#emc | Logs for 2005-11-13

[00:03:45] <SWPadnos> SWPadnos is now known as SWP_Away
[00:08:53] <k4ts> hello
[00:09:00] <k4ts> hi les_w_away
[00:21:24] <les_w_away> hi anna
[00:22:37] <les_w_away> Blah. I'm supposed to be off. Get calls from corporate.They want to do gant charts on my research and stuff. Fsck 'em.
[00:22:39] <k4ts> :-)
[00:23:05] <k4ts> ok
[00:23:10] <les_w_away> haha
[00:25:32] <Jymmm> wtf is a gant chart?
[00:26:08] <les_w_away> basically, a schedule. They want to schedule when i will be innovative.
[00:26:44] <Jymmm> lol
[00:26:47] <Jymmm> fuck em
[00:27:29] <les_w_away> you know, I will be innovative when I fscking feel like it. But, I have to be nice.
[00:27:59] <Jymmm> why?
[00:28:18] <les_w_away> Because i like the color of their money.
[00:28:34] <Jymmm> Oh like they are going to go elsewhere.... please...........
[00:28:55] <les_w_away> It's SOP for the corporate world
[00:29:04] <les_w_away> Guess I'm still in it
[00:29:28] <Jymmm> les_w_away Just simply say... I don't do schedules or timelines.
[00:29:57] <Jymmm> or.... schoduling and timelines will increase my fee 200%
[00:30:14] <les_w_away> It's a game jymmm. The want miracles fsat. I want to milk them for every cent in their deep pockets.
[00:30:15] <Jymmm> PITA Tax
[00:30:20] <les_w_away> It's just life.
[00:30:35] <Jymmm> lol
[00:30:46] <les_w_away> fast not fsat
[00:31:08] <les_w_away> can't type when I'm pissed
[00:31:19] <les_w_away> actually, can't type anyway!
[00:31:30] <Jymmm> tis ok, it's irc
[00:31:36] <Jymmm> not a gant chart
[00:31:37] <les_w_away> oh, ok.
[00:31:42] <les_w_away> ;)
[00:32:45] <les_w_away> I swear Anna I don't usually cuss.
[00:32:52] <Jymmm> les_w_away: I want to order one of thsoe plastic bits, but fuck me... $30/ea and *I* am not sure it'll resolve the chipload issues I'm having (even if the Eng said it would).
[00:33:51] <les_w_away> Well a $60 bit lasts about 2-4 hrs for me.but I can resharpen a few times
[00:34:19] <Jymmm> But you know they work as you need. That's the main difference.
[00:34:33] <les_w_away> yeah
[00:34:33] <Jymmm> I could spend $30 and not work at all.
[00:34:47] <Jymmm> not like I can return it.
[00:34:54] <les_w_away> still trying acrylic and PC?
[00:35:42] <Jymmm> I want to try acrylic, but I know I need different tooling. PC I havne't tried yet (still can't find the piece I have around here).
[00:35:56] <Jymmm> Plus I have an order for some signs
[00:36:22] <Jymmm> had to go buy a bunch of paint today..... mmmmmm purrrrty colors
[00:36:39] <Jymmm> pewter, copper, and perle white
[00:36:43] <les_w_away> We use one shot sign paints
[00:36:52] <Jymmm> one shot?
[00:37:00] <les_w_away> yeah a brand.
[00:37:06] <Jymmm> ah
[00:37:27] <les_w_away> lasts about 3-5 years in direct sun
[00:37:38] <Jymmm> plus PU
[00:37:53] <Jymmm> thought the PU has no UV protection
[00:38:08] <les_w_away> you bought some PU? I get that for free.
[00:38:27] <Jymmm> les_w_away say what?
[00:38:29] <Jymmm> free
[00:38:32] <les_w_away> actually...PU is not good in UV. It degrades quicky
[00:38:57] <Jymmm> really? diidnt know that. Alternative?
[00:39:48] <les_w_away> Acrylics are uv transparent...so they don't degrade much. Trouble is the substate might because it gets to it
[00:40:54] <les_w_away> UV inhibitors work well, but they are of limited lifetime because they protect by being sacraficial.
[00:41:09] <les_w_away> sacrificial
[00:41:13] <les_w_away> dammit
[00:41:19] <Jymmm> lol, damned if I do, damned if I don't.
[00:42:06] <les_w_away> metal flake and colloidal iron oxides are the best.
[00:42:52] <k4ts> night all
[00:43:01] <K`zan> gn es 73
[00:43:17] <Jymmm> never heard of them
[00:44:58] <les_w_away> google em. What kzan?
[00:45:38] <les_w_away> wondering what gn es 73 could possibly mean.
[00:46:01] <Jymmm> gn goodnight ?
[00:46:13] <les_w_away> and the es?
[00:46:40] <Jymmm> 73 well, that's a ham radio greeting here in the us
[00:46:49] <les_w_away> I know that
[00:46:52] <les_w_away> but the es
[00:46:52] <Jymmm> no clue on the es
[00:47:12] <K`zan> Anyone here got a clue about the new PCIx video cards?
[00:47:19] <les_w_away> mystery woman kzan just won't tell us.
[00:47:25] <les_w_away> huh?
[00:47:33] <K`zan> Oh, sorry es == and
[00:47:41] <les_w_away> ohhhh
[00:47:42] <Jymmm> ah
[00:47:49] <K`zan> Yes, gn == Good night.
[00:48:17] <les_w_away> Don't know about pcix
[00:48:17] <K`zan> I think one needs a new MB to use the new PCIx video cards, but nowhere that I looked does it say that.
[00:48:24] <K`zan> Me either :)
[00:48:30] <Jymmm> PCI Express
[00:48:34] <K`zan> Yes
[00:48:43] <les_w_away> video cards arfe basically driven by FS for me.
[00:48:44] <K`zan> seen uit abbreviated PCIx
[00:49:08] <les_w_away> arfe? I am now a dog?
[00:49:09] <K`zan> Yep, same here, the old ATI 9600 Pro/256 is OK, but getting long in the tooth...
[00:49:44] <K`zan> It looks like the Nvidia 6600GT is a pretty good card for the $$$.
[00:50:07] <les_w_away> I need a new card.
[00:50:41] <les_w_away> kinda looking at airplanes too...in that case I would not need a card as much.
[00:52:03] <les_w_away> Les needs a Warrior or Archer.
[00:52:23] <K`zan> LOL
[00:52:27] <les_w_away> Not a Robinson I think...would only live a few weeks.
[00:52:36] <K`zan> I'd be happy with a PA150 :)
[00:52:50] <les_w_away> Have a few hours in those.
[00:52:55] <les_w_away> Spin good.
[00:53:04] <Jymmm> les_w_away: You can get me one of these for christmas ---> http://www.bombardier.com/index.jsp?id=3_0&lang=en&file=/en/3_0/3_2/3_2_1/3_2_1.jsp
[00:53:17] <K`zan> LOL, I'll take one too :)
[00:53:21] <les_w_away> oh pa not cessna...oops
[00:53:57] <K`zan> That would be OK too :), I'd be so easy to please.
[00:54:14] <K`zan> ACtually if I had a choice a B47 or B206 would be nice :)
[00:54:29] <K`zan> ANYTHING :)
[00:54:36] <les_w_away> My heater product is type approved in the lear 50
[00:54:46] <K`zan> Cool
[00:54:55] <K`zan> Very cool
[00:55:02] <les_w_away> Had to go to some FAA meetings
[00:55:08] <les_w_away> It was not too hard
[00:55:10] <K`zan> That had to suck
[00:55:31] <les_w_away> Naw they were pilots too...we just talked
[00:55:40] <les_w_away> no problem
[00:55:46] <K`zan> Not beaurocrap administrators, that is good.
[00:55:51] <les_w_away> yup
[00:56:10] <K`zan> Get the bureaucraps just before the lawyers :)
[00:56:54] <les_w_away> The worst are the aeromedical group
[00:57:00] <les_w_away> for me at least.
[00:57:12] <les_w_away> my eyes and reflexes are going some.
[00:58:00] <K`zan> They do tend to be, what appears to me, a bit pickier than necessary...
[00:58:41] <les_w_away> Ultimately...guess I will end up a blind inanimate unresponsive blob.
[00:59:00] <K`zan> We all do :-( in time :-(
[00:59:25] <les_w_away> Take aspirin regularly? Put it on the FAA form? You are grounded.
[00:59:37] <K`zan> as noted, a BIT overboard.
[00:59:48] <K`zan> </sarcasm>
[00:59:53] <les_w_away> Those...those...
[01:00:05] <les_w_away> ok no more cussin from me tonight.
[01:00:25] <K`zan> :-) Feel free :-)
[01:00:30] <Jymmm> les_w_away till drunk as a skunk
[01:01:00] <Jymmm> or you smell like one. whatever comes first
[01:01:04] <les_w_away> oh it's just those corporate boobs.
[01:01:44] <les_w_away> ants.
[01:02:17] <Jymmm> boob ants?
[01:02:23] <Jymmm> ant boobs?
[01:02:27] <Jymmm> =)
[01:02:39] <les_w_away> jymmm, ants don't have boobs.
[01:04:06] <Jymmm> won't know.... never looked =)
[01:04:15] <les_w_away> ok, I'll slap myself in the face.
[01:04:20] <les_w_away> now I'm ok.
[01:04:40] <K`zan> LOL
[01:04:46] <les_w_away> Hi folks...I love everyone. May I be of service?
[01:04:54] <Jymmm> is it JUST the scheduling, or something else?
[01:05:04] <K`zan> Too many pissants have to be impotent errr important...
[01:05:37] <K`zan> Got a good haul at the thrift store today.
[01:05:52] <les_w_away> nah jymmm, I am just a bit of a wild spirit and also involed in corporate stuff.
[01:05:56] <Jymmm> K`zan: two blue haired old ladies for the price of one?
[01:06:02] <les_w_away> Incompatible.
[01:06:04] <K`zan> Tub full of goodies for $2.17 (after gov bloodsucking).
[01:06:20] <les_w_away> What did you get?
[01:06:35] <K`zan> Sega Dreamcast, Sony Playstation, bunch of decent mice, some heatsinks and fans.
[01:06:46] <K`zan> One of the fans was what I really wanted :)
[01:06:59] <les_w_away> At a thrift store?
[01:07:11] <K`zan> Playstation is prolly really ancient as is the Dreamcast, neither one probably worth keeping.
[01:07:33] <K`zan> Yes, scabattle errr seattle has some really decent thrift stores.
[01:07:51] <les_w_away> I worked at a surplus electronics store during the summer when I was inschool.
[01:07:57] <les_w_away> It was kinda fun.
[01:08:14] <K`zan> I Imagine, you get to think about what you can use all day :)
[01:08:38] <les_w_away> It was near the cape so we got some neat stuff.
[01:08:57] <K`zan> Cool, lived down there about 40 years ago now.
[01:09:04] <les_w_away> cape canaveral that is.
