#emc | Logs for 2007-04-07

[01:07:24] <tomp> re: the graphic hal schematics.
[01:07:32] <tomp> if the user adds a parallel port, the pins might be individually moveable or monolithic, like a db25.
[01:07:38] <tomp> if individual, the user can add his names and locations and wires at will.
[01:07:46] <tomp> i have no auto numbering scheme so added a new attribute 'hal-name' that can be anything user types in.
[01:07:50] <tomp> sufficient? http://imagebin.org/8030
[01:08:54] <tomp> already, communicating with this will require the other person to have the same library, and the graph file ( schematic )
[01:28:13] <jmkasunich> hi tomp
[01:29:15] <jmkasunich> the library will become part of EMC2, so everybody will have that
[01:29:57] <jmkasunich> and the schematic file will be a top-level config item (it becomes the master for a .hal file)
[01:31:45] <tomp> oh! i better get crackin ;)
[01:32:10] <jmkasunich> don't spend enormous amounts of time making component outlines
[01:32:21] <jmkasunich> we should (and will) figure out how to automate that
[01:32:46] <jmkasunich> the general exploration of idea that you are doing now is a good thing
[01:33:08] <jmkasunich> personally, I think each instance of a component should be monolythic
[01:33:18] <jmkasunich> a parport is a big box, with a bunch of pins
[01:33:38] <tomp> no , i focus on pins, the outlines are almost automatic, and the packages can be rotated, but indiv pins cant be moved, and hal names and data types are included
[01:33:48] <tomp> (youre ahead of me... )
[01:34:27] <tomp> so a whole db25 instead of indiv pins? i cant determine the sense of the pins tho...
[01:35:09] <jmkasunich> I would think that "parport.0" would become the "reference designator" for that component (like R1, or U3 in a regular schematic)
[01:35:28] <jmkasunich> and the pins would be named after the part of their HAL names that follows "parport.0"
[01:35:43] <jmkasunich> like "pin-01-in" or "pin-01-in-not"
[01:37:26] <tomp> is the direction (or some directions ) decided in the hal file? i am not parsing anything yet
[01:37:34] <jmkasunich> direction?
[01:37:42] <tomp> in/out
[01:38:05] <jmkasunich> hal files don't care about direction - if you try to connect two outs, it will complain when you run the hal file
[01:38:42] <tomp> so you're thinking these graphs are direction-less, just connected elements?
[01:38:43] <jmkasunich> I don't expect to check for that at the schematic level
[01:39:30] <jmkasunich> if somebody puts two muxes on a page and connects their outputs together with a net called "short-circuit", I'd expect the result to be a hal file that includes "net short-circuit mux.0.out mux.1.out"
[01:39:39] <jmkasunich> that will fail when you run it
[01:40:30] <tomp> ok, i've been thinking/working with direction and data size, i'll just create components then
[01:40:32] <jmkasunich> I dunno if geda has any feature that will warn you if you try to connect two outputs, but if they don't I'm not gonna lose any sleep over it
[01:41:44] <tomp> gEDA doesnt care about these as far as i can tell, and i'm adding 'attributes' for hal names and data types, might be useful later for some checking
[01:42:01] <jmkasunich> attributes to each pin?
[01:42:10] <jmkasunich> or attributes to the component as a whole?
[01:42:17] <tomp> no, to components , it has free form fields
[01:42:35] <jmkasunich> so what names or types are you sticking in attributes?
[01:42:41] <tomp> tho pins also can have such attributes adde3d, like maybe bus width
[01:42:51] <tomp> i added hal-name and hal-data-type
[01:42:59] <jmkasunich> hal name of what?
