
[15:15:14] <etla> it has a mode selector !
[15:15:18] <SWPadnos> heh
[15:15:19] <alex_joni> it does :D
[15:15:22] <alex_joni> * alex_joni put it there
[15:15:42] <etla> and tkemc responds when I press the button in halvcp
[15:15:52] <alex_joni> now that's a surprise :D it really works
[15:16:00] <etla> so it's probably fairly simple to get this demo panel running along with AXIS
[15:16:09] <alex_joni> etla: it also has machine on/off, etc
[15:16:13] <alex_joni> but you probably noticed that
[15:16:22] <etla> yah
[15:16:29] <SWPadnos> now why didn't you mention this earlier? ;)
[15:17:31] <etla> that was a quick 'patch' for AXIS ;)
[15:17:46] <SWPadnos> heh
[15:18:02] <SWPadnos> can someone look at this auction, and tell me if it makes any sense:
[15:18:07] <SWPadnos> http://cgi.ebay.com/ebaymotors/RIDGELINE-RTL-NAVAGATION-LEATHER-DVD-Damaged-2K-miles_W0QQitemZ4647755920QQcategoryZ119378QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
[15:20:10] <SWPadnos> "runs perfect, new truck, right axle broken (and the airbag has popped, and it looks like the seat belt was cut after the accident, whicn looks like a whale fell on it) ...
[15:20:16] <cradek> makes sense in what way?
[15:20:27] <SWPadnos> "perfect ... broken"
[15:20:34] <SWPadnos> it just seemed contradictory
[15:21:40] <cradek> about this style of car in general: how long does the bed have to be in order to allow you to pretend you're driving a rugged pickup truck?
[15:21:47] <SWPadnos> heh
[15:22:06] <cradek> looks like about 3.5 feet?
[15:22:09] <SWPadnos> I don't generally like short-bed trucks - they look like some stupid "me-too" design
[15:22:16] <SWPadnos> 5 actually
[15:22:25] <SWPadnos> but that may be with the gate down ;)
[15:22:56] <cradek> occasionally I wish I had (or could borrow) a real pickup truck
[15:23:04] <cradek> but I can't see driving something like this every day, yuck
[15:23:09] <SWPadnos> it's the top pick from car& driver, consumer reports, and some truck magazine
[15:23:26] <SWPadnos> I agree, but for towing lathes around the country, it's great ;)
[15:23:29] <cradek> top pick for what kind of use?
[15:23:36] <SWPadnos> for a truck
[15:23:37] <etla> alex: what does vcp stand for ?
[15:23:44] <SWPadnos> Visual Control Panel, I think
[15:23:44] <cradek> huh
[15:24:19] <cradek> to me a truck has 4x8' between the wheel wells
[15:24:26] <SWPadnos> yep - I agree
[15:24:47] <SWPadnos> even the long version of my minivan has 4x8 falt capacity with the seats out
[15:24:50] <cradek> this thing looks no more useful than a car with a big trunk (and it doesn't keep the rain off your groceries)
[15:24:51] <SWPadnos> flat
[15:25:07] <SWPadnos> well it does have a large cab - larger than many sedans
[15:25:36] <cradek> I'm sure that's true
[15:25:53] <cradek> although looks like the back seat legroom is a joke
[15:26:03] <SWPadnos> it's pretty good, actually.
[15:26:07] <SWPadnos> the bed is 5 feet
[15:26:30] <SWPadnos> anyway - it loks like the van is repairable, so I should stop looking at these ;)
[15:26:45] <cradek> yay
[15:26:54] <SWPadnos> yep
[15:26:56] <cradek> I've never had a van but they seem really useful
[15:27:01] <SWPadnos> makes it much easier on the pocketbook
[15:27:12] <SWPadnos> well, it sure got a lot of junk to Fest this year ;)
[15:27:14] <cradek> like my bronco (a great compromise vehicle) but better still
[15:27:42] <cradek> yeah, perfect for stuff like that
[15:28:03] <SWPadnos> I've beed surprised at the fuel economy of this 10-year old van - I get 25+ MPG on the highway, and 20 or so in town
[15:28:09] <SWPadnos> been
[15:28:14] <cradek> wow that is surprising
[15:28:15] <robin_sz> robin_sz is now known as stopnow
[15:28:28] <SWPadnos> that's almost as good as the new Accord we got last year
[15:28:30] <stopnow> stopnow is now known as robin_sz
[15:28:34] <cradek> the bronco with a 351 gets maybe 9/11
[15:28:44] <SWPadnos> ugh
[15:28:51] <etla> anyone know if motion.spindle-revs is supposed to work ?
[15:28:56] <cradek> but it'll get you through any amount of snow or mud and still get 9 :-)
[15:29:01] <SWPadnos> I wish I could afford a hybrid vehicle
[15:29:05] <SWPadnos> heh
[15:29:15] <SWPadnos> my economy does go down a bit with the studded tires on
[15:29:24] <cradek> etla: that's used for threading, you hook it to an encoder
[15:29:36] <etla> ok so it's an input
[15:29:44] <cradek> etla: I think so
[15:29:46] <etla> is there an output that would show the programmed revs
[15:30:07] <cradek> etla: yes, I forget what it is
[15:30:19] <cradek> I know the nist-lathe config uses it to drive the spindle speed control
[15:30:34] <etla> iocontrol spindle speed out ?
