01:24:21 robin_s has joined #emc 01:41:30 robin_z has quit 01:58:50 robin_s has quit 02:16:05 picnet has quit 02:22:45 Mer has joined #EMC 02:24:31 Mer has quit 02:44:09 les: still around? 03:22:06 jmkasunich has quit 03:53:18 picnet has joined #emc 04:21:16 almost finished with the new blast cabinet :) 04:23:13 put photos of the thing on my site 04:24:34 tommrow ill add in the vacume unit stuff and mount the lights 04:38:55 picnet has quit 05:53:55 picnet has joined #emc 08:50:47 asdfqwega has joined #emc 09:07:03 picnet has quit 09:54:32 picnet has joined #emc 14:46:41 Imperator_ has joined #emc 15:00:17 thalx has joined #emc 15:04:16 picnet has quit 15:16:30 rayh has joined #emc 15:16:50 * rayh is away: I'm busy 15:18:36 SteveStallings has joined #emc 15:26:08 jmkasunich has joined #emc 15:27:26 * rayh is back (gone 00:10:36) 15:27:38 Hi Guys 15:28:40 hi ray 15:28:40 tap, tap, ....... yes it seems to be on..... duh.... Hi Ray! 15:29:16 What you tapping on steve, the box or the operator. 15:29:35 the microphone.... checking to see if it was dead 15:30:15 IRC seems to go dead on me... silently....then I try to send and it says disconnected 15:31:03 I've had a few times like that only don't get any response. 15:34:00 are we any closer to setting a date for EMC coders meeting? 15:34:14 Mornin, John. Recover from our discussion yesterday? 15:39:33 rayh is now known as rayh_ 15:40:43 rayh_ has quit 15:40:59 rayh has joined #emc 15:47:00 I'll try again, are we any closer to have a date for the EMC programmers meeting? 15:48:58 Don't know what's going on with IRC but John wrote. "Also, I heard from Fred this morning. He has reserved a room 15:48:58 for us from Monday 4/25 thru Thurs 4/28... So the date is now 15:48:58 official." 15:51:52 sorry guys, was away for a bit 15:51:57 yep, the dates are official 15:51:58 Thanks Ray 15:52:09 should I put them on the web site? 15:52:22 good question 15:52:28 this is a limited attendence event 15:52:43 I dunno how we want to publicize it 15:52:50 I still haven't written up an agenda 15:54:18 perhaps it could be listed with a stipulation that there are a limited number of slots with active coders having priority 15:54:57 sounds reasonable 15:55:10 picnet has joined #emc 15:55:14 lemme take a look at the linuxcnc page for a minute 15:57:20 maybe the "upcoming events" part of the "news/history" page should be moved to the top of that page (above the 04 election results) 15:57:25 and the fest added there 15:57:35 (actually a link to a page dedicated to the fest 15:58:07 actually, the upcoming events item needs updated, cabin fever is over now 15:58:16 (how did that go, BTW?) 16:00:55 Interest in CNC is building year by year. We had 4 demo machines this year. Thalx was running a Sherline with EMC from RC46. He was cutting a oval shaped "plunging spiral" pocket into a piece of aluminum as a demo of doing things that would be impossible manually. 16:01:40 cool 16:02:32 Tom Hubin also ran a Sherline from TurboCNC making optics parts and John Guenther ran a 7" mini-lathe making writing pen parts. 16:03:12 I didn't actually run my machine, too busy answering questions, so it just made the motions in air to cut a gear blank. 16:08:41 Roland Friestad is working on a outline of the events for the integration session at his shop. 16:09:19 I think we should have a page for each event, linked from the "history/news/events" page 16:09:29 the details of each page to be revised as we learn mode 16:09:31 more 16:09:37 I'd like to see some of the guys who recently posted about operator panels show up and work on these. 16:13:25 hey 16:13:25 :) 16:13:56 i got my mill running now 16:13:59 with emc2.. 16:13:59 :) 16:14:19 cool 16:15:18 although not working as I want it to yet 16:15:19 :) 16:15:28 trying to find out how to reverse my dir pins.. 16:16:03 gtg.. 16:16:48 reverse the pins? you mean change from active high to active lo, or change which signals are on which pins? 16:18:46 JMK - stub page now exists, all yours 16:19:06 dave-e has joined #emc 16:23:01 just thought of something... the history page has a link to the board election info 16:23:12 when should we have the next election? 16:23:43 should the term be exactly 1 year (which would work out to august) or should it run from NAMES to NAMES (or something like that)? 16:24:09 1 August 2005 ? After NAMES, Roland's show, and the programmers fest 16:25:36 sounds reasonable - have the existing board handle the major events, then turn over to a new board and give them plenty of time to get ready for the next years events 16:25:50 I'll post something on the board list to start discussion 16:27:17 Sounds good to me. 16:27:43 Matt's term will run a bit longer than a year but .... 16:28:52 Matt's term has run for many years ;-) 16:29:16 SWPadnos has joined #emc 16:29:29 Hello 16:30:49 SWPadnos: Hi 16:31:00 hello SWPadnos, welcome 16:31:07 hi 16:31:28 Thanks. Looks like I found a good time! 16:32:48 yep - regular meetings on Sunday 16:33:06 That's what I thought I remembered :) 16:33:09 hey, the gang's all here! 16:36:37 Hi Dan 16:37:15 hi Ray 16:38:22 I saw a few posts on the CADCAMDRO yahoo group about porting Mach2 to Linux. I had to bust up laughing about that. 16:38:54 fortunately Jon Elson straightened things out a bit 16:41:27 wot. They want the mach2's graphical interface? 16:42:51 without Micro$oft 16:43:16 robin_z has joined #emc 16:44:04 afternoon 16:44:44 hi Robin 16:44:54 I had mentioned EMC and AXIS, but I'm not sure anyone noticed. 16:44:56 mcuh going on? 16:47:18 I have something that I would like to bring up, that I hope doesn't piss too many off 16:47:55 I think that we spend a lot of time spinning wheels over setting up real time linux 16:47:56 fire away ... it probably will, but im sure we'll cope. 16:47:56 Steve Stallings runs to get hip boots.... 16:48:13 * robin_z nods 16:48:13 (better to be pissed off than pissed on) 16:48:39 maybe more time on that than the actual control of machines with the application 16:48:45 yep 16:48:50 while we have a BDI available 16:48:57 I'm all eyes. 16:49:02 that makes it pretty easy to get it up and running 16:49:31 Sorry, but it seems like it's kind of derailing things in some ways 16:49:53 to have to go over the install over and over again 16:50:00 well, 16:50:08 that could be solved very simply 16:50:28 the current BDI is debian based 16:50:46 just publish the rt kernel .debs 16:51:15 maybe rpms as well 16:51:19 work in underway in that very direction 16:51:19 yes, I love the deb based system 16:51:54 I guess what I'm asking is: how do we direct the questions from the outside in a constructive way? 16:51:55 you mean like these debian packages? http://homepage.ntlworld.com/bdi-emc/debian/ 16:52:11 like: I want to set up EMC on gentoo... 16:52:19 or Fedora... 16:52:22 or ... 16:52:41 while the main application needs to continue on 16:52:50 danfalck: tell em its not supported, but might well work, when they get it built they can publish a package 16:53:05 that would work 16:53:10 jepler: is thatyour local repository? 16:53:33 robin_z: no, I think it's paul's 16:53:43 I suspect that we will always have to insist on a Debian (or other) distribution as the base. Likewise Mach2 requires W2K or XP. 16:53:45 right 16:53:47 he mentioned it in this mailing list post: 16:53:48 Subject: [Emc-users] New emc package 16:53:48 Date: Tue, 11 Jan 2005 13:22:13 +0000 16:53:53 I've never actually used the packages 16:54:02 it needs to be set up on sourceforge I think 16:54:12 or at least linked to 16:54:21 I gont the RT kernel installed pretty easily on Gentoo. 16:54:32 A FAQ would be pretty straightforward to write 16:54:44 linuxcnc.org could host FAQs or HOWTOs for different distributions, if people want to write them 16:55:29 I think it's mostly the same for any distribution - IF you direct people to compile their own kernels. 