maybe I should get my crew here to contribute some code devel, actually repay in a currency that has some meaning.
what would you contribute, if you could?
documentation? :-) shoot, I dont know. they all write code for DNA analysis.
I'm director of bioinformatics of http://tigr.org
how about a nice assembly for DNA sequence fragments for the mammalian genomes?
the trouble with emc is ... you need a few years of CNC and about a years emc experience before you could contribute much useful to behonest, its a remarkably deep project ... the logical option would be to create some sort fo well-defined add-on tool for emc, liek a configurator for example, somehting discrete would seem the best plan
or some neat way of viewing the HAL xml config in a human friendly form
your guys must be great at visualisation tools
really? we bang on XML all the time.
I think anyone who uses emc and is well-spoken could contribute a lot to documentation
I write so freaking much, if I could contribute documentation I'd be happy to.
or review and improve what's there, instead of just adding more words
since all my people run on goverment funds, bout all I do is write grants.
yes, a lot of it is out of date, or has not been proofread
well, it came from NIST source originally, it would be governmint work still :)
zactly. keeping up with your code base would be hell.
I made a dxf to g-code converter in perl, you all want that? :-)
perl's a language we don't use in emc yet
two places where someone that's not steeped in EMC tradition can help a lot: 1) graphical configuration tools and 2) importers/converters for either old EMC configs or for things like Mach configs ...
perl save the human genome project. we bang a lot of text, basically.
perl and bioinformatics go back a long way
isnt Nat Torkington the bio guy?
dont sound familiar.
I'd better take this as a cue to bail, since otherwise I'll go on some anti-perl tirade
I'm out of here too.
thanks for everything. again.
* robin_sz waves
he moves fast
like a greased rabbit
* robin_sz wonders whether 0 is a positive integer
thats roughly the same conclusion I came to
I could ask my mother, but you'd have to do something very nice for me
watch the sign on most dros as they cross 0, they're of the same inconsistant opinion :)
ive been looking at partitions, quite an interesting subject
for maths anyway
well, according to this page, it's neither positive nor negative: http://www.answers.com/topic/integer-1
yes, it has no sign, becuz it has no value
all the -0 displays are goofy
those make sense, since the measurement resolution is likely higher than the display resolution
so it may be reading -0.00001, but it would need to be -0.00005 to round up to something displayable
hmm .. ok, I gave in and looked it up on sloanes
yes, thats why, but thats not 'making senset, that's 'why it happens' ;)
[00:19:40] <robin_sz> http://www.research.att.com/~njas/sequences/A000027
hmmm. I recognize the characters, but the page makes no sense
ooh, now , this is neat ... very neat
[00:21:23] <robin_sz> http://integrals.wolfram.com/index.jsp
tomp makes link on wiki to the examples the jepler pointed out http://linuxcnc.org/docs/devel/html/hal/vcp/index.html
you have an option to use "myconf" for configuration. It provides distributed state (clients receive notifications when options change), stores things in XML, and has a GUI similar to GNOME's GConf
the screenshot's here: http://mync.sourceforge.net/f2.png
it's a separate library, independent enough
one energy chain mounted, one to farbicated the brackets to go!
CAT5 ran and terminated, just have to get under the SOB table and mount the bitch the wall!
erDiZz: looks nice. can you find a old agie 150mmx150mm to test with? some are being thrown away here in the US, and they're very accurate & easy to interface, just xy but 1/2um step
tomp, to test what?
that's kind'a the only type of machines I have access to
not till this summer though
tomp, my uncle has a small business of replacing broken soviet control units
wow, maybe he can use your control and maybe emc
...on wedm machines mostly
have him contacct Tom Truty at http://www.techstar-edm.com/,
he restores old wedm also, and may have some parts
tomp, yeah, at the moment the goal is to implement all functions that the software he uses has
so that he could use mync as a replacement
that old peace of sowtware I'm talking about was written in modula-2 in late 80-s - early 90s...
i worked a few guys who came from Moscow and specialized in EDM, they work with a 'Dr.Otto' on an anti-arc device ( wow modula-2)
I think they all know each other here
all the ingersoll's sold by Mitsubishi in the US have the 'otto box'
at those times the production was centralized, the head facility for manufacturing CNC controls for wedm was 'Istok' in the town of Fryazino near Moscow
I read some of the journals ( looked at the pictures) there was a lot of good work done there
tomp, yeah, but it was all stuck in the 90s
I was told that there is a whole hangar at Istok with a hundred or more ready-to-work wedm machines
they're just standing there
get your control running & go to Turkey, they're hungry for machines.
will do my best :)
* erDiZz is a bit too patriotic to move anywhere
well, I'm off to bed
dont move, sell them machine, take money home :) good nite
03jepler 07v2_1_branch * 10emc2/docs/src/Submakefile: since html documentation isn't even included in the debs, make it possible to build pdf docs without latex2html intstalled, and get rid of the build-dependency on latex2html.