[15:30:40] <cradek> yeah that's it I bet
[15:31:10] <etla> ok, I get 500 as default when spindle is forward
[15:32:04] <cradek> linkps iocontrol.0.spindle-speed-out => spindle-speed-cmd
[15:35:48] <alex_joni> etla: that's the one
[15:36:04] <alex_joni> etla: default spindle speed on is 500, you can alter with Sxxx in your program
[15:36:38] <etla> alex: I'm trying to fill in the spindle section in your demo vcp file... no luck yet.
[15:37:22] <etla> seems the script does not clean up after it on exit
[15:37:54] <etla> or maybe I just broke it...
[15:38:45] <etla> it doesn't like: LED {
[15:38:46] <etla> halpin = iocontrol.0.spindle-forward
[15:38:46] <etla> on-color = "#00F"
[15:39:27] <SWPadnos> there's already a pin by that name - you need to name it differnetly, then connect it to that pin (I think)
[15:39:29] <etla> in a vcp, does halpin = xxx, automatically create signal xxx
[15:39:31] <etla> ?
[15:39:37] <SWPadnos> it shouldn't
[15:39:51] <SWPadnos> it creates a pin, which you connect later
[15:40:55] <etla> yeah, they are linked with linkpp in the hal file
[15:41:05] <SWPadnos> right
[15:41:27] <etla> so you cant do the linking in the vcp file directly, need to create a pin and then link in the hal file
[15:41:28] <SWPadnos> you may want that line to read "halpin = fwd-led"
[15:41:35] <SWPadnos> corerct, I believe
[15:41:45] <SWPadnos> halvcp makes the UI and pins, halcmd makes the connections
[15:41:54] <etla> how does the vcp file determine the type (bit/float etc) of the pin to be created
[15:42:02] <etla> or is a LED pin always bit etc
[15:42:04] <SWPadnos> LEDs are bits
[15:42:12] <etla> right
[15:42:13] <SWPadnos> yep - it's implicit in the control type
[15:43:17] <SWPadnos> the only place where it wouldn't be is int vs float, for a slider / meter or the like
[15:43:25] <SWPadnos> (but those aren't there yet, so ... ;) )
[15:46:11] <etla> hmm... halmeter shows the iocontrol pin changing, but my newly created LED does not change
[15:46:41] <SWPadnos> did you linkpp/linksp it?
[15:46:56] <etla> yes, linkpp led.spindle-forward iocontrol.0.spindle-forward
[15:47:17] <SWPadnos> hmmm
[15:47:51] <etla> the LED is halpin = led.spindle-forward
[15:48:21] <SWPadnos> and halcmd show sig shows the connection? (did you re-start halvcp?)
[15:49:28] <etla> I don't see the connection with show sig...
[15:50:04] <SWPadnos> ah. did you unload then reload halvcp (along with the .hal file that connects it to everything else)?
[15:50:26] <etla> I just pressed the X in the right top corner to close it
[15:50:51] <SWPadnos> ok. you need to reload the hal file as well to make the new connections
[15:53:07] <etla> closing the vcp window only seems to kill the app, not delete all the old signals and connections
[15:53:13] <SWPadnos> that's true
[15:53:17] <NickServ> This nickname is owned by someone else
[15:53:17] <NickServ> If this is your nickname, type /msg NickServ IDENTIFY <password>
[15:53:24] <etla> then when I re-run the vcp start script HAL complains about duplicate signals etc etc
[15:53:42] <SWPadnos> yep - too bad there's no "revert this hal file" function in halcmd ;)
[15:54:13] <etla> I'll restart emc
[15:54:22] <SWPadnos> that may be the simplest way :(
[15:56:11] <etla> yep. now my spindle-forward led works
[15:56:26] <etla> VCP is the answer to completing 'crippled' jog-pendants
[16:13:01] <alex_joni> http://lerch.no-ip.com/ChangFa_Gen/ <- seen this?
[16:23:13] <etla> the frames don't work in firefox ??
[16:24:24] <etla> alex: are there more widgets implemented in VCP, or only LED and BUTTON ?
[16:25:23] <alex_joni> etla: not sure, I think those are the only ones for now
[16:25:29] <alex_joni> this guy is crazy: http://lerch.no-ip.com/atm/
[16:26:29] <etla> crazy a bit like the emc team ?
[16:27:34] <alex_joni> that's beyond our crazy I'd say :D
[16:30:23] <etla> how could I load the vcp shell script everytime emc starts ?
[16:31:27] <alex_joni> add it to the end of the emc runscript
[16:31:41] <alex_joni> or use 'loadusr' in the normal hal file
[16:31:47] <SWPadnos> can halcmd loadusr be used to run the script from within the ini file?
[16:31:48] <etla> thats scripts/emc
[16:31:58] <SWPadnos> I guess so ;)
[16:33:16] <skunkworks> always wanted to gring my own mirror
[16:33:20] <skunkworks> grind
[16:33:28] <etla> do it with a cnc machine ??
[16:33:30] <etla> ;)
[16:33:32] <skunkworks> (father did)
[16:34:54] <alex_joni> SWPadnos: yeah
[16:36:44] <bill203> neat: http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=7626230885&
[16:36:51] <bill203> I could make some really small parts. :-)
[16:38:52] <alex_joni> http://www.topromanian.com/c/index.php/50/Dacia+Logan+@+TopGear.html
[16:39:59] <etla> where do I check the screen resolution in ubuntu ??