16:55:31 FAQ would be a great thing to have, but even there we sometimes have internal disagreements about how to support people who insist on rolling their own. 16:55:54 danfalck: personally, I think .debs are the way to go. id drop the BDI stuff. just publish the .debs, tell people to do a debian network install and then add the single source line to apt/sources.conf and type apt-get install emc 16:56:24 robin_z: I think bdi-live has a place for people who use windows but want free CNC software 16:56:30 We need to be able to support people who do not have the skills to compile their own kernel. I think the best way is to insist that they use the "approved" distribution. 16:56:39 a bdi-that-you-install, less so 16:56:47 danfalck: a lot of time and effort goes into the BDI, but really, its a problem thats solved a million ways by other people. 16:57:12 that's what I'd like to see. Insist on it to minimize the impact on the rest of development 16:57:25 SteveStallings: suits me. debian seems to be the way forward to me, fedora has far too short a lifecycle 16:57:43 Sure - there are two types of people to support - those that will use a BDI install, and those that want to use their favorite distro. 16:58:10 As someone mentioned on the list, if you have a favorite distro, then you probably know how to compile a kernel 16:58:14 we do need a way to install on a machine without a net connection 16:59:00 dave_e: you must have a connection (or a friend with one) to get the files. Once someone has the correct files, they can be burned to CD. 16:59:00 We cannot afford to constantly devert valuable developer time to solve install problems on alternate distrubitions. Let them form a sub-group to solve problems. 16:59:04 dave-e: wel, heave them do a CD based disc install ,a dn then just add the emc .debs 16:59:43 SteveStallings: lets call it "emc-users" and allow developers to conecentrate on development in a seperate group, lets call that "emc-dev" 16:59:56 gotta go, lunch with wife... 17:00:01 apt and dpkg also have the ability to use a local directory as a deb repository. 17:00:29 yep, apt-get is great! 17:00:36 dave-e: my point is, installers for debian exist a plenty, disc based, net based, mix and match, just let the user choose and use one, then add our stuff 17:00:51 SteveStallings: Exactly - so a FAQ will point knowledgeable people in the right direction, and that's the extent of developere support for them. The BDI is supported more fully. 17:00:57 as long as I can transport on a cd I'm fine 17:01:25 except if I have to compile, 17:01:40 users shouldnt have to compile 17:01:50 that's right 17:01:51 I've not gotten a usable driver to the vital board yet dispite clean compiles 17:02:01 oh. 17:02:14 * robin_z scratches 'buy a vital board' from his to-do list 17:02:20 the same software compiled on Ray's machine and dist as a bin works better 17:02:20 Early adopters like Dave will always need to compile or have a friend in the compile community. 17:02:22 you're a developer then :) 17:02:57 not a developer, just slightly teched 17:03:01 well, the .debs have great dependency tracking 17:03:20 There is a neitherworld between user and developer. 17:03:32 indeed 17:03:46 (beta-tester :) ) 17:03:51 if we get the dependency tree right, the emc-dev pseudo package should make sure your debian has all the right bits of the toolchain, the right compiler, tk libs tcl libs etc etc 17:04:17 The trick is to find a developer that WILL produce experimental debs for these folk. 17:04:32 neitherworld is filled with those evil implementers 8-) 17:04:47 robin..don't get me wrong ... I think the vital board has great potential 17:04:53 Or teach these early adopters to compile for themselves. 17:05:39 IMO, the issue with these pci based boards is timing between pci and the fpga they use and emc. 17:05:57 wish I could prove that 17:06:13 dave-e: yeah, im sure it does. but I cant spend money on non-working kit right now 17:06:33 just hang in there and wait awhile 17:06:34 In an ISA the signals were all subservient to the EMC running. 17:06:48 actually, that answers another question for me. sadly in a negative way. 17:06:54 In the fpga the loops are often programmed to run independent. 17:06:56 FAQ suggestion - I will start building a FAQ if people will e-mail content to webmaster@linuxcnc.org (real content, not just "we need to...") 17:07:04 so vital isnt "stable" yet then? 17:07:23 there a many vital boards out there....and I may (?) be the only one having problems 17:07:29 or I'm just picky 17:07:35 I wouldn't say that. Abdul and I ran it successfully at his shop. 17:07:49 does it run, bug free for days and weeks on end yet? 17:07:52 and I've come close... 17:08:05 not for me 17:08:08 Within the parameters of a Sherline mill running 100 ipm. 17:08:37 it may be that I should chane MB's 17:08:40 so ... its still "development" then 17:08:42 And shutting down EMC when he goes home. 17:08:42 change 17:09:03 the STG card is expensive and ISA 17:09:08 so thats dead 17:09:24 not if it is the only thing that runs well 17:09:35 I cant by PCs with ISA 17:09:47 ISA and PC104 are very alive in the industrial setting. 17:10:02 actually I can't say that... ppmc runs well when it runs...but I tend to break it .... 17:10:42 SteveStallings, i'll help where i can and find time 17:11:20 asdfqwega has quit 17:11:25 SteveStallings: you plan to work from the existing FAQ or what? 17:11:38 robin .... email Don Lemke and see how he is doing with his vital board 17:11:48 I must admit that I need to look at the existing FAQ....... 17:12:10 rayh: so I could buy an industrial PC and an STG card for .. about 2K GBP .. or I could buy a plug and play rackmount motion controller for about .. err 2K 17:12:12 but the idea is to make it more visible on the web site 17:12:36 is the FAQ visible on the website? 17:12:38 Henkka was maintaining that FAQ for a while. 17:12:42 rayh: ISA is a seriously dead format for home and semi-industrial users 17:13:01 I believe that there is a lyx doc with the existing in sf. 17:13:01 rayh: which is why we need the PCI vital solution 17:13:20 PCI shouldn't be an issue, if the cards get set up correctly. 17:13:24 The Vital solution is very attractive. 17:13:44 its a lot more card than 99% of our users will need I suspect 17:13:49 The fpga setup solution is not a single sided EMC issue. 17:14:15 a 3 axis and 16 IO card for a lot less dollars would be nicer 17:14:37 Anyone with a parallel port can get Jon Elson's board for $250. 17:14:51 (and a PCI parallel port for $30) 17:14:55 SWPadnos: doe it drive servos? 17:14:57 Abdul found that the price/capability thing was exponential. 17:15:12 it outputs step/direction signals for a Gecko-like drive. 17:15:21 A nearly bare board cost nearly the same price. 17:15:23 SWPadnos: so thats a 'no' then 17:15:48 rayh: understood 17:16:15 < $220 17:17:52 pci/isa board 17:18:25 Go to pricewatch, click on motherboard, and type ISA into the search. There aren't too many options these days. 17:18:47 true but you only need one good one 17:19:59 Pricewatch is not the place to find these. 17:20:23 That's true, but you do need a certain CPU speed (depending on how you want to drive your motors) 17:20:27 And as robin points out they are spendy.\ 17:20:27 I just searched for 'isa motherboard' 17:20:57 rayh: nope, but that's where a lot of hobbyists would shop (or Dell / Gateway / etc.) 17:21:07 but who want to use stg anymore anyway? 17:21:51 but the alternatives are : ppmc, vigilant, ? 17:22:11 BTW Dave...do you have motenc working on an actual machine? 17:22:32 I just took it off...and am back to stg. 17:22:36 If I needed high speed stepping I'd work with Elson's board. 17:22:52 It worked ok? 17:23:27 it would run but shifted 125 thou in zero on a single run ...g1 x 24 y 8 17:23:32 g1 x 0 y 0 17:24:01 oh..encoder noise problem or whatever? 