14:07:02 Any thoughts on how we can get the "registered" developers more involved ? 14:09:11 How well is it known that EMC2 is now "runnable". More folks might get going, at least on I/O stuff, if they realized that a working platform exists. 14:09:24 Several are there for emc2. I believe that they are following the development by testing. 14:12:38 There is a lot to learn before a body begins to write serious code. 14:12:56 where is the emc2 download located? 14:13:47 cvs @ Sourceforge 14:13:52 checkout the emc2 module 14:17:11 was just looking at emcsegmot.c cvs 14:17:33 I don't think the minFerror bug was fixed there 14:17:53 That is quite possible... 14:17:55 making it fabs() 14:18:16 but it was fixed in emcmot.c? 14:18:53 To be quite honest, having three (or more) variations of emcmot.c makes it hard to keep them all in sync 14:19:02 for sure 14:19:26 It would help if common code was split out... 14:19:45 so that fixes in one would be seen in all the others. 14:20:15 for the time being I am just going to try to untar what I have...do the fix, and tar it back 14:21:46 always get garbage when I try to just ./compile 14:22:03 but thr BDI make thing always works 14:23:46 see 26th July 2004 mail to the emc-commit list for the fix 14:26:25 ok 14:48:05 I see your fix for following error on line 3430 in emcmot.c 14:50:03 Actually I did fabs(emcmotDebug->jointVel[axis])...you fabs() the whole expression. Either should work though. 14:50:40 I will test this on the machine and report back, 14:38:15 I'm surprised at the additions and changes to emc. Just updated my copy of the repository. 14:39:20 when was the last time you did an update ? 14:39:40 'bout a week ago. 14:40:13 emc hasn't changed... emc2 has had a few commits. 14:41:02 Oh. /emc.run/1.10/Mon Aug 9 14:43:19 2004// 14:41:21 Several others in there as well. 14:42:20 I count nine changes in emc/src/emcmot as well. 14:42:44 Unless I'm reading things wrong. 14:44:40 I see an issue with the additions of univpwm.ini and univstep.ini. They were made from version 1.1 of emc,ini and are missing all of the stuff that has been added more recently. 14:45:39 Individual axis vel and accel and step and hold times. The units stuff is all missing. 14:45:45 there were a number of changes at the end of August with Jon's ppmc driver code 14:46:24 Sept 12 I see. 14:47:00 Must not be seeing right. 14:47:45 The web cvs interface at SF shows when and what changed... 14:49:12 It looks to me like univstep.ini and univpwm.ini are nothing more than emc.ini tuned for a specific PC and machine. 14:49:55 probably... 14:50:22 Does Jon's board do setup and hold timings ? 14:50:43 Even the ; IO_BASE_ADDRESS = 0xe000 14:50:44 is pci specific. 14:51:09 Doesn't matter if it does or not. 14:51:45 Seems like this stuff needs to be kept up to date with changes in the base code rather than forking. 14:52:16 A current version of jon's stuff would use individual axis vel and accel. 14:52:44 Servo stuff doesn't use setup or hold either. 14:54:02 Just a couple of ini & run scripts along with comments about the options... 14:56:29 Just clutter and confusion. 14:56:58 Remember a few months back when generic wouldn't run the first time cause the generic.var file didn't match the new interpreter. 14:57:29 must have been while I was up a mountain... 14:58:30 I believe that I set up generic for stepper and emc for servo and got both current. 15:00:48 perhaps a post to the dev list asking if someone would like to take on the task of a cleanup ? 15:02:00 I see that Matt tried to bring lucas up to date last year. 15:02:19 There is a lot of clutter in there. 15:03:53 Much of the run script could be pruned out 15:04:08 I suspect that several of the *.nml files will not run on a current system. 15:05:19 Two samples of *.nml would suffice - one for local, and another for remote GUI's 15:05:56 likewise, single samples of tool tables & *.var 15:07:22 lucas.var is identical to generic.var 15:07:54 But it we took it out, lucas.ini would fault. 15:08:17 would imagine all the *.vars are the same... 15:09:31 minitetra has the offsets for the cable hex. 15:10:18 offsets are in real world coordinates rather that strut lengths ? 