[16:40:35] <cradek> xdpyinfo will show it
[16:40:44] <cradek> % xdpyinfo|grep dimensions
[16:40:44] <cradek> dimensions: 2752x1200 pixels (706x311 millimeters)
[16:41:15] <etla> thanks
[16:41:21] <alex_joni> etla: there is a menu entry under administration? configuration? something like that, where you can select the resolution
[16:42:21] <alex_joni> SWPadnos: that might be a car for you the last one I pasted
[16:42:30] <SWPadnos> watching
[16:42:35] <SWPadnos> the SST? ;)
[16:43:00] <alex_joni> the one from TopGear
[16:43:06] <alex_joni> last one
[16:43:10] <etla> where is the vcp code so I could look it there is other stuff than LED and BUTTON ?
[16:43:14] <SWPadnos> the Renault?
[16:43:40] <alex_joni> it's actually Dacia
[16:43:43] <alex_joni> by Renault :D
[16:43:49] <alex_joni> but it's 3k quid
[16:43:53] <alex_joni> or about 5k $
[16:43:58] <SWPadnos> cheap
[16:44:17] <SWPadnos> have you seen the episode where they try to break a Toyota HiLux pickup truck?
[16:44:19] <alex_joni> better than searching ebay for older ones :D
[16:44:22] <alex_joni> no..
[16:44:23] <SWPadnos> heh
[16:44:32] <SWPadnos> it's great. I'll see if I can find it
[16:46:08] <SWPadnos> http://www.youtube.com/watch?v=nD0UVI99R8Q
[16:48:44] <alex_joni> thx
[16:51:43] <etla> bye guys.
[16:55:15] <alex_joni> SWPadnos: crazy :D
[16:55:20] <SWPadnos> yep
[16:55:27] <SWPadnos> watch the whole thing - it's pretty amazing
[16:55:49] <alex_joni> I'm almost half through :D
[16:56:02] <SWPadnos> what's the last thing they did to it?
[16:56:03] <alex_joni> great comercial for toyota :D
[16:56:11] <alex_joni> they just rescued it from the sea
[16:56:19] <SWPadnos> ah - lots more to come ;)
[16:56:47] <alex_joni> they just dropped a caravan over it
[16:57:01] <SWPadnos> heh. lots more to come ;)
[16:57:02] <alex_joni> one of those trailers
[16:59:29] <SWPadnos> bummer - the sound seems screwed up toward the end
[16:59:42] <alex_joni> fire next ;)
[16:59:52] <SWPadnos> more to come ;)
[17:00:10] <alex_joni> I figured :D
[17:00:46] <SWPadnos> once I saw that video, I thought - "that's a vehicle I could be happy with"
[17:03:04] <alex_joni> sound's gone :(
[17:03:35] <SWPadnos> yeah - it was broken up on my machine - totally unintelligible
[18:57:21] <Jymmm> Jymmm is now known as Red70sShow
[18:57:21] <Red70sShow> Red70sShow is now known as Jymmm
[19:20:59] <les_w> hi all
[19:21:06] <Jymmm> howdy Mr Les
[19:21:17] <les_w> had the meeting today
[19:21:34] <Jymmm> did you get em all drunk first?
[19:21:40] <les_w> tried to sell em on rt linux automation....
[19:21:42] <alex_joni> hi les
[19:21:54] <les_w> They asked how much to get this going
[19:22:16] <les_w> I said Idon't know.....half a million or so
[19:22:25] <les_w> they said.....
[19:22:33] <les_w> ok. that's fine.
[19:22:42] <les_w> I am a happy person.
[19:22:46] <alex_joni> lol
[19:22:49] <cradek> will you get to try emc2 now?
[19:22:50] <alex_joni> nice one les
[19:22:52] <SWPadnos> heh - give me a call sometime ;)
[19:23:12] <les_w> swp I will need lots of help on this
[19:23:17] <Jymmm> les_w shoulda said 3/4 of a million
[19:23:18] <les_w> payng help of course
[19:23:24] <SWPadnos> even better ;)
[19:24:04] <les_w> I'm really sold on using a real time linux factory automation system.
[19:24:10] <Jymmm> * Jymmm holds up digital multi-colored led sign.... "WILL WORK FOR CASH"
[19:24:19] <les_w> haha
[19:24:29] <les_w> gee I will work for cash too
[19:24:43] <les_w> I just hope the damn economy holds together
[19:24:45] <Jymmm> les_w congrats!!!
[19:24:50] <les_w> thx
[19:24:50] <SWPadnos> I think most of us would - if we had the opportunity
[19:25:17] <cradek> will work for servos
[19:25:25] <les_w> oh all the finger burner physics has to work right....
[19:25:28] <les_w> np
[19:25:33] <alex_joni> les_w: does rt linux mean RTLinux ? hope not..
[19:25:38] <les_w> the analyzer is working out great
[19:25:56] <les_w> no alex RTAI from what you folks tell me
[19:26:06] <alex_joni> les_w: that's the way ;)
[19:26:27] <cradek> I bet we could all use a little extra money and might be able to help out in different ways.
[19:26:41] <les_w> I think so
[19:27:27] <les_w> big bucks for writing guis?