[01:09:17] <K`zan> Cocoa Beach, Melbourne, Titusville, Bahama Beach...
[01:09:59] <K`zan> My sister in law worked for NASA, got passes for all the launches and some really neat day-of-launch envelopes with mission patch.
[01:10:03] <les_w_away> I lived in Palm bay and worked at Harris as a photolithography engineer.
[01:10:13] <les_w_away> I grew up in Orlando.
[01:10:16] <K`zan> Good ole Palm Bay :)
[01:10:39] <K`zan> Ahh, orlands, still probably the most stop light populous city in the world :)
[01:10:43] <K`zan> orlando
[01:11:07] <les_w_away> We used to go to the "printed circuit inn" after work to have a few
[01:11:15] <K`zan> :-)
[01:11:36] <K`zan> Miss Bill & Bettys
[01:11:40] <les_w_away> or drunken Shrimping on the Indian river
[01:11:49] <K`zan> LOL, done my share of that.
[01:11:58] <les_w_away> mather's bridge.
[01:12:07] <K`zan> Never got enough, hell, there is no such thing as enough shrimp :)
[01:12:31] <les_w_away> I would drink 6 beers and get one shrimp.
[01:12:53] <K`zan> Compared to the east coast the west, at least here in WA, seafood is an obscure food.
[01:13:13] <K`zan> LOL, I like shrimp better than beer, even decent homebrew :)
[01:13:19] <les_w_away> One time one of the girls was hospitalized. Sand flies.
[01:13:34] <K`zan> Nasty little buggers.
[01:13:40] <les_w_away> yeah.
[01:14:21] <K`zan> Never liked stepping on stingrays either, once was QUITE enough.
[01:14:31] <les_w_away> Well, I ought to go to the music room. I could play my sitar, or listen to stuff.
[01:15:02] <K`zan> I'm trying to decide whether to go back down to the thrift store and spend $15 on a box of peripheral for the Dreamcast.
[01:15:13] <les_w_away> why not
[01:15:28] <K`zan> Don't know the first thing about game consoles.
[01:15:38] <les_w_away> I just bought 150 lb of aluminum remnants today
[01:15:41] <K`zan> Last game console I had was an odessy :-)
[01:16:13] <K`zan> Besides pissing away $$$ on game crap delays the filling of of milk can for the mill :)
[01:16:39] <les_w_away> I just do FS. I have Orbiter too.
[01:17:01] <K`zan> Need to read the manual on orbiter, not worth a damn with it,
[01:17:13] <K`zan> did you see the shuttle challenge for FS9?
[01:17:50] <les_w_away> it's hard to reach low earth orbit by seat of the pants...much harder to do entry at a specific point
[01:18:01] <les_w_away> no did not see
[01:19:01] <les_w_away> on orbiter...I burn at apoalsis....try to make 30 km at about texas
[01:19:16] <les_w_away> to hit florida
[01:19:57] <K`zan> Search simviation for it, supposed to be pretty good. Someday I'll boot winditz again and try it :). Now that I have X-Plane running under linux, I do tend to use it.
[01:20:03] <les_w_away> I usually miss by a couple hundred km
[01:20:18] <les_w_away> X plane has a good flight model
[01:20:20] <K`zan> LOL, I don't even get out of orbit :-( <blush>
[01:20:39] <les_w_away> FS is harder than the real thing.
[01:21:05] <K`zan> Yes, it does, but that is about all. Not so crazy about the eye candy and that models in one version don't work in the next, but it is good for flight model.
[01:21:24] <K`zan> Yep, no "grab ya by the arse" feel to it at all...
[01:21:35] <K`zan> Esp with helos, which is my pref.
[01:22:03] <les_w_away> As you know, I was looking at Robinson R22s
[01:22:33] <les_w_away> I have concluded that the thing would kill me real quick
[01:22:56] <les_w_away> but it sure would look nice parked in the yard.
[01:24:21] <les_w_away> Anyway a rotorcraft rating added to my certificate would take at least 30 hours
[01:24:33] <les_w_away> then I would kill myself.
[01:24:49] <K`zan> LOL, I doubt that.
[01:25:17] <K`zan> Not an affectionado of the Robinson stuff, around here they seem to fall out of the sky too often.
[01:25:33] <K`zan> Want an old Bell 47 :-)
[01:25:41] <les_w_away> I checked...lose engine...you have one second to drop the collective. Or you're dead.
[01:25:42] <K`zan> Or OH-23
[01:25:54] <les_w_away> I had engine failure inan Archer.
[01:25:54] <K`zan> That tight? Wow
[01:26:21] <K`zan> But still, first thing to do is drop collective, but a sec is pretty tight.
[01:26:30] <les_w_away> I didn't even know it it for 15 sec or so!
[01:26:38] <K`zan> Ouch
[01:26:51] <K`zan> Helos sure as hell are a challenge.
[01:27:13] <les_w_away> prop was windmilling...didn't know it till the next time I added power
[01:27:26] <les_w_away> and not a lot happened.
[01:27:26] <CIA-5> 03fenn * 10emc2/src/hal/utils/halcmd.c: added linkpp and exit commands, condensed help overview
[01:28:15] <les_w_away> Had plenty of airspeed though
[01:28:20] <les_w_away> no problem
[01:28:25] <les_w_away> made the field
[01:28:31] <K`zan> Yep, f/w are a LOT more forgiving. You didn't notice the noise stopped?
[01:28:41] <les_w_away> in a robinson I would have been a rock
[01:28:56] <K`zan> You would have been a gooey pancake...
[01:28:58] <les_w_away> no, i didn't
[01:29:14] <K`zan> Must have a damn GOOD muffler on those.
[01:29:27] <les_w_away> I was at about 1800...throttled back already
[01:29:33] <les_w_away> on downwind
[01:29:33] <K`zan> Ah
[01:30:01] <les_w_away> 1800 rpm
[01:30:11] <K`zan> Makes sense now.
[01:30:27] <K`zan> At idle with just the prop spinning it would be quiet.
[01:30:46] <les_w_away> wanted to add a little manifold pressure on base to final. Nothing.
[01:31:22] <les_w_away> I aborted base and headed straight for the field at a 45
[01:31:26] <les_w_away> no problem
[01:31:43] <les_w_away> I even had to slip it in a little...I was hot.
[01:32:44] <les_w_away> ate up a lot of runway.
[01:33:08] <K`zan> Good choice
[01:33:09] <les_w_away> squeeked about halfway down
[01:33:10] <K`zan> :-)
[01:33:32] <K`zan> Any one you can walk away from ... :-)
[01:33:40] <les_w_away> it's just what i was taught...worked...
[01:35:23] <les_w_away> well, i'm off to the music room. 'll forget about corporate BS.
[01:35:29] <les_w_away> later!
[01:45:09] <Jymmm> G'Night ls
[06:09:42] <CIA-5> 03paul_c * 10emc2-auto/wiki/ (8 files in 7 dirs): "Auto update wiki from a cron job. Sat Nov 12 05:30:01 GMT 2005 "
[06:25:34] <fenn> yo bobby
[06:26:08] <fenn> made anything that works yet?
[06:53:34] <_dreamer> hi, how are you? and the answer is sadly no, just dreaming still
[06:53:41] <_dreamer> getting my new home ready
[06:54:36] <_dreamer> have you got your hexapod up and working?
[07:07:20] <fenn> no.. "The ability to do nothing is very powerful, but it must not be abused!"
[07:07:55] <_dreamer> lol
[07:07:55] <fenn> i spent a couple months in oregon riding my bike and rasslin bears 'n such
[07:08:00] <_dreamer> whoa
[07:08:07] <_dreamer> that's not doin' nothing
[07:08:12] <_dreamer> sounds darn tootin'!
[07:08:22] <_dreamer> what lives should be made out of
[07:08:44] <fenn> well, riding your bike every day from nowhere to nowhere gets boring after a couple weeks
[07:09:34] <fenn> right now i am merging my old bookmarks into my current bookmarks
[07:09:47] <fenn> i have way too many bookmarks
[07:12:58] <fenn> it looks better in xml, but here is what i've done so far in html: http://fenn.freeshell.org/bookmarks.html
[09:23:57] <fenn> fun with the wife, huh
[09:55:48] <alex_joni> morning
[09:57:49] <fenn> mornin
[09:58:02] <fenn> i'm bored and lazy - gimme somethin to do
[09:58:23] <alex_joni> hrmm.. lots to do ;)
[09:58:28] <alex_joni> you got emc2?
[09:58:36] <fenn> of course
[09:58:42] <alex_joni> ok.. try make install
[09:58:47] <alex_joni> and see how that suits you..
[09:59:31] <alex_joni> first look at it, and ask what isn't clear
[10:01:49] <fenn> why are realtime modules inserted at boot time?
[10:01:57] <alex_joni> they are not
[10:02:11] <alex_joni> but the realtime script gets installed in /etc/init.d/
[10:02:20] <alex_joni> so you can install them at boot-time.. if you like
[10:02:39] <alex_joni> anyways.. the realtime stuff should be PC-related
[10:02:46] <alex_joni> not emc-related
[10:03:52] <fenn> heh not much for uninstall
[10:03:59] <alex_joni> nope
[10:04:19] <alex_joni> before I do that, I want make install to work properly
[10:04:35] <alex_joni> if you mess with make install, use install.run
[10:04:39] <alex_joni> not emc.run
[10:05:47] <fenn> ever heard of checkinstall?
[10:05:58] <alex_joni> nope.. what's it do?
[10:06:22] <fenn> uses a make install script to generate an rpm/deb/slack package, i think
[10:06:50] <fenn> " Some times - very few - the program's
[10:06:50] <fenn> author adds an "uninstall" rule to the Makefile, but that's not usually the
[10:06:50] <fenn> case. That's my primary reason to write checkinstall."
[10:07:06] <alex_joni> heh
[10:07:55] <alex_joni> sounds goo
[10:07:57] <alex_joni> good
[10:10:40] <alex_joni> but that means.. add just another dependency
[10:10:48] <alex_joni> or include the code.. which is just as bad
[10:11:00] <fenn> i was thinking of using it as a means of making rpms
[10:11:11] <alex_joni> I see.. well, why not
[10:11:14] <fenn> and so i can remove your goobery install after it screws up my disk :)
[10:11:23] <alex_joni> btw, there is a branch with work on rpm stuff
[10:11:38] <fenn> i dont really get how branches work in cvs
[10:11:49] <alex_joni> you need to checkout with an aditional flag
[10:12:00] <alex_joni> cvs checkout -r foo_branch emc2
[10:29:52] <fenn> http://pastebin.com/426424
[10:30:36] <alex_joni> looks ok ;)
[10:30:44] <alex_joni> except the RPM build which fails
[10:32:00] <fenn> ah it wants a version number
[10:32:05] <alex_joni> you could sort that CVS file sort out..
[10:32:10] <alex_joni> folder
[10:32:25] <alex_joni> * alex_joni didn't succeed..