03jepler 07v2_1_branch * 10emc2/debian/control.in: since html documentation isn't even included in the debs, make it possible to build pdf docs without latex2html intstalled, and get rid of the build-dependency on latex2html.
03jepler 07v2_1_branch * 10emc2/src/ (Makefile.inc.in configure configure.in): since html documentation isn't even included in the debs, make it possible to build pdf docs without latex2html intstalled, and get rid of the build-dependency on latex2html.
03fjungclaus 07TRUNK * 10emc2/src/po/de_axis.po:
- Done some more translations
- Added translations for new MDI history stuff
Going to bed now. CU
thanks for translations! goodnight
jmkasunich_ is now known as jmkasunich
I can't understand a word my calc prof says
mehnos this, mehnos that
owhite: you're back
with a problem.
I think I broke something with the last make.
[02:56:55] <owhite> http://pastebin.ca/346147
owhite: ipcrm -M 1004
whats that? a command?
owhite: emc uses shared memory buffers to communicate between different parts of the program
owhite: one of the buffers has a different size specified in 2.1 and earlier than in 2.2
owhite: emc is supposed to remove the buffer when it exits, but at least sometimes it doesn't.
owhite: when it doesn't remove the buffer, and you switch to a version with the other size specified, you get this error message.
owhite: you can fix it by manually removing the shared memory segment, the ipcrm command I showed earlier
except the right number may be 1005
that sounded real plausible :-) but I tried it and....oh.
* BufferLine: B toolSts SHMEM localhost 8192 0 0 5 16 1005 TCP=5005 xdr
now honestly, how can you be so smart?
if I was smart I'd fix the bug
new problem. something broke my tkemc.tcl, but....
I will try to fix that. :-)
I think it leaves those droppings only when it fails to start up right
crap the last build put lines like ">>>>>>" into my code.
no, cvs up with conflicts does that
cvs did that
that's a cvs thing, right?
owhite: that means you and another developer modified the same lines in that file
I live for conflicts.
it shows you your lines, the other developer's lines, and lets you sort it out
you made a change in your local copy that conflicted with another change in cvs
that's why it's called a conflict
how could another developer POSSIBLY know have a better idea than I do with my tkemc.tcl code?
well that's a good question
I've done all kinds of horrible things to tkemc.tcl, I knew this day was coming. :-)
all this because you guys talked me into using 2.1 :-)
not that anyone should care, but if I can get pyvcp to then I will just use the standard tkemc. I hacked it a lot to work on my laser system.
er, get pyvcp _to work_ ....
[03:09:02] <cradek> http://timeguy.com/cradek-files/emc/scara-backplot.png
owhite: I'd suggest renaming your tkemc, then do a cvs up to fetch the standard one
(if cvs sees the file missing, it will reload the original)
cvs up -C tkemc.tcl
thanks good idea.
I suspect owhite doesn't want to give up his changes!
I said "rename it"
not "delete it"
I guess both do preserve the old file
oh if you have to hand merge anyway, the <<< >>> are as easy as anything
cradek: fun ain't it?
(or as much of a pain as anything)
yep, it's very cool
has anyone told the list about this?
no, I don't think so
try a setp scaragui.joint4 30 while its cutting
(like between the EMC2 and the AXIS)
cradek: jmkasunich: re the scara-backplot... when you get the kinematics for the operator worked out, you can program him too :)
shouldn't one of ABC turn the tool around? (that's the R axis maybe?)
its CNC - the operator walked away for a cup of coffee
C turns the tool
out of curiousity, what are you guys talking about?
no it doesn't
but emc won't let us configure C unless we do A and B as well
owhite: EMC2 controlling a pretend robot
jmkasunich: ok, so I can't control C, but it's in the kins
I even think its correct
it would be cool to see it keep the point in the same place
but its untested at the moment
you mean jog C, and watch the rest of the arm move as the tool orbits its own tip?
that's how it should work right?
alex mentioned a possible real-life use for this config
is EMC linked to software that's then rendering that picture?
I'm keeping an eye out for parts to build one
if the angled tool was a plasma cutter, you could cut out pieces pre-beveled for welding
you guys are really amazing.
owhite: the picture moves to represent emc's calculations for all the joints
owhite: if yo've successfully compiled the TRUNK with AXIS, you can try configs/scara
very very cool.
oh that's scara. nice.
there's also a puma robot, but it works ... a little less well
[03:21:47] <jepler> http://www.myscienceproject.org/j-wall.html
it has a cool gripper though
the model works fine, but the emc kinematics for puma aren't quite right
I would settle for just getting a vcp panel to turn on my ventilation :-)
jepler: hahaha ramen jello
this morning there was a link posted to a nice homebrew head for a 5 axis machine, it was in cnczone & i dont join. found it...http://www.cnczone.com/gallery/showgallery.php/ppuser/1117/cat/500
the tool doesnt quite hold still while the arm moves, it spirals over the surface of a sphere, so very similar
hm this is not a very good helical path, look at the facets .. http://imagebin.org/7208
(it's from that person's gallery)
i think his 'finishing' passes are ... longitudinal, yes scalloppde
jeez, get a lathe
he might be beginning the spiral from bot to top, dunno
from the looks of the coolant I think he's cutting a spiral
there's a pyvcp_demo.hal in configs/sim, does anyone know what launches the "demo"?
