#emc | Logs for 2006-09-18

[00:04:13] <jmkasunich> its faster to download at 300 baud
[00:30:58] <cradek> I hope people don't think they should download that iso for anything but an initial install
[00:31:39] <cradek> downloading an iso (and reinstalling your entire OS) to get a new version of emc or some other update is crazy
[00:40:36] <jmkasunich> I know
[00:41:02] <jmkasunich> thats why I'm less than fond of things like "emc live cds"
[01:00:58] <cradek> this is good for a first install - the rest is just a matter of education I guess
[01:01:45] <cradek> one obvious way we can keep people from downloading an iso and reinstalling their OS when we release emc 2.0.4 is to just not make a new cd
[01:02:17] <cradek> perhaps this will be the last (only) cd we will make from dapper, which will keep us going until 2011 if all goes as planned
[01:02:28] <SWPadnos> they'll get the "update" icon anyway (if there's an internet connection onthe machine)
[01:02:42] <cradek> yes
[01:03:14] <SWPadnos> that's the most likely way that a user would learn of an update, I think
[01:03:25] <cradek> also for these reasons I did NOT put a version number on the "distribution"
[01:03:56] <cradek> we'll see how it goes I guess
[01:04:07] <cradek> hope it's the right thing to do
[01:04:25] <cradek> I have to say it's nice to be able to boot a machine from cd and see if X, realtime, opengl, etc work for emc.
[01:04:40] <SWPadnos> hmmm. is there a reasonably trivial script or procedure that a user can follow to update from a CD (or at least a file downloaded on a different machine)?
[01:05:01] <cradek> you mean a non-net-conencted machine?
[01:05:05] <SWPadnos> yep
[01:05:14] <cradek> sure, get the deb, dpkg -i the-deb
[01:05:15] <SWPadnos> download on one, update on another
[01:05:35] <cradek> someone (dave-e?) used a thumb drive to do this
[01:05:49] <SWPadnos> ok, that's for an upgrade?
[01:05:53] <cradek> yep
[01:06:07] <SWPadnos> good deal. even I could write the directions for that ;)
[01:06:15] <cradek> it's pretty easy
[01:06:34] <cradek> if you plug in the thumb drive, I think a file browser appears, and you can click on the deb
[01:06:53] <cradek> it says "do you want me to run the package installer thingy?" and you click ok
[01:07:24] <SWPadnos> ok. that's easy enough
[01:07:41] <SWPadnos> I wasn't sure, since I'm justa bout always net-connected
[01:07:42] <cradek> this cd is a real bonus for dialup folks
[01:07:58] <cradek> you can get a friend to make you one, and it's all you need
[01:08:08] <cradek> emc updates are simple to download over dialup
[01:08:23] <SWPadnos> yep. that's a bonus. along with the testing features you mentioned earlier
[01:08:26] <skunkworks> All I can say - the current ubuntu setup make emc2 easy as pie.
[01:08:41] <cradek> skunkworks: it's easier now :-)
[01:10:04] <skunkworks> I heard about the live cd. Cool
[01:11:04] <skunkworks> but still - it isn't like the bdi - where you had to install a bunch of dependencies and then get the csv. (talking out of my arse a bit)
[01:11:24] <cradek> the biggest difference is bdi is not a live cd
[01:11:28] <skunkworks> install ubuntu - update it if you need to - run the emc2 script
[01:12:09] <skunkworks> * skunkworks says painless again
[01:21:51] <LawrenceG> cradek: bummer... image is 655mb and all I have is old 650mb cdr's :{
[01:22:14] <LawrenceG> time to spend 0.50 :}
[01:36:57] <cradek> SWPadnos: alex wanted me to ask you to make sure he set the permissions on the wiki correctly
[01:37:41] <cradek> LawrenceG: I have some of those old ones too...
[01:40:45] <skunkworks> cradek: do you have any videos of your servo setup running yet?
[01:40:57] <cradek> jepler and I made one, but it sucked
[01:41:06] <skunkworks> aww
[01:43:24] <cradek> http://cfn13.com/StoryHeadline.aspx?id=18686
[01:44:43] <jepler> thank gosh they had a box of tissues
[01:45:07] <skunkworks> that as bad as "uncontrolled accelleration"
[01:45:50] <jepler> here's the video such as it was -- too bad you can't see the threads it makes. http://emergent.unpy.net/files/sandbox/threading.avi
[01:45:54] <cradek> is that when someone WAY too old to drive pushes the wrong pedal to stop?
[01:49:54] <skunkworks> Where something "happens" to the car so that the gas stuck pushed all the way down. people get into accidents because they don't know what to do.
[01:50:46] <cradek> except for being out of focus, that's an ok video I guess
[01:51:29] <skunkworks> 15kbs hosting porn also? ;)
[01:51:46] <jepler> skunkworks: leave my poor DSL alone
[01:51:46] <cradek> I was downloading it too, but I stopped mine
[01:52:05] <jepler> or at least don't bad-mouth it
[01:52:34] <skunkworks> sorry.
[01:53:06] <skunkworks> I will send you some bandwidth. whats your email?
[01:53:20] <skunkworks> do you need any electricity? I could fax that to you.
[01:53:38] <cradek> fax me a pizza would you?
[01:53:59] <skunkworks> also I ment 15KBs
[01:54:45] <jepler> mmm pizza
[01:55:07] <skunkworks> you do pizza? veggi pizza?
[01:55:08] <cradek> jepler: we just had falafel - he's open 'til 10, hurry
[01:55:26] <cradek> if by 'do' you mean eat, sure
[01:57:25] <cradek> my cat jumped from the top of my monitor, over my head, onto the back of my chair
[01:57:35] <skunkworks> we just had a early thank-giving meal. Wife dad goes out to montana so we have a big feast as he won't be here on thanks giving
[01:57:38] <cradek> I ducked a bit, but it was still impressive
[01:57:52] <skunkworks> Nice
[01:58:10] <cradek> oh man, is it almost that time of year again?!
[01:58:16] <jepler> cradek: yep
[01:58:20] <cradek> curses
[01:58:21] <jepler> haven't you noticed it's gotten cold outside?
[01:58:31] <cradek> SHUT UP
[01:58:34] <cradek> LALALALALALA
[01:58:46] <cradek> oh sorry
[01:59:01] <cradek> but no, I'm pretty much in denial still
[02:00:06] <cradek> did you guys watch 'the secret life of machines' in the early 90s?
[02:00:24] <skunkworks> was that a pbs series?
[02:00:26] <cradek> it was a fun little 'how things work' series
[02:00:42] <cradek> the author has nicely made them all available with bittorrent
[02:00:48] <cradek> they're fun to watch again
[02:00:54] <cradek> http://www.timhunkin.com/control/n_tv_index.htm
[02:01:00] <skunkworks> was that where at the end of the show there was a robot arm doing work - then it unplugs itself?
[02:01:16] <cradek> there was always something funny at the end of each episode
[02:01:29] <skunkworks> or a I thinking of a totally different show? have to look
[02:01:31] <cradek> I don't remember that one but I bet so
[02:01:39] <cradek> that's very much his style
[02:01:45] <skunkworks> :)
[02:01:49] <cradek> he was a cartoonist too, and a lot of the stuff is cartoons
[02:06:49] <skunkworks> well - time for bed.. I will have to watch the video tomorrow as my windows media player is acting stupid.
[02:06:53] <skunkworks> thanks though
[10:53:10] <SWPadnos_> SWPadnos_ is now known as SWPadnos
[12:17:09] <Lerneaen_Hydra> does anyone here know of a page that has values for UNC/UNF threads?
[12:17:21] <Lerneaen_Hydra> like pitch, total depth, outer diameter and so on?
[12:17:56] <SWPadnos> Machinery's Handbook?
