#emc-devel | Logs for 2007-08-31

[00:00:31] <jepler> skunkworks: I'd like to do one tonight but I can't remember what project was waiting on a circuit board :-P
[00:00:58] <skunkworks> heh
[00:01:37] <skunkworks> when did you fix the com timing issue with the pluto? is that on the livecd?
[00:01:56] <jepler> yes, it was fixed by the 2.1.2 release
[00:02:01] <skunkworks> ok
[00:36:26] <jmkasunich> jepler: I've noticed in my webserver logs that several robots hit me on a daily basis
[00:37:36] <fenn> jepler you might have to mill the table flat
[00:37:49] <jmkasunich> fenn: that will take a very long time
[00:37:52] <fenn> yep
[00:38:12] <fenn> "flat"
[00:39:15] <fenn> well, you could put some foam on top, mill that flat, and use it as a disposable table surface.. should cut faster at least
[00:39:24] <fenn> pink styrofoam
[00:39:30] <jmkasunich> his spindle is a dremel
[00:39:47] <fenn> there's dremel milling cutters
[00:39:56] <jmkasunich> its still a dremel
[00:40:00] <fenn> so?
[00:40:01] <jmkasunich> not very stiff
[00:40:36] <fenn> i cut plexiglass with a dremel before.. it was horrible but it did cut it
[00:41:05] <SWPadnos> jmkasunich, can you look at a paragraph in the AD7656 datasheet and tell me if it means what I think it means?
[00:41:19] <jmkasunich> sure
[00:41:55] <cradek> fenn: it seems like it would just melt it, not cut
[00:42:31] <cradek> maybe jepler could map the table and use kinematics ... nah
[00:43:07] <jmkasunich> how bad is it?
[00:43:17] <cradek> don't know yet
[00:43:27] <fenn> you could put a lathe bit on it and pretend it's a shaper :)
[00:43:39] <cradek> oh hey did I tell you guys we found the probe for max?
[00:43:46] <SWPadnos> maybe there's a flycutter for the dremel
[00:43:51] <cradek> SWPadnos: haha
[00:43:53] <SWPadnos> hahahahaha - I kill me
[00:44:11] <SWPadnos> "new, 0.3" flycutter"
[00:44:12] <fenn> i have a 3/8" endmill/bur
[00:44:31] <fenn> it came with my dad's dremel
[00:44:44] <cradek> the table is attached on two ends. I'd recommend shimming it as well as possible and call it good
[00:44:44] <SWPadnos> there are relatively large grinding bits, but tha's probably not the greatest for plastic
[00:45:04] <fenn> isnt the table like fiber board anyway?
[00:45:16] <cradek> it's that fancy plywood
[00:45:38] <cradek> 11-ply I think, it's thick
[00:46:01] <fenn> i bet a roughing endmill like used on circuit boards would work well enough
[00:48:19] <fenn> oh it has steel inserts
[00:53:14] <cradek> another problem is it can't cut to the edge of the table no matter what
[00:54:46] <fenn> styrofoam is a good material for a machine like that
[00:55:34] <jmkasunich> a wooden table?
[00:55:56] <fenn> ground rods and a dremel
[00:56:33] <fenn> i think grinding bits give the best finish, and it turns the chips into dust instead of fuzzy fluffy crap that sticks to everything
[00:56:50] <fenn> double cut burs might be better though, havent tried
[00:57:19] <jmkasunich> fuzzy fluffy crap....
[00:57:24] <jmkasunich> * jmkasunich shakes his head
[00:57:30] <fenn> what?
[00:57:35] <jmkasunich> I prefer chips that go clink
[01:05:32] <fenn> hey jepler what do you think of making a standard ieee1284 header file?
[01:06:12] <fenn> i mean instead of being a part of pluto-*
[01:08:19] <jepler> fenn: I'm not against it .. I doubt it'll go right to the top of my list though
[12:37:07] <skunkworks_> yes - I like chips that ricochet off the ceiling.
[12:37:22] <skunkworks_> you start cutting and have to walk away so you don't get hurt
[12:37:48] <skunkworks_> guard - we don't need no stinkin guards
[12:50:49] <jepler> hi skunkworks_
[12:51:40] <jepler> never did find a circuit board to mill last night, though I did bleed on the machine a little
[12:51:51] <skunkworks_> eww
[12:52:08] <jepler> inside the electronics box there was a pin header sticking up, and I scratched open a knuckle on it -- ouch
[12:52:48] <skunkworks_> yeck. I hate that.