17:24:03 finally it took off at full speed...no control toward +y and I used the estop 17:24:17 same cabling with stg works perfectly 17:24:24 uh oh 17:24:51 This guy that called me bought a motenc near here 17:24:54 like I said before...check with Don Lemke...I think he is running one without problems 17:25:12 ah yes ...Buddy 17:25:14 the one that wants hardware feed override 17:25:37 He has got Abdul working on it I guess 17:25:46 where is don? 17:26:07 I think there will be a vital board at the coding fest and at the integration session 17:26:32 Don is MI maybe 17:26:47 look for his posts on emc-user 17:26:58 ok 17:27:37 I obviously want to go to the vital board on future intallations 17:27:41 17:27:41 "Donald V. Lemke" 17:28:15 thanks 17:28:29 wonder why yours acts up 17:28:39 wish I knew 17:28:50 perhaps it's just a defective card 17:28:53 I really suspect it is software 17:29:04 or my motherboard 17:29:49 I'm going to use the stg for now...just to be able to concentrate on making parts rather than fighting the machine 17:29:55 if it runs away sounds like i/o isn't getting written 17:30:21 well it ran away just after a g1 x0 y 0 at 100 ipm 17:30:25 Well the stg IS solid....I'll agree 17:30:26 I'm of the opinion that we should not expect to take on the responsibility for making a manufacturer's board work with EMC. 17:30:29 no pending motion 17:30:33 using it 30 hrs a week 17:31:03 I agree and it isn't that Abdul has not been helpful 17:31:27 We have tended to do that with stg and parport. 17:31:29 we really missed the boat when Abdul was sick and couldn't make it to fest 17:31:43 ray: agreed...but helping out a manufacturer should be ok 17:32:04 Absolutely. 17:32:08 With him taking on the product responsibility of course 17:32:20 And testing and all the rest. 17:32:59 But a non-functioning, flakey board should not reflect badly on us. 17:33:11 A low cost board that works would really make servo emc take off 17:33:36 Right. And I think Abdul's board is very close to that. 17:34:01 I'm still waiting for someone to do a Mauch/Kaluga with a dac 17:34:15 Someone mentioned a 3 axis low cost unit, but in small quan it would cost as much as 8 17:34:26 but that implies that you don't need i/o 17:34:41 SteveStallings, the first faq submission is on the way 17:34:45 the boards dont cost much to make 17:34:50 well with plenty of dio 17:34:56 probaly sub 50 dollars all in 17:34:57 Yep. The little logic i put on the rutex mb for Smithy cost near a $100 17:35:17 right robin...it's development costs 17:35:23 yep 17:35:28 especially for low numbers 17:35:41 but if you have high resale, you get low numbers ... 17:35:44 catch 22 17:35:56 yeah 17:36:04 vital if people can make it work is still a good deal 17:36:35 the development cost of trimming a proven 8 axis board down to 3 + 16 IO is small 17:36:40 I need to talk to abdul about distributing a bin for rc46...that might fix the problems 17:36:49 good enough to start some serious interest with commercial machine builders 17:37:19 that would get some volume that could lower the price for all 17:37:48 This also assumes that emc has a good trajectory planner 17:37:58 ahh :) 17:38:00 Chris may have made that happen 17:38:10 lets hope so ... 17:38:10 les: What volume do you think? 17:38:33 I wish I could do some serious testing of it but just can't right now 17:38:36 les: I have the latest CVS head built, I'll try it tonight! 17:39:07 pemmet has joined #emc 17:39:09 SWPadnos: at least hundreds anually I wou;ld think 17:39:28 robin: a full servo one? 17:39:51 Chris has tested steper 17:39:56 it would be interesting to know how many boards stg has sold 17:40:08 yeah 17:40:12 hello 17:40:18 prob quite a few but not for emc 17:40:22 afik, ppmc and vital are pretty low numbers 17:40:38 hey chris 17:41:09 I sure missed the start of the meeting today 17:41:17 For Vital to make any profit right now they need to persue non emc aps 17:41:30 to pay for the thing 17:41:32 slomo - got it, thanks 17:42:03 sure, just let me know how else i can help 17:43:40 les: I disagree. 17:43:52 hmm how so? 17:44:03 Glad you asked. 17:44:08 haha 17:44:10 17:44:27 I think that we need to get the word out. 17:44:55 If it works well for motion and IO, I believe that many projects will use it. 17:45:22 It does need to be proven. 17:45:27 I would love to use emc for client's machines 17:45:40 Yep. 17:45:40 we were talking about that yesterday 17:46:07 like the 75 ft foam cutter I was working on yesterday 17:46:45 The client is not a technical type and emc might be a real problem for him 17:46:47 emc needs two tinkgs to be useable in that sort of app 17:46:49 I know of half a dozen integrators who would use it if we can prove to them that it is as good as a commercial device. 17:47:11 I'm one of them 17:47:13 rayh: its MILES off commercial stuff, but it costs less too. 17:47:47 rayh: all the commercial stuff has click-and-drool configurators and GUI builders 17:47:51 There are a lot of applications that don't necessarily involve a lot of man-machine interfacing. 17:48:00 Maybe but I've deconstructed some of the commercial stuff and it ain't miles. 17:48:12 robin: if Chris' stuff works I think it can be a competitive modern control 17:48:14 Think about things like one-off feffects for theater (or the film industry) 17:48:16 rayh: motion wise, no .. we're pretty good 17:48:30 rayh: but customisabele wise? we suck im afraid 17:49:04 les: yes, control wise, but the front end is not easy to customise without programming 17:49:10 Could you list those issues. 17:49:11 And it's solid (for me) I have never had a crash or lockup with original emc 17:49:32 it does need some help on the TC enc 17:49:34 end 17:49:37 Try customizing an AB, Fanuc, Siemens. 17:49:56 ie...plc 17:49:56 rayh: have a look at the Baldor MINT and Workbench stuff 17:49:57 Try customizing a Mach. 17:50:05 rayh: Mach? 17:50:06 Ah...a conversational aux program that does stuff like .ini files? 17:50:14 What the hell we trying to pull over on ourselves. 17:50:25 rayh: whats a Mach? 17:50:38 2, 5 17:50:48 rayh: you're kidding surely? 17:51:03 rayh: have you tried customising mach2? 17:51:16 Just being a devil. 17:51:32 Are you referring to ini setup? 17:52:00 rayh: no, GUI stuff, custom code for buttons, operatiosn etc 17:52:04 That ain't hardly customizing. Thats integration. 17:52:38 Mach now supports lots of user extensions and adaptations to machine interfaces. It ain't all painless, but it is there. 17:53:05 rayh: compare adding say a new screen and a dozen extra buttons to Mach2 and EMC, I can do that in 20 minutes on Mach2 with just click and drool 17:53:30 What's the new screen do? 17:53:35 rayh: I can add custom code for M-codes with the same ease 17:53:36 Ray, please help me understand your customizing vs. integration 17:53:43 Is Axis configurable? 17:53:54 rayh: whatever I want, interacting with the core, new displays of parmeters 17:54:25 subroutines? 17:54:37 rayh: if we could have the flexibility of Mach2 with EMC we'd have it gripped. a world-class product 17:54:37 You mean you can get a slider readout of spindle load in 20 mins with mach? 17:54:51 yes 17:54:53 assuming I could get the value in there, yes 17:54:57 easy 17:55:26 with NO coding either 17:55:30 assuming I could get the value in there, yes. And therein lies the rub. 17:55:55 I hate to poke and run but the fam calls for lunch. 17:56:06 'k 17:56:10 later 17:56:11 les: not without programming 17:56:29 I see 17:56:29 rayh: do me one favour ... download mach2 and try the screen builder 17:56:33 les: I'm not sure how any frontend could be 17:56:47 les: sure you could drag a button into place but then what? what does it do? 17:57:08 cradek: link it to a variable internally so it changes state as the var chages 17:57:15 Well it seems like labview like cofigurability would be a tall order 17:57:23 a lot of work 17:57:24 cradek: have it call a function when you press it 17:57:39 * rayh is away: I'm busy 17:57:50 LabView isn't something to aspire to, in my opinion 17:57:51 mach2s motion sucks though 17:58:12 I built the stepper factory test units with DAQFactory 17:58:12 Mach stuff is built on Visual Basic - like programming with objects 17:58:24 yep, works OK 17:58:35 robin_z: but those things are called programming in my book 17:58:51 You drag and drop a button or bar and write a script in a window that pops up 17:59:02 they do not rely on the user running a compiler 17:59:06 dont even write a script 17:59:14 hmm 17:59:20 just pop into a field the name of the var it calls 17:59:26 fuction 17:59:28 whatever 17:59:45 so can you guys always read my emails to the list? I have been PGP signing them and I wonder if it causes problems 17:59:50 it seems like I get fewer responses than I expect 17:59:58 especially when answering questions for people 17:59:59 yeah, DAQFactory did that stuff too 18:00:22 Chris: I get your list posts 18:00:32 cradek: I have no problems w/ your emails 18:00:36 cradek - they will not preview in Outlook, require specific open to read 18:00:38 if it needs functions that dont exist, write them in VB, drop them in a file, call it M901.vbs, link the button to M code M901 .. done 18:00:45 seems like I answer lots of questions, never to hear back from the questioner 18:00:54 SteveStallings: grrr 18:01:24 SteveStallings: well MIME PGP signed messages have only been the standard for about 10 years - no big surprise that MS doesn't do it right yet 18:01:41 cradek: Remember when I grumped about this on an unrelated mailing list? 18:01:41 Where is the screen builder for Mach2? 