#emc-devel | Logs for 2011-06-14

[00:01:51] -!- theorbtwo has quit [Read error: Operation timed out]
[00:23:44] -!- crazy_imp has quit [Read error: Operation timed out]
[00:26:41] -!- chester88 has quit [Ping timeout: 240 seconds]
[00:38:38] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has parted #emc-devel
[00:42:24] <jepler> looks like the doc building machine hasn't built emc 2.5 docs since the recent merge of ngcgui. I got around to fixing the problem (locally missing package) just now.
[00:42:29] -!- chester88 [chester88!~chris@S010688ae1d61f51a.no.shawcable.net] has joined #emc-devel
[00:43:40] * jepler disappears again
[00:48:47] -!- servos4ever has quit [Quit: ChatZilla 0.9.85 [SeaMonkey 2.0.11/20101206162726]]
[00:54:12] <rayh> Someone asked in an email if there would be a presentation on the Architecture of EMC2 at the CNC workshop..
[00:54:54] <rayh> We had some drawings a while back. Do they exist someplace? Are they reasonably current?
[00:54:57] -!- Techrat_CT has quit [Ping timeout: 255 seconds]
[01:06:33] <JT-Shop> rayh: everything should still be there
[01:17:51] <rayh> Thanks. I didn't see them in the 2.4 manuals. I'll look around the wiki a bit.
[01:53:05] -!- Aggrav8d has quit [Ping timeout: 276 seconds]
[02:02:44] <cradek> the diagrams I've seen are pre-HAL
[02:03:36] <cradek> jepler: thanks!
[02:08:31] <rayh> Okay. Looks like I'll have to try to work something up.
[02:08:43] -!- Tom_L has quit []
[02:09:50] <cradek> http://linuxcnc.org/docs/2.5/html/config/emc2hal.html
[02:10:21] <cradek> following from this page, a diagram of how EMC hooks to the various parts of a real machine (through HAL) seems like it would be useful
[02:10:47] <rayh> looking
[02:11:53] <cradek> IMO that really is the overview of what an integrator has to understand
[02:13:09] <cradek> although ... I'm not sure a diagram could be clearer than this doc page
[02:13:53] <rayh> Yep.
[02:14:02] <cradek> do you know the general goals of the guy(s?) asking about an architectural overview? do they mean for contributing to development, or for integrating for a certain machine, or something else?
[02:15:09] <chester88> Hey Chris do you know if anyone had ideas why hostmot2 has some problems with some machines in 2.5 versus 2.4?
[02:15:32] <cradek> chester88: do you have more information than that?
[02:16:45] <chester88> What do you want to know. I have the problem. When I move I get following errors. I can hear missed steps as i jog sometimes
[02:16:46] <chester88> if i switch to 2.4 no problems
[02:16:59] <cradek> so you're talking about stepgen behavior?
[02:17:07] <chester88> yes
[02:17:34] <cradek> if you can hear missed/rough steps you should be able to capture it on halscope and ask for help on emc-devel
[02:17:50] <cradek> I don't use hostmot2 stepgen so I haven't seen/heard this
[02:17:52] <chester88> I have. I did
[02:18:15] <chester88> others have had this problem too according to the forum.
[02:18:25] <cradek> hm, how long ago was it on the list?
[02:18:57] <cradek> this might be very important - maybe a recap/reminder is needed?
[02:18:58] <chester88> first time i'm sure a year ago or more. second 6 months maybe.
[02:19:27] <chester88> maybe i will send to mail list again.
[02:19:43] <cradek> I remember a huge forum thread full of terrible misleading information
[02:20:16] <chester88> could be. i gave all information that was asked of me I am pretty sure.
[02:20:33] <chester88> including halscope.
[02:20:39] <cradek> I'm not blaming anyone, just trying to think of what you could do that might help
[02:20:54] <chester88> ok anyways just wasn't sure if anyone had come up with something.
[02:20:58] <cradek> there aren't a lot of people set up to test or work on it unfortunately
[02:20:59] <rayh> cradek: I suspect that the requester was fairly new to EMC2 and would like to have a visual representation of how the manor parts relate to each other.
[02:21:15] <rayh> major
[02:21:50] <chester88> What is an easy way to see what has changed with hostmot2 from 2.4 to 2.5 ? with git?