[01:43:25] <tomp> whatever, like parport-0.iforgettheconvention
[01:43:43] <jmkasunich> the component mux has 4 pins, 3 of type float and one of type bit - you cannot attach a type attribute to the _component_ and have it make any sense
[01:45:00] <tomp> right, but the component 'constant' has an element 'value' that can have such an attribute ( hal-name was common to components, hal-data-type was common to pins)
[01:45:19] <tomp> it's all very new to me
[01:46:15] <jmkasunich> constant is a horrible example
[01:46:23] <jmkasunich> since it has only one pin and one parameter
[01:46:57] <jmkasunich> that will lead you astray
[01:47:14] <jmkasunich> lets talk about mux
[01:47:18] <tomp> yes, maybe theres a serialized hal-comp-name and a 'sub'-serialized hal-pin-name
[01:47:21] <tomp> ok
[01:47:29] <jmkasunich> it has 4 pins: in0, in1, sel, and out
[01:47:39] <jmkasunich> those pin names should be the names of the GEDA pins
[01:47:59] <tomp> ? gain0 gain1 (6 pins?)
[01:48:12] <jmkasunich> mux doesn't have gain pins
[01:49:14] <tomp> ok (fictitious example i invented, sorry )
[01:49:23] <jmkasunich> man mux2
[01:49:36] <tomp> got the hal user manual
[01:49:45] <tomp> ok 4 pins yep
[01:50:01] <tomp> and those are the names i'd use
[01:50:18] <jmkasunich> if GEDA allows you to associate attributes with specific pins, then you could attach the type, and the in/out, to each pin
[01:50:43] <jmkasunich> if it only allows attributes for the component as a whole, then types and directions should be completely ignored, NOT stuck on the compoent
[01:51:24] <jmkasunich> man mux2 is important - because it points out what can be automated - that man page is automatically generated by comp
[01:51:38] <jmkasunich> the geda symbols should also be generated the same way
[01:52:15] <tomp> gEDA allows pintypes in out & io, (and others we might not need like clock or hiZ ) and allows the pin to have custom attributes, as many as you like
[01:52:27] <jmkasunich> ok, thats good
[01:52:35] <jmkasunich> so we can have pin attributes of direction and type
[01:54:23] <tomp> ok, will use man <comp> to clarify info for now, and use the list from the src/hal/components directory as my table of contents
[01:54:41] <jmkasunich> sounds like a plan
[01:54:49] <tomp> k, thanks
[01:56:34] <tomp> question: there's no place for functions ( nor need as far as i can see in a graph) and parameters look like pins to me ( same graphical representation ), is that cool?
[01:57:00] <jmkasunich> functions are something that needs treated separately
[01:57:24] <jmkasunich> it would be nice if parameters weren't on the outside of the component, they represent "internal" adjustments
[01:57:51] <jmkasunich> I had thought about using attributes for parameters - I dunno if geda lets a user click on an attribute and change its value
[01:58:01] <tomp> i'll see what can be done to make them internalized
[01:59:13] <tomp> theres some attribute editors for gEDA, they say its like a spreadsheet, i say its like VB or VC attributes, a dialog pops up and you change values for given params
[01:59:16] <tomp> they come in the geda-utils package
[02:00:02] <tomp> http://www.geda.seul.org/tools/gattrib/gattrib.png
[02:00:27] <tomp> back to it :)
[03:39:42] <ds3> Hmmm
[04:47:20] <tomp> hal component 'constant' with attributes from 'man constant' and naming conventions like hal http://imagebin.org/8034, name & value easily edited in gEDA schematic capture app 'gschem'
[05:08:58] <toast> toast is now known as toastydeath
[07:23:27] <Jymmm> Jymmm has changed the topic to: Welcome! EMC (Enhanced Machine Controller) is a linux-based CNC control. | Latest release: EMC 2.1.4 | http://www.linuxcnc.org | http://wiki.linuxcnc.org || http://tavmjong.free.fr/INKSCAPE/DRAWINGS/clock_plain.svg
[08:33:31] <alex_joni> alex_joni has changed the topic to: "Welcome! EMC (Enhanced Machine Controller) is a linux-based CNC control. | Latest release: EMC 2.1.4 | http://www.linuxcnc.org | http://wiki.linuxcnc.org"
[08:34:49] <alex_joni> alex_joni has changed the topic to: Welcome! EMC (Enhanced Machine Controller) is a linux-based opensource CNC control. | Latest release: EMC 2.1.4 | http://www.linuxcnc.org | http://wiki.linuxcnc.org
[08:44:50] <Jymmm> Jymmm has changed the topic to: Welcome! EMC (Enhanced Machine Controller) is a linux-based opensource CNC control. | Latest release: EMC 2.1.4 | http://www.linuxcnc.org | http://wiki.linuxcnc.org || http://tavmjong.free.fr/INKSCAPE/DRAWINGS/clock_plain.svg
[08:49:38] <alex_joni> you got a thing for that clock don't you?