15:11:57 I remember something about that. We set joint homes at 20 something then offset for a world zero. 15:14:12 So how would we preserve individual or specific machine stuff without cluttering the emc directory? 15:15:16 Add comments around machine specific variations 15:16:25 And put them in a separate directory? 15:17:46 I can see that even remove_realtime_base is getting dated. 15:19:44 doesn't include adeos or rtai_hal in the module list 15:19:27 I can see why there is a desire to abandon emc in favor of emc2. 15:20:22 I noticed that a while back on Live and Knoppix both. 15:22:22 the run scripts in emc2 are a little cleaner 15:24:55 Inflamitory statement follows -- I just hope that we are not using the same if/then/else/endif mindset in the design of 2 15:25:19 for the run script ? 15:26:07 For handling multiple os versions. 15:26:24 much of that has been removed. 15:26:56 however, the remains are still pretty much the same. 15:27:25 That needs to be looked at and overhauled at some point. 15:29:18 Back to var files - We can add comments to one and they get ignored at run time. 15:30:01 We don't use any tickle stuff in 2 at the moment but when we do, we will need to add back in the cleanup stuff. 15:30:24 emc2 uses tkemc 15:30:40 Any comments will kill the run in which that var is called and remove them during exit. 15:30:58 That happened with the new interpreter. 15:33:40 I see also in 2 a hard coded tkemc as the graphical interface. 15:33:53 This is a problem left over from emc as well. 15:33:55 Added some random comments to generic.var and it is running fine 15:34:38 See if they persist. 15:34:53 No they don't 15:35:50 but is that really a problem ? 15:36:23 Only if a run version gets returned to a repository. 15:37:05 but then that would also be a problem with ini files 15:37:13 Back in the old days, I'd add variables to the var with comments about what they did. 15:37:39 As soon as the new interp came along all of that went away. 15:38:06 No you can do anything you want with ini file comments. 15:38:22 OK, so we need to look at the var_dump routines to see what changed. 15:39:14 Or just ignore it. Would be nice to be able to add machine specific variables in there. 15:39:55 And be able to comment about what they are. 15:43:18 Nice if we could do something like 1000 2.1235 (x tool change position) 15:43:33 Or however they used to comment. 15:43:26 Pity there isn't an easy way to make some of the files read only in CVS 15:44:11 Very true but the process to limit access is awkward, isn't it. 15:44:34 on specific files, very. 15:45:21 My guess is that we need a general cleanup of the emc directory and a broadcast statement of the nature of files that should be placed in there. 15:45:44 yup 15:46:03 At the same time we should allow for machine or project specific inis and runs and such someplace else. 15:46:42 May even be possible to set a filter up so that any ini/var/run changes trigger a warning message to who ever did the commit 15:47:12 Can we do that? 15:47:36 and add in tbl and nml as well. 15:47:58 would need to look into it and try it out here.. 15:48:21 What might we name a directory below emc that would house this kind of stuff. 15:48:50 We have scripts. Could we change the nature of it a bit. 15:48:50 emc/sample_configs ? 15:49:08 Okay. I'm easy. 15:50:02 How would just emc/configs be? 15:51:07 That's what we've got in emc2 15:51:36 Paul, for your information.... as of the commits you did today, I can now get a clean compile of EMC2 again. Thanks. I am slowly moving toward using EMC2 to demo my products at shows. 15:53:33 paul_c: Gotta get with the fam. If I can get a few minutes I'll write a suggested cleanup for emc and post it to developers and see if I get someone who will help. 15:54:17 OK and I'll look into cvs foo to see what can be done to limit future changes to the files 15:54:23 in question. 15:54:50 Sounds like a bit of progress. 16:00:25 Paul, would you like for the LinuxCNC web site to start promoting EMC2 as ready for alpha testing? Is it too soon to provide binaries for download? 16:02:06 I'd be inclined to wait for a few days and ask John 16:02:12 OK 16:02:27 I'm probably not the best person to ask at the mo... 16:03:10 One thing that was suggested yesterday though... 