[02:22:41] -!- pcw_home [pcw_home!~chatzilla@ip-66-80-167-54.sjc.megapath.net] has joined #emc-devel
[02:22:54] <cradek> gitk v2.4_branch..v2.5_branch -- src/hal/drivers/mesa-hostmot2
[02:23:26] <chester88> k thanks
[02:23:42] <pcw_home> Chester88: is this possibly a ubuntu 8.04 vs 10.04 issue?
[02:24:11] <chester88> no . both 2.4 and 2.5 were on U 8.04
[02:24:18] <cradek> if you can reproduce it reliably and bisect it to identify the bad commit, that would be the best thing possible you could do to get it fixed
[02:24:50] <chester88> problem is i don't know what causes it.
[02:24:54] <pcw_home> Do you have accelertion limited?
[02:25:06] <cradek> you mean it's not consistent?
[02:25:13] <pcw_home> acceleration
[02:25:29] <chester88> no I mean there are alot of difference between 2.4 and 2.5
[02:25:55] <cradek> that's why you use bisect, to do a binary search of the relevant changes
[02:25:59] <pcw_home> do you have acceleration limited?
[02:26:07] <chester88> Peter can you send me a list of things you would need to know to help?
[02:26:38] <chester88> I will have to check Peter been a while since I delved into it.
[02:27:05] <pcw_home> based on my current theory:
[02:27:07] <chester88> I was using 2.4 so didn't bother me. Now i want 2.5 features
[02:27:07] <pcw_home> #1 is accelaration limited? (non 0)
[02:27:09] <pcw_home> #2 what is the jitter on both systems
[02:27:19] <chester88> same system
[02:27:33] <chester88> same configuration
[02:27:38] <chester88> just recompile
[02:27:39] <pcw_home> i should not even try to spell after one pale ale...
[02:28:11] <pcw_home> but latencies could be different
[02:28:34] <chester88> same computer
[02:28:35] <cradek> if same system/configuration you can use git bisect and find it in a few tests. there are < 32 relevant changes.
[02:29:08] <chester88> i would have to revert my 2.5 to before each commit?
[02:29:13] <cradek> that's like 5 builds
[02:29:23] <chester88> how do you do that?
[02:29:35] <cradek> http://kernel.org/pub/software/scm/git/docs/git-bisect.html
[02:30:15] <cradek> you'll start with 'git bisect start -- src/hal/drivers/mesa-hostmot2'
[02:31:01] <cradek> then give it known good and known bad refs, then test and report the results it asks for
[02:32:21] <chester88> ok I will read up and see what i can come up with. Thnaks.
[02:32:38] <pcw_home> chester88 do you know your jitter figures?
[02:32:58] <chester88> not off hand no
[02:33:27] <pcw_home> Thats seemed to be a common point with systems that had trouble
[02:33:34] <pcw_home> that
[02:34:44] <cradek> pcw_home: you think large jitter + unlimited accel is the magic combination?
[02:35:07] <chester88> what do you consider large jitter?
[02:35:54] <cradek> not sure if you're asking me, but personally I think > 25k is large
[02:36:54] <chester88> ok my jitter is not that large
[02:37:20] <chester88> i believe around 9000
[02:38:11] <pcw_home> Yes, I think the basic stepgen software architecture (from the software stepgen) is not really appropriate for a hardware step gen
[02:38:13] <pcw_home> I also think the problem with the base thread causing stepgen trouble is related.
[02:38:14] <pcw_home> chester88 can you check again under 2.5?
[02:38:16] <pcw_home>
[02:38:46] <cradek> if a change in the hostmot2 directory of the emc source caused it, and you can run in 2.4 and 2.5 with the same configuration, you will find it in about 5 build/test cycles
[02:39:15] <chester88> acceleration is limited to 6 inches per second squared
[02:39:27] <chester88> 3 on another.
[02:39:29] <cradek> someone who can do the methodical testing is what we need
[02:39:48] <chester88> yes I will take a stab at it latter.
[02:40:46] <pcw_home> which card is this?
[02:40:53] <chester88> 5i20
[02:40:59] <chester88> two actually
[02:41:12] <cradek> can you share the complete config you are using to test in the two versions?
[02:41:22] <chester88> yes
[02:41:59] <chester88> but not at this moment. I will test and share the config and findings later.
[02:42:03] <cradek> ok, cool
[02:42:10] <pcw_home> Can you try a different card? threr may also be a issue with clock accuracy that is not handled properly
[02:42:14] <chester88> Or at least try :)
[02:42:25] <pcw_home> there may be
[02:42:35] <chester88> Yes I could try that too
[02:42:48] <chester88> ok this is what i needed a plan!