[08:50:12] <Jymmm> it's animated, I think Chris would like it and I wwouldn't mind leanring to CNC them
[08:51:22] <Jymmm> Since it's SVG, I was wondering if it could be enlarged then CNC'ed and hold accurate time.
[08:52:31] <alex_joni> hmm.. you just need the ratio
[08:52:42] <alex_joni> and find/machine the proper gears
[08:52:59] <Jymmm> just not sure on the seconds
[08:54:29] <alex_joni> Dallur: wow.. just seen http://imagebin.org/8006
[09:10:54] <robin_sz> meep?
[09:12:28] <alex_joni> hi robin
[09:13:06] <alex_joni> robin_sz: here's what was lacking from emc the last time you tried plasma: http://imagebin.org/8006
[09:25:54] <robin_sz> mmm ... nice!
[09:26:16] <robin_sz> lots of work gone on there then
[09:26:39] <Jymmm> ok I'm outta here
[09:27:29] <alex_joni> robin_sz: this is all HAL.. so it's up to the user/integrator how they use it
[09:28:13] <alex_joni> bbl..
[09:28:18] <robin_sz> 'k
[09:28:27] <robin_sz> I can't imagine many users figureing that lot out
[09:28:59] <robin_sz> I guess we'll have to make up pre-configure HAL/VCP setups like that for specific applications
[09:29:38] <robin_sz> it seems to have everyting though ... I see torch up/down is bit based ...
[09:30:03] <robin_sz> I always fancied analogue in
[09:30:13] <robin_sz> and some sort of PID loop
[11:32:36] <Rugludallur> hmm is it not possible to run pyvcp standalone any more ?
[11:33:14] <Rugludallur> or did I misunderstand, it still requires HAL even in standalone mode ?
[11:46:38] <Martin_Lundstrom> hello everyone
[11:46:50] <Martin_Lundstrom> Dallur, are you around?
[11:46:51] <Rugludallur> hey
[11:46:58] <Rugludallur> Sure am :D
[11:47:12] <Martin_Lundstrom> hey, how is the work going?
[11:47:28] <Rugludallur> Martin: I have not had time to do anything for the last 3 weeks but I'm using easter to work on it
[11:47:52] <Rugludallur> Martin: http://imagebin.org/8006
[11:48:17] <Rugludallur> Martin: i'm finishing the PyVCP and HAL stuff today, based on that design
[11:48:21] <Martin_Lundstrom> ok, I got the last HTD wheels for my table
[11:48:49] <Rugludallur> Martin_Lundstrom: great, when do you think they will be operational
[11:50:00] <Martin_Lundstrom> hopefully partly operational during next week
[11:50:27] <Rugludallur> Martin: great
[12:00:01] <Rugludallur> hmm has anyone here used "labelframe" in pyvcp, for some reason it's not rendered nor is anything inside it, wondering if it might be a python version thing
[12:10:44] <Rugludallur> Martin_Lundstrom: welcome back
[12:14:59] <Martin_Lundstrom> thanx
[12:15:45] <Rugludallur> Martin_Lundstrom: btw, if there is any more info or features you want for the new config now would be a good time to air them
[12:16:14] <Rugludallur> Martin_Lundstrom: if you want some info shown in Axis UI or anything
[12:16:21] <Martin_Lundstrom> I think we already aired them!?