[19:27:31] <les_w> ha
[19:27:49] <cradek> gui is already done, you can give us big bucks for it if you want
[19:27:57] <cradek> we'll release a special gpl-licensed version to you
[19:28:04] <les_w> heh
[19:28:04] <alex_joni> lol
[19:28:15] <cradek> that means you get the source code too!
[19:28:22] <alex_joni> cradek: release a special BSD one ;)
[19:29:19] <les_w> well thgere is some motion control involved but a lot of it is a very fast trim to frequrncy and Q of a fourth order transducer
[19:29:29] <les_w> frequrncy
[19:29:50] <les_w> ah it wasn't me...kwyboard is stuck!!
[19:30:00] <les_w> bleh
[19:30:01] <cradek> maybe you can afford a new keyboard
[19:30:03] <SWPadnos> heh
[19:30:04] <les_w> yeah
[19:30:46] <cradek> brb
[19:30:48] <les_w> These types of automation would used adaptive control
[19:30:50] <les_w> k
[19:31:18] <les_w> so feedforward coeficients turn into feedforward functions
[19:32:07] <SWPadnos> you'd probably end up with a DSP or othercontroller doing the actual laser control
[19:32:18] <SWPadnos> other controller
[19:32:45] <les_w> From what I can see it could be done native to the pc
[19:32:47] <alex_joni> laser?
[19:32:57] <SWPadnos> it may not be needed if there's a suitably speedy general-purpose motherboard that's been tuned for low latency operation
[19:33:00] <alex_joni> did someone say "giant laser" ?
[19:33:03] <les_w> I mean we have 10k servo updates/sec already right?
[19:33:10] <SWPadnos> sometimes, yes
[19:33:43] <SWPadnos> this would be a single-purpose machine, so some liberties could be taken with cache control and that sort of thing
[19:34:03] <SWPadnos> (or multiple processors - like a dual-core with one core always running the RT code)
[19:34:31] <les_w> we would just whack the device with an impulse....do some FFts, crunch some more numbers based on history, and ablat with about a 50 watt YAG
[19:34:43] <SWPadnos> how many points in the FFT?
[19:34:46] <SWPadnos> (roughly)
[19:34:59] <les_w> hmmm
[19:35:01] <les_w> not much
[19:35:03] <les_w> 1024
[19:35:07] <les_w> or so
[19:35:34] <les_w> might be more or less since its a Heisenburg thing
[19:35:58] <les_w> I only need a few hundred PPM frequency accuracy
[19:36:00] <SWPadnos> I don't think you'd want to run the numbers to adapt the algorithm (number of points) in realtime
[19:36:15] <les_w> no it would not need to be
[19:36:27] <les_w> prob only direct beam control
[19:37:32] <SWPadnos> complex or real?
[19:37:36] <les_w> this will be fast enough that heating will perturb the results
[19:37:51] <les_w> FFTs are always complex!!!
[19:37:53] <les_w> haha
[19:38:04] <les_w> but yeah a full complex fft
[19:38:05] <SWPadnos> heh - the results are, but the source may not be
[19:38:23] <les_w> right
[19:38:31] <les_w> that would be the case here
[19:39:47] <les_w> The key is to make them adaptive
[19:39:58] <SWPadnos> adaptive to what?
[19:40:02] <les_w> the old dos/c systems I made did that
[19:40:10] <les_w> they get smarter and smarter
[19:40:29] <SWPadnos> ah - "lst time I had to burn off this much, and this looks about the same ..."
[19:40:39] <SWPadnos> s/lst/last/
[19:41:21] <les_w> haha
[19:41:49] <les_w> For me some parts of the plant had big delays
[19:41:56] <les_w> like part is made
[19:42:00] <SWPadnos> that's a slow function anyway - aggregate a few tens or hundreds more samples, then twiddle with the coefficients
[19:42:13] <SWPadnos> (the adaptive control)
[19:42:14] <les_w> cooked on conveyor
[19:42:20] <les_w> and then measured
[19:42:36] <les_w> but the transfer fuction of the delay was in the code
[19:42:52] <les_w> so it new when to apply changes
[19:43:01] <les_w> in phase to keep atability
[19:43:07] <les_w> that sort of thing
[19:43:16] <les_w> stability
[19:43:55] <les_w> like a servo with a time delay on the encoder.....
[19:44:20] <SWPadnos> right
[19:44:36] <SWPadnos> I think phase lag isn't taken into account in the existing emc codebase
[19:45:05] <les_w> Hi lawrence
[19:45:38] <les_w> Well , its doable. I can write the system functions anyway
[19:45:54] <SWPadnos> that's good. the CPUs can run the numbers just fine ;)
[19:46:01] <les_w> yeah.
[19:46:25] <les_w> When I see a machine, I start doing Laplace transforms.
[19:46:27] <les_w> heh
[19:46:32] <SWPadnos> heh
[19:46:44] <SWPadnos> I stopped doing those - they can be bad for your health
[19:47:17] <alex_joni> * alex_joni used to like nabla ;)
[19:47:37] <les_w> Yesterday I was contemplating the shunt inductance I needed to get the power factor in on the transducer....
[19:48:37] <alex_joni> les_w: seen this: http://lerch.no-ip.com/ChangFa_Gen/ ?
[19:48:44] <les_w> then with the analyzer I found I could put a mechanical spring like device on that synthesized the electrical inductance.
[19:48:51] <les_w> looking
[19:48:53] <SWPadnos> looking at a particular benchmark, it seems that a 1024-point double precision complex FFT takes around 70k clock cycles, using the fastest library available
[19:49:51] <les_w> 70k sounds fine
[19:50:10] <les_w> milliseconds?