[02:43:04] <cradek> pcw_home: but if 2.4 vs 2.5 is REALLY what makes the difference...?
[02:44:18] <pcw_home> another option is to use the stepgen in velocity mode and add the PID loop
[02:44:20] <pcw_home> Not sure about 2.4/2.5 I didn't even know anything changed i n the stepgen software
[02:44:56] <cradek> there were 3 or so very minor changes in stepgen.c between 2.4/2.5
[02:45:22] <pcw_home> (velociy mode and PID loop basically skips the whole stepgen control mess)
[02:45:48] <cradek> our goal should be to determine whether 2.5 stepgen is broken, and if so, fix it
[02:46:09] <cradek> I don't care one bit about a workaround that makes chester88's machine work :-)
[02:46:28] <cradek> (he may feel different!)
[02:46:32] <chester88> lol nice. Actually either do I
[02:46:53] <chester88> I can make 2.4 do most of what i wanted.
[02:47:03] <chester88> anyways ttyl
[02:47:07] <cradek> good luck testing
[02:49:02] <pcw_home> I waan't thinking so much of a workaround as a basis for better driver software (more robust vs jitter for example)
[02:49:34] <cradek> I'd be really sad if we couldn't have position mode stepgen in hm2
[02:50:40] -!- JT-Shop has quit [Read error: Connection reset by peer]
[02:51:09] -!- JT-Shop [JT-Shop!~chatzilla@216-41-156-49.semo.net] has joined #emc-devel
[02:51:17] <cradek> otoh, it might give us index homing
[02:51:32] <pcw_home> I'm just suspicious that its a basic driver error that shows up in certain situations (base thread triggers it, excessive jitter triggers it)
[02:52:02] <cradek> I see - you mean like a fundamental design problem
[02:56:34] <pcw_home> Yes, the driver was basically taken from the software stepgen, but ther are some things thats are quite different, for example
[02:56:36] <pcw_home> 1. The software stepgen gen no long term timebase errors but the hardware one does (its clock is not the thread clock)
[02:56:38] <pcw_home> 2. The hardware stepgen has excellent short term timing, so that errors in the contro loop (mainly from thread jitter)
[02:56:39] <pcw_home> should be passed through a long time constant low pas filter (in other words for high frequencies believe the hardware, for low the thread
[02:57:06] <pcw_home> no more beer for me
[03:12:28] -!- ve7it has quit [Remote host closed the connection]
[03:22:45] -!- Birdman3131 has quit [Read error: Connection reset by peer]
[03:26:32] <KimK> chester88: I have never used git bisect before yesterday/today, and it was easy. It does almost everything for you. Three to start: "git bisect good nnnnnn", "git bisect bad nnnnnn", "git bisect start". Then it leads you through, you build after each bisect and report the good|bad results with "git bisect good|bad". When done, "git bisect reset" puts everything back the way it was.
[03:28:10] -!- ries has quit [Quit: ries]
[03:30:58] <SWPadnos> you can even tell it to run a script for you, to check whether it was a good or bad build
[03:31:41] <KimK> Really? Those git guys thought of everything, lol.
[03:32:11] <SWPadnos> many things, yes they did
[03:33:17] <KimK> I can see where that would be handy, if something big had to run overnight, or over a weekend, or some such.
[03:37:05] <SWPadnos> or if you can automate it anyway - might as well let the computer do all the work
[03:58:43] -!- ries has quit [Quit: ries]
[04:37:13] -!- davidf has quit [Ping timeout: 240 seconds]
[04:49:41] -!- Calyp has quit [Ping timeout: 240 seconds]
[05:06:11] -!- vladimirek [vladimirek!~vladimire@bband-dyn79.178-41-183.t-com.sk] has joined #emc-devel
[05:47:17] -!- scanf has quit [Ping timeout: 276 seconds]
[05:58:44] -!- vladimirek has quit [Remote host closed the connection]
[06:36:36] -!- mk0 has quit [Quit: Ухожу я от вас (xchat 2.4.5 или старше)]
[06:59:40] -!- e-jones [e-jones!~jkastner@ip-84-242-102-36.net.upcbroadband.cz] has joined #emc-devel
[07:16:35] -!- psha[work] [psha[work]!~psha@] has joined #emc-devel
[07:22:40] -!- awallin [awallin!~quassel@cs27057046.pp.htv.fi] has joined #emc-devel
[07:43:01] <KimK> Hi psha, have you got any troubleshooting advice for me on the docs build process? I've done a lot of editing on the 2.5 docs, but of the 5 English and 4 French docs, I'm only getting 4 English and 0 French PDFs. Any advice about where to look for troubleshooting info would be helpful. I'm saving the outputs of "make" to a log file for review. Anything else I can do? You're at work and I'm going to sleep soon. Any advice appreciated. Thanks.