[12:16:48] <Rugludallur> Martin_Lundstrom: We did, just a final check :D
[12:17:13] <Martin_Lundstrom> why dont we start with a basic config that has all basic necesary features
[12:17:25] <Rugludallur> Martni_Lundstrom: We can always add later :D
[12:17:28] <Martin_Lundstrom> and when that works well
[12:17:38] <Martin_Lundstrom> add more
[12:18:21] <Martin_Lundstrom> did you have a look at my code?
[12:18:44] <Rugludallur> Martin_Lundstrom: yup, I did , I think I have seen some of it before also
[12:19:16] <Martin_Lundstrom> I just use a giant IF to check for wich status
[12:19:32] <Rugludallur> Martin_Lundstrom: have you had any problems because it's not realtime ?
[12:19:47] <Martin_Lundstrom> ?
[12:20:03] <Martin_Lundstrom> its realtime enough I think
[12:20:06] <Rugludallur> Martin_Lundstrom: Because shell scripts run in user space, not realtime
[12:20:27] <Martin_Lundstrom> the shell part is not realtime
[12:20:54] <Martin_Lundstrom> but the shell part is not resonsible to turn the Z axis
[12:21:17] <Martin_Lundstrom> its the HAL component
[12:22:03] <Rugludallur> Martin_Lundstrom: yes, I was just wondering if having the flow in user space caused any problems
[12:22:25] <Martin_Lundstrom> its just part of the flow
[12:22:26] <Rugludallur> Martin_Lundstrom: I'm so scared of user space I do everything in realtime :P
[12:22:37] <Martin_Lundstrom> great ;)
[12:23:35] <Martin_Lundstrom> the shell scripts are a great easy way to comunicate M codes to hal components
[12:24:15] <Martin_Lundstrom> then the execution is managed by a HAL component
[12:24:42] <Rugludallur> Martin_Lundstrom: Yeah, since the script is called from the gcode then it does not need to be realtime
[12:25:02] <Martin_Lundstrom> I guess usually not
[12:25:17] <Rugludallur> Martin_Lundstrom: in my case the gcode does not know anything about the THC, it's transparent so I have to halt gcode execution in realtime
[12:25:54] <Martin_Lundstrom> cool, then we dont need my M103 and M105
[12:26:10] <Martin_Lundstrom> How is the height set?
[12:26:29] <Rugludallur> Martin_Lundstrom: no don't need any extra gcode, you just turn on the spindle and the rest is handled automatically
[12:27:06] <Martin_Lundstrom> but it must be possible to set the torch height
[12:27:17] <Martin_Lundstrom> oups
[12:27:17] <Rugludallur> Martin_Lundstrom: Spindle_ON == Torch_ON, just turn it on and the system automatically moves the Z down untill a probe switch is trigggered
[12:27:47] <Martin_Lundstrom> I forgot our thc can not be set externaly
[12:27:48] <Rugludallur> Martin_Lundstrom: the cutting height is then set as voltage to maintain on the arc
[12:27:56] <Martin_Lundstrom> yep
[12:28:23] <Rugludallur> Martin_Lundstrom The inital pierce height is configurable though
[12:28:38] <Martin_Lundstrom> how to set that one?
[12:28:47] <Rugludallur> Martin_Lundstrom: from the Axis GUI
[12:28:55] <Martin_Lundstrom> nice
[12:29:01] <Rugludallur> Martin_Lundstrom along with Corner Height Lock % and a bunch of other stuff
[12:29:26] <Martin_Lundstrom> ok, I wish you good luck
[12:29:51] <Rugludallur> Martin_Lundstrom: thanks :D this will take a while to get running but I think it is worth it
[12:30:27] <Martin_Lundstrom> how do you control state?
[12:30:36] <Rugludallur> Martin_Lundstrom: state '
[12:30:43] <Rugludallur> Martin_Lundstrom: State ?