[12:54:54] <skunkworks_> I tried to cut a circuit board last night - but using the tooling that came with it - the smallest tip is .015 - so I need to make up a board that has enough space between traces
[12:56:44] <skunkworks_> I actually tried to cut your quadture divider :)
[12:57:38] <jepler> the attiny one?
[12:58:34] <skunkworks_> yes - the surface mount one. I had to use a cutter diameter of .007 just to get all the taces to show up.
[13:02:40] <skunkworks_> so cutting with a .015 cutter made for some interesting traces ;)
[13:03:50] <jepler> on this machine I have a slightly different version that passes drc when I set it to 15/15 trace/sep
[13:05:09] <jepler> it's very similar, it looks like I moved the "R" programming point a bit and have a ground trace instead of a ground "plane"
[13:07:37] <jepler> it's also got two holes (vias) on the ground trace; I think I envisioned having different division tables in the AVR firmware which could be selected by adding a jumper
[13:07:56] <jepler> but there's only one pin to be used in this way ("S") so I wonder why I have two holes on GND ...
[13:08:40] <jepler> ("R" is connected to reset; required for programming but not too useful as a gpio)
[13:09:11] <jepler> if you want to try it, I put the 15/15 version here: http://emergent.unpy.net/files/sandbox/quadrature_divider.brd http://emergent.unpy.net/files/sandbox/quadrature_divider.sch
[13:09:21] <jepler> * jepler begins to wonder whether he's talking to himself
[13:11:19] <skunkworks_> sorry - had to fix some problems :
[13:11:20] <skunkworks_> :)
[13:11:47] <skunkworks_> and raid the vending machine
[13:15:57] <jepler> you have a RAID vending machine?
[13:15:59] <jepler> that's pretty cool
[13:16:12] <skunkworks_> heh
[13:16:40] <jepler> bbl
[13:48:27] <cradek> can someone who has not run cl during this reboot please try running demo_sim_cl?
[13:48:53] <jepler> I doubt I have; I'll try it
[13:49:13] <cradek> be sure you're updated first
[13:49:25] <jepler> cradek: feel free to reboot install8 or whatever number it came up as this time, too
[13:49:52] <cradek> oh good idea, I had only tried it in -simulator
[13:51:21] <jepler> it works on india / simulator
[13:51:44] <cradek> strange, I must still have droppings somewhere
[13:53:03] <jepler> looks like shared memory segment 0x434c522b is classicladder's
[13:53:08] <jepler> ('CLR+')
[13:53:31] <jepler> when I exited normally it did get cleaned up
[13:54:29] <cradek> yay, that fixed it, thanks
[13:54:37] <cradek> glad I didn't email him (I almost did)
[13:55:28] <cradek> darn, I thought we were getting signal names in the gui
[13:55:37] <cradek> that's the feature I really want
[13:55:39] <skunkworks_> I thought fenn had done that..
[13:55:51] <skunkworks_> a while ago
[13:55:54] <skunkworks_> or did I just dream that.
[13:56:06] <cradek> hmm the window with all those checkbuttons is gone too
[13:56:23] <cradek> oh I get it if I poke the vars button
[13:57:18] <cradek> Symbols?
[14:08:08] <cradek> jepler: can AXIS tell if CL is running, and if so, add a menu entry to start the CL gui?
[14:08:38] <skunkworks_> skunkworks_ is now known as skunkworks
[14:10:07] <jepler> cradek: good questions
[14:13:16] <cradek> symbols.... wonder what they are
[14:13:38] <cradek> they seem to be a way to name %B0 => CLinB0 but what I really want is %B0 => estop
[14:16:10] <SWPadnos> I had a problem with the symbols window when I tried changing things
[14:16:20] <SWPadnos> the new names always got sorted to the bottom, out of view
[14:16:42] <cradek> I can't figure out how to make the symbols window do anything
[14:16:57] <cradek> (it's just empty)
[14:16:58] <SWPadnos> so it looked like things were just disappearing and/or keeping their blank names
[14:17:05] <SWPadnos> can you scroll down?
