I read that nobody saw anything bad with my classicladder patch.
Does that mean you will allow it?
did you read where somebody nominated you as the official CL maintianer for EMC?
yes something like that
do you have commit access to the cvs server?
I'll do what I can
I have to apply right?
just send the pertinent info to cradek
I'm looking for the pertinent info now
oh that was the next quest
wtf... sourceforge is at it again (or somebody made a typo somewhere)
[00:37:49] <jmkasunich> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?CVS
thats the first step
there's a link to an archived list message there with the real details
but the link now goes to the wrong place
is there current info about it on the wikki?
oh oops thats what you are talking about
I should read the whole line :)
the link right before the .cvsrc section header is the one that should give the answers
and instead gives some stupid spam about italy
arg... sourceforge is spiraling downhill fast
mail archives page takes forever to load, but they make damned sure the ads work
Ur not a fan, I take it
yes of course
[00:46:07] <jmkasunich> http://jmkasunich.dyndns.org/pics/sourceforge_sucks.png
which page are you looking for? I seem to be getting through
I'm trying to search the archives
a retry got me 98 messages, I'm trying to find the right one
sourceforge wins again
use the gmane list archives - they work right
I've now found two independent places (one in the wiki, one a message in the archives) that are suppose to point to your message about how to submit a key
[00:53:10] <cradek> http://news.gmane.org/gmane.linux.distributions.emc.devel
they both point to something completely different
[00:53:25] <jmkasunich> http://sourceforge.net/mailarchive/message.php?msg_id=15301001
does gmane use the same message IDs I wonder?
well I can't find it either
the strategy is this:
send me your ssh public key
I'll make a cvs account with a username that matches your sourceforge username
jmkasunich will add you as a developer on sourceforge
that's all :-)
ok. but what is a ssh public key :)
the post that I can't find explained how to generate the key and everything
[00:59:10] <cradek> http://sourceforge.net/mailarchive/message.php?msg_id=20060404143141.GP9018%40timeguy.com
truncate it at the .GP and you get the italy spam
it looks like our friends at SF have broken old URLs that refer to the archives
I'll put the new link into the wiki page
I fixed our wiki page
what I can't fix is their screwed up email censoring
cvs -d:ext:anon@cv...:/cvs co emc2
that's what they screwed up my email to say
but many people (wrongly) think that's a feature I suppose
did you see the screenshot of their search results I posted a few mins ago?
ChrisMorley: I have to run for a few hours, but I'll get your account set up tonight still
sounds good when ever is conveient
you have a sf account right?
I applied but I could not log in. i haven't tried lately
ok, we'll have to figure that out too
to be honest, I don't know if there is still any reason to have an SF account
jmk can help with that, he has all the powers I do
it has nothing to do with access to CVS anymore
jmkasunich: to assign stuff in the bug tracker is one big reason that comes to mind
oh, I guess it lets you modify bug trackers
I have to run, bbl
ChrisMorley: getting an account at SF is the first step (if you want to be able to do things with trackers)
cvs access is independent of SF now, and chris will set you up there
for SF, you gotta get the account working yourself, then tell me your SF user name and I can give you project access for EMC
ok I am trying to find all the info again for source forge
I think I'm gonna copy the info from the message directly into the wiki, instead of waiting for SF to bork it up again
I think good Idea!!
Ok I have a working acount at sourceforge. Chester8888
ok, you are now listed as a developer
great! now I send chris the ssh key and all is good?
you'll need to do a developer checkout, apply your changes to that checkout, and commit from there
you might want to do a test commit first
then I can commit changes. What is standard procedure as not to screw up other people?
depends on what you are doing
unless you are making very radical changes, just commit them on trunk (the main dev version)
yes that is what I meant. anything big of course I wil confir first
if you want to commit partially complete work that will break the build until you commit the rest, you should use a branch
oh, one last thing - if you aren't subscribed to the commit list, you should do that
in addition to keeping you up to date, if you break something, the compile farm FAIL messages go out on that list
by partailly complete do you mean some thing that will not compile properly or something that does not work?
certainly if it breaks the compile, don't commit it (or use a branch)
if it breaks a function that few people use, its a gray area
ok. I will try not to break anything!