[12:31:24] <Martin_Lundstrom> ok, eh like wich step the process would be in
[12:32:03] <Rugludallur> Martin_Lundstrom: By monitoring ArcOK and Torch ON
[12:32:13] <Martin_Lundstrom> hmmm
[12:32:27] <Rugludallur> Martin: it also means that if the arc dies in middle of cut, the sequence is re-started
[12:32:36] <Martin_Lundstrom> cool
[12:33:48] <Martin_Lundstrom> anyway in webprogramming there is no state, therefore a variable could be sent each time to check wich "step" to preform
[12:36:09] <Martin_Lundstrom> if look at my c spagetti, you can see in the fat if clause a variable that changes to a higher value all the time when its making progress in the process
[12:36:38] <Martin_Lundstrom> I use that one to control state
[12:40:34] <Martin_Lundstrom> (then you might save some code to check what values you would have in arcok and torch on etc...)
[12:41:35] <Martin_Lundstrom> (I think it great to work with when using a timer)
[12:42:24] <Rugludallur> Martin: all my stuff is hal so :D
[12:43:48] <Martin_Lundstrom> I not a great programmer but im also talking about hal :p
[12:44:45] <Rugludallur> Martin_Lundstrom: I'm not a programmer, which is why I use hal :D
[12:49:26] <Martin_Lundstrom> A... you are going to use alot of exicting components to solve it !?
[12:49:58] <Martin_Lundstrom> not so much of your own C code?
[12:50:14] <Rugludallur> Martin_Lundstrom: not 1 line of gcode :D
[12:50:49] <Martin_Lundstrom> gcode or C code????
[12:50:51] <Rugludallur> Martin_Lundstrom: This is just a diagram of how to connect the components -> http://imagebin.org/8006
[12:51:05] <Rugludallur> Martin_Lundstrom: err s/gcode/c code/
[12:51:13] <Martin_Lundstrom> :)
[12:51:20] <Martin_Lundstrom> it looks nice
[12:51:42] <Rugludallur> Martin_Lundstrom: all the logic is just using those hal componenets and hooking them up as shown
[12:51:53] <Martin_Lundstrom> cool
[13:17:50] <Rugludallur> man the logic function is really cool, I have never used it before but it rocks :D
[14:54:24] <alex_joni> Rugludallur: what software did you use to draw that?
[15:11:11] <Rugludallur> alex_joni: openoffice draw
[15:11:43] <Rugludallur> alex_joni: The layout is pretty good, If in the future someone creates a gui editor for hal I think something along those lines would make sense
[15:12:55] <alex_joni> Rugludallur: seen the work of tomp lately?
[15:13:04] <alex_joni> also on imagebin..
[15:13:07] <Rugludallur> alex_joni: nope,hmm let me check
[15:13:11] <alex_joni> Rugludallur: great stuff..
[15:13:20] <alex_joni> (your schem I mean :P)
[15:13:32] <alex_joni> you have 2-3 problems in there.. but I'm sure you figure them out
[15:13:44] <Rugludallur> alex_joni: hmm problems ,, please elaborate :D
[15:14:03] <Rugludallur> alex_joni: If there is something i missed I would really like to hear about it
[15:14:27] <Rugludallur> alex_joni: I was talking to tomp earlier, I think he was going to use my diagram for some of his work
[15:14:56] <tomp> Rugludallur: i cant get mine as pretty as yours, but i can rearrange elements and use a library of components and drivers, making the variants of parport now
[15:15:14] <Rugludallur> tomp: is that an editor ?
[15:15:24] <tomp> yes, gEDA
[15:15:51] <Rugludallur> tomp: hmm let me rephrase, does that generate hal files ?
[15:16:39] <tomp> no, hope to do that later with the fixed text format that is used, can be parsed, and enum'd
[15:17:01] <Rugludallur> tomp: :D brilliant
[15:17:11] <tomp> hal->graph or graph -> hal
[15:18:42] <Rugludallur> tomp: The ultimate feature would be to hook it to hal and show signals in realtime
[15:20:08] <tomp> the 'wires' are hal signals, they may be anim9able later, i havent seen anyone do that in gEDA yet, but's its common in these kind of apps ( like OrCad )
[15:20:20] <tomp> anim8able
[15:22:02] <tomp> realtime? like some process monitor? maybe possible ( isnt that redundant? like maybe maybe :)
[15:22:22] <alex_joni> tomp: the idea would be to show the current HAL configuration
[15:22:40] <alex_joni> even HAL values and params, etc
[15:22:57] <alex_joni> Rugludallur: for once teh parport.0 has some pins backward
[15:23:23] <tomp> thats the initial idea, to make the hal file understandable (like comic books versus text books :)
[15:24:00] <Rugludallur> alex_joni: you mean the DirX or the in-not ?