[19:50:20] <les_w> nice generator.
[19:50:25] <SWPadnos> cycles - atound 70 uS for a 1 GHz processor
[19:50:29] <SWPadnos> around
[19:50:29] <les_w> Need Biodeisel
[19:50:59] <les_w> ok I'll wnat a 4096 pt fft then!
[19:51:04] <SWPadnos> heh
[19:51:07] <les_w> oh...floats
[19:51:17] <les_w> np right?
[19:51:20] <SWPadnos> right - that's just the algo, with a hot cache
[19:51:25] <SWPadnos> np?
[19:51:31] <les_w> no problem?
[19:51:45] <SWPadnos> right - this was double precision (8-byte) complex values
[19:51:54] <les_w> oh...fine
[19:52:33] <SWPadnos> but in a real situation, you'd have to load the coefficients from somewhere (like the analyzer), then get them into the CPU cache
[19:52:41] <SWPadnos> that probably takes longer than the FFT
[19:52:49] <les_w> yeah
[19:53:41] <les_w> even with that, mechanicals like shoving around parts will dominate the time
[19:53:45] <SWPadnos> jmk and I noticed that the step generator, which takes very few cycles to run, and has very little data, took an extra 1000-9000 cycles to run when the cache was cold
[19:54:01] <les_w> hmm
[19:54:10] <SWPadnos> a few hundred with hot cache, a few thousand with cold cache, dependent mostly on the period
[19:54:24] <SWPadnos> and that's for ~256 bytes of code and data
[19:54:51] <les_w> well, caches and pipelines and hard real time are not the best mix....
[19:55:23] <SWPadnos> remember that the coefficients will be in memory that is accessed at ~700 MB/sec (more on an Opteron), and you want 1024 * 8 * 2 = 16384 bytes of data, plus the code to manipulate it
[19:55:31] <les_w> but I see no real problems
[19:55:36] <les_w> at all
[19:55:48] <les_w> and the funding flows
[19:56:05] <SWPadnos> they're friends for speed, but not necessarily for deadline calculations ;)
[19:56:05] <les_w> as long as the economy holds
[19:56:16] <SWPadnos> heh - one can't really ask for more than that
[19:56:24] <les_w> yeah
[19:57:03] <les_w> well it's only 4 but I think i''l call it a day
[19:57:09] <les_w> need to celibrate
[19:57:16] <SWPadnos> heh. meetings are amazingly tiring
[19:57:37] <les_w> I know...I'll bush hog a few acres with the tractor
[19:57:55] <les_w> oh this meeting was short
[19:57:58] <SWPadnos> err - yeah, that sounds fun ;)
[19:58:06] <SWPadnos> short and sweet, it sounds
[19:58:47] <les_w> Well many is the time I have flown to california or europe for a one hour meeting
[19:58:59] <SWPadnos> yep. done that too.
[19:59:10] <SWPadnos> at least you get to go to California or Europe or whatever
[19:59:20] <les_w> yeah
[19:59:33] <les_w> and stay the weekend or even the next week
[19:59:34] <alex_joni> * alex_joni needs to go to greece for a meeting
[19:59:38] <alex_joni> in 2 weeks
[19:59:43] <SWPadnos> lucky you!
[19:59:49] <alex_joni> should be nice I heard :)
[19:59:53] <SWPadnos> I'd love to go there, even if I had to go to a meeting ;)
[19:59:54] <les_w> yeah
[20:02:21] <les_w> Well I hope we can do some stuff
[20:02:29] <les_w> there is a lot of talent here
[20:03:16] <SWPadnos> indeed there is
[20:03:36] <alex_joni> you just need to lure it out
[20:03:49] <les_w> yeah
[20:03:50] <les_w> haha
[20:04:33] <LawrenceG> les_w: Hi there your self..... sounds like you are having fun with the HP unit
[20:04:45] <les_w> oh yeah lawrence
[20:04:52] <les_w> it's great
[20:05:04] <LawrenceG> I like test gear that draws pretty pictures!
[20:05:13] <LawrenceG> got any screen shots?
[20:05:27] <les_w> push a button....bing an equvalent circuit with values pops up on the screen
[20:05:42] <les_w> yeah. can you dcc?
[20:06:19] <LawrenceG> probably not... send to ve7it @ shaw . ca
[20:06:30] <les_w> ok just a sec
[20:06:36] <LawrenceG> thanks
[20:07:32] <LawrenceG> I have been working on a small servo amp project here.... a voltage controlled current source using a 10amp power op amp
[20:07:48] <LawrenceG> linear... no fets
[20:08:33] <etla> lawrence: for brushless servos ?
[20:09:14] <LawrenceG> no... that project is still in a box.... this is for driving some small brushed 12v dc motors
[20:09:44] <LawrenceG> (actually car heater fan motors....)
[20:10:13] <etla> maybe you saw my prototype dspic+irams ?