[07:45:52] <KimK> Oh, I guess I should ring the doorbell. Here goes: psha[work]
[07:48:50] -!- pjm__ has quit [Ping timeout: 276 seconds]
[07:49:51] -!- robh__ [robh__!~robert@5ace7020.bb.sky.com] has joined #emc-devel
[07:54:56] -!- scanf has quit [Quit: Leaving]
[07:55:30] -!- Dannyboy has quit [Remote host closed the connection]
[08:16:54] -!- mhaberler [mhaberler!~mhaberler@extern-182.stiwoll.mah.priv.at] has joined #emc-devel
[08:33:02] -!- awallin_ [awallin_!~quassel@2001:708:110:1020:224:7eff:feda:7c7d] has joined #emc-devel
[08:39:17] -!- rooks has quit [Read error: Operation timed out]
[08:48:46] <psha[work]> KimK: request pending...
[09:03:10] -!- mhaberler has quit [Quit: mhaberler]
[09:08:10] -!- mhaberler [mhaberler!~mhaberler@macbook.stiwoll.mah.priv.at] has joined #emc-devel
[09:36:17] -!- ewidance [ewidance!~ewidance@LPuteaux-156-15-52-242.w82-127.abo.wanadoo.fr] has joined #emc-devel
[09:37:34] -!- ewidance [ewidance!~ewidance@LPuteaux-156-15-52-242.w82-127.abo.wanadoo.fr] has parted #emc-devel
[10:57:27] -!- kljsdfhklj has quit [Quit: kljsdfhklj]
[11:27:08] -!- crazy_imp has quit [Quit: Reconnecting]
[11:52:51] -!- Birdman3131 has quit [Ping timeout: 255 seconds]
[12:16:10] -!- mhaberler has quit [Quit: mhaberler]
[12:47:25] -!- factor has quit [Quit: Leaving]
[12:58:49] -!- kljsdfhklj has quit [Quit: kljsdfhklj]
[13:04:47] -!- izua has quit [Ping timeout: 244 seconds]
[13:04:57] -!- acemi has quit [Read error: Operation timed out]
[13:17:59] -!- mk0 has quit [Quit: поздравляю всех с кредитом. халява продолжается.]
[13:28:22] -!- izua has quit [Changing host]
[13:37:08] -!- Valen has quit [Quit: Leaving.]
[13:47:46] -!- El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]
[14:08:20] -!- awallin_ has quit [Remote host closed the connection]
[15:06:56] -!- nullie has quit [Quit: Ex-Chat]
[15:13:20] -!- scanf has quit [Ping timeout: 260 seconds]
[15:14:04] -!- psha[work] has quit [Quit: leaving]
[15:36:18] Tech_Talk is now known as Birdman3131
[15:46:51] -!- mhaberler [mhaberler!~mhaberler@macbook.stiwoll.mah.priv.at] has joined #emc-devel
[15:52:38] -!- L84Supper has quit [Quit: puff of smoke]
[15:52:57] -!- syyl has quit [Ping timeout: 252 seconds]
[16:27:51] -!- Connor [Connor!~Connor@] has parted #emc-devel
[16:30:02] -!- Connor [Connor!~Connor@] has joined #emc-devel
[16:46:47] -!- motioncontrol has quit [Quit: Sto andando via]
[16:48:42] -!- psha [psha!~psha@] has joined #emc-devel
[16:53:27] -!- ve7it [ve7it!~LawrenceG@S0106009027972e37.pk.shawcable.net] has joined #emc-devel
[17:06:35] -!- mrsunshine_ has quit [Ping timeout: 244 seconds]
[17:08:24] -!- syyl_ has quit [Ping timeout: 252 seconds]
[17:09:26] -!- elmo40 has quit [Ping timeout: 260 seconds]
[17:13:52] -!- nullie has quit [Quit: Ex-Chat]
[17:50:15] -!- n2diy has quit [Read error: Connection reset by peer]
[17:50:25] -!- IchGuckLive has quit [Quit: ChatZilla 0.9.86 [Firefox 3.6.3/20100423140709]]
[17:52:01] -!- Loetmichel has quit [Ping timeout: 240 seconds]
[17:55:54] -!- vladimirek [vladimirek!~vladimire@bband-dyn79.178-41-183.t-com.sk] has joined #emc-devel
[17:56:03] -!- mrsunshine__ has quit [Quit: Leaving]
[18:37:06] -!- isssy has quit [Quit: Visitor from www.linuxcnc.org]
[18:42:32] -!- andypugh [andypugh!~andy2@cpc2-basl1-0-0-cust1037.basl.cable.virginmedia.com] has joined #emc-devel
[19:12:29] -!- MarkusBec has quit [Ping timeout: 240 seconds]
[19:12:42] -!- lilalinux has quit [Ping timeout: 252 seconds]
[19:28:59] -!- cassmodiah has quit [Ping timeout: 250 seconds]
[19:34:38] -!- skunkworks_ [skunkworks_!~chatzilla@184-158-0-129.dyn.centurytel.net] has joined #emc-devel
[19:52:27] -!- cpresser has quit [Read error: Operation timed out]
[19:54:10] -!- cpresser [cpresser!~cpresser@rstenpresser.de] has joined #emc-devel
[20:13:07] -!- e-jones has quit [Quit: Ex-Chat]
[20:42:43] <JT-Shop> cradek: do you need someone to host git while your in transit?