[10:32:30] <alex_joni> well I didn't try hard enough
[10:32:35] <fenn> er, what?
[10:32:42] <fenn> "sort that cvs file sort out"?
[10:32:49] <alex_joni> well.. I do some cp foo/*
[10:33:02] <alex_joni> and that * expands to the CVS folders inside the checkout
[10:33:10] <fenn> ah
[10:33:12] <alex_joni> and I don't want those getting installed
[10:33:16] <fenn> right
[10:33:22] <alex_joni> so I added -f (only files)
[10:33:33] <alex_joni> and the cp barfs on the folder
[10:33:48] <alex_joni> so it fails, that's why I added a - in front of it
[10:33:54] <alex_joni> so make continues, even if the command fails
[10:33:59] <alex_joni> but that's a nasty hack :(
[10:34:01] <fenn> is that what the error 1's are?
[10:34:09] <alex_joni> yes
[10:40:10] <alex_joni> fenn: jepler suggested : globnocvs () { find . -maxdepth 1 -not -name CVS -name "${1-*}" }
[10:40:41] <alex_joni> or smthg like: $(foreach i,$FILES,install ... $i) or whatever the format is
[10:46:29] <fenn> man cp says that -f is "force"
[10:46:46] <alex_joni> probably..
[10:46:59] <alex_joni> but it doesn't have a flag to not copy dirs
[10:47:26] <fenn> that is the default here
[10:47:32] <fenn> cp -r is for directories
[10:47:43] <alex_joni> I know.. same here
[10:47:49] <alex_joni> but a cp * will fail
[10:47:58] <alex_joni> when the * gets expanded to a dir
[10:48:08] <alex_joni> and it should only disregard it, not fail..
[10:52:27] <fenn> what is the "${1-*}"
[10:52:42] <alex_joni> dunnp
[10:52:44] <alex_joni> dunno
[10:54:41] <fenn> filenames starting with .
[10:54:57] <fenn> no idea how that works though
[10:55:19] <alex_joni> probably all without *
[10:55:29] <alex_joni> because * doesn't expand to . files I think
[10:58:07] <fenn> does make always use bash?
[10:58:30] <alex_joni> dunno :(
[10:58:51] <fenn> well, anyone who is picky enough to use a different shell probably knows how to fix it
[11:03:00] <alex_joni> yup
[11:11:32] <fenn> hint: don't run make indent
[11:12:17] <alex_joni> on what?
[11:12:30] <fenn> it indents the binary files!!!
[11:12:42] <fenn> as well as everything else
[11:12:42] <alex_joni> huh.. that's not supposed to be happening
[11:13:45] <alex_joni> # Reformat the sources using long format names to aid readability.
[11:13:45] <alex_joni> # This is applied ONLY to C source and header files - indent does
[11:13:45] <alex_joni> # not work correctly for C++ files
[11:14:08] <alex_joni> def_indent:
[11:14:08] <alex_joni> @ echo "Re-format `pwd` $(RT_SRCS) $(C_SRCS) $(C_HDRS) $(C_PHDRS)"
[11:14:08] <alex_joni> @ (if [ ! -z "$(RT_SRCS)$(C_SRCS)$(C_HDRS)$(C_PHDRS)" ] ; then \
[11:14:47] <alex_joni> look at src/Make.rules
[11:34:23] <fenn> why are all the variable names in ()'s?
[11:34:28] <fenn> like $(DESTDIR)
[11:34:36] <fenn> isn't that the same as $DESTDIR
[11:34:55] <alex_joni> not in make
[11:35:18] <alex_joni> $DESTDIR won't work in make
[11:35:27] <fenn> weird
[11:35:36] <alex_joni> because it's not using the shell stuff, but searches for it's own
[11:43:07] <alex_joni_> alex_joni_ is now known as alex_joni
[12:06:21] <fenn> * fenn goes to bed..
[12:06:27] <alex_joni> night
[12:06:34] <alex_joni> any luck with make install?
[12:06:54] <fenn> make didn't like my function
[12:06:55] <alex_joni> I mean.. did it worked as you'd expect?
[12:07:03] <alex_joni> :(
[12:07:30] <fenn> i'm prolly using functions wrong.. havent done any scripting with functions before
[12:07:48] <alex_joni> same here.. well I'll look at it
[12:07:53] <alex_joni> how about the whole thing?
[12:08:02] <alex_joni> did it install, did it run afterwards?
[12:08:18] <fenn> er, uh, i didn't actually install it :0
[12:09:50] <alex_joni> ok ;-)
[12:10:03] <fenn> the rpm didn't complain.. looking to see where it put stuff
[12:10:40] <alex_joni> I usually do:
[12:10:50] <alex_joni> DESTDIR=/tmp/foo && sudo make install
[12:13:30] <fenn> Can't find variable PARAMETER_FILE in section [RS274NGC] of file /usr/local/etc/emc.ini
[12:14:20] <alex_joni> strange..
[12:14:35] <fenn> PARAMETER_FILE = emc.var
[12:14:54] <fenn> how does it know where emc.var is?
[12:14:58] <alex_joni> and /usr/local/etc/emc.ini is actually there?
[12:15:24] <fenn> yeah
[12:15:29] <alex_joni> it cd's to the dir first
[12:15:32] <alex_joni> and runs from there
[12:15:35] <alex_joni> emc.run that is..
[12:15:55] <fenn> to what dir, /usr/local/etc/ ?
[12:15:55] <alex_joni> so emc.var should be in the same dir
[12:15:59] <fenn> yeah it is
[12:17:52] <alex_joni> but if it says it can't find the variable, it means smthg is borked INSIDE the emc.ini
[12:20:52] <fenn> sudo halcmd -fk
[12:20:52] <fenn> halcmd: error while loading shared libraries: libnml.so: cannot open shared object file: No such file or directory
[12:22:34] <alex_joni> libnml??
[12:23:56] <alex_joni> I'll look at that..
[12:24:06] <alex_joni> seems halcmd uses some parts of nml (for ini-stuff)
[12:24:27] <fenn> i guess /usr/local/lib isnt in my path
[12:24:43] <fenn> i thought ld automagiclly did stuff to find lib directories
[12:29:00] <fenn> what is /etc/.runinfo exactly?
[12:29:32] <alex_joni> * alex_joni is still thinking where to place that..
[12:29:37] <alex_joni> it's a file read by rtai-load
[12:29:43] <alex_joni> to know what modules to load
[12:29:53] <alex_joni> I think /lib/modules/... is more appropiate
[12:30:29] <fenn> maybe in /usr/realtime?
[12:30:37] <alex_joni> nope
[12:30:46] <alex_joni> that would break other rtai-stuff
[12:30:55] <alex_joni> if you have any other rtai applications on your pc
[12:33:04] <fenn> ok bedtime for real now
[12:33:12] <alex_joni> night
[13:36:09] <les_w_away> Well, morning and afternoon all.
[13:36:40] <les_w_away> les_w_away is now known as les_w
[13:38:18] <alex_joni> morning..
[13:38:32] <alex_joni> and in case I don't see you.. good day, good evening and good night
[13:38:33] <alex_joni> ;)
[13:41:03] <anonimasu> morning
[13:41:14] <anonimasu> * anonimasu is till pondering about a jog wheel
[13:44:00] <les_w> I just had a thought yeaterday about that Mic-6 al plate i bought yeasterday
[13:44:20] <les_w> that stuff makes very nice machine tables
[13:44:37] <les_w> cut into squares or hexagons
[13:44:54] <les_w> the 30mm stuff in particular
[13:47:01] <alex_joni> how big?
[13:52:13] <les_w> well the stuff is actually 1.25 inches by 18.5 inches diameter...a circle
[13:52:19] <les_w> ground and polished
[13:52:25] <les_w> on the faces
[13:52:35] <alex_joni> so how does that transform to a machine table?
[13:52:51] <les_w> and about 1/3 normal price
[13:53:13] <les_w> cut squares or hexagons and tile them together
[13:53:24] <alex_joni> I see...
[13:54:13] <les_w> I ordered some...if it looks good I will order a bunch shipped on pallets
[13:55:40] <les_w> I want to see if they are dinged up....normally mic6 has self adhesive plastic on the faces to protect the polished surfaces
[13:55:52] <les_w> this has been removed...
[13:56:07] <les_w> but a few dings won't hurt anything really
[13:58:35] <alex_joni> wot's dings? scratches?
[14:02:59] <les_w> indentations
[14:03:09] <les_w> with raised areas around
[14:03:09] <alex_joni> I see
[14:06:49] <alex_joni> lunch time
[14:06:50] <alex_joni> bbl
[14:06:54] <les_w> k
[14:23:21] <anonimasu> hm
[14:23:35] <anonimasu> looks like I'll turn the spare machine into a cnc machine instead if the rails are ok..
[14:23:50] <anonimasu> /rails/slides
[14:23:53] <anonimasu> ways.
[14:23:55] <anonimasu> whatever..
[14:57:55] <les_w> pictures pictures!
[15:01:45] <anonimasu> * anonimasu still has the camera cable at work :)
[15:30:53] <rayh> Hi Alex
[15:31:01] <alex_joni> hey rayh
[15:39:37] <alex_joni_> alex_joni_ is now known as alex_joni
[15:41:19] <rayh> was fenn able to get anywhere with linkpp yesterday?
[15:41:30] <SWPadnos> he committed a change
[15:41:47] <rayh> Hi SWPadnos
[15:41:56] <SWPadnos> hiya
[15:42:34] <alex_joni> SWPadnos: he did? I missed that..
[15:42:34] <alex_joni> let me look
[15:42:54] <SWPadnos> there was no message to the commit list, but cia picked it up
[15:43:13] <alex_joni> probably he's not subscribed to the commit list
[15:43:27] <alex_joni> which means his message is awaiting moderator approval
[15:43:29] <alex_joni> either jmk or paul_c
[15:43:30] <SWPadnos> ah
[15:43:46] <SWPadnos> it was at 19:49 last night, my time (currently 10:05)
[15:45:08] <rayh> quite a few since my last up.
[15:45:21] <SWPadnos> time to re-up, I guess :)
[15:45:45] <alex_joni> hrmmm.. looking at the code
[15:45:45] <alex_joni> I agree mostly..
[15:46:31] <SWPadnos> there was another change that jmk wanted along with this - make the save command use linkpp if a signal has the same name as a connected pin
[15:46:40] <SWPadnos> (output linkpp, that is)
[15:51:01] <rayh> linkpp was not added to the man page
[15:51:14] <alex_joni> no it wasn't ..
[15:51:32] <alex_joni> but he added it to the help function of halcmd
[15:51:42] <rayh> ah.
[15:51:49] <rayh> Hi Dan
[15:51:53] <dan_falck> hi ray
[15:52:13] <rayh> How's the family?
[15:52:37] <dan_falck> doing good. Robin just went to a scout leader get together
[15:52:48] <rayh> We're looking at the latest HAL stuff.