16:03:33 A page of stuff made with EMC powered machines 16:03:34 I am just hoping to shift some of the cleanup and other effort towards EMC2. 16:05:03 I would be happy to provide a gallery of EMC implementations if people would provide photos, descriptions, and links. 16:05:18 Steve, I added some more links last night. 16:05:35 I just found Dean Hedin's page a moment ago. 16:05:48 Added Les' page last night 16:06:16 As the Documentation team start adding tags to the source code, I'd like to try setting up a web server here to host the doxygen stuff on 16:06:38 and maybe crosslink from linuxcnc.org 16:12:28 Let me know about the doxygen site when it is running. Recently at work I told my lead engineer to check out doxygen as a way to make sense of some code we turned over to a customer who hacked on it for five years and then wanted us to cleanup and enhance it. It took some formatting work, but doxygen helped him a lot. 16:12:38 is somebody testing EMC2 with servos ???? 16:13:16 Not yet - No drivers 16:13:26 Hello all, first 16:13:58 Les - I see your links, but think we need a little work on organizing. High time we made a place for hardware that supports EMC. 16:16:55 I've looked at some of the German EMC sites to gather more links 16:17:08 I kinda have the machining links buried a bit 16:18:08 let me see which link dan put on the site 16:19:33 oh ok that's the direct link 16:22:38 I could also supply pictures of emc made parts...signs, the encoder product line prototypes, test stations, and others 16:23:15 That would be great. 16:23:49 Have some work this week using (gulp) .02" d endmill 16:23:56 looks kinda silly 16:25:20 Just added "hardware" links category and moved a few others around a bit. 16:26:07 It seems like about a year ago I asked for things that people were doing with EMC and got one response from David Munro 16:26:09 Didn't Paul create another drivers besides Vital? 16:26:28 Vigilant 16:26:39 ok I'll check it out 16:26:43 Going to retro a Bridgeport with emc as soon as I find one close by 16:27:03 I will try to document that 16:27:49 Question: Does the Vital card work reliably for servo emc? 16:28:43 In other words has it been tested in a real cnc machine 16:28:49 Dave Engvall would be the one to ask 16:29:10 ok will get in touch with him 16:29:13 He has it running on a Mazak 16:29:22 oh good 16:30:09 * paul_c needs to pull a finger or two out and get the lathe running 16:30:58 Viligent Products added to hardware list. Thanks Paul. 16:31:31 Vigilant 16:32:45 Ok, I see it. 16:33:43 I checked out "Tim's first part made with EMC" and all the photos are gone. I will remove it. 16:35:53 OK, but we do need to be careful with two of us editing as there is nothing to prevent us both from haveing a page "checked out" at the same time. 16:36:17 having --- gee my spelling gets worse and worse 16:37:20 Steve, go ahead and remove Tim's link. I'll hold off 16:38:52 Done 16:40:17 Steve, check this out: 16:40:23 http://www.hedingrips.com/ 16:40:47 and: 16:40:50 http://users.pandora.be/engineering/index_files/page0001.html 16:48:02 I have had some email conversations with the latter 16:48:47 as for the first link...#$%# wish I had thought of that! 16:49:40 I end up with a lot of short cut-offs of exotic wood 16:50:35 Added these two sites, but we should check back as they are both "buried" links. 16:55:04 Weather is pretty here.... time for some yard work. 16:55:23 well i'm off...need to hit some little white dimpled balls in to the woods 16:55:31 Same here... house painting. I will leave irc on to log. 18:43:13 Hi Paul. Did we pretty well wrap up the emc directory stuff? 18:43:35 I think we covered most of it. 18:43:57 Thanks for clearing up my thinking there. 18:44:23 Have read up on the CVS stuff and I think we could trap commits of specific files and trigger an error message 18:44:54 Would need to do some testing first though. 18:45:57 Ought we to leave emc.(run,ini,var,tbl,nml) and generic.(run,ini) in the base emc directory? 18:47:16 Yes - Just two basic sets along with the run script(s) 18:47:48 emc.run (or generic.run) can be reduced to a few lines. 