[14:17:15] <cradek> no
[14:17:28] <SWPadnos> ok, I may be thinking of a different window - it's been a while
[14:18:03] <cradek> http://timeguy.com/cradek-files/emc/symbols-names.png
[14:18:34] <SWPadnos> err - ok, that looks a bit sparse
[14:25:24] <cradek> it's broken, it isn't supposed to do what I want anyway, or I don't know how to use it: pick some or all of the above
[14:25:51] <cradek> oh hmm, jepler already fixed it
[14:26:13] <jepler> it looks like hell though
[14:26:17] <jepler> not enough room for the labels
[14:26:28] <cradek> hmm
[14:27:16] <SWPadnos> was one of the goals to make the pin names in CL human-readable?
[14:27:18] <cradek> the window resizes nicely
[14:27:27] <cradek> damn, that's so cool, just what I wanted
[14:27:30] <cradek> thanks jepler
[14:27:53] <jepler> cradek: your task is to test that I got it right for %X (integers)
[14:28:13] <jepler> the in and out might be reversed(?)
[14:28:14] <cradek> ummmm
[14:29:00] <SWPadnos> what happens if a pin is dinconnected then connected to a different signal?
[14:29:29] <SWPadnos> uh - disconnected
[14:29:30] <jepler> SWPadnos: I *think* that what I wrote checks the name at every redraw but I didn't actually test what happens then
[14:29:58] <SWPadnos> ok. I guess it seems inefficient to do that, but if you don't, you get sig name droppings
[14:30:06] <SWPadnos> no perfect solution there
[14:37:51] <jepler> cradek: in which menu does the 'open classicladder gui' item go?
[14:39:00] <cradek> I guess it should go with halmeter/halscope/etc
[14:39:14] <jepler> that menu needs reamed
[14:39:42] <cradek> yeah, at least that configuration/debugging section sould be split out
[14:40:32] <cradek> hm maybe it needs more than that even
[14:41:42] <cradek> since things enable and disable, we'll introduce bugs if we change the menus, won't we
[14:43:59] <jepler> oh probably -- but at least the enable and disable code refers to menu texts instead of indexes now
[14:44:11] <cradek> oh I forgot - that makes it pretty safe then
[14:44:15] <cradek> it was bad before
[14:47:06] <cradek> wow
[14:47:15] <cradek> * cradek tries to think of more features to request this morning
[14:47:46] <jepler> hah
[14:47:48] <cradek> vars.has_ladder.set(hal.component_exists('classicladder_rt'))
[14:47:49] <cradek> AttributeError: Variables instance has no attribute 'has_ladder'
[14:47:55] <jepler> argh
[14:47:56] <jepler> but it worked for me!
[14:48:19] <jepler> of course it is because I am incompetent
[14:49:32] <cradek> jepler: AXIS is frozen while the ladder editor is up
[14:52:38] <jepler> geez nothing makes you happy does it
[14:54:49] <jepler> I'm glad I have you to test my code since I apparently don't bother -- I only looked whether it was greyed out or not
[14:54:52] <jepler> and I guess I got that wrong too
[14:58:01] <cradek> I wonder if there's a good way to tell an in from an out
[14:58:10] <cradek> maybe label color?
[14:58:24] <cradek> no, that would suck
[14:59:01] <cradek> <some input
[14:59:02] <cradek> some output>
[14:59:08] <jepler> =>signame or signame=>
[14:59:10] <jepler> I dunno
[14:59:33] <SWPadnos> signame_{in,out} ?
[15:00:09] <SWPadnos> this is in the ladder display?
[15:00:12] <cradek> yes
[15:00:31] <SWPadnos> on the diagram or in list-type dialogs?