18:01:42 heh 18:01:48 don't shoot the messenger, please.... 18:01:49 jepler: I guess so 18:01:51 cradek: you could solve that ... just drop all mail from Outlook clients, then you want be tempted to write answers to them :) 18:01:54 cradek: outlook + mailman + pgp = problem 18:01:55 SteveStallings: sorry! I'm not. 18:02:03 SWPadnos: distributed with mach2 18:02:06 well I'll quit doing it then. 18:02:33 robin_z: don't tempt me 18:03:16 jepler: what is mailman? 18:03:21 last post I see is to Buddy...the feedoveride guy 18:03:23 cradek: the mailing list software 18:03:31 no reply though 18:03:34 ok 18:03:45 I just answered a bunch for the foamcutter guy too 18:04:00 didn't get my copy yet though 18:04:07 ? 18:04:13 ok, there it is 18:05:14 cradek: that response about foamcutter is not a signed message? 18:05:19 no 18:05:33 I'm not going to sign anymore 18:05:37 I just got one foamcutting response (like 2 minutes ago) 18:05:41 you're going to let microsoft win? 18:05:49 no, I'm going to be pragmatic 18:06:04 if I want to help people use emc, they should be able to read my messages 18:06:10 cradek...much cleaner without all the yellow banners 18:06:23 **sigh** 18:06:43 but you don't know it's actually from me anymore 18:06:58 if I say something stupid now, I'll blame it on jepler 18:07:12 but how mnay people are going to spoof you? 18:07:20 If they're running EMC, they should have a Linux machine ... 18:07:30 I se the "how best to upgrade" thread 18:07:31 (and therefore a non-microsoft email client :) ) 18:07:36 I happen to agree .. but 18:07:53 SWPadnos: sure, but that doesn't reflect the actual state of the world. 18:08:02 but again no replies to Chris' answer 18:08:28 indedd - it is a sad state. 18:08:55 Chris: perhaps you answers are so good that nothing else remains to be said! ; ) 18:09:04 les: haha you're very generous 18:09:18 heh 18:15:14 I'm going to bail...have fun 18:15:28 dave-e has quit 18:16:57 yeah me too...have to assemble some circuit boards 18:17:21 circuit boards, you mean not wooden boards? 18:17:43 cradek: how hard is it to alter the gui in Axis? 18:17:56 I am so stacked with work that I have to do the engineering stuff on weekends 18:18:13 danfalck: The GUI is described by the file "axis.nf" which is really a tcl script 18:18:45 so you just change the axis.nf file and it works the next time you run axis? 18:18:47 danfalck: a couple of blocks in the axis.py script list widgets used in the Python code, variables, and commands that are implemented in Python but called by tcl code 18:18:58 no compiling? 18:19:02 danfalck: You could modify the installed copy of axis.nf, or edit the one in tcl/ and run "setup.py install" again. 18:19:12 ok 18:19:46 just wondering how hard it is, in relation to the question earlier about Mach2 18:19:56 The axis.nf file is actually written by an unrealeased GUI designer program for Tk, so editing it by hand is likely to break the GUI designer .. but nobody but me cares about that. 18:20:13 which ide? 18:20:30 Here's a web page about a 1-year-old version of the GUI designer: http://unpy.net/~jepler/nf/ 18:21:00 it's been used internally by a commercial company for about 2 1/2 years now, i've never gotten around to severing all the ties to that environment and releasing it to the world.. 18:21:38 hm, that version is actually older than I thought, but the program isn't undergoing much active development 18:21:52 cool your own GUI builder... 18:22:04 it was a fun project 18:22:23 but, the user can twiddle with Axis stand alone 18:22:49 sure, if she understands tcl, tk, and Python... 18:22:57 yea 18:23:47 time for lunch .. see you guys later 18:24:09 danfalck: if you have something specific you're interested in doing in AXIS, let's talk about it sometime, or on the mailing list... 18:24:22 danfalck: the problem with axis is that it works perfectly for me and for chris, so we're not developing it anymore... just using i.\ 18:24:25 er, it. 18:24:57 Mer has joined #EMC 18:25:30 jepler: I don't have any specifics to alter right now, but will later on.... 18:26:35 I also like to understand what's going on in it. Learning experience ;) 18:31:44 So, about this EMC2 / Kernel 2.6 thing :) 18:32:23 you need JMK or Paul for that, and neither is here just now...... 18:32:41 Oh, well. Too late I guess. :( 18:33:01 JMK will likely be back late today, Paul is traveling I think 18:34:16 cool - I'll leave this open and check from time to time. 18:34:34 hello 18:36:35 I need to go out to the shop. I'll talk to you guys later. Thanks. 18:36:47 jmkasunich: change high to low 18:37:23 jmkasunich has quit 18:38:06 jmkasunich has joined #emc 18:38:35 jmkasunich: change high to low 18:38:35 :) 18:38:52 re: programmers fest - how does one sign up, what facilities are available (lodging, etc.)? 18:53:30 planning still in process, not sure how many we will be able to accomodate, there are plenty of hotels in the area including one fairly cheap one, facilities will probably be just a conference room, perhaps a Sherline with limit switches if NIST sees fit to make it available, otherwise just a hard core software developers conference 18:54:19 sounds god to me. 18:54:41 JMK will flesh out the details as he gets time 18:55:18 OK. It's still several months away, but flights and hotel reservations creep up fast. 19:05:55 Mer has quit 19:10:47 picnet has quit 19:11:11 * rayh is back (gone 01:13:31) 19:15:48 alex_joni has joined #emc 19:16:27 greetings 19:20:01 jmkasunich has quit 19:23:37 no one around? 19:24:13 just catching my breath..... shoveling snow.... 19:24:31 heh.. I just escaped the snow... 19:24:43 came back to the city 19:24:48 I had about 80 cm in the mountains ;) 19:25:19 a bit more than we had 8-0 19:25:39 what did I miss? 19:28:06 discussions about ease of installings, configuring, extending EMC and other stuff about altenatives to STG card 19:28:20 * alex_joni needs to read the logs... 19:30:00 lol 19:30:13 I just read about: "Tarquin Fintimlinbinwhinbimlim Bus Stop F'tang F'tang Ole Biscuit-Barrel" 19:30:43 ????? 19:30:59 robin-sz's name for a compile script ;) 19:31:40 Hey there. 19:32:06 hello 19:32:12 perhaps you can discuss kernel 2.6 compilation issues?? 19:32:22 * alex_joni hides 19:32:31 I am kidding.. shoot 19:32:45 oh come on :) It's just EMC2, kernel 2.6, and Jon Elson's USC card. 19:32:49 What could be simpler? 19:32:56 * alex_joni wonders 19:32:59 dave-e has joined #emc 19:33:17 well.. what's the problem? 19:33:26 besides.. that it's not working ( I assume ) 19:33:35 Compilation of EMC2 would be the firsst problem. 19:33:51 yes... 19:33:56 (others are solvable once that one goes away) 19:34:05 to get that done a lot of tinkering with the Makefiles is needed 19:34:19 besides source modify 19:34:38 I have some limited knowledge about emc2, but less about 2.6 19:34:40 Actually, I'm looking at an excellent "getting your module to compile with kernel 2.6" article at http://www.captain.at/programming/kernel-2.6/ 19:34:53 I wish I'd have the time to install a 2.6 kernel 19:35:52 It looks like all you need (for a standard, non-RT kernel) is to do make -C (wherever the kernel is) SUBDIRS="eherever mh module is" modules, with your module name (no .o or .ko) as a target. 