[20:10:14] <les_w> sent
[20:10:46] <LawrenceG> yea... looks great.... how is the irams working out.... I havent tried using one of those yet
[20:10:55] <etla> it spins the motor now - still some bug in commutation that needs to be ironed out
[20:11:09] <etla> the irams works just fine. just need lots of cooling
[20:11:33] <etla> it's the irams10up60b. will probably use the iramx16 for the 1kw servos
[20:11:38] <LawrenceG> if you need tested epp code, the stuff I posted on the barnes web site seems to work
[20:12:00] <etla> yes, I tried it and it works. but i decided on a mesa m5i20 card instead
[20:12:21] <etla> can't do everything so I decided I'm not going to write emc drivers at this stage
[20:12:22] <LawrenceG> I like the mesa card as well
[20:13:18] <etla> got a jogwheel working yesterday also
[20:14:20] <LawrenceG> great.... I need to get a high speed spindle built for my mill.... I was making a circuit board yesterday for the other servo amp project and 3000rpm on a tiny cutter is pretty useless
[20:14:43] <cradek> even 10k is pretty slow for that
[20:15:10] <etla> yep. I'll have the 1kw servo as the spindle. it has a max of 3000rpm but I'm hoping it can be geared up for small tools
[20:15:24] <cradek> spec on the pcb tools I use is >20krpm
[20:15:27] <etla> need to figure out a smart way to gear up/down depending on cutter radius
[20:16:01] <cradek> make a collet to hold a dremel tool? (not seriously)
[20:16:29] <LawrenceG> les_w: got the pix... nice toy.... quite a beast for current generation test gear
[20:16:35] <etla> I guess 3mm diameter would be the smallest I would use
[20:16:41] <les_w> I have all the hardware to do the S word for the spindle but have not hooked it up
[20:17:19] <les_w> I can go 0-24,000
[20:17:24] <les_w> rpm
[20:17:37] <les_w> still slow for making pcbs
[20:18:11] <LawrenceG> for most of my stuff, a trim router on a new Z axis assembly would be much better than the current morse 3 spindle
[20:18:20] <les_w> Lawrence the tinkg is huge
[20:18:29] <les_w> and $44k?
[20:18:32] <les_w> why?
[20:19:12] <les_w> it's just a pc and I guess some flash a/d
[20:19:21] <LawrenceG> les_w: well looking at the specs, I suspect it is all related to the accuracy guarantee
[20:19:26] <les_w> yeah
[20:19:31] <les_w> it's way good
[20:19:56] <LawrenceG> I mean a laptop and a few op amps plus software is all one needs ... right?
[20:20:12] <les_w> screenshot look kinda like a quartz crystal?
[20:20:35] <LawrenceG> les_w: that poor scope cart looks unhappy
[20:20:36] <les_w> It's a ceramic resonator
[20:20:51] <les_w> oh I had to make a bracket
[20:20:52] <etla> what are u looking at ?
[20:21:00] <les_w> for an old Tek cart
[20:21:07] <les_w> I have a bunch of them
[20:21:26] <les_w> oh, a picture I sent etla
[20:21:40] <les_w> I'ts a piece of test gear I just got
[20:21:54] <LawrenceG> they would be good for pushing tool boxes around the shop.... you have the one with the scope on it nicely setup
[20:22:18] <les_w> see the old digital tek scope next to it?
[20:22:26] <les_w> old but works just fine
[20:22:54] <etla> * etla likes lecroy
[20:23:05] <les_w> yes they are good
[20:23:25] <LawrenceG> yea.. I have a collection of tek scopes here..... they have been working for 35 years, so the chances are good they will do a few more in my shop
[20:23:30] <etla> except the new cheap ones which are windows based. got one of them to crash once a day literally
[20:23:39] <les_w> heh
[20:24:34] <les_w> Well this piece of gear is an impedance ananlyzer they sent me. It's the most expensive piece of gear in the shop.
[20:24:57] <etla> what are you going to be doing with it ?
[20:24:59] <les_w> Next is the green B&k audio analyzer
[20:25:16] <les_w> I am designing ultrasonic transducers
[20:25:32] <etla> wow, that's pretty close to what our lab does
[20:25:34] <Jymmm> bka fingerburners
[20:25:47] <les_w> I would be great for analyzing machine structures too
[20:26:02] <les_w> really etla
[20:26:06] <Jymmm> Jymmm is now known as Red70sShow
[20:26:06] <Red70sShow> Red70sShow is now known as Jymmm
[20:26:30] <etla> les: yeah, materials characterization: paper, wood, blood vessels etc.
[20:26:39] <etla> US based that is
[20:26:50] <les_w> oh neat
[20:27:08] <etla> what's your transducers for ?
[20:27:22] <les_w> I was glueing up a tranducer with epoxy and just watched the inpedance curve to see it cure
[20:27:47] <Jymmm> Jymmm is now known as CodesWithIdiots
[20:27:59] <les_w> it's kind of an energy harvesting thing
[20:28:06] <etla> shear or longitudinal wave ?
[20:28:24] <les_w> can't say much more..... lawyers won't let me
[20:28:30] <les_w> wish I could though
[20:28:31] <etla> ok...
[20:28:40] <alex_joni> etla: it's diagonal wave
[20:28:45] <les_w> haha
[20:29:00] <les_w> longitudinal...I can say that
[20:29:39] <les_w> In another project we did trapped torsional shear waves though
[20:29:44] <etla> our lab once did this lost energy measurement thing, it was for detecting ice on airplane wings... but the idea never 'took off'
[20:29:59] <CodesWithIdiots> <rimshot>
[20:30:00] <les_w> neat
[20:30:32] <etla> I think some big player bought the project and killed it.... happens all the time
[20:30:38] <les_w> yup
[20:31:12] <les_w> Well I paint alcohol on transducers to change the mass and watch the parameters change as it evaporates
[20:31:47] <etla> like qcm
[20:31:54] <etla> quartz microbalance
[20:32:05] <les_w> the electrical impedance changes a lot even if some object gets close to it
[20:32:19] <les_w> yeah
[20:32:36] <les_w> I can measure mass fairly accurately I guess
[20:32:59] <etla> I know they do very precise layer thickness measurements that way in vapor deposition of metals etc.