[12:18:06] <cradek> I've looked for a website and always end up getting out the book
[12:18:16] <Lerneaen_Hydra> I don't have a yankee table in mine
[12:18:20] <Lerneaen_Hydra> only metric
[12:18:26] <Lerneaen_Hydra> (at least for the values I want)
[12:18:28] <cradek> and either the book doesn't explain it well, or I'm dense, or both
[12:18:39] <alex_joni> cradek: denser than water?
[12:18:43] <cradek> could one of us could look it up?
[12:20:08] <alex_joni> cradek: it's either too early for you, or too late for me.. that didn't make much sense
[12:20:26] <Lerneaen_Hydra> the only thing I know is that it's 1/2" UNF, and that is 20TPI
[12:20:50] <Lerneaen_Hydra> what I need to know is the total radial depth
[12:21:01] <Lerneaen_Hydra> ie how deep to go on the final cut
[12:22:47] <Lerneaen_Hydra> could ID 0.445" be correct?
[12:22:49] <cradek> 2A2C;./'l
[12:22:59] <Lerneaen_Hydra> ...
[12:23:09] <alex_joni> cradek: your cat is not making much sense today ;)
[12:23:13] <Lerneaen_Hydra> I can't say I understood that
[12:23:20] <Lerneaen_Hydra> * Lerneaen_Hydra agrees with alex
[12:23:27] <cradek> she stepped in the sink and got her foot wet
[12:24:02] <Lerneaen_Hydra> oh, my cat would not like that
[12:25:15] <Lerneaen_Hydra> http://homepages.tesco.net/~A10bsa/unfgo.htm <-- seems to say that I should go down to 0.4387"
[12:25:19] <Lerneaen_Hydra> for 1/2*
[12:26:13] <cradek> hey I'm bookmarking that
[12:26:47] <cradek> well they don't say what measurement on the drawing that number is
[12:27:02] <cradek> if you have a sharp pointed tool you may have to cut deeper
[12:27:20] <Lerneaen_Hydra> yeah that's what I was thinking
[12:27:48] <Lerneaen_Hydra> the tool is quite sharp, though the chuck that's going to sit on the axel had quite shallow threads
[12:27:54] <Lerneaen_Hydra> so it should work
[12:28:14] <cradek> oh you have the part? then cut to fit!
[12:29:05] <cradek> otherwise I can look in the book for you tonight if you haven't found it
[12:30:20] <SWPadnos> oh man - one needs to be careful with babelfish translations :)
[12:31:31] <SWPadnos> here's a somewhat strongly worded request for item status, translated English -> Italian -> English:
[12:31:35] <SWPadnos> Hello. Still I have not felt to you null as far as this article. If I do not feel myself from you in 3 days, will obtain a reimbursement, beginning the procedures of PayPal and eBay of claim and leave the answers negative. I pray says to me that what the condition of this article is. Thanks
[12:32:04] <SWPadnos> at least the hello and thanks are correct
[12:46:11] <CIA-8> 03jepler 07HEAD * 10emc2/src/hal/classicladder/classicladder.c: with this change, 'halcmd unloadusr' can be used on the nonrt part of classicladder
[12:51:18] <Lerneaen_Hydra> cradek: I cut it and it fits great :)
[12:55:34] <skunkworks> logger_aj: bookmark
[12:55:34] <skunkworks> See
[13:17:25] <Lerneaen_Hydra> uh, is there interest in having my lathe and mill edgecam post processor in the wiki?
[13:23:58] <jepler> sure
[13:25:04] <Lerneaen_Hydra> oh, random question, how many decimals does the interpreter use?
[13:25:11] <CIA-8> 03jepler 07HEAD * 10emc2/src/hal/utils/halcmd.c:
[13:25:11] <CIA-8> create "unload" which can unload a component whether it is realtime or userspace.
[13:25:11] <CIA-8> improve online documentation.
[13:25:19] <CIA-8> 03jepler 07HEAD * 10emc2/docs/man/man1/halcmd.1: document unloadusr and unload
[13:25:23] <Lerneaen_Hydra> say for example I set K for threading to 1.2345678901234567
[13:25:46] <Lerneaen_Hydra> when doing *really* ( ;) ) long threads then one would get use of all the decimals
[13:26:35] <jepler> Lerneaen_Hydra: some internal values are stored as doubles (17 digits) and others as floats (8 digits).
[13:26:45] <Lerneaen_Hydra> oh, ok
[13:26:49] <Lerneaen_Hydra> I
[13:26:52] <Lerneaen_Hydra> 'll go with 8 then
[13:26:54] <Lerneaen_Hydra> gah
[13:28:16] <Lerneaen_Hydra> is there a page for CAM post processors?
[13:28:21] <Lerneaen_Hydra> (on the wiki)
[13:28:39] <jepler> not yet
[13:29:51] <SWPadnos> you should be able to add one - the ownership problem was fixed this morning
[13:31:03] <skunkworks> jepler: what codec was that video encoded in?
[13:31:24] <jepler> skunkworks: mpeg4 by mencoder
[13:32:51] <skunkworks> jepler: thanks
[13:38:53] <Lerneaen_Hydra> how do you make new pages in the wiki?
[13:39:01] <Lerneaen_Hydra> oh..
[13:39:03] <Lerneaen_Hydra> right
[13:39:11] <Lerneaen_Hydra> if I were to read first...
[13:56:17] <Lerneaen_Hydra> uh, who has admin acsess to the wiki?
[13:56:24] <Lerneaen_Hydra> the posts are >200kb
[13:56:28] <SWPadnos> I suppose I do. what's up?
[13:56:33] <Lerneaen_Hydra> which means I can't uplaod them
[13:56:33] <SWPadnos> hmmm - one sec
[13:57:09] <Lerneaen_Hydra> you get grab them in ftp://basic:basic@lerneaenhydra.shacknet.nu
[13:57:13] <Lerneaen_Hydra> in EMC related
[13:57:23] <Lerneaen_Hydra> EMC2_lathe and EMC2_mill
[13:58:18] <SWPadnos> hmmm. there's a "MaxPost" setting, but that should be for the page itselt. I'm not sure that it applies to uploads as well
[13:58:40] <SWPadnos> itself, not itselt
[13:59:38] <Lerneaen_Hydra> oh.
[13:59:40] <Lerneaen_Hydra> hmm
[13:59:44] <Lerneaen_Hydra> can you upload bigger files?
[14:00:11] <SWPadnos> I'm not sure. I see some that are 211k, but they may be just under the max size
[14:00:30] <SWPadnos> I also don't see how to upload a file (vs. create/edit a page)
[14:00:45] <Lerneaen_Hydra> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl/emcinfo.pl?action=upload ?
[14:01:46] <SWPadnos> ok. I can change that. what are the sizes of your post files?
[14:01:49] <rayh> It does apply to uploads
[14:02:03] <SWPadnos> I see that on the upload page (very helpful, that ;) )
[14:03:05] <Lerneaen_Hydra> the biggest one is just over 1mb
[14:03:13] <SWPadnos> ok
[14:03:29] <SWPadnos> how much over? ;)
[14:03:35] <Lerneaen_Hydra> IIRC it's 1.04
[14:03:39] <Lerneaen_Hydra> let me check...
[14:03:46] <SWPadnos> in bytes would be better
[14:03:56] <Lerneaen_Hydra> 1,065,622 bytes
[14:04:49] <Lerneaen_Hydra> in EMC related
[14:05:10] <rayh> The wiki is much more approachable now that DH is hosting. Thanks SWP.
[14:05:14] <Lerneaen_Hydra> EMC2_lathe and EMC2_mill
[14:05:15] <SWPadnos> ok. I'm not sure I want to increase it that far. I'll ftp the files into the upload directory
[14:05:20] <SWPadnos> no problem ray
[14:05:20] <Lerneaen_Hydra> ok
[14:05:23] <Lerneaen_Hydra> sounds sane
[14:05:45] <Lerneaen_Hydra> (.cgd)
[14:06:20] <Lerneaen_Hydra> can't go into the dir?