[15:00:45] <cradek> http://timeguy.com/cradek-files/emc/ladderdisp.png
[15:00:58] <cradek> estop-out is an input
[15:01:03] <cradek> estop-enable-in is an output
[15:01:17] <cradek> (to/from cl)
[15:01:44] <SWPadnos> hmmm. I was thinking it could be as simple as left side=inputs, right side=outputs
[15:01:47] <SWPadnos> but I was wrong :)
[15:01:59] <cradek> yeah it's not quite like that
[15:02:18] <SWPadnos> the "right side=outputs" assertion is more or less correct, but that doesn't help when the same thing has multiple contacts in the diagram
[15:02:30] <SWPadnos> like estop-enable-in
[15:02:45] <cradek> yes coils not marked %B are outputs
[15:02:50] <cradek> that's all you can know for sure I think
[15:03:25] <SWPadnos> arrows may be the best bet
[15:03:30] <cradek> I think so too
[15:04:01] <SWPadnos> I wonder if there are any arrow characters, like the old PC ASCII chars 3-6 or thereabouts
[15:04:19] <SWPadnos> (in the default fonts on Ubuntu, that is)
[15:04:33] <cradek> surely there are « and »
[15:04:51] <jepler> \N{LEFTWARDS ARROW} U+2190 <-
[15:04:54] <jepler> \N{RIGHTWARDS ARROW} U+2192 ->
[15:04:56] <SWPadnos> A << A >> is more or less what I got there :) (plus an accent, and a double-angle char)
[15:05:11] <jepler> hmm looks like irssi helpfully transcribed that paste into ASCII
[15:05:16] <cradek> oh, looks right here :-)
[15:05:17] <jepler> I saw cradek's characters
[15:05:32] <SWPadnos> all you need is the right arrow - before for input, after for output
[15:05:43] <cradek> true, I think
[15:06:17] <SWPadnos> chard U+2191 and U+2193 are probably up/down arrows, which can be used at some point to show transitions
[15:06:21] <SWPadnos> chars
[15:06:41] <jepler> the question is how non-ASCII chars are to be represented in "C" source for gtk+ ..
[15:06:46] <cradek> yeah
[15:06:56] <SWPadnos> oh, well that could be an issue
[15:07:13] <cradek> or we could use >
[15:08:17] <SWPadnos> now where have I seen that before?
[15:08:30] <SWPadnos> how about -> ;)
[15:09:09] <cradek> or maybe » would just work
[15:09:41] <SWPadnos> it could, hard to tell on Chatzilla / Windows ...
[15:09:49] <jepler> http://library.gnome.org/devel/gtk/unstable/gtk-question-index.html#id2975698
[15:10:11] <jepler> looks like the most reliable way would be to write \-escapes of the UTF-8 encoding in the source code
[15:10:37] <cradek> that's easy enough
[15:10:46] <cradek> least likely for someone's editor to screw up too
[15:11:41] <cradek> isn't writing \xa8abcd going to give an endian problem?
[15:12:09] <cradek> (I'm not sure how \x works for things longer than a byte)
[15:12:20] <cradek> brb
[15:12:40] <jepler> '\xe2\x86\x90'
[15:13:01] <jepler> the order of the bytes in a byte stream is well defined for utf-8
[15:44:29] <jepler> unfortunately, with the font gtk uses on my machine, the -> unicode arrow is ugly
[15:47:04] <jepler> RIGHTWARDS DOUBLE ARROW looks good though
[15:47:15] <SWPadnos> it seems a little long
[15:47:39] <SWPadnos> RIGHTWARDS DOUBLE ARROW estop-reset might not fit on low resolution screens ;)
[15:47:52] <jepler> http://emergent.unpy.net/files/sandbox/cl-arrows.png smart ass
[15:48:06] <cradek> looks great to me
[15:48:19] <SWPadnos> yeah - that's a nice character
[15:49:10] <cradek> but ... it's still kind of confusing
[15:49:35] <cradek> =>signal could mean that we're writing to the signal
[15:49:41] <jepler> OK so what do you want?
[15:49:59] <cradek> thinking...
[15:50:20] <jepler> I guess I should have waited to check it in ..
[15:50:31] <cradek> my fault
[15:50:40] <SWPadnos> I think the label should fade between the signal name and "(input)" or "(output)", with a user-settable cross-fade time
[15:51:08] <cradek> I think maybe it should be opposite what you did
[15:51:08] <jepler> if it's at the right of course it's an output; if it's not at the right, does it matter?
[15:51:13] <cradek> * cradek kicks SWPadnos
[15:51:16] <SWPadnos> ouch
[15:51:18] <SWPadnos> :)
[15:52:03] <SWPadnos> I think it's right in the screenshot
[15:52:07] <cradek> jepler: yeah I think it matters
[15:52:11] <cradek> SWPadnos: hmm.