[15:53:15] <dan_falck> I will watch and read
[15:53:50] <rayh> fenn just added a "linkpp" command so that we can connect pins directly.
[15:53:58] <dan_falck> great
[15:54:12] <rayh> well more like it hides the intermediate signals between pins
[15:56:05] <SWPadnos> it's a macro for addsig / linksp / linksp
[15:56:19] <SWPadnos> the signal is still there
[15:56:48] <SWPadnos> what branch is it in? I don't see the change in webcvs
[15:57:20] <rayh> Is there any checking for duplicate generated signal names?
[15:57:32] <SWPadnos> I'm doing an update - I haven't seen the cde yet
[15:57:36] <SWPadnos> code
[15:57:56] <alex_joni> rayh: it calls the addsignal code
[15:58:03] <alex_joni> which was properly written by jmk ;)
[15:58:09] <alex_joni> so yes, it does check ..
[15:59:43] <SWPadnos> hmmm - it looks like the type parameter is wrong in the call to do_newsig_cmd
[16:00:06] <SWPadnos> linkpp should call hal_signal_new directly, since it already has the HAL_TYPE
[16:01:03] <alex_joni> yup
[16:01:08] <alex_joni> * alex_joni agrees
[16:01:12] <alex_joni> sloppy code by fenn ;)
[16:01:34] <SWPadnos> I'll commit a change in a few minutes, with a couple of other cleanups
[16:01:46] <alex_joni> change the manpage aswell ;)
[16:03:25] <SWPadnos> gah - too much work :)
[16:03:45] <alex_joni> linkpp should not have taken more than 20 minutes to do
[16:04:03] <alex_joni> including documentation
[16:04:03] <alex_joni> checkout, modification and commit
[16:04:15] <SWPadnos> what do you think about removing the lase 'else' from the pin checks, since either of the earlier cases results in a return?
[16:04:38] <alex_joni> it confused me a bit.. but mostly because of the indent which is screwed
[16:04:41] <SWPadnos> where are the manpages in the cvs tree?
[16:05:02] <SWPadnos> ah - docs/man
[16:05:05] <alex_joni> docs/man
[16:06:26] <SWPadnos> hmmm - what format is the manpage closest to? (I can't see much in all those slashes :) )
[16:06:40] <alex_joni> not sure ;)
[16:07:01] <rayh> If you copy it to /usr/share/man/man1 you can read it.
[16:07:11] <rayh> with the man command.
[16:07:16] <SWPadnos> not on my Windows machine ;)
[16:07:20] <rayh> troff is the format, I think
[16:07:24] <SWPadnos> OK
[16:07:45] <SWPadnos> bummer - programmers notepad doesn't have that
[16:07:48] <alex_joni> SWPadnos: mc reads it ok with view
[16:07:58] <SWPadnos> on Windows?
[16:08:03] <alex_joni> no
[16:08:32] <rayh> whatchadoinonwinders?
[16:08:52] <SWPadnos> igottalottaprogramsthatsucksoIhavetouseit
[16:10:18] <rayh> It really is difficult to type that way.
[16:10:37] <SWPadnos> any stylistic preference regarding retval vs. direct checks for HAL_SUCCESS
[16:10:40] <SWPadnos> harder to read
[16:10:42] <SWPadnos> ie:
[16:11:00] <SWPadnos> nevermind
[16:16:29] <SWPadnos> well - I can't compile it (due to platform problems), but I can commit it :)
[16:16:50] <alex_joni> hopefully no \n\r crap in there
[16:16:59] <SWPadnos> nope - PN is smart enough for that
[16:17:08] <alex_joni> * alex_joni wouldn't commit untested code...
[16:17:13] <alex_joni> even if it reads nicely
[16:17:16] <SWPadnos> but thanks for the reminder - I'll make it explicitly unix line ends
[16:17:18] <alex_joni> but.. that's just me
[16:17:32] <SWPadnos> me either - chall I email the two files to you? (or dcc)
[16:17:36] <SWPadnos> shall
[16:17:48] <alex_joni> pastebin them
[16:18:00] <SWPadnos> pastebin.org?
[16:18:15] <alex_joni> com
[16:18:20] <SWPadnos> ah - OK
[16:20:20] <alex_joni> * alex_joni started his develbox
[16:20:27] <alex_joni> I hope you realize what that means..
[16:20:32] <alex_joni> I had to get out of bed
[16:20:38] <alex_joni> and move about 10 feet
[16:20:44] <SWPadnos> sorry - I wouldn't wish that on anyone
[16:21:18] <alex_joni> now I need to find my mouse, and I'm ready
[16:21:18] <alex_joni> :D
[16:21:47] <SWPadnos> hopefully, pastebin preserves line endings :)
[16:22:45] <SWPadnos> OK - they're both there
[16:22:59] <SWPadnos> halcmd: http://pastebin.com/426638
[16:23:08] <SWPadnos> manpage: http://pastebin.com/426639
[16:23:24] <alex_joni> I about finished halcmd.. but looking now
[16:26:19] <SWPadnos> OK - the manpage looks acceptable (odd, but acceptable)
[16:27:34] <SWPadnos> I added in a check for the other arrow direction as well (so linkpp pin => pin or linkpp pin <= pin will work)) though that wasn't necessary, I guess (unless someone does >= ) :)
[16:27:45] <SWPadnos> so you could take that out
[16:28:09] <SWPadnos> (not enough coffee yet)
[16:28:13] <alex_joni> heh.. I already had the other direction ;)
[16:28:32] <SWPadnos> the = is the other direction - it occurred to me as I was typing that last line :)
[16:29:04] <SWPadnos> the whole arrow check is a little - hmm - basic, shall we say? ;)
[16:30:01] <alex_joni> it's ok..
[16:30:23] <SWPadnos> sure, but "=-awalkinthepark" qualifies :)
[16:30:33] <alex_joni> so?
[16:30:34] <alex_joni> :D
[16:30:40] <SWPadnos> or just =
[16:30:51] <SWPadnos> (which could possibly be confusing some time :)
[16:31:01] <alex_joni> brainiac is fun =))
[16:31:06] <SWPadnos> heh
[16:31:15] <alex_joni> babes with explosives right now
[16:31:24] <SWPadnos> dude!
[16:31:24] <alex_joni> lol
[16:32:01] <alex_joni> "when it comes to big bangs this group has lots of experience"
[16:32:02] <alex_joni> ROFLMAO
[16:32:06] <SWPadnos> so - doies that compile (or work)?
[16:32:07] <SWPadnos> heh
[16:32:15] <alex_joni> hang on..
[16:37:00] <SWPadnos> I wonder if spammers have figured out how to make a filter that converts ' AT ' to '@', and ' DOT ' to '.'
[16:37:01] <alex_joni> it compiled now..
[16:37:09] <alex_joni> SWPadnos: long time ago
[16:37:14] <SWPadnos> cool - what had to change?
[16:37:27] <SWPadnos> (I was just seeing jmk's address at the bottom of the manpage :)
[16:37:36] <alex_joni> I added some aditional checking
[16:37:40] <SWPadnos> ah
[16:37:42] <alex_joni> that the pin types are the same
[16:37:47] <alex_joni> before the sig was created
[16:38:00] <SWPadnos> right - I guess it makes sense to not create the signal if they're different
[16:38:16] <SWPadnos> I was also thinking about the question of deleting the signal if weither connect fails
[16:38:20] <SWPadnos> either
[16:43:52] <rayh> Interesting -- there is no tcl syntax highlighting at pastebin.
[16:44:00] <SWPadnos> nor troff
[16:44:17] <CIA-5> 03alex_joni * 10emc2/docs/man/man1/halcmd.1: added info for linkpp, thanks to fenn & swp
[16:47:13] <SWPadnos> the other hilighter often missing is assembly language (of whatever flavor)
[16:47:23] <SWPadnos> though pastebin has that
[16:47:24] <alex_joni> seems it's working ok..
[16:47:28] <SWPadnos> cool
[16:47:39] <SWPadnos> the nest thing would be the change to save
[16:48:08] <alex_joni> in a minute ;)
[16:48:21] <SWPadnos> that may be harder, because there's no guarantee that the pin for which the signal is named will be first (and therefore a little harder to detect that it's already been output)
[16:48:34] <SWPadnos> waiting patiently (still on a Win machine, after all :) )
[16:48:44] <CIA-5> 03alex_joni * 10emc2/src/hal/utils/halcmd.c: cleaned up and added some aditional checks to linkpp
[16:49:49] <alex_joni> how does that look?
[16:50:04] <SWPadnos> good question. the manpage text looks better
[16:50:30] <SWPadnos> (at least, it does in the commit message)
[16:54:19] <SWPadnos> weird - the merge tool shows two rtapi_mutex_get lines, which clearly wouldn't work (unless it's smart enough to know that the same routine is asking again)
[16:56:40] <SWPadnos> you know - all those hal error lines should probably print out the line where the error occurs (unless run from the command line)
[16:56:58] <alex_joni> yeah.. probably so
[16:57:11] <alex_joni> but halcmd always runs from command line :)
[16:57:16] <SWPadnos> that implies a HAL_ERROR_PRINT function that takes a string showing the actual error, and decides the format
[16:57:30] <alex_joni> if you have a .hal file.. it's actually halcmd -kf < foo.hal
[16:57:35] <SWPadnos> yeah yeah - parameters given on the command line (bonehead) ;)
[16:57:48] <SWPadnos> Mr. spelling errors :)
[16:58:02] <alex_joni> so you'd need to count the commands... I think
[16:58:18] <alex_joni> but what if you feed it 2 files?
[16:58:18] <alex_joni> :D
[16:58:20] <SWPadnos> no - if using a file, then count blank or comment lines
[16:58:22] <dmess> isn #'s as such
[16:58:43] <alex_joni> SWPadnos: how would you know? that it's a file?
[16:58:52] <SWPadnos> in that case, print the currently executing filenamre
[16:58:58] <SWPadnos> ...name..
[16:59:05] <SWPadnos> halcmd knows if it
[16:59:19] <SWPadnos> halcmd knows if it's looping through a file or using an explicit command line
[16:59:29] <SWPadnos> there can be a flag
[17:07:38] <alex_joni> you sure halcmd knows?
[17:08:54] <SWPadnos> it can know. the section of main that deals with it can just set a (global) flag
[17:09:05] <SWPadnos> but you're right - now it doesn't keep track
[17:09:19] <alex_joni> it needs some modifications, and it could know..
[17:09:32] <alex_joni> save is pretty strange
[17:09:51] <SWPadnos> there would still need to be at least two versions of the "error print" function, to keep it down to a single call
[17:09:56] <alex_joni> it saves signals first
[17:09:58] <alex_joni> then pins
[17:09:59] <SWPadnos> different numbers of parameters
[17:10:24] <SWPadnos> does it not save a new signal, then the pins connected to it immediately thereafter?