[20:33:13] <etla> one atom-layer accuracy
[20:33:13] <les_w> yes
[20:33:37] <les_w> I used to get near that with an ellipsometer
[20:33:42] <les_w> optical device
[20:34:38] <les_w> once upon a time I was a photolithography engineer for an IC line
[20:34:51] <les_w> at Harris/Intersil
[20:35:25] <etla> cool. except thosy guys probably work 3-shift hours ?
[20:36:41] <les_w> Oh, It was a long time ago. I made op amps and PROMS
[20:36:49] <les_w> We had 3 shifts a day
[20:36:52] <les_w> 24 hrs
[20:37:18] <les_w> back then it was 5 micron stuff...huge
[20:40:26] <les_w> Now I have one optical and two ultrasound projects
[20:40:36] <les_w> same math really
[20:41:47] <etla> we did time of flight measurements in carbon nanotube paper a couple of weeks ago !
[20:41:59] <les_w> wow
[20:42:08] <etla> basic science, nowhere near applications yet...
[20:42:09] <les_w> I would like to see that paper
[20:42:15] <etla> it's black.
[20:42:22] <les_w> heh
[20:43:02] <etla> anyway, they want to make it stronger by irradiating it, so the guys measured youngs modulus before and after irradiation
[20:43:12] <les_w> I see
[20:43:20] <les_w> hmmm
[20:43:31] <les_w> alpha radiation I guess
[20:43:48] <les_w> now we'll get into something you can't talk about!
[20:43:50] <les_w> haha
[20:43:52] <etla> some ions similar to carbon I think
[20:44:14] <etla> it forms bonds between the tubes and that way makes it stronger
[20:44:19] <les_w> donors and acceptors
[20:44:36] <les_w> haha
[20:44:54] <etla> I work at uni so I can talk pretty much about everything... one advantage compared to all the nokia and ericsson guys around here
[20:45:04] <les_w> yeah
[20:45:38] <les_w> I get muffled up by the lawyers pretty bad
[20:46:43] <etla> an NDA is the first thing they get you to sign just when you visit nokia research...
[20:47:18] <les_w> yeah I have to have a lot of NDAs
[20:48:51] <les_w> I'm still thinking though....ion implanting bonding agents for carbon nanotubes
[20:48:58] <les_w> silazanes?
[20:49:07] <les_w> hmm
[20:50:38] <les_w> I'm surprised nanotubes even have any dangling bonds to work with
[20:50:56] <les_w> oh well blast em with high energy ions and they might
[20:52:17] <etla> yes I think the energy is pretty high, the ion replaces a carbon in the tube.
[20:52:24] <CodesWithIdiots> CodesWithIdiots is now known as Jymmm
[20:52:26] <les_w> right
[20:52:39] <les_w> then you can hydrogen bond to that
[20:52:39] <etla> there are some MD simulatio papers on this but now they want to show it experimentaly also
[20:54:11] <les_w> not sure, but I think there are double carbon bonds in nanotubes and buckyballs
[20:55:14] <CIA-8> 03alex_joni 07HEAD * 10documents/lyx/emc2/ini_homing.lyx: added Homing description from the EMC2_Code_notes
[20:55:40] <etla> need to sleep now, bye.
[20:56:10] <les_w> bye etla
[20:57:01] <CIA-8> 03alex_joni 07HEAD * 10documents/lyx/emc2/ini_config.lyx: removed indexes for now, as they cause problems outputting a pdf, it seems that no _ or other special characters are allowed, and lower case also seems like a problem. (will reenable Idx when we'll have a clue about it)
[21:06:11] <alex_joni> * alex_joni heads to bed
[21:06:13] <alex_joni> night all
[21:06:52] <les_w> night alex
[21:18:51] <K4ts> hello
[22:02:10] <Jymmm> les_w: ok, when do we report to work?
[22:19:34] <davidf> hi..
[22:20:34] <davidf> Can anyone tell me where to find a channel for live help with ubuntu - installing a modem?
[22:21:02] <davidf> Have ubuntu installed but having trouble connecting to IP.
[22:22:14] <jepler> "#ubuntu" is the channel on this irc server
[22:23:37] <davidf> thanks.
[22:25:27] <K4ts> hight
[22:25:52] <K4ts> nght
[22:25:57] <K4ts> sorry
[22:32:20] <xpc> hi
[22:32:49] <Guest727> hi again.
[22:33:27] <Guest732> hello?
[22:33:49] <Guest732> am i davidf again?
[22:33:59] <Guest732> nope. but its me.
[22:34:06] <SWPadnos> heh
[22:34:16] <SWPadnos> do you have DSL or cable?
[22:34:43] <Guest732> jepler: thanks for the tip but that cahnnel is so busy i dont have much hope of getting help there.
[22:35:14] <Guest732> SWPadnos: adsl phone with dual connect westell modem usb & ethernet.