but yeah, we sure have better arcs than that
owhite: I know sim/lathe has a vcp panel
there's a picture of the path in another image -- it goes from the south pole to the north pole in a single curve
yah I spotted that but I cant find anything that launches pyvcp_demo.
owhite: run the demo with: halrun pyvcp_demo.hal
hm looks like the hal file has bitrotted
03jepler 07TRUNK * 10emc2/configs/sim/pyvcp_demo.hal: fix demo to work with pyvcp changes
owhite: after you "cvs up", 'halrun pyvcp_demo.hal' shold work properly
cvs up, or cvs -z5 update -Pd ?
you should put the "-z5" and "update -pD" in your ~/.cvsrc so that you don't have to remember them each time
that also means you don't have to type it right each time
good defaults for ~/.cvsrc are suggested on the wiki: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?CVS
and do I run make after I cvs up?
it can't hurt and doesn't take long, but in this case (an updated .hal file) it is not strictly necessary
.cvsrc is working. pyvcp_demo.hal is working.
what motherboards work well with emc? ie, minimizes RT-delays...
dunno, there some thing that bother rt ( on board vid with unstoppable interrupts ), mainly stuff that is incompatible with rt, viewed from 'it does stuff when it wants to and wont abide with the scheduler'... so we dotn know whats good, we know whats bad
if I delete a file, then run cvs up, it'll restore that file, right?
so you all may be too busy to look at this, I'm wondering about pyvcp_demo -- I dont get how pyvcp_demo.hal connects to the check button name chbutton-sample in the .xml file.
lemme take a look
most nice of youze.
(gotta update first, hang on a bit)
you can use cvs -z5 update -Pd, or augment your .cvsrc file :-)
<--- incredibly knowledgable.
I have a couple of scripts, update and build
one does the cvs up -dP, and configure
the other does the make
both run from the top level directory to save doing cd
okay. but how about that chbutton? :-)
which config are you running again?
there are about 7 configs under sim
which one has the vcp?
sorry. halrun pyvcp_demo.hal
ok, got it running
what do you want to know
I dont get how pyvcp_demo.hal connects to the check button name chbutton-sample in the .xml file.
jmkasunich: owhite: pyvcp connects the checkbutton to the halpin with <halpin>"chbutton-sample"</halpin> the halpin gets it's name this way
yeah but if that's a pin, what other pin connects to it? I dont see anything in pyvcp_demo.hal
ah, yes, thisis only 1 end of the circuit
nothing is connected to the checkbutton in the demo
the regular button is connected to the led
when you run the demo, you get a halcmd prompt
type "show sig" at that prompt
you'll see three signals
one is connecting the button to the led
one is connecting the scale to the number, and one is connecting the spinbox to the bar
now type "show pin"
you'll see that the checkbutton pin haws nothing connected to it
but if you click the button, and then do show pin again, you can see it change
right. could I use linksp as in other .hal files?
try this for example:
that disconnects the button from the signal "bit1" that goes to the led
click the button to confirm that it no longer blinkes the led
now "linksp bit1 pyvcp.chbutton-sample"
actually that didnt work.
the unlink didn't work?
yep. didnt work.
show pin still shows button-sample ==> bit1 ?
use the up-arrow key to scroll back in the history, check to see if you typed it right
I get this:
halcmd: show sig bit1
Type Value Name
bit FALSE bit1
halcmd: unlinkp pyvcp.button-sample
halcmd: show sig bit1
Type Value Name
bit FALSE bit1
oh sorry I put that in the .hal file.
the very best way to get comfortable with HAL (in my opinion) is to use halcmd interactively
it still hasnt hit me that halcmd is like an interpretter.
its exactly like an interpreter
exactly. I havent gotten that point.
in fact some of the devel have started referring to "the halcmd language"
i just got images for pyvcp http://imagebin.org/7209
an image widget
okay. so you I connected the checkbox to the LED.
er, I connected the checkbox to the LED.
it works now? (clicking toggles the led)?
now - you can (still using it like an interp) load the parport driver
as an excersize, the user shall try to put that into the .hal file. :-)
what!? oh that is too cool.
loadrt parport cfg="0x0378"
(or whatever address your port is at ;-)
loadrt hal_parport cfg="0x9800"
(I forgot the hal_ the first time)
show pin now gives you a much longer list
the parport code isn't running though, because you have no realtime threads
loadrt threads name1=foo period1=1000000
that creates a 1mS thread called foo
will list the available realtime functions
addf parport.0.write foo
will add the functiont that writes the port to the thread
addf parport.0.read foo
adds the function that reads the port
halcmd: addf parport.0.write foo
halcmd: addf parport.0.read foo
halcmd: show thread
Period FP Name (Time, Max-Time)
999849 YES foo ( 0, 0 )
show thread lets you see the result
the thread isn't running yet
do you have something connected to your parport that is safe to play with?
time and maxtime have changed.