[14:06:25] <SWPadnos> nope
[14:06:31] <SWPadnos> the space kills me
[14:06:35] <Lerneaen_Hydra> test now
[14:06:39] <Lerneaen_Hydra> replaced with a _
[14:07:07] <SWPadnos> yep - I see that
[14:07:09] <SWPadnos> thanks
[14:07:24] <Lerneaen_Hydra> ooh, nice
[14:08:35] <SWPadnos> maybe I should have used ncftp instead (I like the status bar)
[14:08:51] <Lerneaen_Hydra> according to my server the file is done
[14:09:11] <SWPadnos> according to my telnet session, it's opening a dat achannel ... :(
[14:09:36] <Lerneaen_Hydra> :(
[14:09:43] <Lerneaen_Hydra> you want me to upload it somewhere?
[14:09:52] <Lerneaen_Hydra> user/pass basic
[14:09:52] <SWPadnos> one sec - I'm trying with ncftp
[14:09:55] <Lerneaen_Hydra> ok
[14:13:05] <SWPadnos> interesting - I got a "connection reset by peer" message
[14:14:31] <SWPadnos> Lerneaen_Hydra, can you md5sum those so I can check after the transfer
[14:14:41] <Lerneaen_Hydra> yeah
[14:15:07] <Lerneaen_Hydra> 19700444c01e5de738e81a4f67959f96 *EMC2_mill.cgd fca549b70df3bf95348f33035c50f20b *EMC2_lathe.cgd
[14:15:31] <SWPadnos> ok. it doesn't seem to be going too smoothly.
[14:15:38] <Lerneaen_Hydra> strange...
[14:15:52] <Lerneaen_Hydra> should I upload them somewhere?
[14:16:01] <SWPadnos> possibly. let me fiddle for a bit
[14:17:28] <Lerneaen_Hydra> ok
[14:17:57] <Lerneaen_Hydra> (000006) 9/18/2006 16:19:48 PM - basic (> ôÿABOR
[14:17:59] <Lerneaen_Hydra> (000006) 9/18/2006 16:19:48 PM - basic (> 500 Syntax error, command unrecognized.
[14:18:00] <Lerneaen_Hydra> doesn't seem to good
[14:18:25] <Lerneaen_Hydra> (mainly the ôÿ)
[14:18:25] <SWPadnos> yeah. I'll log in again. I think there may be a problem with passive transfer or something
[14:18:36] <Lerneaen_Hydra> that should work
[14:18:52] <Lerneaen_Hydra> though I've set it to use ports in the 5000-5100 range
[14:19:05] <Lerneaen_Hydra> maybe that doesn't work well with your config
[14:19:43] <SWPadnos> is this a text file?
[14:19:54] <Lerneaen_Hydra> nope
[14:19:59] <SWPadnos> ok
[14:20:12] <Lerneaen_Hydra> lots of bad chars and EOL
[14:20:56] <Lerneaen_Hydra> seems to work
[14:20:57] <SWPadnos> ok. one more try. binary, no passive transfer, overwrite the partial files
[14:21:15] <SWPadnos> nope, connection reset by peer
[14:21:19] <SWPadnos> on both files
[14:21:30] <Lerneaen_Hydra> funny
[14:21:49] <Lerneaen_Hydra> my server reported the transfer as succsesfully completed
[14:22:14] <SWPadnos> I wonder if there's some shadow or proxy thing happening
[14:22:38] <Lerneaen_Hydra> maybe
[14:22:51] <Lerneaen_Hydra> AFAIK there's nothing like that on my end
[14:23:01] <Lerneaen_Hydra> at least nothing that I have control of or have ever noticed
[14:23:18] <Lerneaen_Hydra> I could mail it or something
[14:23:47] <SWPadnos> I have lathe, just trying for mill now
[14:24:27] <SWPadnos> I had 759155 bytes out of 436990, so something's definitely weird somewhere
[14:24:43] <Lerneaen_Hydra> yeah
[14:24:58] <Lerneaen_Hydra> you have gftp?
[14:26:07] <SWPadnos> that's the gnome client, right?
[14:26:44] <Lerneaen_Hydra> yeah
[14:26:51] <SWPadnos> if so, the answer is no, since I'm on an SSH terminal connection
[14:26:55] <Lerneaen_Hydra> it's worked for me so far
[14:26:56] <Lerneaen_Hydra> oh
[14:27:00] <Lerneaen_Hydra> what about wget?
[14:27:09] <SWPadnos> yep
[14:27:56] <Lerneaen_Hydra> wget ftp://basic:basic@lerneaenhydra.shacknet.nu/EMC_related/EMC2_lathe.cgd
[14:28:07] <Lerneaen_Hydra> wget ftp://basic:basic@lerneaenhydra.shacknet.nu/EMC_related/EMC2_mill.cgd
[14:28:57] <SWPadnos> strange
[14:29:40] <Lerneaen_Hydra> now it's attempting to resume
[14:29:41] <SWPadnos> I wonder how many times wget will try that ;)
[14:29:50] <Lerneaen_Hydra> I'll remove resume suppoty
[14:29:51] <Lerneaen_Hydra> support
[14:30:25] <SWPadnos> I stopped it
[14:30:57] <Lerneaen_Hydra> would an http link be easier?
[14:31:18] <SWPadnos> hard to say. I'm not sure why FTP isn't working
[14:32:35] <Lerneaen_Hydra> now they're in web.student.chalmers.se/~lock/<filename>
[14:32:37] <SWPadnos> it seems to stop at nearly the same location every time - not exactly the same, but within a couple k (376k)
[14:33:26] <Lerneaen_Hydra> test the http adress
[14:34:18] <Lerneaen_Hydra> from what I saw you just downloaded the whole file
[14:34:20] <Lerneaen_Hydra> from ftp
[14:34:29] <Lerneaen_Hydra> oh
[14:34:29] <SWPadnos> only I didn't get the whole file.
[14:34:33] <Lerneaen_Hydra> no you didn't
[14:34:35] <Lerneaen_Hydra> you resumed
[14:34:47] <Lerneaen_Hydra> http://web.student.chalmers.se/~lock/EMC2_lathe.cgd
[14:34:49] <SWPadnos> yep - wget lost the connection, and tried to auto-resume
[14:34:52] <Lerneaen_Hydra> http://web.student.chalmers.se/~lock/EMC2_mill.cgd
[14:35:12] <SWPadnos> just for the heck of it, I'll try from here
[14:39:19] <SWPadnos> interesting. using filezilla on my local machine, the transfers hung until I used wget on the http addresses, at which time both transfers worked fine
[14:39:36] <SWPadnos> both the wger from dreamhost and the ftp on my windows machine
[14:39:53] <Lerneaen_Hydra> sounds like an issue at your end then
[14:40:25] <SWPadnos> except that the DH transfer has nothing to do with my end
[14:40:33] <Lerneaen_Hydra> oh, right
[14:40:34] <SWPadnos> other than the terminal being shown here
[14:40:35] <Lerneaen_Hydra> funny
[14:50:56] <Lerneaen_Hydra> SWPadnos: are the files in the wiki?
[14:54:18] <SWPadnos> they're in the upload directory, yes
[14:55:38] <Rugludallur> hmmm was wait-for-pin removed ?
[14:56:30] <SWPadnos> I think it's only in HEAD, not a released version
[14:57:55] <Rugludallur> just updated from cvs and my config broke :P
[14:58:41] <SWPadnos> bummer - I dunno what happened then
[14:58:59] <Rugludallur> ok, im checking :)
[14:59:17] <Rugludallur> been a while since I got latest so ..
[14:59:21] <SWPadnos> heh
[14:59:43] <SWPadnos> do you have a kinematics module specified in your ini? (or hal files)
[15:00:24] <Lerneaen_Hydra> SWPadnos: what are the files called?
[15:00:28] <Lerneaen_Hydra> unchanged name?