[15:52:30] <cradek> I guess I can handle either way
[15:52:31] <SWPadnos> there's always a problem with the frame of reference
[15:53:03] <cradek> I was seeing 'into' the signal, 'from' the signal
[15:53:32] <SWPadnos> yeah, I can see the way you're seeing it
[15:53:41] <SWPadnos> "this goes to estop"
[15:53:44] <SWPadnos> err
[15:53:49] <SWPadnos> "this goes to ext-estop"
[15:53:55] <cradek> yeah
[15:54:30] <SWPadnos> it doesn't help that the names are also backwards 2/3 of the time
[15:54:32] <jepler> make it whichever way you can agree makes more sense
[15:54:45] <jepler> if you can't agree, just make it random
[15:55:09] <cradek> SWPadnos: in this example, the names suck - it's not a general problem
[15:55:22] <SWPadnos> heh
[15:55:31] <SWPadnos> it's when the names suck that the arrows are needed ;)
[15:55:53] <cradek> could be
[15:56:43] <SWPadnos> for anything with a "-in" or "-out" name, it sill be confusing at least 50% of the time
[15:57:02] <SWPadnos> since it would connect to at least two components, and for one of them the name will be wrong
[15:57:20] <cradek> yeah those are dumb names
[15:57:23] <jepler> so maybe that's the wrong name for a signal
[15:57:49] <SWPadnos> maybe some naming convention should be on the discussion list for post-2.2
[15:57:50] <jepler> I'm not sure what the right names are
[15:58:30] <SWPadnos> luckily, they're pretty easy to changee
[16:00:46] <cradek> this sure makes ladder easier to use
[16:00:50] <cradek> it was really a pain
[16:01:43] <jepler> how, uh, does it work when you are editing rungs?
[16:02:57] <jepler> you still have to figure out %B0 or whatever
[16:03:13] <cradek> argh
[16:03:20] <cradek> yeah you really have to turn off the symbols
[16:03:38] <jepler> then it replaces it with the "other" string, and then when you hit 'apply' again it errors
[16:03:46] <jepler> it sucks
[16:03:58] <jepler> (the other string including the arrow you can't possibly type)
[16:03:59] <jepler> sigh
[16:04:11] <jepler> I (opposite of heart) classicladder
[16:04:27] <cradek> if I turn off the symbols I can edit fine
[16:05:37] <jepler> how do you "turn off the symbols"?
[16:05:49] <cradek> checkbutton at the top of the main window
[16:06:26] <jepler> ah
[16:11:51] <fenn> hey that's great - signal names. i couldnt figure out how to get the names out of hal
[16:14:21] <cradek> I'm not sure the new names are perfect, but they're less confusing
[16:15:27] <fenn> i was thinking you could display an abbreviation if the signal name was too long, and the whole thing would pop up as a tooltip when you mouse over the coil/contact
[16:15:55] <cradek> I also thought about tooltips with the signal name, but I like this much better
[16:16:04] <fenn> yeah
[16:16:05] <cradek> otherwise you'd have to waggle the mouse over everything to see what's going on
[16:19:25] <cradek> I wonder if when (whoever) wrote this there weren't edge triggers
[16:19:58] <cradek> I replaced estop-strobe (which comes from a oneshot hal component) with a -|^|- hooked to gui-estop (%I0)
[16:23:43] <jepler> dunno
[16:24:42] <cradek> whee
[16:24:53] <cradek> this is great
[16:43:39] <SWPadnos> heh - that commit message is pretty funny
[16:44:21] <SWPadnos> fix confusing signal names
[16:44:22] <SWPadnos> ...
[16:44:23] <SWPadnos> -0-0-0/0 , 0-0-0/0 , 1-1-50/4 , 99-0-0/0 , 10-0-0/2 , 0-0-0/0 , 0-0-0/0 , 0-0-0/0 , 0-0-0/0 , 0-0-60/2
[16:44:25] <SWPadnos> +0-0-0/0 , 0-0-0/0 , 1-1-50/4 , 99-0-0/0 , 10-0-0/2 , 103-0-0/0 , 103-0-0/0 , 103-0-0/0 , 103-0-0/0 , 103-0-60/9
[16:44:26] <SWPadnos> ...
[16:44:29] <SWPadnos> like that's less confusing ;)
[16:44:58] <cradek> yeah I have no idea what that is
[16:46:47] <fenn> looks like a .clp file
[16:47:46] <SWPadnos> yeah - it's the clp file
[16:48:07] <SWPadnos> it just looked funny when I saw the first 15 lines or so of the message
[18:14:25] <cradek> when I turn machine off, I see emcLubeOff() is called, but iocontrol.0.lube only occasionally turns off in response
[18:16:10] <cradek> and I don't see how to turn on the debug prints for iotask... seems like there aren't any?