I sure do :-)
a little relay driver.
right - that is the time execution time of the thread
ok, what pin is a relay attached to?
lets do pin 3
ok, that will be parport.0.pin-07-out
we already have a signal called bit1, controlled by the checkbutton
linksp bit1 parport.0.pin-07-out
and (if everything is working right) you should be able to click the checkbutton and turn the relay on and off
got any inputs on that port? limit switches or something?
nope. all output.
frankenstein is alive (but write only)
so then I can roll all the above commands into a .hal file and be ready to roll?
if you had an input we could make another signal, disconnect the led from "bit1", connect it to the new signal, and connect that to a parport.0.pin-??-in
yeah that'd be somehting.
would you anticipate....
you can also tell halcmd "save" and it will generate a list of hal commands that re-create the current config
...I use the univstep stuff. do you know of any problems with using that along with this system?
save <filename> saves that to a file
not sure I follow
the univstep needs its own parport
sorry I am going with a two parallel port system.
but if you have a second one, then yes, you can run hal stuff thru it
one will be for univstep, the other will be for this.
remember - the univstep is also using hal
if you start your emc, then open another shell, you can run halcmd there
$ halcmd -kf
will get you the halcmd prompt for interactive stuff
yep. just wondering if the threads or timing would somehow be incompatible.
show pin and show sig will give you _much_ longer lists when EMC is running
you don't use the threads command
emc already has threads
just add the "parport.1.read" and "parport.1.write" functions to the servo thread
oh. I wont use: loadrt threads name1=foo period1=1000000
well this is just amazing.
if you do "show thread" when EMC is running, you'll see two threads
servo and base
on a univstep system, base is probably empty
(its normally used for software stepgen, etc)
servo is a 1mS thread just like the foo thread that we created
halcmd also has help
who wrote it? youze?
originally, but many folks have added to it in the last couple years
would it be good to have slower,not so important thread for user buttons, lights...
got it. the hal stuff is just so great.
tomp: you could
but simple stuff uses _very_ little time
on my 1.6GHz box, 1mS is 1,600,000 clocks
my foo thread (with parport read and write) is using an average of 10,000, and a max of 40,000
okay. I have everything logged. I'll start working up a .xml file for my little application. I'l l be driving all kinds of appliances now.
ventilation, motor power supplies, that sort of thing.
the xml is just to define the pyvcp panel
yah I got that.
then you use a hal file to connect ot
take a look at the indow dressing http://imagebin.org/7209
earlier today they told me that are deprecating halvcp, and I should get with the program and do pyvcp.
really great stuff.
so. jmkasunich I do appreciate the help. I'm heading out.
bedtime here too
hey, this ubuntu wont recognize my paper tape punch ;)! http://imagebin.org/7212
steve_stallings is now known as steves_logging
Anyone able to help getting a PCI parallel port going?
How many people here are running the latest version?
tray: I dont see why that would be a problem
The PCI parallel port is a ITE chip, and when the computer turns on, the output of pin 2 is on straight away. I have a resistor & led across to pin25 (gnd) to indicate the status
any suggestions on testing the configuration of the parallel port
I used one a long time ago..
a very long time
shame but if you think of anything I'm here. ;)
I've got a c file that I believe is the driver, but not sure how to load the driver.
dont you have a kernel module?
the file is parport_pc.c
I got it from this site: http://www.emegatech.com.tw/download.htm
It's the driver for the ITE chipset
isnt there one already with the kernel?
I suspected that but I'm don't know how to configure it.
or if I can configure it?
righto, enough for today. Try again tommorrow. Take care. Thank you.
SWPadnos you familure with PS/2 ?
thinking of making a gift for the GF... something off the mill.
anonimasu, got any idears?
I have this... http://lloydleung.com/gallery3/Past%20Machining%20Projects/2004-03-04_Cube_1.5%5E3/
why not one like that?
write her name in the center ;)
I dont have tools long enough to reach the center... only to bore.
how big is that cube
1.5" ^ 3
I have no idea
maybe make a four tiny ones... like 1/2"... and make use thin thread to make 4, as little earrings.
that's pretty large
dont you think?
yeah... I was holding my fingers about 1/2" apart
then I thought about maybe 6mm.
then I thought about... how the heck am I gonna bore it out.
pcb mill ;)
anonimasu,... btw, that's my gallery, that's mine... it's sitting ontop of my centre speaker right now
anonimasu, was gonna just turn it out
on the lathe, much easier.