[15:00:42] <SWPadnos> same as you had them: EMC2_lathe.cgd and EMC2_mill.cgd
[15:00:55] <Rugludallur> no kine here :(
[15:01:10] <SWPadnos> that's probably it then
[15:01:11] <Rugludallur> unless loadrt trivkins is
[15:01:20] <Rugludallur> should be there probably
[15:01:32] <SWPadnos> it shuold be something like that
[15:01:57] <SWPadnos> was the error something like unknown symbol "kinematcs_forward" ... ?
[15:03:30] <Rugludallur> wierd, just put the wait-for-pin line in again
[15:03:34] <Rugludallur> and now there is no error given
[15:03:56] <Rugludallur> going to check for some sort of cache or odd things
[15:03:58] <SWPadnos> problem solved! (sort of :) )
[15:04:14] <Lerneaen_Hydra> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cam_Post#EdgeCAM
[15:04:20] <Lerneaen_Hydra> which pages should link to that?
[15:04:34] <Rugludallur> straight to HAL:4: ERROR: Can't find program 'wait-for-pin'
[15:04:49] <Rugludallur> loadusr -w wait-for-pin led.senseZUP
[15:05:55] <SWPadnos> LH: maybe this one: http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?Cam
[15:06:54] <jepler> Rugludallur: wait-for-pin was removed. instead, components should call hal_ready(), and 'loadusr -W' be used instead.
[15:07:48] <Rugludallur> ok, in my case it was because of vcp
[15:07:55] <Rugludallur> I will try it :D
[15:07:58] <Rugludallur> thanks
[15:08:31] <jepler> +loadusr -W halvcp halui.vcp
[15:08:39] <jepler> maybe you just want to use this command ^^^
[15:10:21] <Lerneaen_Hydra> there, post processors are up on the wiki and linked and all :) woo
[15:10:35] <jepler> with the name of your own .vcp of course
[15:15:36] <Rugludallur> jepler: thanks much cleaner way that wait-for-pin :D
[15:17:31] <CIA-8> 03jepler 07HEAD * 10emc2/configs/demo_mazak/tester.hal: use loadusr -W instead of sleep
[15:18:28] <Lerneaen_Hydra> jepler: done anything with image-to-gcode recently?
[15:20:45] <jepler> Lerneaen_Hydra: nope
[15:20:49] <CIA-8> 03jepler 07HEAD * 10emc2/scripts/emc.in: assigning PATH earlier prevents an error that 'axis-remote' is not on the PATH when cleaning up an old running emc
[15:21:09] <Lerneaen_Hydra> ok. do you plan to add anything else to it?
[15:21:23] <Lerneaen_Hydra> like opening 3d models natively or something
[15:22:52] <jepler> oh some day I'll get back to it
[15:23:29] <Lerneaen_Hydra> haha
[15:23:35] <jepler> hi acemi
[15:23:41] <acemi> hi
[15:43:33] <CIA-8> 03jepler 07HEAD * 10emc2/scripts/emc.in: (log message trimmed)
[15:43:33] <CIA-8> simplify cleanup:
[15:43:33] <CIA-8> * because halshow is run with emcsh it is killed without an extra block of commands
[15:43:33] <CIA-8> * assume rtapi and hal are in use
[15:43:33] <CIA-8> * 'halcmd unload all' will unload all components (including halcmd and halmeter)
[15:43:34] <CIA-8> by process ID, no need to parse the output of 'halcmd show'
[15:43:36] <CIA-8> bugfix:
[15:52:08] <CIA-8> 03jepler 07HEAD * 10emc2/scripts/emc.in:
[15:52:08] <CIA-8> when killing another session, first kill just the UI programs and see if the
[15:52:08] <CIA-8> lock file goes away. This improves a race condition where the new session
[15:52:08] <CIA-8> would perform cleanup (including removing realtime modules), then run
[15:52:08] <CIA-8> 'realtime start', all before the old session reached its 'unload all' and
[15:52:09] <CIA-8> 'realtime stop' commands. When this happened, the new session would fail to
[15:52:13] <CIA-8> start.
[16:02:22] <jepler> hi Jymmm
[16:02:33] <Jymmm> Mornin jepler
[16:03:08] <jepler> http://www.myconfinedspace.com/wp-content/uploads/2006/09/airforce-id-chart.jpg
[16:03:56] <Jymmm> heh
[16:04:20] <SWPadnos> I like the "Swamp gas" one :)
[16:05:33] <Jymmm> You dont remember those storys from way back in the news?
[16:06:06] <SWPadnos> I was thinking of "Men In Black"
[16:06:31] <Jymmm> "35 people reporting siting a ufo... AF confirmed it was swamp gas"
[16:07:09] <SWPadnos> it's a good thing the AF is expert on swamp gas, isn't it
[16:07:30] <Jymmm> Too bad there are no swamps within 200 miles of Area 51
[16:07:43] <Jymmm> Thus the "weather ballons"
[16:08:25] <SWPadnos> I could tell you what's happening, but then I'd have to kill you
[16:09:20] <Jymmm> Not at all... what they do now is if you get too close they send in the FBI to raid/search your home now.
[16:09:32] <bill20r3> that's a chance I'm willing to take.
[16:10:06] <Jymmm> No joke, there was a show on Discovery Channel about Area 51
[16:10:52] <SWPadnos> unfortunately, Discovery channel != the pinnacle of scientific discourse
[16:10:56] <Jymmm> these ppl where "stalking" area 51 far too much for their comfort, so they tracked their asses down and sent in the FBI to search/raid each of th epersons home.
[16:11:57] <jepler> As anyone knows, there are no aliens at area 51. the "greys" are just a result of genetic manipulation experiments, but the popular belief that they are aliens was allowed to continue by the Illuminati and Majestic 12 becaue of it is value in controlling the fringe elements
[16:12:24] <jepler> (everything I know about conspiracy theories I learned from x-files and the game "deus ex")
[16:12:25] <SWPadnos> greys are wusses, depending on your TL
[16:13:10] <Jymmm> Guess I never watched enough x-files, or just dont recall "greys"
[16:13:21] <SWPadnos> I only know them from GURPS
[16:13:42] <jepler> in a few episodes they showed grey-like aliens
[16:13:51] <jepler> but mostly I'm telling you the story from "deus ex"
[16:14:01] <Jymmm> ah
[16:14:15] <jepler> good game if you like 1999-level polygon graphics
[16:14:30] <Jymmm> I'm not much of a gamer
[16:16:28] <SWPadnos> GURPS is a non-computer game (if you can believe that)
[16:16:48] <SWPadnos> Generic Universal Role-Playing System (similar to D&D, but much more generic)
[16:18:57] <Rugludallur> If someone with some programming experience could give me a quick pointer it would be much appreciated
[16:19:07] <SWPadnos> a->b
[16:19:17] <bill20r3> \$quick
[16:19:34] <Rugludallur> I'm trying to update the VCP spin_button so that it sets the hal pin to the default value upon loading the widget
[16:20:17] <Rugludallur> I would think spin_button_update is for this but it does not seem to work for me :P
[16:20:51] <SWPadnos> are you trying to set the text box value, or the HAL pin (or both)?
[16:21:17] <Rugludallur> the spin button value is set upon init of the widget without any problems but the hal pin is still at 0
[16:21:50] <SWPadnos> and it isn't updated "automatically" at the first update cycle?
[16:22:57] <Rugludallur> untill the value in the spin button is chagned
[16:22:58] <Rugludallur> err changed, so what I need to do is do an initial "forced updated" to set the hal pin to the spin_button value
[16:22:58] <Rugludallur> I would think with the signal handler in place the correct way to do this would be to initiate an update from the spin_button widget ?