[17:10:34] <alex_joni> nope
[17:10:35] <SWPadnos> or is it all signals, then all pin connections
[17:10:35] <alex_joni> ;)
[17:10:38] <SWPadnos> bummer
[17:10:39] <alex_joni> yup
[17:11:31] <SWPadnos> hmmm - this should go on the list of "connection based" modifications to halcmd
[17:11:48] <SWPadnos> (like listing pins exported by a component)
[17:12:25] <alex_joni> well it does that already
[17:12:32] <alex_joni> halcmd show pin iocontrol
[17:12:43] <alex_joni> and it matches all pins started with iocontrol*
[17:12:45] <SWPadnos> that only works when the pins are named the same as the component
[17:12:49] <SWPadnos> look at motmod
[17:12:55] <SWPadnos> (axis.0. ...)
[17:13:01] <alex_joni> I see..
[17:14:00] <SWPadnos> jmk and I were discussion extending the halcmd syntax to accept name1:name2 to restrict {list,show} to a single component (of name1)
[17:14:09] <SWPadnos> discussing - man, I need more coffee
[17:14:10] <rayh> jmk and I had this discussion a couple days back.
[17:14:17] <SWPadnos> yep - I was there :)
[17:14:24] <rayh> Right.
[17:14:55] <SWPadnos> it's a language issue more than anything else - making a restricted lster is no problem at all
[17:14:58] <SWPadnos> lister
[17:15:31] <rayh> could you explain that a bit more?
[17:15:33] <SWPadnos> (it's as easy as just checking thing->owner, and making sure it's the one you want)
[17:16:19] <SWPadnos> each item that halcmd prints out the name of with list or show has a structure that includes the text of the name, in addition to a pointer to whatever "owns" that item
[17:16:56] <SWPadnos> if you want only motmod-related pins, you can use the normal pin-searching function, but do an additional check that thispin->owner== motmod before printing
[17:17:19] <rayh> and owner is the module like iocontrol
[17:17:23] <SWPadnos> sure
[17:17:32] <rayh> okay
[17:17:43] <alex_joni> SWPadnos: you are aware that all this functionality will eventually be for a handfull of users
[17:17:54] <SWPadnos> like Ray ;)
[17:17:55] <alex_joni> most of them won't ever touch this..
[17:18:08] <alex_joni> give ray an halgui, and he won't type anymore
[17:18:16] <alex_joni> I know I wouldn't...
[17:18:30] <SWPadnos> well - right now you can list all of the pins, but you can't tell which component they're associated with unles they were named the same
[17:18:36] <rayh> Ya never know what rayh might do!
[17:18:57] <alex_joni> rayh: maybe you're not the best example :)
[17:19:13] <SWPadnos> rayh can't write halgui (and neither can anyone else) without that, unless they duplicate the functionality of halcmd (which seems pointless)
[17:19:30] <alex_joni> halgui should use halcmd
[17:19:40] <alex_joni> I see where you are pointing..
[17:19:43] <dmess> of course
[17:19:46] <rayh> I was working on a tickle based hal pin, signal, and link viewer.
[17:19:52] <SWPadnos> exactly, and right now, halgui can't tell what pins belong to what component
[17:20:00] <SWPadnos> because halcmd can't tell it
[17:20:10] <rayh> genau
[17:21:01] <SWPadnos> uh-oh - my wife finished with the vacuum. I'd better run to Home Depot while I have the chance.
[17:21:15] <SWPadnos> see ya later :)
[17:21:19] <SWPadnos> SWPadnos is now known as SWP_Away
[17:42:50] <CIA-5> 03 07 * 10emc/: rm -fR emc*
[17:45:10] <alex_joni> CIA-5: what's that?
[17:45:31] <SWP_Away> yeah - that's an odd message
[18:30:38] <alex_joni> hey jymmm
[18:30:50] <les_w> hi
[18:31:08] <les_w> taking a break from grinding leaves on the little tractor
[18:31:28] <les_w> now i'll get on the big tractor and really grind leaves
[18:46:39] <Jymmm> leae grinding
[18:47:04] <Jymmm> les_w you dont burn them?
[18:49:01] <rayh> If I want an or2 block, do I both loadrt and add it to a thread?
[18:49:51] <alex_joni> rayh: yes
[18:50:10] <CIA-5> 03alex_joni * 10emc2/src/hal/utils/meter.c: fixed bug #1020123, now halmeter pops up the selection dialog with the old selected value
[18:51:01] <rayh> Thanks
[18:51:33] <alex_joni> you need to loadrt it for it to exist
[18:51:47] <alex_joni> and you need to add it to a thread if you want it to update the value ;)
[19:07:31] <dmess> ground makes good topsoil.. ;)
[19:08:01] <dmess> burnt makes acidic compost
[19:09:29] <dmess> or maybe its basic... too much ash makes it cautic... (basic)
[19:11:06] <dmess> ash + bat GUANO comes out relatively correct for the garden... if you can handle the smell of the guano
[19:12:49] <dmess> ive been workin on the mechanics of the paraglider simulator... comin' along
[19:17:49] <alex_joni_> alex_joni_ is now known as alex_joni
[19:40:08] <les_w> I need ashes a lot here...soil ph is often 5.5
[19:40:23] <alex_joni_> that's bad for golf balls
[19:40:27] <alex_joni_> ROFL
[19:40:27] <les_w> haha
[19:40:41] <alex_joni_> alex_joni_ is now known as alex_joni
[19:41:18] <les_w> pretty nice day for fall outside work here
[19:41:34] <les_w> 60f, no clouds, little wind
[19:41:42] <alex_joni> nice
[19:42:06] <les_w> "severe clear" air quality
[19:42:25] <alex_joni> I love severe clear air ;)
[19:42:34] <alex_joni> specially during winter..
[19:42:38] <les_w> 40 km+ visibility
[19:44:00] <les_w> anyway, back out to attack some hedgerows with the front end laoder. Was just in taking a break
[19:45:32] <dmess> WHOO golfs??? anyway...
[19:57:09] <anonimasu> :)
[19:57:18] <alex_joni> hey anders
[19:57:24] <anonimasu> hey
[19:57:31] <alex_joni> what's up?
[19:58:16] <anonimasu> not much
[19:58:19] <anonimasu> I am a bit tired
[19:58:31] <alex_joni> work? or too much fun?
[20:00:29] <anonimasu> neither
[20:00:43] <alex_joni> hrmmm.. then?
[20:01:05] <les_w> tired of being admired?
[20:01:12] <anonimasu> haha
[20:01:15] <anonimasu> actually that's a good thing
[20:01:20] <anonimasu> :D
[20:04:17] <anonimasu> alex_joni: been slacking too much I guess
[20:05:05] <alex_joni> yes you have..
[20:05:09] <alex_joni> you need to get some work done
[20:05:18] <alex_joni> no more excuses
[20:05:29] <anonimasu> *hides*
[20:05:49] <dmess> under "her" dress
[20:06:04] <alex_joni> dmess: right on
[20:06:30] <dmess> if it smells good EAT it
[20:06:42] <alex_joni> dmess: that's sick..
[20:07:16] <dmess> thats not what she says...
[20:07:37] <dmess> she calls me GOD and stuff...
[20:07:47] <anonimasu> this scares me
[20:07:51] <alex_joni> dmess: your mum?
[20:08:05] <dmess> mo... my wife
[20:08:23] <dmess> my kids mum though
[20:08:39] <dmess> i still love her... ; )
[20:10:14] <dmess> be brave Anon..
[20:11:40] <anonimasu> haha
[20:12:35] <dmess> they only LOOK wild
[20:13:18] <rcsu> //trash is the server name
[20:16:57] <alex_joni> rcsu: sorry?
[20:17:14] <rcsu> ehem, sorry wrong channel
[20:17:36] <alex_joni> yet I totally agree :)
[20:18:32] <anonimasu> alex_joni: have you got any clue on connecting a jogwheel to emc?
[20:18:50] <alex_joni> depends on what it needs to do..
[20:19:07] <anonimasu> heh
[20:19:11] <anonimasu> jog the machine
[20:19:17] <alex_joni> which axis?
[20:19:20] <anonimasu> any..
[20:19:22] <anonimasu> x/y/z
[20:19:28] <alex_joni> and how do you select?
[20:19:32] <anonimasu> with a pushbutton..
[20:19:37] <anonimasu> how do commercial machines do it+
[20:19:38] <alex_joni> on the jogwheel?
[20:19:43] <anonimasu> yeah..
[20:19:49] <anonimasu> besides the jogwhee´┐Żl
[20:19:51] <anonimasu> wheel
[20:19:51] <alex_joni> I have no idea.. that's why I'm asking..
[20:20:07] <anonimasu> dmess: still there?
[20:20:24] <anonimasu> * anonimasu pokes dmess
[20:20:36] <anonimasu> alex_joni: I think that's how they do it
[20:20:42] <alex_joni> right..
[20:20:49] <alex_joni> well.. I'm thinking about it
[20:20:55] <alex_joni> first you need a HAL component
[20:20:59] <alex_joni> that talks to the hardware
[20:21:13] <alex_joni> that might be the encoder module in HAL
[20:21:21] <alex_joni> along with some free pins on the parport
[20:21:26] <alex_joni> or whereever
[20:21:30] <anonimasu> hm, I am thinking about a avr since it's a bit faster..
[20:21:41] <anonimasu> and doing the velocity calcs in hardware..
[20:21:45] <anonimasu> err accel/vel
[20:21:48] <alex_joni> velocity?
[20:21:58] <alex_joni> what for?
[20:22:08] <anonimasu> hm, of the wheel..
[20:22:18] <alex_joni> doesn't it do only position?
[20:22:25] <alex_joni> and emc takes care of accel/speed?
[20:22:40] <alex_joni> like jogging on the GUI ?
[20:22:44] <anonimasu> hm, is that so I thought a jog wheel worked just like a handwheel..
[20:22:55] <anonimasu> the faster you move it the faster the machine goes there..
[20:23:07] <anonimasu> or is that just a matter of accelerating to the position...
[20:23:10] <alex_joni> well.. the faster you move it.. the greater the distance
[20:23:18] <anonimasu> ah yeah true
[20:23:20] <alex_joni> and the machine can accelerate ...
[20:24:02] <anonimasu> hm, might be even easier then I thought..
[20:24:09] <alex_joni> so I think userlever app is enough
[20:24:13] <rayh> is axis.0.neg-lim-sw-in a path that I can use from parport to set a limit switch condition.
[20:24:14] <alex_joni> and NML to send position
[20:24:23] <anonimasu> like the jogwheel in emc1?
[20:24:23] <alex_joni> rayh: yes
[20:24:29] <alex_joni> anonimasu: exactly
[20:24:55] <anonimasu> maybe I should just go that route..
[20:25:16] <anonimasu> would it be easier then making it a hal module?
[20:25:33] <rayh> If i want a single all axis neg limit like emc's parport I'd have to or them all?