18:48:04 Along with the tickle config, axis comp, and rt install and remove. 18:48:21 yes. 18:48:43 That cleans it up a lot. 18:53:12 paul_c * emc2/configs/emc.ini: Comment about comments added 18:53:23 And we have a new friend on this channel 18:53:43 Introduction? 18:53:54 It will be a nice suprise for John when he gets back... 18:54:32 CIA-9 is a bot with a feed from the SF commit messages 18:54:37 I'm all ears. 18:54:46 and sends a notification to us on irc 18:55:09 Really. 18:55:11 So each time a commit is made, we get a realtime notification here. 18:55:23 I gotta try this. 18:55:38 Thought you might want to play ;-) 18:55:53 pick a file, any file.... 18:59:54 Gotta update documents and then change scratch. 19:04:10 paul_c * emc2/src/emc/rs274ngc/rs274ngc_pre.cc: Small comment about splitting the files up 19:04:10 incomming 19:04:27 buggerme - That thing is fast... 19:04:46 rayhenry * documents/scratch.txt: bot test 19:05:27 That should add some incentive perhaps... 19:05:49 Yes it is. That confirms my need for a full time connection. 19:06:11 Anyone got 156 miles of two conductor? 19:06:31 Is cable (or isdn) viable in your neck of the woods ? 19:07:12 Nah. Dish at $80 or so. 19:08:01 Something for the squirrels to play on 19:08:20 Right. 19:08:43 Was naming it CIA your choice. 19:08:53 nope 19:09:52 http://cia.navi.cx 19:12:00 * paul_c notes rayhenry@sf isn't on the commit list.... 19:13:23 Want me to add you while I'm bouncing messages ? 19:13:46 Certainly. But I've been getting commit messages. 19:14:01 CIA is a nice project. 19:14:25 You still need to register with your SF usr addy - A note was sent out last week... 19:14:49 Added it anyway. 19:15:20 Note?? I seem to be missing things now and then. 19:15:27 and checked the nomail box so that you don't commit notices twice. 19:16:01 I think that you are loosing me. 19:18:04 Just forwarded a copy of the original notice to you 19:18:51 The emc-commit list has you listed twice now. Once with your SF account, and again with the regular addy. 19:19:33 On the SF entry, it is set so that you dont get any notices sent to that account. 19:21:38 I got it figured out now -- I think. 19:23:00 Hopefully it will encourage the likes of JE to subscribe 19:23:21 I understand. 19:23:33 Hi Robin. 19:23:38 Hi Ray 19:25:11 so .. did I miss much? 19:26:42 Got here not long ago myself. 19:27:05 You building flame tables? 19:30:57 well .... 19:31:02 sort of ... 19:31:31 I'm wondering whether just laser cutting parts for people might be more profitable and a whole lot easier 19:34:53 I might go out to the factory in a bit and continue prepping the laser ... spent 3 days cleaning it up so far, it was filthy!! 19:35:33 note to previous owner: if yuo do insist on running it in a factory full of powder coating paint, ffs keep the lid closed. 19:37:34 Ah. Laser table. 19:37:58 yep ... I sort of acquired another one. 19:38:08 the first is an EMC candidate I think 19:38:32 the second is fully fettled and ready to rock ... well, when I put it in position anyway 19:42:40 http://www.rapidcut.co.uk/images/LY2500.jpg 19:43:39 the controller on that one is a Bosch, quite nice, lots and lots of IO, but only about 100K of program memory 19:45:02 the other one I have is an EMC candidate I think ... awful controller using a 6502 and lots of TTL ... I thnk it works, but its poor, and destined to blow up at an inconvenient moment 19:46:02 also, I need to do "interesting" things with the laser I want to put on it .. so EMC is the baby for that I think 19:51:36 That should work. It would be trivial with EMC. 19:52:41 and anyway ... EMC doesn't have aproblem with program memory ... we can have huge ones for minimal cost .. :) 19:52:55 * robin_sz lurves PC based controls 19:53:04 What kinds of interesting things are you wanting to do with EMC control of a laser. 19:53:23 well, the laser I have has several control inputs ... 19:53:27 nothing weird there 19:53:32 standby, 19:53:38 pre-heat 19:53:44 shutter 19:53:46 and beam on 19:53:51 .. that bits easy 19:54:04 it also has an 8 bit digital command interface 19:54:26 set a new cmnd word and it spots that stuff has changed and does it ... 