19:36:11 (SUBDIRS = "wherever my modules is" ) 19:36:23 (are - damn - I'm starting to type like Bush. 19:36:26 ) 19:36:30 lol.. yeah 19:36:39 that's kinda only half the trouble 19:36:52 you still have some include order to sort out in the sources 19:38:02 * rayh is away: for a bit 19:38:20 the Makefile needs to be redone 19:38:36 so that it can be used to compile stuff both on (2.2)2.4 and on 2.6 19:38:42 true. I solved one problem by putting uaccess after time.h (in rtai_rtapi.c) 19:39:38 Maybe different makefiles would be good, with the .configure script choosing between them 19:39:58 that could be done.. but I think that needs to be decided first 19:40:07 (or at least the "makefile.inc") 19:40:19 makefile.inc is generated by ./configure 19:40:36 can you send me a makefile.inc that would work for 2.6? 19:40:50 I would change ./configure to output one for 2.6 systems 19:40:59 right - it may be possible to make any kernel-specific stuff go there 19:41:08 Once I get one working, I'll let you know :) 19:41:20 coo 19:41:27 you can send to the devel list 19:41:33 that would be easiest 19:41:36 yep. 19:41:44 and you'll get a lot of competent feedback 19:41:59 yep^2 19:42:53 maybe I'll step onto the 2.6 train .. I might be able to help ;) 19:43:12 That would be excellent. 19:43:17 you could try to talk to paul_c, he did the new BDI (emc1 2.6 kernel) 19:44:05 I had no real difficulty in installing Adeos and RTAI (Fusion) on my Gentoo box. It was just a matter of finding the right installation files. 19:44:30 shouldn't be... 19:44:50 I mean you shouldn't get into troubles 19:45:09 you could check out libnml from emc2 CVS (tag pc_2_6_test) 19:45:15 that should compile on 2.6 19:45:30 maybe get some ideas from the code paul_c modified there 19:45:39 Good idea. 19:46:18 I'm not a CVS wizard yet, so I'm still figuring out how to manage different changesets, annd keep whatever changes I make locally. 19:46:51 unless you commit (cvs commit -m "message" file) it's only local 19:47:21 and when you do a (cvs up -dP) it will take from CVS what's new (and maybe merge your local stuff into the local copy) 19:47:57 you need (cvs co emc2 -r pc_2_6_test) to checkout the pc_2_6_test tag 19:48:26 ah - cool. I tried to get to the CVS manual linked from the coding style document, but it's a dead link (and I could only find a Subversion manual on that site now) 19:55:25 jmkasunich has joined #emc 19:56:03 * jmkasunich is back (I think) 19:56:13 hey john 19:56:30 hi... had IRC problems while I was away, don't know how much I missed 19:56:38 looks like there was a good discussion going 19:57:30 a little 19:57:35 a bit about emc2 / kernel 2.6 19:57:36 logger_aj, bookmark 19:57:36 See http://193.226.12.129/irc/irc.freenode.net:6667/emc/2005-01-30#T19-57-36 19:57:49 jmk: can read there about it 19:57:59 asdfqwega has joined #emc 19:58:00 I am still catching up what I missed yesterday and today 19:58:00 thanks 20:01:06 * rayh is back (gone 00:23:04) 20:01:17 hey rayh 20:01:26 Hi alex. 20:01:40 Nice sledding here today. 20:01:56 yeah... I did a little bit of skiing today 20:02:00 and yesterday 20:02:24 had a little bit of snow in the mountains (around 80 cm) 20:03:04 Nice. 20:03:19 Which mountains? 20:04:21 eastern romanian carpatians 20:07:17 picnet has joined #emc 20:07:50 alex. I've used the config stuff with emc1 and rc46. Works swell. 20:08:13 dave-e has quit 20:13:53 glad to hear that 20:14:06 I've sadly read about people not knowing where/if it is 20:14:16 me! 20:14:20 I think we should advertise it a bit more... 20:14:29 yes you, robin ;) 20:14:41 alex_joni: is thatyour work? 20:15:01 robin: we (paul_c & myself) did it this way (rcslib/etc) because that should be the first way you gobefore compiling rcslib 20:15:06 and before compiling emc 20:15:10 right ... 20:15:17 so , question 1 20:15:39 why is configure in etc/ and not in / 20:15:52 I'm also not sure it works on all platforms, so directly in rcslib/ could confuse users 20:16:07 they see that ./configure doesn't work.. they go away 20:16:32 well, every other project ive seen has it in / 20:17:10 robin: btw, I'm working on a PC104 card.. not fully done yet 20:17:41 I would argue that people cd rcslib, , see no configure, type make, it doesnt work, go away 20:18:08 it will have 2 LS7266 and 2 82C55 20:18:30 add a couple DACs too 20:18:47 yeah.. 20:18:57 a README is a minimu 20:18:59 m 20:19:03 they could at least read the README 20:19:16 if there was one, yeah 20:19:42 maybe update the README.. ? (I should do that) 20:19:48 no 20:19:56 create a README 20:20:26 there is no README in rcslib 20:21:04 ok.. I'll add a README in rcslib/ 20:21:48 excellent :) 20:21:54 describing the new (easy) way of ./configure 20:21:54 and the old (hard) way of make PLAT=whatever_foo 20:21:59 yes 20:22:41 personally id favour having configure in / and instructions on what to do when it borks in the README 20:22:57 ok.. question 2 20:23:14 * robin_z tries to remember what questions 2 was 20:23:17 oh yeah 20:23:51 actually, thats all the questiosn for rcslib 20:24:02 the only ther niggle was in emc 20:24:43 firs time you run it does the "RTL Version?" 20:24:47 prompt ,, 20:25:20 it wasnt even obvious to me it was a prompt ... we need that to be more obvious 20:25:27 and/or automatic anyway 20:25:44 it should/could do: 20:26:07 You must specify the rt linux kernel extensions you wis to use: 20:26:15 I have found: 20:26:22 1) RTAI_2.1 20:26:29 2) rtlinux 3_0 20:26:31 RTL is kinda deprecated ;) 20:26:42 3) rtlinux 3_1 20:26:52 choose: 20:27:02 no .. its not deprecated 20:27:11 it did it to me 20:28:31 RTAI is the way to go now... 20:28:31 ./configure does search for the installed RT 20:28:31 and complains if more than one is found 20:28:31 and echo's the one that will be used 20:28:31 with a message that another can be used by specifying (--with-rtlinux=... or --with-rtai=...) 20:28:32 where ... is the path to the RT 20:28:33 robin: it's a big PITA to figure out what RT is installed 20:28:36 they are so unconsistent from version to version... 20:29:23 no reason for amking the user prompt so obscure 20:29:52 anyway, I didnt find a configure in emc anyway 20:30:28 from rtai-24.1.13 to rtai-3.0r1 a lot of stuff changed 20:31:03 there is none in emc 20:31:08 because there is no need for one 20:31:27 well, it asked me that question 20:31:33 you compile emc exactly like you compile rcslib 20:31:33 so you do ./configure in rcslib/etc 20:31:44 and that creates realtime.def and nonrealtime.def 20:32:06 Are there any known problems with the "Fusion" (development) RTAI version? 20:32:10 then you "make PLAT=realtime all" and "make PLAT=nonrealtime all" inside rcslib and emc 20:33:40 didn't try that... 20:33:40 wouldn't neither ;) 20:33:40 as it is right now there are a lot of RT-version we need to handle 20:33:40 wouldn't wanna add unstable ones aswell 20:34:08 indeed. :) 20:34:12 I hope you see why ;) 20:34:30 That's the one I installed, and the test programs all seem to work fine. 20:34:47 not because of the work.. but more because of the lot of installs 20:34:56 I'm not sure if that's part of my compilation problem (but I don't think so - it seems header-related) 20:35:19 and a lot of code to configure.in (which generates ./configure), thus making it unreadable 20:35:30 SWPadnos: the 2.6 is a big issue 20:35:32 So - I didn't do any separate download of rcslib..., Maybe I should start there. 20:36:02 which needs addressing 20:36:02 jmk: still arouns? 20:36:02 jmk: still around? 20:36:24 yeah 20:36:38 alex_joni: so, I dont understand. you say emc doesnt need a configure, yet mine ended up not knowing which version ot RT to use 20:36:44 for a couple mins anyway... wife has things for me to do ;-) 20:37:14 I say that means it needs some sort of configure, even if it only pops up a screen asking you to go away and configure rcslib 20:37:20 SWPadnos: you are facing more possibilities 20:37:20 1. try emc1 (emc & rcslib) 20:37:20 probably won't work on 2.6 ;) 20:37:20 2. try emc2 (the new development beeing done right now, not fully functional, but can be used) 20:38:04 probably won't compile on 2.6 either 20:38:04 3. get paul_c's version of emc1 from the emc2 CVS 20:38:04 that probably will compile on 2.6, although not many drivers are there... (only freqmod AFAIK) 20:38:05 emc2 won't work on 2.6 either, as far as I know 20:39:15 robin_z: no ideea what you did, to make it pop you that question ;) 20:39:18 I'm working on that as we speak. 