[22:35:48] <Guest732> dohc didnt install when I installed ubuntu 5.10
[22:35:53] <SWPadnos> ok, do you have to provide login information, or is it supposed to work with DHCP?
[22:36:06] <SWPadnos> DHCP or DOHC (never seen DOHC before)
[22:36:22] <SWPadnos> except in engines ;)
[22:36:35] <Jymmm> Dual Over Head Cam
[22:36:49] <Guest732> I tried setting up eth0 & had some success with the initial setting of ip etc. but I cant connect.
[22:37:06] <Guest732> haha. right gues it is dhcp.
[22:37:52] <Guest732> I dont know if i have to provide login or not or even what dhcp is. :(
[22:37:56] <SWPadnos> was the computer connected during the install
[22:37:57] <Jymmm> Dynamic Host Configuration Protocol
[22:38:08] <Guest732> no.
[22:38:48] <Guest732> I thought later I should have had the modem connected to the ethernet card, right?
[22:39:16] <Guest732> probably would have detected the connection then on install?
[22:39:26] <SWPadnos> that's probably best, though USB may work fine
[22:39:39] <SWPadnos> I suspect it's best to have it connected at install
[22:40:09] <SWPadnos> the other possibility is that the modem (if it was connected to a different machine) may not want to provide the new ethernet card with an address -it may be limited to one client machine
[22:40:10] <Guest732> ubuntu site says usb may be difficult or impossible without drivers etc. All I have is windows driver disk.
[22:40:26] <SWPadnos> ok -ethernet is usually better anyway
[22:41:27] <Guest732> SWPadnos: My guess is the modem wont care. My ip only allows one connection at a time but I disconnected from other machine first/
[22:42:01] <SWPadnos> a lot of cablemodems, and also many DSL modems, will only assign an IP to the first MAC address they see
[22:42:07] <Jymmm> ethernet router saves SO MUCH trouble
[22:42:39] <SWPadnos> you could try disconnecting the computer, turning the modem off for a few minutes, then turning it on
[22:42:51] <SWPadnos> just wait until it's ready before pluggint the computer in again
[22:42:59] <Guest732> I can call my IP tech sup and get that issue straight first. Should I do that now & then go from there?
[22:43:06] <SWPadnos> there is another possibility before that though
[22:43:16] <SWPadnos> they probably won't know anything
[22:43:34] <SWPadnos> usually when you have a problem, the first question is "does it work with Windows"
[22:43:42] <Guest732> sw: I did that re disconnecting power & cycling the modem.
[22:43:43] <SWPadnos> and the next statement is "sorry - we can't help you then"
[22:43:49] <SWPadnos> ok. try this:
[22:44:08] <SWPadnos> from a command prompt, issue the command sudo /etc/init.d/networking restart
[22:44:13] <Guest732> SWPadnos: heh. I know that answer well.
[22:44:43] <Guest732> SWPadnos
[22:44:58] <Jymmm> I would highly recommend a ethernet router... They are not that expensive these days, can share the connection with any number of computers, serves as a NAT firewall.
[22:44:58] <SWPadnos> yes? ... ;)
[22:45:06] <SWPadnos> I second that
[22:45:35] <Jymmm> and would save you all these headaches your having right now.
[22:45:36] <SWPadnos> there's almost no reason to do without one these days
[22:46:01] <Guest732> I only have one modem connection, & right now it's on my win 98 box, wjich Im using here. Do I need to connect to the linux box to try this?
[22:46:36] <SWPadnos> well, the linux box would need to be connected to try the command I gave you - otherwise it's no good
[22:46:43] <Guest732> ok I can fork out the $30.00 or so for one (that about right?)
[22:47:04] <SWPadnos> but you'll probably save a lot of hassle an dtime if you just frive to the nearest Best Buy or whatever, and bet a 4-port NAT router
[22:47:04] <Jymmm> Guest732: Yep, or a tad more and you can have a wireless AP too
[22:47:58] <Guest732> ok. if that'll work it will save a lot of greif (for you kind folks as well.) :)
[22:48:47] <Jymmm> Guest732 do you have a laptop?
[22:48:52] <Guest732> I'll try SWPadnos' suggestion & if that doesn't get me going I'll frive to the Best Buy * bet one. :)
[22:49:35] <Guest732> Jymmm: no, a desk top.
[22:49:48] <Jymmm> Guest732: Does anyone in your home own a laptop?
[22:49:55] <Guest732> no.
[22:50:07] <Guest732> why?
[22:50:35] <Jymmm> IF you THINK you might get a laptop in the future, you should get a ethernet router with wifi, then you can use the laptop anywhere - no wires =)
[22:51:05] <Jymmm> else you spend $35 today, and then have to buy one tomorrow with wifi for $50
[22:51:34] <SWPadnos> most laptops also have ethernet these days, so that's not really an issue
[22:51:45] <SWPadnos> the portability is, but not connectivity
[22:53:26] <Guest732> OK. I don't intend to buy a laptop any time soon. If I can ever afford one then I can afford another router with wifi too. Thanks for the tip though.
[22:54:20] <Guest732> OK. I'll sign off now & try SWPadnos' suggestion etc. Thanks for all the help.
[22:54:27] <SWPadnos> have fun
[22:54:41] <Guest732> SWPadnos: yeah, right. :)
[22:54:53] <SWPadnos> heh
[22:55:49] <Guest732> bye.