19:54:45 thats the bit that would be neat with EMC 19:55:33 the digital cmnd word sets things like pulse rate, duty cycle, power level and such 19:55:52 You could automate all of the controls and mix them right into g-code using M101+. 19:56:40 yep ... and write a nice custom driver to map the cmnd words to messages in emc 20:11:08 Robin, did you already have a lot of custom laser/plasma stuff worked out in EMC? 20:11:19 yes 20:11:49 2 basic areas. 20:11:59 no .. 3 basic ares 20:12:07 1) torch ignition 20:12:30 basically fire torch, wait x seconds 20:12:43 if you get an arc good signal then continue 20:13:34 if no signal in x seconds, then do something like stop program, note current line and X Y position, lift torch, come to table front position etc 20:14:01 2) tec table. basically like the tool table 20:14:52 but with things like pierce delay time, cutting speed, corner correction coeff., amperage etc for a whole list of materials 20:14:57 and finally ... 20:15:15 3) corner correction 20:15:27 plasma is a 'floppy tool' right? 20:15:32 ok 20:15:43 you have to slow down on corners or they flare 20:16:00 as the bendy flame drags around the curve, lagging behind 20:16:47 so ... corner speed = programmed speed x ( coefficient x fn(radius) ) 20:17:14 most or even close to all of that was all done ... 20:17:16 somewhere 20:17:41 but the structure of EMC1 drove me insane 20:18:00 would emc2 be better? 20:18:04 I might port it to EMC2, as there are glints of sanity in it :) 20:18:43 I re-wrote the GUI as well, again , its not finished, but its about 80% done .. 20:19:42 ummm ... now where is that 20:19:54 ah ha! ... http://www.redpoint.org.uk/cgi-bin/cvsweb.cgi/tkemc/ 20:20:05 ok I'll check it out 20:20:24 its still in tcl 20:20:30 but more ... structured. 20:20:35 needs finishing 20:21:05 if you look in 'components' directory you'll get the idea 20:22:05 I'll see if I can bring it up on a machine with working tk/tcl a bit later.... 20:22:20 I'm on a mac w/ os x right now. 20:22:25 I think it was left in a sort of working state 20:22:38 basically I was working to remove all the globals 20:22:41 it surfs the net and plays music well.... 20:22:47 of which there were 'undreds 20:23:25 and encapsulate each set of things as a component ... eg Axes displays 20:23:30 command buttons 20:23:31 etc 20:23:58 you then can build flexible generic GUI's from these components ratehr than a big, hardwired script 20:24:23 my key principle was that the GUI should still work with any component removed 20:25:18 so you could build a remote display by taking the standard GUI and removing the buttons component, and it should work just fine, without any mods to the code 20:25:33 that would be good 20:25:45 thats about how far I got 20:26:05 I played with Tkemc a bit a couple years ago 20:26:20 but I am a bit slow 20:26:36 and I broke a lot of things on my system 20:26:38 its the worlds worst language 20:26:41 playing around 20:27:02 truly awful programming language ... I hate it muchly 20:27:11 but, it does its job .. just. 20:27:33 I'd prefer to get to grips with QT ... but thats a big learning curve 20:27:44 kind of like a 2n op lathe next to the main machine... 20:28:08 yeah .. or in the office so you can monitor severla machines ... 20:28:23 * paul_c decides to pull out a BDI-2.xx box... 20:29:09 it would be great for the OEMs 20:29:19 say you want a toolchanger 20:29:23 no problem 20:29:34 add a 'toolchanger' component to the GUI 20:29:44 slide it into the framework ... 20:29:58 it knows what messages to listen for coming up from emc 20:30:10 it knows what messages to send down to emc when you click button 20:30:51 it needs someone to finish it ... 20:30:56 Im crap at finishing 20:31:08 (so my wife tells me anyway ;) 20:31:17 can I play with it? 20:31:21 sure ... 20:31:29 thanks. 20:31:45 click the tarball or zip links ... 20:31:53 bottom of page 20:32:25 ok 20:32:32 got the tar file 20:37:15 hm, did anybody know if John joins also today ??? 20:37:37 have some problems wirering my servo with EMC2 20:39:11 my driver for my motion control card works now, and i can see the encoder counting on the gui 20:39:13 John is on his way back from Georgia 20:39:22 uh huh ...