[13:09:53] <a-l-p-h-a> http://lloydleung.com/gallery3/Past%20Machining%20Projects/2004-03-04_Cube_1.5%5E3/Cube%20xvid%204%20fps%2010%20no%20sound.avi
went to ubunut channel by mistake. Gary is there. odd ;)
do you know him?
* ejholmgren winces
250MB across 10baseT
jepler: awallin: fyi, an image widget http://imagebin.org/7209 http://pastebin.ca/347263
ejholmgren: over 33 minutes
friend alex joni, you know about contact a mavati?
skunkworks, should be about 4 minutes
friend alex joni, you know about contact a mavati?
[16:21:34] <alex_joni> http://membres.lycos.fr/mavati/
SWPadnos: really? I thought it was 10 megaBITS per second
oops - may have done the math wrong
yep. that gives roughly 1 megaBYTE per second, and there are 240 seconds in 4 minutes
yep - didn't devide by 10
fryen. is har talk witt he?
Giraya: sorry, I don't speak that language
fryend. is hard talk witt he?
I sent him an email last time, and got an answer in a couple of days. so .. no, not really
you get me your email (marc le)?
Marc Le Douarain
28, vallée du LochTour Défense 2000 / Studio 32
56390 BRANDIVY23, Av. Louis Pouey
Tél: 02.97.56.03.76 (W.E.) 92800 PUTEAUX
marc.le-douarain AT laposte.net
[16:37:55] <alex_joni> http://sourceforge.net/users/mavati/
alex_joni; been stalking again?
lerneaen_hydra: no, but I learned how to read
watt is laposte.net, irc?
Giraya: ever heard of email?
AT = @
ok i dont konw AT=@
"Electronic mail (abbreviated "e-mail" or, often, "email") is a store and forward method of composing, sending, storing, and receiving messages over electronic communication systems. The term "e-mail" (as a noun or verb) applies both to the Internet e-mail system based on the Simple Mail Transfer Protocol (SMTP) and to intranet systems allowing users within one organization to e-mail each other. Often these workgroup collaboration organizatio
alex_joni; I see a truncated message
lerneaen_hydra: truncated where?
amigo alex joni vai te catar seu grosso
Often these workgroup collaboration organizatio
yeah, that's what I see
Often these workgroup collaboration organizations may use the Internet protocols for internal e-mail service.
[16:42:50] <skunkworks> http://www.youtube.com/watch?v=9107SknOFl0
friend i know the internet you not need this
Giraya: then I don't understand what you want..
you asked how to contact mavati, I linked you to 2 of his pages (which you didn't read, because they had his email address there)
then I pasted you his email address, and that still was not good enough
Giraya: and it's not nice to be disrespectfull, even if it's in a language I don't understand
ok i stress
the work is hard and pression
i need this clp
yeah I think alex here only speaks romanian
Giraya: if I were I would stop wasting time, and write an email to mavati
jepler: google helps with translations these days
* alex_joni estou passado. vou dar um role
i want to say for you one
'I am passed. I go to give one rolls'?
we a strange
and i like yor work
well.. close enough
we machines to made yor program
I am tired. I am going away..
but the impression is that you dont like
y can email for mavati
thanks for all answer
perhaps you dont interessanting in help for me
lerneaen_hydra,? what time is it?
lerneaen_hydra, had dinner?
damn you... I wanted to frag you.
unless you want to go a few rounds... and let me mop the floor with you...
Mr. Rogers seems good 14/20
favourite it... as I used that server a lot
I didn't quite get how to aim
heloo friends you see one configuraration for me?
03flo-h 07v2_1_branch * 10emc2/src/emc/usr_intf/axis/scripts/ (image-to-gcode.py axis.py hal_manualtoolchange.py): more i18n
rafa, can you post the file you want us to look at on http://pastebin.ca
my friend in work its strees
its you help
sorry for he ok?
pastebin = www.pastebin.ca/347759
this idea no for emc
i want to only load classicladder
if you only need classic ladder, you should download from mavati's page
there is no need to use HAL and emc for this
but i dont know ..
a correct funcyonaly
i need realtime
realtime need for hal
classicladder should be realtime without HAL
i trought that needed hal..
it does when you want to use it with EMC
because i wiky for emc..
[17:48:29] <SWPadnos> http://sourceforge.net/projects/classicladder/
that page says that classicladder supports RTLinux, RTAI, and Xenomai
[17:48:57] <rafa> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Classicladder
the wiki page is only related to EMC and the customized version of classicladder that we include
in tehe page says to use hal
" halcm -kf"
if you only need ladder, then use the sourceforge calssicladder project
halcm for hal or not?
there are two versions of classic ladder
one that comes with EMC, and one that does not
you do not need to use the version with EMC to use classicladder
yes i need
none of us are experts with classicladder - you will need to get help from them
because i made one process for automation
so you need to go to the help forum for classicladder
sorry, we can't help you
please one question
because classicladder in emc?
classicladder was modified to work with EMC, but none of us knows much about it
lh, you stink. :) "8" points. hehe.
what are you playing?