I dunno about other developers, but I do a cvs up / configure / make each day before I start working on things
so I'm always up to data
if trunk doesn't compile though....
I have to request to make a branch?
no, you can do it yourself using cvs commands
need to read the online docs
ok I will read up
we don't use branches that often
if I have something half done, I just don't commit it
well I can test till it works then commit
I almost always do a 'cvs diff -u' before a 'cvs commit -m "some message" '
the diff will show you all the changes you are about to commit
your position is not to comfit something that will break old configs till major revisions
sometimes thats more than you think
yes I have seen
trunk is the working area - if there is a good reason to change a config, we discuss on IRC or list, and then commit it to trunk
we do NOT do changes to 2.0.x or 2.1.x that will break configs
but there can be (and will be) config changes needed to move from 2.1.x to the upcoming 2.2
yes ok thats what I thought. I will not be doing anything like that till I talked to you guys anyways.
2.2 will branch off of trunk shortly before we're ready to release, then we test for a while, and finally do the 2.2.0 release
just minor stuff
have you looked at the webcvs?
[01:29:07] <jmkasunich> http://cvs.linuxcnc.org/cvs/emc2/
its a really nice tool
especially version graphs
[01:29:35] <jmkasunich> http://cvs.linuxcnc.org/cvs/emc2/VERSION?graph=1
VERSION is the file that tells a running EMC what version it is
Help -> About prints the contents of VERSION, for example
so it gets touched for every release
"normal" files tend to get touched more on the trunk, like this:
[01:31:13] <jmkasunich> http://cvs.linuxcnc.org/cvs/emc2/src/emc/motion/command.c?graph=1
if you scroll down to version 1.76, you can see a branch that chris made to work on rigid tapping
and below that the 2.1 release branch
yes I see it is very clear.
what is the longer term vision for emc beyond 2.2?
major revisions add new features, minor ones are bugfix only
I had read back a bit about cnaging to using more c++
since we always want to add new features, there will always be major branches comeing out
I don't see major changes in language, but who knows
I think a peter was talking about that
different parts are in different languages
but there is no major change like emc 3.0 in the forseeable future ?
we really don't have the manpower to do radical redesigns
every developer has pieces that he thinks should be totally redone
no developer has the time and energy to actually totally redo a large part of the program
if all the developers agree on which piece needs redone it might happen
but everybody has a different part they hate
what do you hate?
the parts I don't understand
which is task mostly
well then i hate lots....:)
some parts I don't understand but can accept on faith
other parts I understand just enough to not like them
It's amasing really that it wirks at all
the interpreter and the GUIs are the former
task is the latter
Thank you for your help John but I must go now.
[01:57:38] <jmkasunich> https://sourceforge.net/tracker/?func=detail&atid=200001&aid=1726272&group_id=1
not that it will do any good...
just for fun: http://sourceforge.net/tracker/index.php?func=detail&aid=1726276&group_id=1&atid=200001
at least the page didn't time out because one of the ad servers was slow, that drives me nuts
of course, I could have an ad blocker, and I'm too lazy for that
I really liked SF when I first started on this project
their business model never really made sense though
give away storage, what a plan
You mean like Ubuntu giving away from Cd's for the asking, postage included?
jmkasunich: your bug reports are sad and entertaining at the same time
then I accomplished my mission
I'm glad I only have to run stuff for one project
* Jymmm swore jmkasunich's mission in life was to be the King of all Dumpster divers and to build an Empire on the wealth he has gathered.
yeah, they're are inundated... but many of their problems are their own doing
i found a 49,000 dollar bandsaw
Hey is there a viso like utility to create electrical diagrams that anyone knows of?
Okey, this is scarry.... http://edc.sourceforge.net/english.htm
thats kinda a silly lookin npn
works i guess
I meant the cardboard box holding the 120VAC relays
does emc have native support for remote displays, or should I use normal x remote displays?
Um yes, sorta kina
You can run mini on a seperate computer and control your mill remotely
you can even ue keystick via terminal
what about axis?
what about it?
i would like to run axis on a separate machine and talk to the emc machine over the network
try it =)
ah... i will ;)
I haven;t tried remote axis, but I have the other GUI interfaces.
it's kinda cool actually.