20:39:24 (could be a while, though) 20:39:31 might be coded somewhere in the Makefiles or in generic.def 20:39:52 I'm guessing Robin either had no RT installed, or more than one 20:39:55 alex_joni: I didnt do a configure in rcslib, I did the standard makes in src in emc and rcslib 20:40:06 SWPadnos: working on what? emc2 on 2.6? 20:40:28 The question that robin got comes from a bdi-2xx. 20:40:39 rayh: no it doesnt 20:40:45 It is a start up emc problem rather than a compile problem. 20:40:48 this is CVS head 20:41:04 never got bdi-2xx installed... 20:41:12 But the date of the cvs is not the issue 20:41:24 It is the date of the linux that you are running under. 20:41:28 had HW-problems whith that 20:41:47 anyway ... 20:41:48 I believe that is the case anyway. 20:42:01 i agree. its a initial run problem 20:42:03 anyways... 20:42:15 This was something added when rtai did not report itself properly during the early days. 20:42:18 robin: the README is a must, I'll do that 20:42:24 jmkasunich is now known as jmk_away 20:42:31 yeah. agreed. just 10 lines will do it. 20:42:42 * alex_joni goes away for half an hour 20:42:58 I'll leave this open, and catch up in while 20:44:22 just trying that linux/w32 driver for the G2002 20:49:42 OK, the first tiny steps of putting the FAQ directly on the web page are done. In trying to break it up into categories I quickly realized that this is going to be a big effort. I see the FAQ as something for people just starting out. As such I feel we should steer people to the newest and most mainstream stuff rather than emphasize all the details of getting the compile process to run. Perhaps move this to a "knowledge base". 20:54:08 So FAQ would be current or recent things and KB would be older or more specific? 20:54:19 errm 20:54:29 why not just one place 20:54:33 the faq 20:54:44 because eyes glaze over when they see all the compile issues 20:54:48 with sections for installing, building, running, 20:55:00 look at my stub 20:55:07 SteveStallings: poeple go to faqs for answers 20:55:19 if they dont find answers they dump it and move on 20:56:08 The installation related information can be links from the main FAQ page 20:56:24 (ie, For Gentoo installation information, click here) 20:56:43 1 faq page, sections for everyting 20:56:55 all the info in the faq. 20:57:02 if the consensus is to keep it all in one place, then that is what will happen, just stating my feelings about scaring people away 20:57:19 why would it scare anyone away? 20:57:24 It can get pretty big if you put everything on one page. 20:57:30 it could be just a chapter in the faq, q line 20:57:45 SWPadnos: no one would put it all on one page. 20:57:46 Yes - it can be scary to see a 100-page FAQ, when all you need are the 10 lines that relate to your situation. 20:58:10 SWPadnos: consider an index 20:58:15 right - I uncderstand you now. Ignore my previous ingoramus line :) 20:58:20 (understand) 20:58:25 ;) 20:58:51 you dont have to litter the index with stuff, just make a section at the end ;; compiling. 20:58:59 there could be loads of subchapters. 20:59:03 * robin_z shuts up 20:59:19 over half the people that I mention EMC to have looked at it and turned away because they have real about the installation difficulty. They never even tried because of the perceiption, not the reality. 20:59:29 read 21:00:25 well, that is either wrong or wont change either way 21:00:32 we have no FAQ yet, 21:00:39 so they didnt read it there 21:01:05 all the more reason to generate a FAQ 21:01:18 if it is difficult to install, they'll still end up having difficulty and still ask questions 21:01:22 quite 21:01:43 OK, lets say we do what SteveStallings suggests. 21:01:45 from my perspective, break it up into differant sections, installation, config, running 21:01:51 again - go to wwwlinuxcnc.org and look at the FAQ stuff 21:01:56 a FAQ and a knowledgebase 21:02:21 can we call them FAQ part1 and FAQ Part2 then please :) 21:03:22 chapters, parts, indexes, whatever.... I just want peoples first exposure to be something other than the intensely technical compile issues 21:03:33 sure. thats fine 21:03:48 have a HOWTO 21:03:51 and a FAQ 21:03:56 thats normal 21:04:13 the howto is 1 page on how to install from .debs or whatever 21:04:20 the FAQ is the rest 21:07:58 SteveStallings: did you consider a wiki for the FAQ? 21:09:17 I must confess to being ignorant about wikis. There is a German one for EMC, but I cannot read German. I also know nothing about setting up a wiki, but feel that it would probably be a good idea. 21:09:44 I do still feel that a conventions FAQ should exist for the basics. 21:09:45 probably yeah. at least then people can update it , without bothering you 21:13:14 We ought to look at wiki again for this sort of thing. 21:14:28 I think you need a 1 page HOWTO, what it does, what system,sa are supprted, where to get it, how to install it, very short and brief. 2 sides of A4 21:14:43 trying to look into wiki engines now, but if I put it on my server, it will be M$ stuff 21:15:05 oh, you have perl? 21:15:46 there are severl wikis written in perl 21:15:59 so long as your server runs perl, theyll be fine 21:16:15 perl is possible, but not currently implemented on my server 21:17:34 * alex_joni is back 21:17:42 ive never run a MS server, dont really know how they work 21:18:07 the shame is I must say the same for Apache 21:18:20 you mean linux 21:18:56 theres is apache for win32, 21:19:10 I do stumble around a bit on Linux, but have not set up a server. My server runs lots of special stuff like Cold Fusion. 21:19:22 euww. :) 21:19:42 I thought Cold Fusion was perl based? 21:20:27 I am not an IT type and have too little time to become competent with this stuff, I just do as little as possible to get the job done. CF is based on an extension of HTML language. 21:20:39 right 21:22:04 * CfWiki -- For AllaireColdFusion server. Still in development but functional. 21:22:04 * SeedWiki -- A wiki that can use pluggable components and has provision for WysiWyg editing. 21:23:00 My first perference would be for someone else to set up a wiki and I could just 21:23:05 provide a link. 21:23:40 sure 21:25:08 UseMod or MT moveable type ...? 21:25:13 who's going to FEST? 21:25:21 SeedWiki looks interesting, but if effort is going into something new, it should be in the direction of opensource and movable to Linux in the future. 21:25:31 * alex_joni has a request... 21:25:34 The German EMC users have a wiki going. 21:25:40 I hope to go 21:25:58 German wiki looks sort of empty to me? 21:26:29 rayh: going to FEST? 21:26:41 both if I can 21:27:05 oops, I didn't scroll down far enough on German wiki, perhaps scared of the language 8-) 21:27:27 ok, installing CGI::Wiki::Simple then? 21:29:23 toolate, I typed install now :) 21:29:30 any chance we could just open an English language section on the existing German wiki? 21:29:41 dunno, ask em 21:29:56 I keep trying to delegate.... 8-) 21:33:07 you want me to set this up? 21:33:10 or not? 21:33:16 * robin_z doesnt mind either way 21:33:31 robin: set it up 21:33:47 don't ask ... 21:33:47 ;) 21:33:55 go for it 21:34:52 later if I get better situated with server capabilities we might consider moving it on LinuxCNC.org 21:38:54 btw, I forgot to tell you guys... 21:38:54 * alex_joni ran a Java GUI on emc the other day ;) 21:44:48 I noticed that java emc directory - was wondering what you were doing 21:44:49 asdfqwega has quit 21:44:53 asdfqwega has joined #emc 21:46:13 done 21:46:14 http://www.redpoint.org.uk/cgi-bin/wiki.pl?HomePage 21:46:39 * alex_joni prods an0n 21:47:47 ready for it to be linked, or do you want to establish its identity first? 21:48:37 robin: how can I add a page? 21:49:54 alex_joni: easy ... click 'rdit the text of this page' 21:50:02 type some crap 21:50:14 and use a word with StudlyCaps 21:50:19 save it ... 21:50:29 ok so far? 21:51:13 not really.. my connection is crap 21:51:19 ;) 21:52:02 well, just edit a page and use a word with StudlyCaps like AlexsPage 21:52:37 * alex_joni tries that 21:52:58 when youve done that, youll now see the page has someting like 21:53:15 seen that 21:53:19 AlexsPage[?] 