03flo-h 07TRUNK * 10emc2/src/emc/usr_intf/axis/scripts/ (image-to-gcode.py axis.py hal_manualtoolchange.py): more i18n
what tools are used on ubuntu dapper to burn data cd's.
right-click the ISO and select "write to disc" ...
sorry, data files, not an iso image. i imagine the answer is 'write to disk', so will try that
oh - well you can create an ISO just by using the loopback device and copying files, but there are easier methods, I'm sure
ok, will try to make an iso first then use the rt click/write to disk method.
I'm sure that's not the easiest way, but I don't recall what program(s) are installed by default
for making discs
it has 'serpentine' but only for audio, and the drag to disk method built into the file browser you spoke of. I had problems with lotsa coasters from k3b, so wanted to try the 'right' way, so i could find what the error was.
I know there's a gnome equivalent to k3b, but I don't remember the name
G-something I bet :)
gnomebaker or something
I just fire up synaptic and search for burner or burning
yep, in there, gtoaster
also accessories -> gnomebaker
dunno if that's stock or if I added it though
random question, time dilation is proportional to speed, right?
tomp: usually the reason you get coasters is that the cd drive cant handle burning at high speed.. turn it down to 4x or so
as you bo faster - reletive time slows down. (time reletive to something stationary)
right, k3b's dialog is at 53x, and i want 8x and its grey:( i know
lerneaen_hydra: it has an asymptote at c, so time will go 10x faster at .9c than at .99c
hmm, I see
tomp: I'd suggest stuffing a blank cd into the burner. Ubuntu will pop up a widget asking what kind of cd you wish to burn.
the formula is 1/sqrt(1-v/c), where v is your velocity and c is the speed of light in the medium
hmm, I see
that's funny though
it would seem that the speed through the worldline is not constant then
speed_time + speed_spacial
err, speed_temporal + speed_spacial
are you thinking of a 4-D velocity vector?
almost, a 1d spacial and 1d temporal universe
so 2d essentiall
but yes, that idea
rayh: thanks, that part works, and it tries to burn at a rate that it wont let me change, the dialog button is greyed out, thanks. i just installed gtoaster and it doesnt find the cdrw at all. i think i hav eto remove the newly installed 8255 & m2i50 boards and recheck ( it used to work )
you have to understand the frame of reference thing
both time and space change, by the same factor
hmm, care to explain more?
oh, so the total speed is still conserved?
I wouldn't say "conserved", and it depends on whether you're multiplying or dividing :)
the total "speed" through the universe
plus you have to get around the whole infinate mass as you reach the speed of light.
[18:38:47] <SWPadnos> http://en.wikipedia.org/wiki/Time_dilation
skunkworks; yeah but that's another matter
I'm thinking of speeds between 0 and <c
that only affects acceleration - you need infinite force to get that last little bit if speed
lerneaen_hydra: you have to specify what time frame the "speed" is in relation to
ah, that's true
SWPadnos: not infinite force, infinite energy
same thing, once integrated over time :)
or at least that's the way things look from <0.0001c experimental setups
i think they're mostly just misinterpreting the equations
or the equations aren't a perfect representation of reality
if you're transfering momentum to a particle in a particle accelerator using magnets, you're actually using an electromagnetic wave to push it.. but the wave cant travel faster than C
that doesnt mean that if you have a spaceship traveling at some fraction of C it will require more energy to accelerate
if only someone knew how fields actually work ...
lots of people claim to
I don't think so
people claim to know what they do, but not the underlying mechanism that makes them work
it's a lot of work to overthrow a scientific paradigm, especially when you dont really know all the details
like "gravitons" ...
"graviton" is as semantically meaningless as "gravitational field"
and I'm not sure anyone knows how gravity is communicated with the fabric of space-time
or B/E fields
i still dont see how the michelson-morley experiment "disproved" the ether
I don't even remember what that experiment was - it's been so long since I've done it
there's a lot of similar concepts to the ether theory in "dragging" of intertial reference frames
basically it showed that the earth's rotation doesn't affect the speed of light
it can be shown that special relativity doesn't deny the existing of an absolute frame of reference
but that "absolute frame" can not be detected anyway
not if you dont look for it
absolute rotation is easy to demonstrate
it was about that theory of representing the universe as a cellular automation
imagine that the automate is an absolute frame of reference
then we have a constant speed of impact transferring
when a body achieves high velocity, the processes in it become slower, looks like "time dillation"
and "the twins paradox" is a matter of simple paperwork too in this case
why is that called a paradox?