[16:22:59] <jepler> Rugludallur: I think there's a way to "emit" the "changed" signal on the spin button
[16:22:59] <jepler> let me see if I can find it in the gtk documentation
[16:23:14] <Rugludallur> SWPadnos: nope
[16:25:48] <jepler> Rugludallur: ah, I think you need to call gtk_adjustment_set_value: http://gtk.org/api/2.6/gtk/GtkAdjustment.html#gtk-adjustment-set-value
[16:25:55] <jepler> probably *after* connecting the "changed" signal
[16:26:17] <Rugludallur> jepler: nope, that works fine
[16:26:56] <jepler> hm, ok
[16:26:56] <Rugludallur> jepler: Everything works, the only problem is that the initial value assigned to the spin button does not get assigned to the hal pin, only after updating the value does the hal pin get set
[16:27:36] <jepler> you put your call to gtk_adjustment_set_value *after* the call to gtk_signal_connect(..., "changed", ...)
[16:27:39] <jepler> ?
[16:27:39] <Rugludallur> jepler: My C skills are very limited so this might be very simple :P
[16:28:44] <Rugludallur> jepler: you mean to override the initial value of the spin button with the default value again to trigger an update ?
[16:28:45] <jepler> or maybe you want to make these two lines at the end of init_spin_button be unconditional:
[16:28:49] <jepler> pd->pin_state = gtk_spin_button_get_value_as_float (GTK_SPIN_BUTTON(gwp)
[16:28:51] <jepler> );
[16:28:54] <jepler> *(hd->pin) = pd->pin_state;
[16:30:51] <Rugludallur> jepler: hmm that makes sense :D
[16:34:38] <Jymmm> jepler have you ever dev python on win?
[16:35:10] <Rugludallur> jepler: probably best to only set the hal pin to the value of the spin button if the hal pin is 0, otherwise It might override a value it should not ,, right ?
[16:35:14] <cradek> what's win?
[16:35:21] <Jymmm> cradek M$
[16:36:43] <Jymmm> aka microsoft
[16:37:12] <cradek> oh that substandard OS that costs money and has a restrictive license agreement?
[16:37:49] <cradek> (just a warning, jepler will probably troll even worse than me)
[16:40:07] <Jymmm> I have an application in mind, but I've never done any GUI, so I was looking for something that was cross platform
[16:40:31] <SWPadnos> Java: "Write Once, Run Anywhere" It's been proved!
[16:40:44] <cradek> ah, let me know what you find (seriously)
[16:40:55] <cradek> that's a hard problem I think
[16:41:09] <Jymmm> SWPadnos: I'm old skool... I like a songle EXE file for the application without any need to install anything else.
[16:41:34] <cradek> that's incompatible with cross-platform
[16:41:43] <SWPadnos> python is similar to java in that respect - it's normally interpreted, but you can make executables with the right tools
[16:42:34] <Jymmm> SWPadnos I just dont know what any of those are =)
[16:42:58] <SWPadnos> me either, especially on Windows
[16:43:11] <SWPadnos> I have python for windows, but have done no development with it
[16:43:22] <Jymmm> SWPadnos: GUI ?
[16:44:15] <SWPadnos> it's a windowed CLI, IIRC
[16:44:28] <SWPadnos> but that's not the software you would write, it's just the "console"
[16:45:35] <Jymmm> ah, I'm looking for some lang for GUI dev. there is php-gtk, but the $400 compiler only works with php-gtk 1, not 2
[16:45:53] <SWPadnos> and gtk for windows is relatively pathetic, I think
[16:46:03] <Jymmm> it what sense?
[16:46:23] <Jymmm> cludgy? buggy? limited?
[16:46:28] <SWPadnos> the language won't be the main problem. it'll be the cross-platform GUI widget set that's critical
[16:46:34] <Jymmm> ah
[16:46:55] <Jymmm> Well, it doesn't HAVE to be cross-platform. Would just prefer it to be.
[16:47:04] <SWPadnos> I don't think it uses native widgets, so it won't behave just like other WinSoftware
[16:47:53] <Jymmm> cradek will do.
[16:47:56] <SWPadnos> I'm not positive though, I use BC++ Builder on Windows
[16:48:17] <Jymmm> wow, haven't heard borland come up in a while.
[16:48:28] <Jymmm> the name I mean.
[16:48:31] <SWPadnos> I'm old skool ;)
[16:49:10] <Jymmm> I have the full box of v3 (maybe 4) I think. Got it before I left Borland.
[16:49:45] <SWPadnos> I like the DOS version 3.1 the best - I was very used to the editor, and I like the extensible tool filter system
[16:50:06] <Jymmm> I have Turbo C++ aroudn here somehwere
[16:50:09] <SWPadnos> I have BC++ Builder 5 for 32-bit Windows stuff
[16:53:50] <Jymmm> The only lang I really know is PHP, but I dont' want to get 20% into a project then go WTF?! and have to start over somewhere else. I know python is pretty good, just never really touched it.
[16:54:51] <SWPadnos> fora GUI, it's the widgets that matter. you could probably write the "backend" code in C, which is very cross-platform
[16:55:22] <SWPadnos> the front end has to deal with different named widgets, and even wholly different methodologies. therein lies the nightmare
[16:56:36] <Jymmm> ah. Well I've never done any GUI so, I'm clueless in that respect. I was even thinking of writing in flash for the gui aspects and shits and giggles factor =)
[17:00:48] <SWPadnos> you make me sick
[17:00:52] <SWPadnos> err - you're making me sick ;)
[17:02:28] <bill20r3> * bill20r3 vomits a little in his mouth.
[17:03:16] <CIA-8> 03jarl.stefansson 07HEAD * 10emc2/src/hal/user_comps/vcp/vcp_widgets.c:
[17:03:16] <CIA-8> When initializing spin buttons the associated hal pin is checked for a zero value. If the hal pin value is zero it is changed to the default value of
[17:03:16] <CIA-8> the spin button.
[17:07:33] <Jymmm> SWPadnos: You're seeing Flash as all those fscking banner ads. But you can take a flash applet (SWF) and convert to a self contained EXE file. You can embed images, sounds, video, pdf files, and a lot of other content as well. Uses ActionScript (JS) and can be crested using python or even c from the command line.
[17:07:50] <SWPadnos> I still think flash sucks ;)
[17:08:11] <Jymmm> SWPadnos Only becasue flash on nix sucks.
[17:08:20] <SWPadnos> no, because I think flash sucks
[17:08:26] <Jymmm> SWPadnos why?
[17:08:27] <SWPadnos> I use Windows most of the time
[17:09:08] <SWPadnos> I think it has the problem of having been designed for one thing, then morphing into something else, with little or no design thought given to the transformation
[17:09:51] <SWPadnos> also, there's exactly one company that controls the standard (and the program creation tools, AFAIK)
[17:10:17] <Jymmm> They released the specifications a few years ago
[17:10:41] <Jymmm> There are a LOT of tools out there (non Adobe/Macromedia) now
[17:10:52] <Jymmm> many open source too
[17:11:14] <Jymmm> One example... http://www.swftools.org/
[17:11:16] <SWPadnos> well, perhaps I'm just biased then. but since there's no real need for me to use it, I stay that way ;)
[17:11:50] <Jymmm> Well, if your only real exposure to it is banner ads, I can see the reservations.
[17:12:23] <SWPadnos> no, I've seen some apps in Flash. they aren't all bad
[17:12:32] <SWPadnos> mostly web-delivered apps though
[17:13:06] <SWPadnos> though I don't think I've ever seen a flash app that existed without external support (like a browser / flash player)
[17:13:09] <Jymmm> The one drawback to flash is it's security sandbox. But I found an alternative
[17:14:04] <Jymmm> Well, if you wanted (lets say) a tutorial, you can create a swf and convert it to a stand-alone EXE
[17:14:47] <SWPadnos> and a standalone ELF executable for Linux, or a (whatever they are) for Mac?
[17:15:14] <Jymmm> SWPadnos: *I* don't know, could be I guess.