[18:16:17] <cradek> * cradek wishes alex were here
[18:28:11] <cradek> turning on rtapi debug prints confirms that iotask is often not getting the message
[18:29:30] <SWPadnos> err - that's a problem
[18:29:53] <SWPadnos> is there any change in the ordering of messages?
[18:30:06] <cradek> hmm something smells around line 815 of ioControl.cc
[18:30:53] <cradek> oh, maybe not, the comments explain? it
[18:31:14] <jepler> beats me!
[18:42:53] <cradek> while(...) if(...) continue; else break;
[18:43:05] <SWPadnos> heh
[18:43:15] <cradek> crack monkey
[18:57:21] <cradek> any bug involving rcs/nml is sure muddy to me
[18:58:16] <SWPadnos> more like quicksand for me
[18:58:19] <alex_joni> hi
[18:58:25] <cradek> hi alex
[18:58:28] <SWPadnos> speak of the devil^Wangel
[18:58:35] <SWPadnos> hi alex
[18:58:35] <alex_joni> were you looking for me?
[18:59:17] <alex_joni> SWPadnos: :P
[18:59:18] <alex_joni> hi
[19:00:39] <alex_joni> SWPadnos: did you fix that for skunkworks ?
[19:00:52] <SWPadnos> oh - no, I haven't
[19:00:59] <alex_joni> * alex_joni tries to
[19:01:05] <SWPadnos> it's just filtered with robots.txt
[19:01:09] <SWPadnos> I think
[19:01:40] <alex_joni> wasn't it dinero?
[19:02:05] <SWPadnos> well, it seems it doesn't matter, because I can't get to linuxcnc anyway ~:
[19:02:08] <SWPadnos> :~
[19:02:14] <SWPadnos> ~)
[19:02:16] <SWPadnos> hmm
[19:02:25] <SWPadnos> anyway :/
[19:02:27] <SWPadnos> there
[19:02:48] <SWPadnos> ok, now it's connected
[19:03:12] <skunkworks> alex_joni: thank you very much
[19:04:00] <alex_joni> skunkworks: done
[19:04:06] <alex_joni> rdf next
[19:04:33] <skunkworks> thank you - (that was a stupid mistake)
[19:04:36] <alex_joni> skunkworks: please check if it's ok
[19:05:17] <skunkworks> yes
[19:05:26] <skunkworks> gone
[19:05:45] <cradek> wow this bug has me beat
[19:06:00] <cradek> sometimes an io nml message will fire twice, or zero times
[19:06:19] <alex_joni> cradek: really?
[19:06:21] <cradek> yep
[19:06:23] <alex_joni> can you give me some more info?
[19:06:26] <skunkworks> wasn't maddash seeing that?
[19:06:44] <cradek> alex_joni: iocontrol.0.lube is supposed to go on and off with 'machine on'
[19:06:47] <cradek> it always turns on
[19:06:55] <cradek> it rarely, but sometimes, turns off
[19:07:31] <cradek> I see the messages are EMC_LUBE_ON, _OFF
[19:07:54] <alex_joni> right
[19:07:59] <cradek> I see EMC_LUBE_OFF being sent with a fresh serial number, but it never shows up at iocontrol.cc:main
[19:08:55] <cradek> sometimes when EMC_LUBE_ON is sent, iocontrol.cc:main runs it twice
[19:10:06] <cradek> I also sometimes see EMC_TOOL_ABORT firing twice when I hit 'machine off'
[19:10:10] <alex_joni> did you have some estop before that?
[19:10:39] <cradek> no if I just hit F2 over and over, sometimes I'll get EMC_TOOL_ABORT twice
[19:10:56] <alex_joni> that's task again I think
[19:11:12] <alex_joni> (there are some commends in task about aborting twice I think)
[19:11:44] <cradek> occasionally I can get EMC_LUBE_ON twice, I suspect they are the same symptom
[19:12:00] <alex_joni> LUBE doesn't get cleared by TOOL_ABORT
[19:19:22] <alex_joni> shame I can't really help from here :/
[19:19:40] <alex_joni> cradek: if you can't nail it.. do report it and assign it to me
[19:20:26] <cradek> I did report it :-)
[19:20:47] <cradek> I'll keep trying though
[19:20:51] <cradek> brb
[19:51:35] <alex_joni> bye all
[19:51:40] <alex_joni> back way later