[20:26:05] <alex_joni> rayh: yes, but in hardware
[20:26:30] <alex_joni> inside emc2 you simply connect all three to that pin
[20:27:26] <rayh> okay
[20:27:40] <alex_joni> anonimasu: you probably need only minor modifications to handwheel, to make it work with emc2
[20:27:51] <rayh> Why do we get 8 axes regardless of need?
[20:28:10] <alex_joni> rayh: good question ;)
[20:28:22] <anonimasu> alex_joni: it uses the gameport, so it'd require a bit..
[20:28:22] <rayh> is that something that will be handled by refactor?
[20:28:26] <alex_joni> they get created at insmod time..
[20:28:40] <alex_joni> and afterwards you can send motion through SHM how many axes you really have
[20:28:47] <alex_joni> so yes, it's related to the refactor
[20:29:01] <alex_joni> anonimasu: change it so it works with hal-pins
[20:29:02] <rayh> k Thanks
[20:29:10] <alex_joni> anonimasu: but like I said.. only minor changes
[20:29:21] <anonimasu> alex_joni: hmm maybe I should get into hal ;)
[20:33:18] <alex_joni> maybe you should
[20:33:25] <alex_joni> rayh did ;)
[20:33:31] <alex_joni> so what are you waiting for?
[20:33:34] <alex_joni> :-P
[20:33:40] <anonimasu> end of the world?
[20:34:12] <alex_joni> well.. then it won't matter,now would it?
[20:34:21] <anonimasu> just kidding
[20:34:29] <alex_joni> yeah, me too
[20:34:35] <anonimasu> I dont know
[20:34:43] <anonimasu> I am waiting for a good corner to start in.
[20:34:45] <anonimasu> seriously..
[20:34:54] <anonimasu> :)
[20:34:54] <alex_joni> corner?
[20:35:05] <anonimasu> like implementing a jogwheel ;)
[20:35:16] <anonimasu> yeah what end to start in..
[20:35:35] <alex_joni> here's an end.. http://blueballfixed.ytmnd.com/
[20:35:46] <alex_joni> follow the blue ball to the end..
[20:36:16] <anonimasu> lol
[20:37:06] <rcsu> funny :)
[20:38:04] <anonimasu> yeah
[20:38:12] <anonimasu> *digs for the hal-introduction docs*
[20:39:46] <anonimasu> alex_joni: I guess now is as good as ever
[20:39:52] <rayh> alex_joni: How much effort to add a iocontrol.0.cycle-start
[20:39:55] <anonimasu> does anyone have a idea if you can rebuild worn ways?
[20:40:02] <alex_joni> rayh: what should that do?
[20:40:28] <rayh> Send an NML cycle start message to task
[20:40:38] <alex_joni> ahh.. input
[20:40:53] <alex_joni> not much trouble...
[20:40:59] <alex_joni> can you test?
[20:41:06] <alex_joni> * alex_joni can code..
[20:41:13] <rayh> I'd like to add one to the step_cl example
[20:41:23] <rayh> Need to add one to the Mazak.
[20:41:43] <alex_joni> does half an hour sound ok?
[20:41:53] <rayh> sounds fantastic.
[20:42:16] <alex_joni> ok then, do it in half an hour
[20:42:17] <alex_joni> =))
[20:42:36] <rcsu> alex_joni: thats not fair :(
[20:42:37] <alex_joni> kidding.. hang on, I'll probably have things to ask ;)
[20:43:16] <alex_joni> rcsu: what's not fair? kidding?
[20:43:16] <alex_joni> or the blue ball?
[20:44:02] <alex_joni> rayh: can you tell me what cycle start does?
[20:44:20] <rcsu> alex_joni: that kind of kidding
[20:45:00] <alex_joni> rcsu: ok.. take it back :(
[20:45:07] <anonimasu> *starts digging*
[20:45:30] <rcsu> anonimasu: what i have read so far, its fantastic
[20:45:58] <anonimasu> rcsu: hal?
[20:46:04] <rcsu> anonimasu: yep
[20:46:06] <anonimasu> or are you referring to the blue ball?
[20:46:08] <anonimasu> hm
[20:46:11] <anonimasu> where are you reading?
[20:46:13] <anonimasu> the intro
[20:46:25] <rcsu> Hal_Introduction.pdf
[20:46:30] <anonimasu> nice
[20:46:33] <anonimasu> I am looking for it now
[20:46:33] <anonimasu> :)
[20:46:44] <alex_joni> www.linuxcnc.org/Hal_Introduction.pdf
[20:46:59] <anonimasu> nice
[20:47:00] <anonimasu> :)
[20:47:09] <rcsu> ah thanks, i lost my bookmarks today
[20:47:28] <alex_joni> rayh: what's cycle start? can't seem to see such a NML message
[20:53:38] <rayh> It would be run in auto mode
[20:54:13] <alex_joni> I see.. let me find that
[20:54:16] <rayh> sorry was working cl on the other box.
[20:54:37] <rayh> From tcl it is a run button through emcsh
[20:56:25] <alex_joni> yeah.. following that path now
[20:59:40] <alex_joni> EMC_TASK_PLAN_RUN
[20:59:43] <alex_joni> that's the message
[21:00:01] <alex_joni> but we need some preconditions for that. right?
[21:00:53] <rayh> Yes. I'm guessing that that is true.
[21:02:08] <rayh> I'd rather not force a kludge in task right now
[21:05:05] <alex_joni> well.. not in task
[21:05:11] <alex_joni> it's handled in task allright
[21:05:16] <alex_joni> but in iocontrol
[21:05:29] <alex_joni> although.. the channel is not right
[21:05:45] <alex_joni> iocontrol talks to task through the emcStatus NML channel
[21:06:38] <alex_joni> and EMC_TASK_PLAN_RUN goes through a different channel
[21:07:32] <alex_joni> emcCommand
[21:07:56] <rayh> okay. this can wait if we need to write such a file.
[21:08:42] <alex_joni> I really like your (or was it jmk?) idea.. the more I think about it, the more I like it...
[21:08:57] <alex_joni> have the GUI talk through hal pins to emcsh
[21:09:10] <alex_joni> and those pins could be taken over by an hardware panel
[21:09:20] <alex_joni> or even classicladder.. or whatever
[21:09:49] <alex_joni> but no need for realtime stuff, only userlevel
[21:09:58] <rayh> How would this affect the way that Axis does it's interfacing?
[21:10:23] <alex_joni> well maybe not at that level..
[21:11:16] <rayh> What is the slowest thread we could connect a cl to?
[21:11:25] <alex_joni> how do you mean that?
[21:11:48] <alex_joni> you can create a slow thread.. and connect only cl to that
[21:12:03] <rayh> Could we create a thread that is only handled about the rate that we would want a hardware panel polled.
[21:12:18] <rayh> Can't we start more than one cl
[21:12:21] <alex_joni> sure.. or have that part in userspace, not realtime
[21:12:30] <alex_joni> not right now, (more than one cl)
[21:12:39] <alex_joni> at least, not that I know of
[21:12:57] <alex_joni> but would you really need 2 cl's ?
[21:13:06] <alex_joni> you can take care of stuff with only one..
[21:13:06] <rayh> wondered cause we have the classicladder.0.xxx
[21:13:21] <alex_joni> yes.. some day maybe there'll be more than one
[21:13:35] <rayh> Certainly I can do that but it does it at servo rate
[21:13:40] <alex_joni> so?
[21:13:48] <alex_joni> not much overhead I think
[21:13:55] <rayh> so no worries?
[21:14:09] <alex_joni> guess so.. but if, then another cl might solve that
[21:14:43] <rayh> What we would be doing then is writing an equivalent of iocontrol for emcsh.
[21:15:00] <alex_joni> an emcsh equivalent, yes
[21:15:33] <rayh> I see it would be like emcsh. It would not be emcsh.
[21:20:18] <alex_joni> I wonder how usefull that would be..
[21:20:29] <alex_joni> probably pretty usefull
[21:20:40] <alex_joni> * alex_joni makes a list
[21:21:51] <alex_joni> pins for mist on/off, flood on/off, brake, spindle
[21:22:34] <alex_joni> param for feed override
[21:23:24] <alex_joni> then inputs for jogging (x,y,z)(+/-)
[21:23:43] <alex_joni> pins for modeselection (manual, mdi, auto)
[21:27:42] <alex_joni> pin for abort
[21:33:37] <rayh> I've got a question re mist and flood.
[21:33:47] <alex_joni> yeah.. shoot
[21:34:01] <rayh> Since I'm already pulling hal pins for these
[21:34:16] <alex_joni> why would it need to go to GUI and task?
[21:34:50] <rayh> Right. The problem is that iocontrol only has an output pin to pull
[21:35:09] <rayh> and when that pin is pulled it is assumed that status should be updated.
[21:35:30] <rayh> If there were two pins in iocontrol one do it and another did it.
[21:35:36] <alex_joni> well.. it wouldn't be very hard to do that..
[21:35:46] <alex_joni> but that would add lots of loops needed
[21:35:51] <rayh> then we could start flood
[21:35:52] <alex_joni> for machines that don't have it
[21:36:06] <rayh> Right. the loopback thing.
[21:36:09] <alex_joni> yup
[21:36:18] <rayh> I don't have a problem with that with cl
[21:36:29] <alex_joni> well.. one for lube, one for flood, one for mist, etc
[21:36:30] <rayh> and it is rather easy to do in hal itself
[21:36:41] <rayh> phone brb
[21:36:43] <alex_joni> yes.. simply one linkpp for each
[21:36:51] <rayh> exactly
[21:39:06] <alex_joni> ok.. say when off the phone
[21:47:50] <etla> hi a_j
[21:48:08] <alex_joni> hey etla
[21:48:29] <etla> do you know if EMC has always been under GPL ?
[21:48:54] <etla> or was it some other license in the very beginning ?
[21:49:04] <alex_joni> emc1 is PD
[21:49:16] <etla> ok
[21:49:19] <alex_joni> with some proprietary stuff in there too.. not very sure..
[21:49:40] <etla> so a commercial company taking emc1 code and developing it to a product need not release the source ?
[21:50:16] <alex_joni> ask the guys at NIST...
[21:50:22] <alex_joni> it's code done by them
[21:52:11] <Imperator_> logger_aj: bookmark
[21:52:11] <Imperator_> See http://solaris.cs.utt.ro/irc/irc.freenode.net:6667/emc/2005-11-13#T21-52-11
[21:52:24] <alex_joni> hey Martin
[21:52:45] <Imperator_> Hi Alex
[21:52:50] <Imperator_> something new ?
[21:52:57] <alex_joni> not much ;)
[21:53:01] <etla> imperator: were you using the mesa board ?
[21:53:36] <Imperator_> let me say i have one
[21:54:01] <etla> ok... I'm researching servo hardware and there are so many options to choose from...