21:53:34 click the ? mark 21:53:51 type the crap for alexs page. save it .. job done 21:54:17 coo 21:54:23 easy huh? 21:54:52 yup 21:55:03 do we need to establish a useful name like "EMC WiKi" for a root page? 21:55:38 if you like 21:56:09 I like. :) 21:56:25 or EmcFaq? 21:56:27 seems that people need to feel that they are in the right place when following a link from elsewhere like LinuxCNC 21:56:35 well 21:56:43 thats noting to do with it :) 21:57:15 the actual link might be C:foobar/intropserviver.dll?xml-barfalot 21:57:54 put whatever text yoyu like on the link, the uRL is unimportant 21:58:00 not commenting about the URL, just the logo/title that shows when you get there 21:58:19 HomePage seems a little generic 21:58:20 oh, yeah 21:58:22 fine. 21:58:30 do you want the word Wiki in it? 21:58:37 might confuse people 21:58:43 EmcFaq? 21:58:51 EmcHowto? 21:59:00 I think so to distingush it from a simpler formal FAQ on the web site. 21:59:13 EMCWiki? 21:59:37 EmcWiked ;) 21:59:42 95% of machinsits wont knwo or need to know what a wiki is 21:59:58 EMCKnowledgeBase 22:00:23 EMCWhatWorks 22:00:29 * alex_joni quotes robin_z: EMCTarquin Fintimlinbinwhinbimlim Bus Stop F'tang F'tang Ole Biscuit-Barrel 22:00:46 It's unlikely that 95% of machinists will be setting up a machine controller as well :) 22:01:07 whats it to be guys? 22:01:15 * robin_z has vi open ... 22:01:30 EmcKnowledgeBase 22:01:31 ? 22:01:46 great idea! :) 22:01:46 or.. EmcExperiences 22:01:49 !wq 22:02:02 something that implies that it is a living knowledge base, WiKi, Blog, whatever 22:02:42 I like calling it EMC WiKi. I can explain what it is on the LinuxCNC page 22:02:46 whatever, create your own :) 22:03:06 hello ya'll 22:03:14 just link to http://www.redpoint.org.uk/cgi-bin/wiki.pl?WhateverYouLike 22:03:16 hello 22:03:18 create that page 22:03:31 But the "HomePage" title is not editable, is it? 22:03:53 yes, its the title of whatever page it is on 22:04:01 OH! "HomePage" is intended to be above "EMC WiKi" 22:04:10 its just the default is set to HomePage 22:04:19 I can set the default to anything 22:04:30 right - a likn like that creates a new page with "WhateverYouTyped" as the title. nice :) 22:04:34 link 22:04:36 do you want to be able to host other WiKis 22:04:56 shrug 22:05:05 if not lets call HomePage >> EMC WiKi 22:06:15 just link to wahtever you like, I'll look at what you linked to in a weeks time 22:06:20 make that the default 22:06:23 easy huh? 22:07:12 we back up the data remotely quite often, so its safe 22:07:37 * alex_joni added some stuff... 22:08:35 excellent 22:08:51 this is what I want to add to the README... 22:09:02 now people can modify it before I commit it to CVS 22:10:17 OK - LinuxCNC.org now links to EMCKnowledgebase on Robin's server 22:10:30 we have gone "live" 22:10:44 lol 22:10:45 yeah 22:10:56 I have just seen that my user is bull ;) 22:11:09 last edited by 81.196.249.18 22:11:50 * robin_z just edited it 22:13:07 Wikis are pretty awesome ... 22:13:08 oh, thanks for the fixes ;) 22:13:12 * alex_joni agrees 22:13:40 I think the whole linuxcnc.org canbe switched to a wiki (perhaps some certain areas user/pass based) 22:13:41 is it reasonable to feed all the old FAQ stuff into the WiKi 22:13:48 have a look at the 'view other versions' stuff 22:13:58 SteveStallings: yeah, some I think 22:14:10 steve: maybe filter the old-faq a bit 22:14:11 SteveStallings: dunno if all those old snippets of emaiksl should be kept 22:14:55 that is why I asked, for now I will leave the archive on LinuxCNC and ask those in the know to port across the stuff that is still useful 22:15:13 ok.. that sounds reasonable 22:15:15 wikis are great for user-created docs. let the community do your documantation! 22:15:22 yay 22:15:23 :D 22:15:37 * alex_joni wishes he had a comunity for his projects too... 22:15:48 I hate writing documentation on my own ;) 22:16:38 well, thats my contribution to the project for today done then :) 22:16:57 THANKS! 22:17:10 robin: it's appreciated 22:17:29 Fantastic job guys. How do I start a new page under this 22:17:46 rayh: two ways 22:17:54 now he pokes his head out ..... 8-) 22:18:02 1 edit a page, ad a link word LikeThis 22:18:06 save it .. 22:18:21 click the ? mark next to the LikeThis[?] 22:19:44 rayh: if you want to create a page that isn't yet linked from anywhere (kinda pointless, but ..) just add the NewPageName at the end of the URL in your browser, edit it and save it, link to it later 22:19:56 just ignore that 22:20:14 create em linked from the front page, the community will sort it out :) 22:22:21 for anyone that cares, theres a 'perl golf' competition a while ago to write a wiki 22:22:36 you know what perl golf is? ... write it as short as possoble 22:22:51 wanna guess how many lines they got it down to? 22:23:45 how many? 22:23:50 1 line 22:23:59 220 bytes I thnk 22:24:11 Long old bugger eh 22:24:15 :) 22:25:09 slomo has quit 22:27:26 added some BasicSteps for using the wiki 22:27:27 ;) 22:27:37 based on robin's advise... 22:28:07 yup 22:28:27 lets review progress in a week :) 22:29:09 yeah 22:30:01 Hi all 22:30:11 hey Martin 22:30:20 Hi Alex 22:30:40 You are trying to establish a EMC Wiki ? 22:31:07 it's up and running 22:31:13 http://www.redpoint.org.uk/cgi-bin/wiki.pl?EmcKnowledgeBase 22:31:39 ok 22:31:53 i tryed that also last year 22:32:50 how did it go? 22:33:21 i have searched for a Wiki engine that can export the content easyly to latex, because Johns point was that he like well formated documents 22:33:48 really, 22:33:57 my prefernce is for information 22:34:02 I tryed the one which is running wikipedia, but that was not that successfull 22:34:24 well, well try this one 22:34:26 usemod 22:34:30 see what gives. 22:34:31 jmk_away has quit 22:35:18 john can always wget it and filter it to produce the latex stuff 22:35:39 html2latex or whatever 22:36:06 maybe there is now one that can do a export to latex, because there was one last year with a export function but it was not finished 22:36:20 a wiki becomes a complex web of interrelated info 22:36:35 how you flatten that to paper I dont know 22:37:02 hmmm 22:37:31 just cut and paste I guess 22:37:50 a wiki is a living thing, once its pasted onto latex, its dead anyway 22:37:59 a automatic way would be fine 22:38:13 for snapshots 22:38:14 robin: you could do a html2pdf from cron ;) 22:38:34 then it is also possible to develop the emc handbook in a wiki 22:38:54 develop 22:38:55 yes 22:39:06 LinuxCNC now has FAQ links sorted out. Archive is now a sub-page to the main FAQ. 22:40:59 is this links and other stuff or what? 22:42:17 very simple FAQ started on LinuxCNC for new users, links to WiKi and old archives 22:42:24 alex_joni: tell me about your pc104 thing 22:42:38 robin: I have some boards designed for pc104 22:42:49 I have a 16In/16Out board 22:43:18 and I'm working on one with 2xLS7266 (4-axis encoders), and 2x82C55 (a lot of IO) 22:43:34 and probably I'll make one with analog in/out 22:43:45 SteveStallings: why not link to a page called GettingStarted in the Wiki ... then its updateable, even if "seperate" from the rest of the wiki 22:43:54 * Imperator_ got two 7266 samples for his EMC isa card 22:44:03 SteveStallings: Found it had to refresh a couple times. 22:44:12 alex_joni: sounds interesting 22:44:19 which one? 22:44:28 alex_joni: the encoder/DAC one 22:44:38 I wanted to make a basic board for EMC 22:44:45 one thing 22:44:46 to be useable by users.. 22:45:02 you have done 2 boards right? 22:45:18 1 is $small IO 22:45:23 the second one is still not 100% 22:45:28 other is servos + lots of IO 22:45:31 right? 22:45:34 kinda 22:45:45 interesting !!! 22:45:48 the first one has optocoupling on all the io's 22:45:54 Hmmm. $9.00/chip for the LS7266 (in 100's) + $2.68 each for the 8255's - you may be better off with a PLD 22:45:59 the second one doesn't because of the space... 22:46:20 woudl (servos + little IO) and a secpnd big IO make more sense? 22:46:24 wouldn't want anything else than a LS7166 or LS7266 for encoder count 22:46:34 ahh.. I forgot 22:46:43 I have another board with 4xLS7166 ;) 22:46:52 4-encoders 22:46:53 and DACS? 22:46:56 no das 22:47:00 oh 22:47:01 no dacs :( 22:47:06 which boards has them? 22:47:20 I plan on a DAC board (3 In / 3 Out) 22:47:28 maybe 4 if the space is there 22:47:33 umm 22:47:46 want my 0.02USD worth? 