[20:43:22] <cradek> JT-Shop: no, I can move it in an orderly way
[20:43:40] <cradek> I will need help from SWP to update DNS when I'm ready
[20:43:56] <JT-Shop> ok, just wanted to check
[20:43:58] <cradek> I expect just a bit of downtime
[20:43:59] <cradek> thanks
[20:44:49] <andypugh> Anyone reproduced my G43 wierdness?
[20:46:24] <cradek> I haven't tried yet, sorry
[20:48:49] -!- awallin has quit [Read error: Connection reset by peer]
[20:51:20] -!- acemi has quit [Ping timeout: 244 seconds]
[20:57:46] -!- OoBIGeye has quit [Remote host closed the connection]
[21:28:18] -!- psha has quit [Quit: Lost terminal]
[21:30:43] -!- FinboySlick has quit [Quit: Leaving.]
[21:32:04] -!- scanf has quit [Ping timeout: 258 seconds]
[21:38:44] <KimK> During what dates is git expected to be down?
[21:38:47] -!- Birdman3131 has quit [Read error: Connection reset by peer]
[21:40:01] -!- syyl__ has quit [Quit: Leaving]
[21:40:10] <KimK> Also, maybe send out one email to the lists when git goes down, and send one when git comes back, just to be sure.
[21:42:23] <KimK> andypugh: Shall I try it on lathe sim, or is that insufficiently helpful?
[21:54:29] -!- Fox_Muldr has quit [Ping timeout: 276 seconds]
[21:54:42] -!- vladimirek has quit [Remote host closed the connection]
[22:10:24] -!- JT-Hardinge [JT-Hardinge!~chatzilla@216-41-156-49.semo.net] has joined #emc-devel
[22:11:42] <JT-Hardinge> andypugh: I'm at my lathe, what was the steps to reproduce the G43 glitch?
[22:28:40] <andypugh> Well, putting a tool change and G43 on the same line in G-code (not MDI) seems to do it for me.
[22:28:58] <andypugh> KimK: I saw the problem on 2.4.6 Lathe Sim
[22:29:12] <andypugh> (there needs to be an offset in the tool table)
[22:32:40] <andypugh> more details in my email to the dev list
[22:32:40] <andypugh> ]
[22:35:18] <JT-Hardinge> is there a link to the email, I don't get it out here in the shop
[22:35:58] <JT-Hardinge> I do Tn M6 G43 in my programs normally with 2.5
[22:51:57] -!- WalterN has quit [Quit: Leaving]
[22:55:51] -!- nullie has quit [Quit: Ex-Chat]
[22:57:31] -!- mhaberler has quit [Read error: Operation timed out]
[23:08:00] -!- mhaberler [mhaberler!~mhaberler@extern-182.stiwoll.mah.priv.at] has joined #emc-devel
[23:21:40] -!- atom1 has quit [Quit: Leaving]
[23:52:47] -!- elmo40 has quit [Ping timeout: 276 seconds]
[23:53:23] -!- mhaberler has quit [Quit: mhaberler]
[23:54:16] -!- mhaberler [mhaberler!~mhaberler@extern-182.stiwoll.mah.priv.at] has joined #emc-devel
[23:58:26] -!- El_Matarife has quit [Quit: Nettalk6 - www.ntalk.de]