[15:24:07] <alex_joni> it's out not in
[15:24:14] <alex_joni> (I hope at least)
[15:24:20] <Rugludallur> alex_joni: ahh :D thx
[15:24:38] <Rugludallur> alex_joni: and not not
[15:26:32] <alex_joni> notnot sounds right
[15:28:22] <alex_joni> Delay Timer: oneshot.0.width <- hardly that's a bit.. right?
[15:28:35] <Rugludallur> alex_joni: true, already fixed :D
[15:29:25] <Rugludallur> alex_joni: updated version: http://imagebin.org/8043
[15:31:07] <alex_joni> hmm.. it
[15:31:13] <alex_joni> hmm.. it's a bit smaller now
[15:31:30] <Rugludallur> alex_joni: hmm I had to enlarge the canvas, let me rexport in a larger format
[15:31:56] <alex_joni> it's a bit hard to read
[15:36:40] <Rugludallur> alex: http://rugludallur.com/fileadmin/user_upload/NewPlasmaDesign.pdf
[15:37:08] <Rugludallur> alex_joni: it seems OOffice automatically selects the resolution based on the size and there is no way to force resolution on export
[15:37:16] <alex_joni> I see..
[15:37:43] <Rugludallur> alex_joni: but pdf is probably better since it supports vectors
[15:39:29] <Rugludallur> Do you guys know if there is a known issue with labelframe in pyvcp ?
[15:39:55] <Rugludallur> No matter what I do labelframe and items inside it are ignored by pyvcp and invisible
[15:40:35] <tomp> lookin
[15:41:03] <Rugludallur> tomp: it might be a python version thing
[15:49:29] <alex_joni> Rugludallur: pdf is nice..
[15:49:40] <alex_joni> I'm a bit disappointed though..
[15:49:48] <alex_joni> you used so little of halui :P
[15:50:47] <Rugludallur> alex_joni: i used more before
[15:50:58] <Rugludallur> alex_joni: there is one feature which I did not put in there
[15:51:25] <Rugludallur> alex_joni: that's the ability to enable "pause" in program execution so a user can "step" through the execution
[15:52:03] <alex_joni> nice
[15:52:25] <alex_joni> Rugludallur: hope you'll use 'net' a lot
[15:52:46] <Rugludallur> alex_joni: Yup, only net and linkpp
[15:52:53] <alex_joni> linkpp is bad
[15:53:03] <Rugludallur> alex_joni: better to use net for all ?
[15:53:06] <alex_joni> use net instead, we plan to remove linkpp
[15:53:07] <alex_joni> yeah
[15:53:29] <Rugludallur> alex_joni: roger, net makes more sense in that you can "name" the auto created signals so makes sense
[15:53:48] <alex_joni> Rugludallur: and it saves you from confusion
[15:53:55] <alex_joni> linkpp was a very bad idea
[15:54:00] <tomp> Rugludallur: label frames should be ok http://imagebin.org/8046
[15:54:42] <Rugludallur> tomp: do you have the xml for that so I can try exact same over here ?
[15:56:28] <tomp> other tabs in that file use some custom pyvcp, there was some talk & the changes werent wanted.. but i'll try to make a simple labelframe and post it
[15:57:47] <tomp> else you can have my pyvcp-widgets files and the 6 files that make that widget panel... up to you but will taint your pyvcp suite
[15:58:43] <Rugludallur> tomp: give me a sec, let me upload my test labelframe to pastebin
[15:58:50] <tomp> k
[15:59:37] <Rugludallur> http://pastebin.ca/428443
[16:00:08] <Rugludallur> tomp: it's the sample from the man docs
[16:03:04] <tomp> looks like what the code says to do http://imagebin.org/8047
[16:03:32] <tomp> someone else try it, maybe my custom pyvcp widgets is 'fixing' it?