[17:15:47] <SWPadnos> well that would be a critical thing for you, sinec cross-platform is what you're after ;)
[17:15:49] <Jymmm> SWPadnos: But if you want to see some CLI flash dev.... http://www.swftools.org/swfc/swfc.html
[17:16:09] <Jymmm> SWPadnos: I'd prefer it, but not manditory.
[17:18:04] <Jymmm> SWPadnos: Here's a nice lil example using trig ... http://www.swftools.org/swfc/swfc.4.html
[17:18:51] <SWPadnos> I guess one of the things I don't like is that it's intended for presentations, not programs
[17:19:43] <SWPadnos> I'm not sure how much support there is for a program that just does stuff and shows you what's happening, vs a program where you choose things to see (from a list of slides, for instance)
[17:19:53] <Jymmm> I don't know... I'm not sure if anyone has ever gone that far with it. I *think* it might have todo with the sandbox.
[17:20:40] <Jymmm> As most applications are designed to store/manipulate data. and that being the limitation with the sandbox.
[17:20:48] <SWPadnos> right
[17:21:03] <Jymmm> But if the data was on a server, that's another story.
[17:21:15] <Jymmm> not a good alternative, but...
[17:21:24] <SWPadnos> shouldn't be, unless it's on the same server the program was launched from
[17:21:30] <SWPadnos> ie, a web app
[17:22:04] <Jymmm> Right, so the limits it from running "offline"
[17:23:11] <bill20r3> once I did some flash using php-ming, I dont reccomend it.
[17:24:27] <Jymmm> SWPadnos: Now, I came across this http://www.northcode.com/ which gets around the security sandbox and has the ability to create a flash based EXE and have SQLite embedded as well.
[17:26:07] <Jymmm> SWPadnos Basically it's a beefed up (self contained with enhancments) player that creates a EXE
[17:26:26] <SWPadnos> that looks to be Windows only
[17:26:36] <Jymmm> SWPadnos: It is.
[17:26:37] <SWPadnos> if that's OK, then I'd just use BC++ ;)
[17:27:20] <SWPadnos> or find one of the cross-platform GUI kits, like Zinc (or possibly a more recent version of Qt or Gtk)
[17:27:24] <Jymmm> Well, I guess nobody else came up with it for nix yet.
[17:27:45] <SWPadnos> from their siite: " SWF Studio V3 incorporates the MacromediaTM Flash 8 player technology (licensed from Macromedia) and is fully compatible with Flash MX, Flash MX 2004 and Flash 8."
[17:28:19] <Jymmm> SWPadnos: Right, some ppl were/are concerned about compatability
[17:28:44] <SWPadnos> Macromedia isn't so Linux-friendly - they don't even have a 64-bit compile of the browser plug-in (maybe not for XP64 either)
[17:29:02] <Jymmm> and MM license restricts the player distribution, not the applet creation.
[17:29:27] <Jymmm> Well, MM is now Adobe, so who know.
[18:05:19] <Rugludallur> hmm
[18:05:43] <Rugludallur> I have just reproduced an old bug that was haunting me a while back but was supposed to be fixed
[18:05:58] <Rugludallur> Can anyone else reproduce issue 1509665 in HEAD ?
[18:05:59] <SWPadnos> you're not supposed to do that!
[18:06:04] <Rugludallur> :D
[18:08:41] <SWPadnos> too bad the tracker doesn't show the commit that "fixes" the problem
[18:08:57] <SWPadnos> of course, that is harder when CVS is no longer at SF
[18:09:05] <Rugludallur> yup
[18:09:18] <Rugludallur> let me check my irc logs for comments
[18:10:15] <SWPadnos> I found the commit, but I can't really look at the code at the moment
[18:10:25] <Rugludallur> :)
[18:10:30] <SWPadnos> it's the only commit from June 20
[19:28:18] <alex_joni> hi all
[19:28:22] <alex_joni> Rugludallur: wb
[19:32:57] <alex_joni> jepler: got a minute?
[19:35:27] <Rugludallur> alex_joni: thx, it's good to be back :D
[19:35:38] <alex_joni> Rugludallur: it seems you're right
[19:35:45] <alex_joni> spindle control is now in motion
[19:36:01] <alex_joni> not in iocontrol, so I\m busy checking where I went wrong with it
[19:36:21] <Rugludallur> alex_joni: it's worse now than before, now there just does not seem to be a way to have estop kill it :P
[19:36:31] <alex_joni> indeed
[19:36:39] <alex_joni> give me half an hour :)
[19:38:59] <alex_joni> * alex_joni is officially intrigued
[19:43:06] <davidf> hi
[19:45:06] <davidf> Couple things:
[19:46:08] <davidf> Lerneaen_Hydra, I found a question you asked on 9-10 while looking for something I'd asked about the printer port.
[19:47:41] <davidf> You asked re the repeatability of my optical switches. It's less than 0.001, but I havent tested extensively. My mill uses triangular threads so it may be actually better. My tests were usually in the .0002 to .0005 range though.
[19:49:39] <alex_joni> hi davidf
[19:49:51] <davidf> Hi alex_joni
[19:50:49] <davidf> I was looking for cradek 's answer to what file to move to cause emc to not grab the parport on bootup, but I cant seem to find the conversation in the logs.
[19:51:04] <alex_joni> davidf: it's inthe wiki too
[19:51:10] <davidf> alex_joni, can you tell me the name/location of that file?
[19:51:10] <alex_joni> under the troubleshooting page
[19:51:14] <alex_joni> hang on
[19:51:32] <davidf> alex_joni, I looked in the wiki at http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?TroubleShooting
[19:51:33] <cradek> /etc/modprobe.d/emc2
[19:52:00] <alex_joni> davidf: right #2 has what you want
[19:52:05] <davidf> thanks cradek . Wonder where the log of that went? Maybe it was earlier date than I thought.
[19:52:36] <davidf> alex_joni, hmm... then I'm clueless. Let me read that again.
[19:53:06] <CIA-8> 03alex_joni 07HEAD * 10emc2/src/emc/task/taskintf.cc: fix bug discovered by Rugludallur: spindle not turning off on ESTOP
[19:53:12] <alex_joni> davidf: quoting: "On Ubuntu systems, you can do this by creating a file in /etc/modprobe.d/ with the one line
[19:53:16] <alex_joni> install parport_pc /bin/true
[19:53:18] <alex_joni> The official packages put this line in the file /etc/modprobe.d/emc2."
[19:53:48] <alex_joni> Rugludallur: can you test that?
[19:56:20] <davidf> alex_joni, that file exists and contains that line. Guess I want 'install parport_pc /bin/false' ?
[19:56:38] <davidf> (I want to enable printing, and disable emc.)
[19:57:02] <davidf> Or just move the file as cradek suggested?
[19:57:05] <jepler> davidf: you restore the functionality by removing or commenting out that line
[19:57:07] <alex_joni> if the line is there simply add a '#' in fron of it
[19:57:07] <Lerneaen_Hydra> davidf: hi
[19:57:18] <jepler> or removing the whole file
[19:57:18] <davidf> Hey Lerneaen_Hydra
[19:57:31] <davidf> ok thanks alex_joni / jepler
[19:57:32] <Lerneaen_Hydra> so not worse than 0.005?
[19:57:33] <alex_joni> jepler: might be problematic when upgrading the package?
[19:57:51] <jepler> alex_joni: yes, I'd recommend removing the line from the file
[19:58:00] <jepler> I know in that case an upgrade won't replace it with the original version of th efile
[19:58:03] <jepler> or else it'll ask
[19:58:05] <alex_joni> I would just comment it out
[19:58:13] <davidf> Lerneaen_Hydra, oh thats a football field. :)
[19:58:22] <alex_joni> that way you can always easily uncomment it back
[19:58:27] <alex_joni> when you need it again
[19:58:36] <davidf> alex_joni, I like that sug.
[19:58:40] <Lerneaen_Hydra> davidf: where'd you get those switches?