fenn, because it's counterituitive
it's hard to imagine that there were two twins, and then suddenly one of them becomes older thatn the other
what if you had two twins, one who lived at sea level and another who lived on everest
the gravitational field strength difference would cause one to age faster than the other
or however you explain that in relativity terms
fenn, general relativity is a hard part with that cellular automate representation
it was claimed to be solved, but I didn't bother to read
its probably more like an octree, with "adaptive resolution", so you get all those weird quantum mechanics bugs
i dont really like the cellular automata explanation though
I have a plan to program a variable-resolution octree for collision detection once I have time
fenn, it's not required to use the cells to explain special relativity in terms of an absolute frame of reference with constant impact speed
the funny part is that the light travels with different speed in different directions there
but it can't be detected
it can be detected
fenn, no, because the moving observer considers those directions identical
ah, gravitation again
why does the light curve if it isnt traveling at different speeds?
fenn, I mean the vacuum case
and index of refraction..
gravitational lensing is the same as varying index of refraction of the vacuum
that's beyond my understanding already
index of refraction is caused by a different speed of light within a material
how can you change the index of refaction of a vaccum?
usually a material with higher density will have a higher index of refraction
skunkworks: it happens in nature around massive objects like stars and black holes
fenn, yep, I've just looked up 'refraction' in the vocabulary
fenn: I see. I wasn't putting togather the vaccum/gravitational lense togather.
i think its easier to say that the speed of light changes, rather than saying that "space is bent"
this gravitation<->cell density idea is interesting
hello all. quick question. when i restart emc (with either gui) the last coordinate shows rather than zero. where is that info stored?
in the .tbl file
fenn, yep. more cells per unit mean smaller speed
erDiZz: ever played mud's?
only saw it once
when you build a mud world, you have to link each "room" to the neighbors
what .tbl? the tool table? in what directory?
oops its not .tbl
What configuration are you running, Roguish?
erDiZz: if you arent careful in your grid layout, sometimes you have to link rooms together so that there's more rooms to go from a to b to c than from a to d to c (that make sense?)
erDiZz: although you're going from point a to c you still need more hops along the way, and so it takes longer
It should be in your copy of the emc2/configs/m5i20 directory
fenn, like there was a heavier gravitational field in d, than in b
yes, the .var file. thanks.
see that 5220 is set to 1
so that it starts with g54 active.
what is 5220?
last position is stored in [TRAJ]POSITION_FILE
taking notes now.
cradek: where did that come from? and why?
what do you mean?
why - because it's useful on a lot of machines
just wondering why it was changed.. it seems like it would cause more problems than its worth
where did it come from - we put it in a while back
if it's not useful for your machine, just don't use it
if you don't want that behavior, don't specify POSITION_FILE, and you get the old behavior (joints start at 0 every time)
it's very useful for stepper machines without switches
i thought the old behavior was to load the position from some variables in the .var file
no, that's never been the case
the coordinate system offsets are saved there, but not the absolute position
where does one specify POSITION_FILE?
the ini file in the TRAJ section
Is this documented someplace?
I see it's in the user manual
POSITION_FILE = position.txt If set to a non-empty value, the joint positions are stored between
runs in this ﬁle. This allows the machine to start with the same coordinates it had on
shutdown.5 If unset, joint positions are not stored and will begin at 0 each time emc is
5 This assumes there was no movement of the machine while powered off. It helps on smaller machines without home
sorry for the bad paste
hm. i wonder why they would bother saving g92 and g54 offsets if the position isnt saved
I see that this is a 2.1 thing.
funny though, i have no POSITION_FILE line in my .ini file. i switched from tkemc to axis and back.
fenn the assumption was that the machine, rather than the interpreter handled this issue with home switches or such.
That's why the .var did not save current position during shutdown.
03flo-h 07v2_1_branch * 10emc2/src/po/de_axis.po: updated for EMC2.1
Roguish: if you have g54 offsets active, and your joint position goes to 0,0,0 you'll see nonzero values upon startup
Roguish: you can clear it with G10L2P1 X0Y0Z0
I just tried the distributed stepper/inch sample configuration, copied to my home directory. It does not have [TRAJ]POSITION_FILE and when started the next time it starts at (0,0,0)
ok that explains it. i put in an offset while in axis.
so it set the .var file.
and tkemc followed suit.
without the POSITION_FILE.