BUT.... if the remote gui looses connection to emc, emc goes bubye
maybe you can use ssh -X
acemi, yes, but that is standard x forwarding, which you can do on a lan without ssh
Jymm: try Eagle, by cadsoft
hello, I am still trying to solve the problem with pwmgen module, and i'm running out of ideas of possible failiures. I am monitoring the signal with the osciloscope, but the signal is always on logical 0. This are my hal file settings: http://pastebin.ca/513862
any idea what could be wrong?
you are not running the pwmgen.update thread
typically that function will be added on the servo thread
er, "the pwmgen.update function"
i added pwmgen.update base-thread to hal configuration and i get error: "pwmgen.update need FP". How can i solve this? What FP (floating point)?
What is this?
kwaj: reread the manpage or docs for pwmgen: it will have two functions to add, one in a slow thread with FP (servo thread) and one in a fast thread without FP (base thread)
Kidding, I have no idea. The ones that do should be here shortly, but this is a 3 day holiday in the US.
Estop Button (Have have additional photos if anyone interested) http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ih=009&sspagename=STRK%3AMEWA%3AIT&viewitem=&item=190115709940&rd=1&rd=1
i solved the problem...i changed the base-thread to servo-thread
* JymmmEMC pokes SWPadnos with a 0xFFFFFFFF
robin_sz: Think I could make $100/wk making these? Cell phone gun http://www.snopes.com/crime/warnings/cellgun.asp
robin_sz: That's BEFORE someone stupid uses one and tells the cops where he got it that is =)
make ceramic guns
you could CNC a mold up right quick...
chr0n1c: Cool, I'll let you be my official tester =)
chr0n1c: "You'll shoot your eye out kid"
* chr0n1c i already did 8 times
chr0n1c: You 8 eyed freak!
same eye.. 8 times
i have one good one
That must be some suck ammo you have there
chr0n1c: You REALLY should stop buying ammo from those $0.25 candy vending machines.
... bullets made from fiberglass body panel filler!
now we are on to something...
er the other way around
ok.. bad idea, don't do it JymmmEMC
chr0n1c: I have no intension of making arms or ammuniction
i think there is laws about it
I suspect, in the current climate, you'll end up in guantanamo bay, just for thinking about it
robin_sz: I ECHELON have KILL no PRESIDENT idea BOMB what HIJACK you're NUCLEAR talking EXPOSION about.
robin_sz: But you might find this interesting "The CALEA compliance deadline remains May 14, 2007" http://en.wikipedia.org/wiki/Communications_Assistance_for_Law_Enforcement_Act
pointless waste of time
robin_sz: Ironically though, a Data Center does not fall under the definition of a "broadbad carrier"
liek any real threat would be organised through plain text
personal meetings and pgp encrypted mails
I suspect PGP has a backdoor though, same way that ie/netscape did
Yeah, I know. But look how the unPatriot Act affected many other unrelated things.
robin_sz: ie/netscape circa 1980's you mean?
2001 it was discovered
ITAR ***REQUIRED*** a back-door to be given to the feds.
back then that is.
No idea about now.
yeah, but then ther ewas the "workfactor reduction field"
never heard of it.
* robin_sz falls of his chair in suprise
you mean reducing the amount of time to decipher?
in multiple veriosns of the https protocol, it sends (128-40) bits of the key encrypted with the NSA public key in the request
so, they can extract most of the key in a jiffy, leaving just 40 bits to find
citations, well, less and less ...
[17:20:13] <robin_sz> http://www.fuckinggoogleit.com/?search=work%20factor%20reduction%field
Hello- I am an EMC noob trying to setup up simulation- can't get axis to work, though tkemc works fine. This is the first error reported:HAL:1: ERROR: Can't find program 'hal_manualtoolchange'
crotchetyGuy: if you are building from source, make sure that you have all the required development libraries installed. that message might occur if you do not have the python development libraries, for instance.
ok - thank you- any idea which libs?
compiled python from source, so..
[18:10:32] <jepler> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Installing_EMC2#Preparing_other_versions_of_Linux_to_compile_emc2
checking it out....
you should also take the time to read the output of "configure": it tells you when things are missing, but generally tries to continue
you can also look at debian/control.in for a list of required packages (using ubuntu names) -- that has a greater chance of being up-to-date than the wiki package lists
ok- btw, using suse 9.2
they don't include python?
yes, just installed a newer version of python, if that's what you mean.