[21:54:24] <Imperator_> we have made here some beakout boards for real servo control
[21:54:37] <Imperator_> i don't like that PWM stuff from mesa
[21:55:03] <etla> i'd like to do direct pwm to the servo power stage
[21:55:09] <anonimasu> hm
[21:55:17] <etla> so the commutation and current loop is closed by the mesa card
[21:55:21] <Imperator_> then take the mesa breakout board
[21:55:36] <etla> do you have an url ?
[21:55:46] <Imperator_> www.mesanet.com
[21:56:13] <etla> hmm... and your breakoutboard is different ?
[21:56:19] <Imperator_> i like to use a real servo amp with a current controller
[21:56:53] <etla> so you transform to an analog +/-10V at some point ?
[21:58:25] <Imperator_> we have made four little boards, one with a quad 16 bit DAC (+-10V output) one IO board for controlling the amp (enable ....) and one board with up to six encoder inputs and a board for limit and ref switches
[21:58:44] <Imperator_> but we have to write our own FPGA software
[21:59:05] <etla> ok, sounds cool
[21:59:27] <etla> will you close the position pid loop in the fpga or will it be in emc (like hostmot by mesa)
[21:59:35] <Imperator_> jep, ask me in about two weeks, then i know if it is working
[22:00:11] <etla> pid loop on the fpga ?
[22:00:29] <rayh> back
[22:00:33] <Imperator_> doing that inside the FPGA is preferred
[22:00:53] <Imperator_> hi rayh
[22:01:02] <etla> yes, faster. keep us posted on your progress !
[22:01:43] <Imperator_> for the first i will use the Mesa card only for input and output
[22:02:31] <alex_joni> rayh: hello
[22:02:49] <rayh> Hi Martin.
[22:02:53] <etla> imperator: will you make your fpga code available ?
[22:02:53] <rayh> alex
[22:03:47] <Imperator_> etla: jep also the pcb layout and so, but i want to test it before
[22:04:25] <etla> great ! I'm looking forward to seeing your work. I might order some 5i20's in the meantime and play with them...
[22:04:31] <alex_joni> think we need to talk to jmk about this
[22:04:41] <alex_joni> also once we agree on something post on the devel list
[22:05:11] <Imperator_> the PCBs are laying arround here, but i ordered the parts at a very cheap supplyer, so i have to wait a week
[22:05:51] <Imperator_> what do you mean Alex ?
[22:05:53] <etla> ok... I'm still waiting for my surpluscenter servos.
[22:06:07] <alex_joni> Imperator_: was talking to rayh a bit sooner
[22:06:16] <alex_joni> about recreating the emcsh functionality with HAL pins
[22:06:19] <Imperator_> ah
[22:06:23] <alex_joni> on the other side (on iocontrol)
[22:07:02] <etla> imperator: what component supplier are you using ?
[22:07:29] <Imperator_> etla: www.elpro.org (germany)
[22:07:47] <Imperator_> where are you located ?
[22:08:09] <etla> ok, I'll have a look. I'm in finland so either www.elfa.se (10eur delivery) www.farnell.com (25eur) uk.digikey.com (?)
[22:08:41] <Imperator_> hm expensive
[22:08:45] <anonimasu> heh
[22:08:53] <anonimasu> etla: elfa is the same even in sweden
[22:08:53] <alex_joni> fernell is good, but expensive as hell
[22:09:15] <etla> rscomponents also works, no delivery if I pick it up at a local store. but they don't have everything...
[22:09:15] <alex_joni> www.elfa.se sounds like sweden allright
[22:09:40] <anonimasu> yep
[22:09:58] <alex_joni> > If a packet hits a pocket on a socket on a port,
[22:09:58] <alex_joni> > And the bus is interrupted as a very last resort,
[22:09:58] <alex_joni> > And the address of the memory makes your floppy disk abort,
[22:09:58] <alex_joni> > Then the socket packet pocket has an error to report!
[22:10:08] <alex_joni> > If your cursor finds a menu item followed by a dash,
[22:10:08] <alex_joni> > And the double-clicking icons put your window in the trash,
[22:10:08] <alex_joni> > And your data is corrupted 'cause the index doesn't hash,
[22:10:08] <alex_joni> > Then your situation's hopeless, and your system's gonna crash!
[22:10:28] <Imperator_> rscomponents delivers at the next day but to expensive for hobbyusers
[22:12:53] <rayh> know what the difference is between soft-limit and hard-limit in HAL is?
[22:13:14] <alex_joni> in HAL? there should be no soft limit in hal
[22:13:24] <etla> I was thinking about a dsPIC based drive, but now I'm thinking that all the logic should be in the fpga. only external power-stage.
[22:14:14] <alex_joni> * alex_joni found the perfect motto for emc2 ;)
[22:14:18] <alex_joni> "Make it idiot proof and someone will make a better idiot."
[22:15:16] <rayh> Um. You trying to say something about rayh trying emc2
[22:16:32] <rayh> logger_aj: is back!
[22:16:32] <rayh> I'm logging. I don't understand 'is back!', rayh. Try /msg logger_aj help
[22:16:51] <alex_joni> logger_aj is stupid
[22:17:00] <alex_joni> see.. he knows it
[22:17:16] <alex_joni> logger_aj: what you mean .(
[22:17:16] <alex_joni> I'm logging. I don't understand 'what you mean .(', alex_joni. Try /msg logger_aj help
[22:17:36] <Imperator_> etla: buy the servo amps at ebay, that is much cheaper
[22:18:33] <etla> Imperator: I'm not sure what you mean. surely it is not cheaper than buying the components... it's probably much faster and less work thoug
[22:19:29] <Imperator_> developing a servo amp will cost much more than the components for one amp
[22:19:35] <alex_joni> it is cheaper than buying the components..
[22:19:51] <anonimasu> :)
[22:19:55] <etla> well, that depends on the price of your time.
[22:20:18] <anonimasu> glad somone mentioned it
[22:20:20] <etla> dsPIC= 7eur, IRAMS power stage = 20eur. some optoisolators = maybe 10eur.
[22:20:29] <anonimasu> LOL
[22:20:53] <Imperator_> we needed nearly a year to make some stubit breakout boards for the mesa card (2 persons)
[22:20:59] <alex_joni> how many power stages you reckon you'll bust?
[22:21:16] <etla> a_j: I don't know, I haven't tried yet.
[22:21:32] <etla> the testing can be done with a low dc bus voltage.
[22:21:32] <alex_joni> * alex_joni tried a simple one once
[22:21:36] <alex_joni> DC-motors
[22:21:51] <alex_joni> took me a few months-person
[22:21:53] <etla> and the irams have internal shutdown
[22:22:18] <rayh> I'm serious here there needs to be a block comment for a .hal file.
[22:22:38] <rayh> not just the # on each line.
[22:23:00] <rayh> If you don't believe me try # out a hundred lines or so.
[22:23:16] <rayh> * rayh goes back to his ^%$#@ hal file.
[22:23:23] <alex_joni> rayh: :(
[22:23:29] <anonimasu> etla: good luck
[22:24:20] <rayh> either that or we need a line number returned with a fail to load.
[22:24:33] <rayh> It may just be a problem with cl.
[22:24:36] <etla> anon: it surely can't be that hard...
[22:25:00] <etla> ok, need to sleep now, bye guys. Imperator: keep us posted on your progress.
[22:25:05] <anonimasu> etla: build one then..
[22:25:11] <anonimasu> it's not that easy..
[22:25:25] <anonimasu> :)
[22:29:08] <SWP_Away> heh - line numbers on HAL file errors - I wonder where I've heard that before ;)
[22:29:12] <SWP_Away> SWP_Away is now known as SWPadnos
[22:30:28] <rayh> SWPadnos: probably me bitching.
[22:30:43] <alex_joni> * alex_joni is looking at that now.. not that trivial
[22:30:54] <SWPadnos> (nope - me suggesting it earlier today. probably based on your bitching though :) )
[22:31:15] <SWPadnos> the -f option doesn't use redirects, as you had earlier
[22:31:39] <SWPadnos> so when -f is used, just set a global flag, and zero a global line counter (or set to 1)
[22:32:33] <SWPadnos> but you need either multiple error print functions, one that can accept variable numbers of arguments, or multi-line error prints (sprintf / err_print) everywhere
[22:32:54] <alex_joni> SWPadnos: if you look at the code, you'll see it's not that trivial
[22:33:08] <alex_joni> the parser function doesn't read one line at once
[22:33:23] <SWPadnos> it seems to - I'll load it up - hold on
[22:33:29] <alex_joni> it's reading a number of tokens.. which can be on a lot of lines
[22:34:13] <SWPadnos> can you have the tokens separated on differrent lines? that would be a problem if you ever missed one
[22:34:23] <alex_joni> * alex_joni isn't really sure
[22:35:11] <SWPadnos> fgets returns up to MAX chars, but stops at newline
[22:35:20] <SWPadnos> so it reads line by line
[22:35:58] <SWPadnos> this is in the else clause of if (srcfile == NULL)
[22:36:48] <rayh> Hey. I've got a working limit switch pin.
[22:36:54] <alex_joni> nice
[22:36:55] <SWPadnos> cool.
[22:37:03] <SWPadnos> through ladder, etc?
[22:38:41] <rayh> Could be. I did the connections to a single hal signal called limits and that sig to pin 15 on the parport.
[22:38:52] <SWPadnos> ah
[22:39:22] <rayh> I could loop it through cl and make a hardware override button.
[22:39:38] <SWPadnos> like in the limit override conversation
[22:39:55] <rayh> yep.
[23:02:27] <alex_joni> SWPadnos: darn
[23:02:33] <alex_joni> it's even tougher.. :/
[23:02:46] <alex_joni> some (most?) error messages are not printed by halcmd
[23:03:01] <alex_joni> they are generated by hal_lib
[23:03:07] <rayh> * rayh hides
[23:03:34] <alex_joni> so I could either:
[23:03:40] <alex_joni> a). hack hal_lib or
[23:03:47] <alex_joni> b). hack rtapi_print_msg
[23:04:02] <alex_joni> either way seems yucky :(
[23:04:12] <alex_joni> or c). add aditional prints all over halcmd
[23:04:47] <SWPadnos> or add a hal_err_print to halcmd
[23:04:57] <alex_joni> you're not hearing me..
[23:05:06] <alex_joni> not halcmd is printing the error messages
[23:05:23] <SWPadnos> ah - *those* errors :)
[23:05:27] <alex_joni> for instance hal_link()
[23:05:40] <SWPadnos> yep - link, create_(whatever), etc.