[20:00:16] <davidf> See, this computer is the one I originally used for emc that doesnt work due to motherboard incompatibility. But I want it for general use, and want to keep my original emc install configs, etc intact for reference. I have a dev emc install on it as well.
[20:00:42] <davidf> Lerneaen_Hydra, They are junk parts from old deskjet printers.
[20:00:56] <davidf> nice price.
[20:01:04] <Lerneaen_Hydra> davidf: oh, so nothing special with them?
[20:01:28] <Lerneaen_Hydra> were they made to be optical switches, or something else?
[20:02:21] <davidf> Lerneaen_Hydra, not with the switches themselves. But I use them with an op-amp comparator circuit so they trip at very consistent and well defined levels
[20:02:59] <davidf> They are made to be optical switches, with a tab interrupting the beam.
[20:02:59] <Lerneaen_Hydra> oh, so they have a linear response over a large length, and you use the op-amp as a threshold
[20:03:20] <davidf> They contain an IR LED and a phototransistor.
[20:03:23] <Lerneaen_Hydra> any idea of the size of the doide?
[20:04:09] <davidf> You need to interface them to an op amp comparator that goes high or low to TTL logic level when interrupted.
[20:04:14] <Lerneaen_Hydra> yeah
[20:04:19] <Lerneaen_Hydra> that's trivial though
[20:04:21] <davidf> current wise?
[20:04:41] <davidf> 5 ma is plenty.
[20:04:42] <Lerneaen_Hydra> I was thinking size more like diamter
[20:04:52] <Lerneaen_Hydra> of the diode
[20:05:58] <davidf> oh. small. the whole unit is like 1/4 inch wide, plasic with the led inside. The slit is really tiny where the beam comes out like .005 to .01 inch.
[20:06:07] <Lerneaen_Hydra> oh, ok
[20:06:14] <Lerneaen_Hydra> so not standard forks
[20:06:31] <Lerneaen_Hydra> I should be able to hack something like that together though
[20:06:35] <davidf> So you see that with a .005 in slirt, that's certainly your max error.
[20:06:40] <Lerneaen_Hydra> yeah
[20:06:54] <davidf> They look pretty standard to me...
[20:07:16] <davidf> by 1/4 inch I meant front to back.
[20:07:29] <Lerneaen_Hydra> the ones i've seen don't have slits in them
[20:08:10] <davidf> The gap between diode and transistor is about 1/4 inch apart, the overall height is maybe 1/2 inch, oAL length about 3/4 inch.
[20:08:59] <davidf> Look in the gap at the faces of the diode case & trans case. No slit?
[20:09:59] <davidf> Lerneaen_Hydra, I think I got these out of some old HP deskjet 500 C's.
[20:10:25] <davidf> I also took apart some lexmark and others, so you might try those as well.
[20:10:27] <Lerneaen_Hydra> I'll probably just take a sheet of metal and try to make a small slit
[20:10:35] <Lerneaen_Hydra> I've got plenty of big forks
[20:10:38] <Lerneaen_Hydra> without slits
[20:11:21] <davidf> I wrote down the part numbers somewhere, I'll look around & you can probably find them from Digi-Key.
[20:11:30] <Lerneaen_Hydra> ooh, that would be really nice
[20:12:19] <davidf> Oh, ok. Well, the comparator should help, but ambient light will probably mess you up a lot due to movement of the table, etc.
[20:12:35] <Lerneaen_Hydra> yeah
[20:13:03] <davidf> I'd try masking them with something with a slit and then amplification if need be.
[20:13:11] <Lerneaen_Hydra> yeah
[20:13:38] <davidf> What country are U in?
[20:13:44] <Lerneaen_Hydra> sweden
[20:13:49] <davidf> ah.
[20:14:00] <Lerneaen_Hydra> digikey is expensive with shipping
[20:14:06] <Lerneaen_Hydra> farnel are much cheaper
[20:15:01] <davidf> I dont know about there, but here, a short drive on the day before trash pickup can net several old printers, or after garage sales they usually end up on the curb. =)
[20:15:32] <davidf> Lots of fun parts, steppers, gears, etc.
[20:15:42] <pier> encoders....
[20:16:03] <davidf> Right. Cant forget those.
[20:16:06] <pier> do you think they might be of any use?
[20:16:21] <pier> got a lot of them
[20:17:15] <pier> I mean... can Emc handle stepper lost steps by means of those?
[20:17:33] <pier> those printer encoders
[20:18:00] <alex_joni> pier: it can see them
[20:18:06] <alex_joni> and probably halt if needed
[20:18:10] <alex_joni> if the error is too big
[20:18:13] <davidf> My first cnc retro experiment was an _antique_ mini lathe, with nothing but junk printer parts and like $8.00 for bipolar Transistors for the drive. I used the gears from the printers to gear the 7.5 degree steppers down 20:1 for better resolution. Very slow but fairly usable.
[20:18:25] <alex_joni> pier: but using a PID on steppers won't really work
[20:18:28] <pier> and catch up for lost steps?
[20:18:32] <alex_joni> pier: no
[20:18:38] <pier> ok
[20:18:42] <alex_joni> once you lost steps it will try to go any faster
[20:18:55] <alex_joni> which will cause a stall, and ultimately serve no purpose :)
[20:19:33] <pier> ok perhaps it was just a bad idea
[20:19:51] <alex_joni> pier: it's a good idea
[20:20:00] <alex_joni> but needs a different method than on dc motors
[20:20:06] <Rugludallur> alex_joni: spindle turns off now :D
[20:20:12] <alex_joni> Rugludallur: nice :)
[20:20:16] <pier> but as I was having a lost of step in my cnc
[20:20:32] <davidf> Lerneaen_Hydra, Im looking for the part numbers on those switches.
[20:20:35] <Rugludallur> alex_joni: I might just have discovered another bug though, not sure if it is critical or just ui
[20:20:35] <alex_joni> Rugludallur: one line change as always :)
[20:20:42] <alex_joni> Rugludallur: do tell
[20:21:04] <Rugludallur> alex_joni: when I get following errors the spindle is not turned off untill I have closed the error window
[20:21:13] <alex_joni> hehe
[20:21:29] <alex_joni> Rugludallur: that's an interesting one :)
[20:21:46] <Rugludallur> alex_joni: looking at it the estop seems to be triggered but the spindle is still on
[20:21:58] <alex_joni> Rugludallur: let me look
[20:22:53] <pier> I tought (when I used to use my home made sw) to write a routine capable of handling steps read by an encoder and catch up for the lost ones
[20:23:12] <pier> the I switche to Emc
[20:23:18] <pier> switched
[20:24:22] <alex_joni> Rugludallur: spindle stays on here
[20:24:28] <alex_joni> but the machine doesn't go to estop
[20:24:33] <alex_joni> it only goes to estop reset
[20:24:50] <alex_joni> e.g. machine off, which doesn't assume spindle off
[20:28:27] <Rugludallur> alex_joni: same
[20:28:41] <alex_joni> Rugludallur: then I guess it's no real error
[20:28:52] <Rugludallur> alex_joni: although I question the logic of leaving the spindle on if there is a following error :D
[20:29:17] <alex_joni> Rugludallur: sometimes it will move slightly when you turn it back on
[20:29:36] <alex_joni> if you have a stopped spindle, and are near a workpiece.. it'll probably break the tool
[20:29:42] <alex_joni> but that's just a lame excuse
[20:30:10] <pier> alex_joni: I didn't mean to use a PID and DC motors... just a trick to get rid of steps loss on cheap CNC like mine or at least keep the number of them under control before halting or signalling anything to the operator
[20:30:33] <pier> perhaps it isn't worth it
[20:30:41] <alex_joni> pier: keeping track is pretty simple
[20:30:48] <pier> indeed
[20:31:14] <pier> alex_joni: and adding steps? in case of loss?