POSITION_FILE and offsets are different
another mystery solved!
glad my ignorance could be useful.
glad you got it figured out.
fenn: because after you home, you get all your coordinate systems back
catch you all later.
another question: the ferror value? if LINEAR_UNITS=INCH and ferror=1.00, then the following error limit is actually = 1.00" ?
yes at max speed it's 1 inch
holly crap, i got something wrong then. i keep getting ferrors at ferror=1.0
it only changes if you also have MIN_FERROR set
if you only specify FERROR, then it's a constant regardless of speed
wouldn't this be an issue with base periods and feeds? Or no stepgen headroom. Assuming a stepping machine.
what version are you using?
is it stepper or servo?
servo, m5i20 with yaskawa ac servo drives and motors
cradek: when using a position file will it remember where you were, if you re home your machine?
maybe it's wildly mistuned?
can you post your ini and .hal files?
anonimasu: POSITION_FILE is mostly intended for machines without home switches
anonimasu: the position file and homing are both ways to reestablish the machine position
does it store your offset if you have home switches?
the big mill at work does that..
yes but homing resets it
even if you power down you start up in the same coordinate offset..
anonimasu, homing is what you use when the machine is running, the position file is what you use to remember the state of the machine between runs ...
they're two ways of establishing machine position
but, does emc store your position from the last run?
if you set POSITION_FILE, then it does
SWPadnos: hm, here's the thing..
that gets screwed up if you move the machine while emc isn't running though
lets say I do a rapid, my machine stalls..
I re-home it, will my offset (workpiece datum) be the same?
the homing process does not change gcode offsets
well, that wouldt be g-code offsets..
machine zero is independent of coordinate offsets
I'd like it to store zero between runs..
or well "touch off" whatever it sets..
I love being able to run parts and then power down everything and re-power up the mill and just go to x0y0z10 and end up at the same place :)
the AXIS touch-off button sets the G54 offset which is stored between runs
assuming that the "var" file is writable
nice that it works like real controls :)
and that emc is shut down cleanly
looks like a major thing I need is home switches
hmmrrnn... what can I do with the 18KV AC psu that I just got
where's the cat when I need him
eholmgren: tesla coils.'
anonimasu: yes -- it doesn't do any good to have emc know that the workpiece is at (1,2,3) from the machine origin, if emc doesn't know how to move to the work origin from an unknown position.
jepler: I need to do like 80 parts on the mill and I'd hate to lose my zero between them..
is there any simple, easy way to get axis.xx.f-error out to a file?
not really - but using halscope will give you a nice plot
the gui output at the bottom of the screen is only to 3 significant figures. any way to increase that?
change the format and recompile?
no, but you can turn up the scale and read it off the plot
yes you could do that I guess
does halsampler/halstreamer work with parameters?
(and is ferror a param?)
yes, it's param
cradek jepler jmkasunich_ : my school has a robotic arm exactly like the one in your simulation. If I pull some strings perhaps they'll let me try out the axis controls on it
crepincdotcom_ is now known as crepincdotcom
alex_joni, my old company is closing down shop in California, and Toronto, to go to Romania!
alex_joni, www.embarcadero.com evil company
what do they do?
they don't seem to know much about short page titles
<title>Embarcadero Technologies, Inc - Provider of strategic data management solutions that enable companies improve the availability, integrity, accessibility and security of corporate data.</title>
they forgot to add releive companies of overflow cash
that would only fit in the title bar on my high-res monitor
not just DB2, "IBM DB2 UDB for LUW"
SWPadnos, the company is a sweatshop.
the guy on the front page does look a bit nervous
I'm working on EMC's gcode interpreter adoption. The question is about GET_EXTERNAL_ ("world-give-information") functions. If I understood correctly from the docs, they are not required to return real-time ("exactly at this moment") information, right?
are they called when nothing is moving only (or not)?
are they used anywhere except after probing? after probing the machine is probably stopped
I think probe is the only move where you don't know where it will end
the state desired is the state after any prior actions have been taken, and before any following actions are taken
that's a good way of saying it
so I can safely assume that they're called when everything is supposed to be stopped?
why not look and see where they're called?
cradek, I did
ok where are they called?
they are called from Interp::read
_and_ from set_probe_data
Interp::read confused me
I think that's one of the places I've been baffled too
it's when _set.probe_flag == ON only
and for the GUI I'll take the stubs that AXIS uses
more confidence is always better, thanks
03flo-h 07v2_1_branch * 10emc2/src/po/de_axis.po: updated, because the head was wrong
[21:50:49] <alex_joni> http://forums.cnet.com/5208-10149_102-0.html?forumID=7&threadID=232718&messageID=2398270&tag=nl.e406
6.493 kWh, yesterday was 6.543, no new record :)
or, nearly 30cents worth ,...
ah, wrong channel
$1.16 worth actually
Were those european numbers or american numbers. (Does the "." represent a 1000's separator or a decimal separator?) :-)
one point one six u.s. dollars
fenn_gohan is now known as fenn
I was referring to the 6.493 kWh. That could be 6 thousand plus kWh... since there was no context, it could be for a house or a town.
wheeeeee Gumstix has the RT extensions
did you pick one of those up?
looks I'll never get rt installed on this laptop
has anyone had problem installing RTAI on a laptop?
I know there are problems with rt and acpi sometimes
ejholmgren_: do you think that disabling acpi could help?
I couldn't hurt
having problems installing rtai ... or using it?
but test programs not
I get an error inserting rtai_hal.ko
I have googled a lot so far..
and tried with different kernels
I'll try and disable acpi
well... night all