22:47:51 always 22:48:08 the encoderd aint much use without dacs 22:48:13 My planning is 2x 7266=4x encoders and one DAC7744=4x16bit DAC and 16xdigital out and 16x digital in 22:48:35 make the boards with as many dacs as encoders 22:48:56 if you can oly fit 2 dacs and 2 encoders on a baord, so be it 22:49:21 I cant think when you would want an encoder without a dac 22:49:22 I'll look into that 22:49:35 I use the encoders along with G340's 22:49:43 and a feedback to emc 22:49:59 hmmm 22:50:25 anyway ... the most common use will be standard servo 22:50:30 endcoder and dac 22:50:45 the format Imperator_ mentioned sounded ideal 22:50:46 yup 22:51:00 4+4 + 16 IO 22:51:13 yup 22:51:15 that floats my boat :) 22:51:21 but you can't fit that into a PC104 22:51:25 never, ever 22:51:28 2 maybe 22:51:33 3 for sure 22:51:36 Woo! You guys are setting up a wiki? 22:51:37 2 22:51:46 asdfqwega: done it 22:51:47 * asdfqwega can't wait to throw in his two cents 22:51:56 go right ahead 22:52:08 hope to get that all on a euro size isa card 22:52:17 asdfqwega: http://www.redpoint.org.uk/cgi-bin/wiki.pl?EmcKnowledgeBase 22:52:29 if it's not bigger than a STG ;) 22:52:43 Imperator_: did you say ISA? 22:52:50 jmk_away has joined #emc 22:53:01 robin_z: yes, sorry 22:53:07 * asdfqwega wishes for a modern mobo w/ ISA 22:53:08 pci maybe later 22:53:14 robin: PC104 is ISA too ;) 22:53:25 jep 22:53:27 asdfqwega: go SBC 22:53:33 alex_joni: electrically, yes 22:53:34 there are some nice VIA ones 22:53:47 i have also a pci developer card here, that is the next step 22:53:48 passive cooling @ 600 MHz.. yey 22:54:00 Imperator_: you have ProtelDXP? 22:54:04 robin: connectionwise, I mean 22:54:07 alex_joni: sounds expensive 22:54:28 robin_z: what's that ??? 22:54:40 asdfqwega: not really 22:54:41 I'll stick with dirt cheap relics 22:54:46 2-300$ 22:54:57 you get a lot of ISA relics ;) 22:55:01 Imperator_: design package, logic anaysis, FPGA desing, the lot .. all in one 22:55:19 * alex_joni uses Orcad ;) 22:55:19 Imperator_: comes with some PCI examples ... 22:55:24 Imperator_: apparently ;) 22:55:32 ah ok, no i don't have it 22:55:37 but Eagle is cool too 22:55:44 yeah, have eagle 22:55:45 and free.. I might mention 22:55:46 I've got a PCI card with a AMCC PCI matchmaker chip on it...I wanna make use of it sometime 22:56:37 Does anyone know where Paul is? I think I remember something about him traveling... 22:56:56 * asdfqwega uses gEDA/pcb 22:57:04 yup 22:57:11 asdfqwega: states ... 22:57:13 he's away ;) 22:57:30 asdfqwega: I tried to install gEDA.. but it failed 22:57:57 Great, another 'british invasion'...:P 22:58:14 What are you installing it on? I've used it on Mandrake and Debian 23:00:32 SuSE 23:00:36 8.2 23:00:46 it does compile, but it segfaults at some time.. 23:00:57 and I was tired, not interested... 23:01:23 jmk_away has quit 23:01:28 is SUSE rpm-based? 23:02:02 yup 23:02:24 * asdfqwega peers into his local repository 23:03:03 Do you think SUSE could tolerate Mandrake rpm's? 23:03:12 SuSE might have a gEDA package... 23:03:40 don't think so... 23:03:40 depends on libs & such 23:04:19 wow. the G2002 linux front end seemed to build and run first time ... 23:04:21 * asdfqwega can break a .rpm system in no time flat 23:04:38 cool 23:04:43 ? 23:04:56 the G2002 stuff ;) 23:05:04 * alex_joni is thinking of switching to debian 23:05:11 Pbbbbbb! 23:05:12 do it 23:05:18 debian rocks 23:05:29 It rolls! 23:05:42 It rolls and rocks! 23:05:54 I'll miss yast ;) 23:06:11 ok... what should I download? 23:06:16 woody? 23:06:20 sarge 23:06:28 No, Sarge 23:06:41 use the 'net install' 23:07:01 robin types faster, listen to him 23:07:03 basic bootup cd, then grabs the rest over the net as it needs it 23:09:09 I haven't tried out Webmin on debian yet, but it was somewhat useful in Mandrake - if you have to have a something Yast-like 23:10:55 Ray - basic FAQ has a few entries in it now, please read and make suggestions 23:11:54 I used Yast for some stuff... 23:12:32 rayh has quit 23:12:50 never used either. just apt-get install does it for me 23:13:33 yast is for config 23:13:42 users/software/hardware 23:13:49 you name it 23:13:49 it's all in one place 23:13:59 rayh has joined #emc 23:14:12 Two things I'd like to see ported from Mandrake to debian - diskdrake and (possibly) harddrake 23:14:12 ah yes 23:14:13 darn phone 23:14:30 I use a similar thing .. its called 'command line' ;) 23:15:05 robin: I use it too, from time to time 23:15:37 let me guess, but you prefer a GUI 23:15:54 do you monitor system load with hot_babe? 23:15:55 nooo 23:16:00 text mode ;) 23:16:02 Oooo! 23:16:17 hot_babe through the aa library? 23:16:57 yay ;) 23:17:04 I think the debian crowd told the guy who wanted to port hot_babe to debian to .. errm .. go away 23:17:13 aa = ascii art? 23:17:39 yeah 23:17:51 theres aaxine 23:17:59 for viewing mpegs in ascii 23:18:20 I've seen somebody hooking up a TV Tuner through aa 23:19:04 making stereograms in aa = too much time on your hands 23:19:45 heh 23:20:48 Don't laugh, I've seen it done 23:21:01 'course, everyone asks "What is it?" 23:21:06 chiao folks 23:21:09 Imperator_ has quit 23:22:14 and.. "what's it good for?" 23:22:54 eye strain 23:24:37 I use an old CRT for that ;) 23:30:39 w 23:31:44 hey anon 23:31:50 heard you got emc running... 23:43:27 yeah 23:43:31 it works quite nicely 23:43:37 but I havent configured my axis:es yet.. 23:43:47 just changed my port to match.. the default pinouts.. 23:44:06 but I need to get my dir outputs inverted 23:44:14 cool 23:44:21 glad to hear you got it running 23:44:29 yeah, but it's a bit away from working good yet :) 23:44:31 with hal it's pretty easy 23:44:49 how do you do that I tried to look at the hal pdf.. 23:45:10 but I didnt find out how.. 23:45:35 I tried addin out-inverted to the pins but it just made emc fail when starting 23:46:22 well.. you have a file which contains the conenctions made with hal 23:46:36 it's defined in emc.ini 23:46:41 standard_pinouts.hal isnt it? 23:47:06 in configs 23:47:07 yup 23:47:17 yeah that much I understand.. 23:47:24 now open that file and change parport.0.pin-03-out 23:47:52 yes.. 23:48:02 to parport.0.pin-03-inverted-out (I'm looking for the exact naming) 23:48:16 I tried changing it to out-inverted.. 23:48:17 :) 23:48:23 I think thats what it says in the pdf.. 23:50:38 and? 23:51:15 parport.%d.pin-%02d-out-invert 23:51:15 invert not inverted 23:51:19 ah 23:51:38 * anonimasu is starting to like hal 23:51:52 I looked at that stepgen thing earlier also.. 23:51:54 it seems neat 23:52:26 but I need to set up my units and things.. 23:52:33 and try to get any speed out of the machine,.. 23:53:11 rayh has quit 23:53:15 yes... 23:53:49 * alex_joni is gone sleeping 23:53:56 bye 23:53:58 yeah 23:53:59 bye. 23:54:17 alex_joni has left #emc 23:54:44 * robin_z is looking at a g2002 23:54:48 see you guys later 23:54:56 with a neat Linux driver by Steve Hardy 23:55:01 :) 23:55:01 nice 23:55:04 SWPadnos has quit 23:55:05 I am at a lan on my way home 23:55:12 heh 23:55:19 but I dont feel like sleeping 23:55:24 geek! 23:55:29 I feel like playing with my mill.. 23:56:07 robin_z: you know the people that call others geeks are as likely to be geeks as the people they call geeks.. 23:56:10 .d 23:56:13 :D 23:56:25 cnc-geeks *grins* 23:56:42 nah 23:56:47 no way. 23:56:54 i am MUCH more geeky than you! 23:57:08 haha 23:57:25 I have ... counts ... at least 6 linux boxes running here :) 23:57:36 heh, I just have my laptop here right now :/ 23:57:53 my latest is an ARP transparent bridge onthe outgoing link ;))) 23:58:06 nice firewall :) 23:58:13 I do that at thwork.. 23:58:43 bugger. 23:58:49 yeah, the bridge is a test run for when I put traffic shaping into my rack 23:59:02 :) 23:59:11 traffic shaping rocks ^_^ 23:59:19 oh yeah, I have a rack in a colo too. thats top geek stuff that :) 23:59:37 I run it at the network here.. to keep the people from using all bandwidth with BT 23:59:51 yep 23:59:57 I'm using Paul's .deb's for the tcl/tk/tclx