[20:31:20] <alex_joni> that's hard
[20:31:28] <Rugludallur> alex_joni: with the plasma table it will leave the torch on until the arc has made a mess big enough so that it extinguishes eeuwww
[20:31:38] <alex_joni> because you sometimes lose steps in stalling
[20:31:44] <Rugludallur> alex_joni: but no prob, I can probably find a way to get around it :D
[20:31:48] <alex_joni> Rugludallur: I can understand that :)
[20:32:07] <alex_joni> Rugludallur: the proper fix is probably miles away though
[20:32:15] <alex_joni> but there's a motion enabled
[20:32:27] <alex_joni> maybe and that with spindle on in HAL and you're done
[20:32:52] <Rugludallur> alex_joni: I can also just use halui, that probably has a machine-on bit which I can wire to the torch
[20:33:13] <alex_joni> right
[20:33:27] <alex_joni> Rugludallur: the proper way of solving this was discussed briefly a while ago
[20:33:32] <alex_joni> let me see if I can find it :)
[20:33:49] <Rugludallur> it's ok , I have to leave now for a couple of hours anyway
[20:33:56] <alex_joni> http://wiki.linuxcnc.org/cgi-bin/emcinfo.pl?RunLevels
[20:34:01] <Rugludallur> thx
[20:34:12] <Rugludallur> tomorrow I expect to have a video to share :D
[20:35:09] <alex_joni> great
[20:35:30] <pier> does Emc send a signal of given period (and frequence) for a certain time or does it count the actual steps sent to the clock pin?
[20:35:30] <Rugludallur> Rugludallur is now known as Rugludallur-Away
[20:35:36] <alex_joni> sometimes I get scared.. I see that page hasn't been edited since october last year, yet I remember most of the talks with jmk :)
[20:35:53] <alex_joni> pier: both :)
[20:35:53] <Lerneaen_Hydra> spooky
[20:36:13] <alex_joni> Lerneaen_Hydra: of course the important stuff I forget right away
[20:36:25] <alex_joni> like where the hell my house keys are :D
[20:38:25] <Lerneaen_Hydra> yeah :D
[20:44:09] <Lerneaen_Hydra> 'night
[20:44:18] <alex_joni> night
[20:45:07] <pier> night all
[20:58:22] <alex_joni> good night all
[20:59:53] <jepler> bye alex
[21:09:31] <CIA-8> 03baslaarhoven 07HEAD * 10emc2/src/emc/motion/ (control.c mot_priv.h motion.c motion.h): Implemented S-curve (smooth) backlash compensation.
[21:09:49] <jepler> sjml: congratulations on your first check-in
[21:10:54] <sjml> tx, now lets see what i've broken ;-)
[21:12:37] <sjml> night all
[21:39:03] <davidf> Hi,
[21:39:46] <davidf> Darn I cant sudo edit /etc/modprob.d/emc2 - owned by root, so is /etc.
[21:39:55] <davidf> Help please anyone?
[21:42:58] <bill20r3> uh, if it wasn't owned by root you'd have a problem.
[21:43:07] <bill20r3> use su.
[21:44:29] <davidf> thanks.
[21:46:05] <davidf> tried this:
[21:46:08] <davidf> dave@ubuntu:~$ su edit /etc/modprob.d/emc2
[21:46:08] <davidf> Unknown id: edit
[21:46:08] <davidf> dave@ubuntu:~$ su
[21:46:08] <davidf> Password:
[21:46:08] <davidf> su: Authentication failure
[21:46:08] <davidf> Sorry.
[21:46:11] <davidf> dave@ubuntu:~$ edit /etc/modprobe.d/emc2
[21:46:13] <davidf> Warning: unknown mime-type for "/etc/modprobe.d/emc2" -- using "application/*"
[21:46:15] <davidf> Error: no write permission for file "/etc/modprobe.d/emc2"
[21:46:17] <davidf> dave@ubuntu:~$
[21:46:50] <bill20r3> that's not a root prompt.
[21:46:59] <bill20r3> try 'sudo su -'
[21:47:04] <bill20r3> then put in the root password.
[21:47:14] <davidf> guess I need help to use su.
[21:47:26] <davidf> ok.
[21:47:36] <bill20r3> Ubuntu protects you from yourself a lot..
[21:48:12] <davidf> dave@ubuntu:~$ sudo su
[21:48:12] <davidf> root@ubuntu:/home/dave#
[21:48:19] <davidf> thanks.
[21:48:28] <bill20r3> be carefull.
[21:49:00] <davidf> root@ubuntu:/home/dave# edit /etc/modprobe.d/emc2
[21:49:01] <davidf> Warning: unknown mime-type for "/etc/modprobe.d/emc2" -- using "application/*"
[21:49:01] <davidf> Error: no "edit" mailcap rules found for type "application/*"
[21:49:26] <davidf> how to set the file type?
[21:49:39] <davidf> (mime-type?)
[21:50:14] <bill20r3> that's just your editor, it's harmless.
[21:50:37] <bill20r3> it's a warning, not an error.
[21:51:05] <davidf> so how do I actually edit the file then?
[21:51:22] <bill20r3> I use vi, but I dont suggest you do.
[21:51:25] <bill20r3> pico maybe
[21:52:14] <davidf> Before with protected files, I just type edit <filename> and edited it in the terminal.
[21:52:27] <bill20r3> and that's not working this time?
[21:52:41] <davidf> cant do that with this file? it is just a text file with one line.
[21:52:49] <bill20r3> you should be able to.
[21:53:07] <davidf> no, I get that error (warning) and a prompt.
[21:53:20] <bill20r3> guh
[21:53:34] <bill20r3> maybe rename it to emc2.txt, edit it, and rename it back.
[21:53:39] <davidf> here's what happens:
[21:53:42] <davidf> dave@ubuntu:~$ sudo su
[21:53:42] <davidf> root@ubuntu:/home/dave#
[21:53:42] <davidf> root@ubuntu:/home/dave# edit /etc/modprobe.d/emc2
[21:53:42] <davidf> Warning: unknown mime-type for "/etc/modprobe.d/emc2" -- using "application/*"
[21:53:42] <davidf> Error: no "edit" mailcap rules found for type "application/*"
[21:53:43] <davidf> root@ubuntu:/home/dave#
[21:53:54] <alex_joni> davidf: gedit
[21:54:09] <davidf> oh now I remember. Thanks alex_joni
[21:54:23] <alex_joni> * alex_joni goes back to sleep
[21:54:50] <davidf> good night. Sleep tight.
[21:56:01] <alex_joni> davidf: for future reference : sudo gedit file
[21:57:19] <davidf> OK, thanks. I got it as root too. Go back to bed. =)
[21:57:38] <davidf> Thanks bill20r3
[21:58:08] <davidf> bye..
[23:21:41] <tomp> i didnt get axis src by any of these 3 procedures.. live cd install, sudo apt-get build-dep emc2-axis, cvs ... -rRelease_2_0_3 ... so configure cannot find it to build it... what to do?
[23:28:39] <jepler> tomp: for emc 2.0.3, the source of axis was separate from emc2. you can get the source by 'apt-get source emc2-axis'.
[23:29:04] <jepler> 'apt-get build-dep' is an important step, because it gets the other packages you need to actually compile emc2-axis
[23:29:37] <tomp> yes, saw that was the support files, not the src
[23:31:13] <tomp> where should i unpack it? in the run-in-place-dir?
[23:32:13] <jepler> is your goal to build your own copy of emc2.0.3 and axis with a run-in-place configuration?
[23:32:48] <tomp> yes
[23:33:14] <jepler> I'm rusty on compiling AXIS with emc 2.0.x, but I think you can put the directory anywhere but you must set the EMCROOT envronment variable when you run setup.py
[23:33:46] <tomp> ok, enuf to chew on for now, thanks jepler
[23:34:00] <jepler> np
[23:48:39] <Rugludallur-Away> Rugludallur-Away is now known as Rugludallur