[23:05:45] <alex_joni> hal_link() checks if the pins are there, and then does the linking
[23:07:44] <SWPadnos> the functions that call the hal_lib funcs could just prein afterwards, by checking return values
[23:07:48] <SWPadnos> so you'd get two lines
[23:07:59] <SWPadnos> HAL: ERROR: Couldn't do something
[23:08:07] <SWPadnos> HALCMD: Error on line foo
[23:08:07] <alex_joni> yes, that's what I'm doing now
[23:08:15] <SWPadnos> that sounds reasonable
[23:08:21] <SWPadnos> (annoying but reasonable)
[23:08:23] <alex_joni> but.. it's still a lot ;)
[23:08:44] <SWPadnos> yeah. I noticed that return values aren't always checked (at least, not in linkpp ;) )
[23:09:29] <SWPadnos> hmmm = most functions seem to return whatever error - maybe just main() can be changed to check for the return from parse_cmd
[23:10:13] <SWPadnos> that also eliminates the clobal variable, since the print is local to the function thats scanning the input file
[23:10:14] <alex_joni> I prefer adding the name of the function that failed
[23:11:00] <SWPadnos> it would be good - it's possible to have a single global char* that each function changes on entry, and main uses in its printf
[23:11:32] <SWPadnos> char *lastfunc_name
[23:14:02] <SWPadnos> or something like errno, but errstr, which just gets printed along with the file name and line number (rtapi_print_msg(RTAPI_MSG_ERROR, "HALCMD: Error in file %s, line %d: %s", filename, lineno, errstr)
[23:14:50] <SWPadnos> but I like the cleanliness of just printing the function name, since it doesn't require local char buffers and sprintfs everywhere
[23:17:17] <alex_joni> who wants to test?
[23:17:59] <SWPadnos> I should be able to soon, once I finish the mods to my USC
[23:21:22] <rayh> I can sneakernet to the other box for testing.
[23:23:45] <rayh> lube timer is working great.
[23:24:00] <alex_joni> nice
[23:24:09] <alex_joni> cvs is sluggish.. doesn't respond here :(
[23:24:15] <rayh> need to add a lube-up pin that resets the lube error message.
[23:25:17] <CIA-5> 03alex_joni * 10emc2/src/hal/utils/halcmd.c: added linenumber error reporting to halcmd, this is usefull on large .hal files
[23:28:47] <alex_joni> rayh: now I'll tell you what I want for halcmd .. *g*
[23:29:02] <alex_joni> * alex_joni wants %include foo.hal%
[23:29:45] <alex_joni> not that hard to implement, now that I think of it..
[23:32:29] <rayh> so that we can use multiple hal files?
[23:32:37] <alex_joni> yes..
[23:32:45] <alex_joni> and not duplicate the code all over..
[23:32:51] <rayh> That would be a very valuable feature.
[23:32:57] <alex_joni> but I guess multiple files can be used from the ini
[23:33:36] <rayh> I was thinking of the loadxx stuff in a separate file so that you could swap in and out whole files while running.
[23:33:52] <rayh> We can do that from the ini now.
[23:33:57] <rayh> ??
[23:34:45] <SWPadnos> you can't change threads once started though, right?
[23:35:00] <alex_joni> rayh: yes
[23:35:58] <alex_joni> rayh: if you look at emc.ini you'll see that core_stepper.hal along with standard_pinouts.hal get loaded
[23:36:06] <alex_joni> you can have up to 10 files like that
[23:39:06] <alex_joni> hi john
[23:39:09] <jmkasunich> hi
[23:39:12] <alex_joni> you turned up just right
[23:39:19] <SWPadnos> heh - indeed
[23:39:23] <jmkasunich> oops ;-)
[23:39:29] <alex_joni> * alex_joni just commited a change to halcm
[23:39:32] <alex_joni> halcmd even
[23:39:54] <alex_joni> hope you like it.. :)
[23:40:09] <jmkasunich> * jmkasunich is catching up on emails
[23:40:18] <alex_joni> * alex_joni waits patiently
[23:40:24] <jmkasunich> ;-)
[23:40:28] <alex_joni> oh.. and halmeter too
[23:43:44] <jmkasunich> dang... you are too fast
[23:43:53] <jmkasunich> I was gonna do linkpp this weekend
[23:43:56] <jmkasunich> (thanks)
[23:43:59] <alex_joni> fenn started that..
[23:44:04] <jmkasunich> and you even updated the man page!
[23:44:08] <SWPadnos> heh
[23:44:10] <alex_joni> * alex_joni & SWP just fixed it
[23:44:49] <SWPadnos> maybe we can add the search by owner stuff this weekend
[23:45:45] <SWPadnos> there was a more far-reaching change that was briefly discussed as well (touches HAL itself, not just halcmd's interface to it)
[23:45:55] <jmkasunich> what is that?
[23:46:31] <alex_joni> jmkasunich: a few topics
[23:46:42] <alex_joni> 1. linkpp (I think you're good with that)
[23:47:00] <alex_joni> 2. halmeter (now opens the select dialog on the selected pin/signal/param)
[23:47:04] <jmkasunich> nice
[23:47:14] <alex_joni> 3. halcmd (added linenumber to the error messages)
[23:47:19] <alex_joni> not sure if you like that..
[23:47:30] <jmkasunich> seems like a good idea to me
[23:47:33] <alex_joni> 4. halcmd (multi-line comments, rayh asked for that)
[23:47:58] <alex_joni> 5. emcsh equivalent in hal (so that hardware UI's can be built easily)
[23:48:06] <jmkasunich> how are those done (multi-line comments)?
[23:48:19] <alex_joni> not done yet.. wanted to talk to you about that
[23:48:21] <jmkasunich> ok
[23:48:26] <SWPadnos> various ways I've seen
[23:48:35] <SWPadnos> COMMENT <name>
[23:48:39] <alex_joni> I think .. /* */ , does that sound ok?
[23:48:40] <jmkasunich> what's wrong with # at the beginning of each line?
[23:48:44] <SWPadnos> when you see <name> next, the comment is over
[23:49:01] <alex_joni> jmkasunich: he complained that he wants to comment out large parts of a file (few doze lines)
[23:49:04] <SWPadnos> makes it hard to comment out (or uncomment) large blocks
[23:49:07] <jmkasunich> I wee
[23:49:09] <jmkasunich> see
[23:49:12] <jmkasunich> ;-)
[23:49:12] <SWPadnos> have fun ;)
[23:49:17] <alex_joni> I wee too.. but it doesn't help
[23:49:18] <jmkasunich> lol
[23:49:21] <SWPadnos> daily - it's good for you
[23:49:49] <SWPadnos> C-style is good too
[23:50:00] <alex_joni> this guys rock: http://www.vidking.com/viewvideo.php?id=843
[23:50:15] <jmkasunich> this is looking more and more like a language isnt it?
[23:50:20] <lerman> Would you support nested block comments?
[23:50:31] <jmkasunich> that isn't so pretty - even in C
[23:50:38] <alex_joni> yeah.. and macro's
[23:50:45] <jmkasunich> how about #if and #endif?
[23:50:46] <alex_joni> lerman: what'cha sayin?
[23:50:57] <jmkasunich> which also leads to #include
[23:51:05] <alex_joni> %include seems nicer :D
[23:51:23] <jmkasunich> oops, you're right, # is already used
[23:51:37] <lerman> When someone already has a block comment in the midst of a hundred lines of code and you want to comment out the hundred lines of code, what do you do? #if and #endif is how I do it in C.
[23:51:42] <jmkasunich> # came from bash - hal's language is pretty mich a scripting language
[23:52:00] <alex_joni> bash doesn't have multi-line comments, afaik
[23:52:02] <lerman> So, you could have %if and %endif.
[23:52:07] <jmkasunich> right, #if 0.. using comments to eliminste code is dangerous
[23:52:25] <alex_joni> but I think the whole thing started from rayh not finding where halcmd failed
[23:52:31] <lerman> But better than the alternatives.
[23:52:31] <jmkasunich> hal came from bash, hence single line comments
[23:52:34] <alex_joni> so maybe the linenumber is enough...
[23:52:50] <alex_joni> rayh: what do you think?
[23:53:00] <lerman> Well, we could require each line to begin with Nnnnnn :-)
[23:53:05] <jmkasunich> what I'd really like (grin) is math expressions
[23:53:10] <SWPadnos> bc
[23:53:35] <lerman> I (used to) write parsers for little languages in my sleep.
[23:53:45] <alex_joni> jmkasunich: I thought about %include foo.hal%, and that's pretty easy to do (spawn another halcmd and you're done)
[23:53:48] <jmkasunich> SWP, good point, bc can do the math,just need to be able to invoke it
[23:53:57] <SWPadnos> yep
[23:54:09] <jmkasunich> I wouldn;'t spawn another one, I'd just do a text include
[23:54:09] <SWPadnos> though variables could be an issue (from invocation to invocation)
[23:54:23] <alex_joni> jmkasunich: maybe something like usrload? but smthg that expects the result?
[23:54:35] <alex_joni> jmkasunich: why not spawn?
[23:54:40] <lerman> Would you allow one include file to include another?
[23:54:40] <jmkasunich> (push current file ptr ont a stack, open new file, when reach end, pop old file ptr and continue)
[23:54:46] <jmkasunich> sure
[23:55:23] <alex_joni> how about recurrent includes? (*big grin*)
[23:55:26] <SWPadnos> right - you can't spawn because of the lock
[23:55:34] <lerman> Well, you also would want to pushd the current file name, line number, etc, but its still pretty easy.
[23:55:38] <SWPadnos> or can you - hmmm
[23:55:52] <jmkasunich> the spawn method would fail if the language was extended to include macros or anything like that, because you could include a file that defines a macro, but when the spawned halcmd exited, the macro definition would vanish
[23:56:06] <SWPadnos> true
[23:56:22] <SWPadnos> some things could be done with environment space though
[23:56:36] <alex_joni> nah.. that's yucky
[23:56:37] <alex_joni> ;)
[23:56:44] <SWPadnos> icky, too :)
[23:57:08] <jmkasunich> A question... we're talking about adding bash-ish or C-ish features to the hal parser
[23:57:25] <alex_joni> I wonder why the diff on the emc-commit list (for the linenum stuff) so short is.. :(
[23:57:36] <jmkasunich> would it make more sense to figure out how to add hal-ish features to another scripting language that already has all the features we want?
[23:57:53] <alex_joni> m4 baby
[23:57:56] <SWPadnos> it depends on what halcmd is supposed to be
[23:58:11] <jmkasunich> there were so many small changes that it would have been very long, so it was suppressed
[23:58:17] <lerman> The only thing harder than writing m4 code is reading it.
[23:58:19] <rayh> Hi john
[23:58:22] <jmkasunich> hi ray
[23:58:35] <jmkasunich> m4 = spawn of the devil
[23:58:38] <alex_joni> rayh: we started talking about multi-line comments
[23:58:41] <SWPadnos> I don't thikn that things like macros are needed, because (assuming that halcmd can expose and control every aspect of a HAL setup), another scripting language can use halcmd
[23:59:00] <alex_joni> rayh: do you think those are still needed, if you have linenumber printed on errors?
[23:59:13] <rayh> No.
[23:59:25] <SWPadnos> as long as halcmd can handle a piped stdin, you'd have the best of all possible worlds, I think
[23:59:46] <rayh> Once I have something I can use to see where in the file the error is I should be fine.
[23:59:54] <SWPadnos> hmmm - can halcmd handle "halcmd < foo.hal