I will check configure output now too...
result: Python version too old (2.4 or newer required)
however: Python 2.4.3 (#5, Jul 12 2006, 21:12:27)
is what I have now.
old version of 2.3 in /usr/lib
you should see this in "./configure --help"
ok - put the correct path in?
ok - that got it - thanks jepler:)
03jepler 07TRUNK * 10emc2/src/ (configure configure.in): By not specifying the 4th argument to AC_PATH_PROG, the user's path will be respected by default. Add "sbin"-type directories to the end of PATH for certain programs which searched them before
03cmorley 07TRUNK * 10emc2/src/hal/classicladder/projects_examples/vcp.hal: This is test commit changed a comment line in vcp.hal of classicladder src file
humm, got problem, emc is controllable from the keyboard for run/stop etc. But its not moving either the displayed position or the machine, and all the power leds are on normally. Anybody have an idea?
please be more specific: you hit F1 to come out of estop... is there any visible sign (on screen or elsewhere) that it came out? then you hit F1 to turn machine on... is there any visible sign that it responded? then you try to move (how, jog key?)
oh, and what GUI are you using?
that all responds like normal, and its axis, emc2.15, uptime for ewmc is weeks
jog keysw dead, both on scren and at motors
I have that option to save positions turned on, so I'm going to restart it, yes, steppers.
so that rules out the entire stepgen/parport/drives/motors section
it was running my program, on about the 20 pass thru the loop & it just stopped
thats kind of important - I thought you were describing some kind of config problem
"stopped in the middle" ain't the same as "won
as "won't start"
nope, config seems to be good
so you were executing g-code when it stopped?
yes, about the 20th pass thru the loop, doiing a final finsih cut
and then you put it in manual mode and tried to jog, and that didn't work either
Also, even though its showing stopped on the display, the exit etc options are greyed out in the menu.
what do you mean by "showing stopped"
yes, manual is dead, as if I'm not pushing the key(s)
you're not answering my questions
alex_joni I heard him say axis
yes, axis, plotting display is also frozen
bet axis is completely frozen
sorta looks that way from here
so then how did you put it into manual mode?
thats your baby isn't it Alex? Only the backplot part of the display is frozen,
not my baby gene
but I think that you can still "save" emc by opening a terminal
and running a second GUI to connect to it
cradek and jepler wrote axis
alex_joni why do you think its locked up? have you seen that happen before?
jmkasunich: I had it 1-2 times but for limited time
couple seconds at most
well before having him do anything, lets get the facts straight
well.. I'll hand it over to you
I'm heading for bed anyways
good night all
I'm running axis right now... when executing a program, the axis letters, jog mode, home button, and touch-off button on the manual tab are grayed out
so, gene_ when it died, did those remain greyed out, or did they turn black again?
this was for about 15 minutes. I finally got it to stop and restart and it came back up showing the correct co-ordinates for its present location, but when I hit run, it went the wrong way, dunno if I have a bit left or not but when I hit
Everything except the backplot and the macinbe motion was looking normal.
jmkasunich: good luck
you posted that you had a problem at 6:29, and asked for help... I responded within a minute... have you been messing with it since then?
its 20 minutes later, and A) I still don't have answers to any of my questions, and B) you'll never be able to answer them if you've been messing with stuff in the meantime
I may need it. I finally got it to admit it was stopped, so the exit in the pulldown wasn't greyed out, so I stopped it, and that took about 40 seconds before I got my prrompt back in that shell, and it was displaying the single figure '2111' so I had to uparrow twice to rerun it.
at this point there isn't a damn thing I can do to help you
(I might not have been able to help 20 minutes ago either, but at least I'd have had a chance)
* jmkasunich goes to eat dimmer
Now its moving aagain, but it ran into every stop and slipped the steppers when I hit r for run. Oh, oh, I know what that was, it was going to exec the axis logo, I forgot to reload my program.
So I may be alright now, IF I can get my home locations reset, thats always a cast iron bitch for y in particular for this job.
humm, how do dimers taste? :)
humm, how do dimmers taste? :)
enjoy your lunch :)