[16:05:04] <Rugludallur> tomp: Nothing at all is rendered over here
[16:06:20] <Rugludallur> 2.6.12-magma / Python 2.4.2 / EMC 2.1.1 (updated and built this morning)
[16:06:31] <alex_joni> Rugludallur: why not 2.1.4 ?
[16:06:36] <alex_joni> there are a few bugfixes
[16:07:23] <tomp> python 2.4.3 here
[16:07:40] <alex_joni> Rugludallur: that's breezy.. right?
[16:08:45] <Rugludallur> alex_joni: yup
[16:09:01] <Rugludallur> alex_joni: haven't bothered upgrading since everything was working fine
[16:09:37] <tomp> me Ubuntu 6.06 LTS - the Dapper Drake - and emc is maybe 2 weeks old
[16:09:37] <Rugludallur> I'm getting 2.1.4 as we speak,
[16:10:15] <Rugludallur> going to test that
[16:11:15] <Rugludallur> and trunk
[16:11:19] <alex_joni> Rugludallur: support for breezy ends on 13 April..
[16:11:34] <Rugludallur> alex_joni: I'll upgrade :D
[16:11:40] <alex_joni> but if you don't plan to keep the box online.. there's no problem
[16:12:06] <cradek> I'll continue to make emc2 packages for breezy through 2.1.x
[16:12:09] <Rugludallur> alex_joni: I just wanted to get everything stable before upgrading
[16:12:27] <alex_joni> hi chris
[16:12:37] <cradek> I won't make breezy for 2.2.x (assuming I'm the release manager)
[16:12:40] <cradek> hi
[16:12:53] <alex_joni> http://upload.wikimedia.org/wikipedia/commons/d/df/Ubuntusupport.svg
[16:13:00] <cradek> http://imagebin.org/8046
[16:13:02] <cradek> http://imagebin.org/8046
[16:13:08] <cradek> sorry
[16:14:20] <tomp> cradek: thanks
[16:15:34] <tomp> ? oh :) i thought you had tried the code on your sys ,,, never mind
[16:17:59] <cradek> no I accidenatlly pasted it (twice)
[16:18:23] <cradek> http://www.defectiveyeti.com/iacaptchas/
[16:18:29] <cradek> ^^ love this idea
[16:19:39] <Rugludallur> now I remember why I didn't upgrade, gantrykins :D
[16:19:41] <alex_joni> hahaha
[16:20:05] <cradek> alex_joni: reload a few times
[16:20:08] <Jymmm> Enjoy the porn!
[16:20:27] <alex_joni> lose vs. loose (LOL)
[16:21:01] <cradek> is it there? I was hoping so but I haven't seen it
[16:21:25] <cradek> yay!
[16:21:40] <tomp> try entering bother/help/stop
[16:22:26] <alex_joni> I've seen it twice
[16:22:48] <alex_joni> This is *there, their, they're* party.
[16:22:58] <alex_joni> even 3 versions :P
[16:23:04] <Jymmm> I'd rather see questions like: "Typical transmission speeds/distances limits of Multi-mode optical fiber are _________ up to _______."
[16:24:48] <alex_joni> so.. did you hide your eggs?
[16:29:52] <Rugludallur> hmm why hide when you can eat
[16:40:18] <Rugludallur> tomp: labelframe works in head, not in 2.1.1 or 2.1.4,
[16:42:20] <tomp> labelframe must not be in their pyvcp-widgets.py , not much to it, you could cut & paste that tiny bit o code if you want a nontstandard sys like mine ;)
[16:45:52] <Rugludallur> tomp: I'll just stick with head for development, thx
[16:54:10] <Jymmm> Okey, this is awesome... There is just so much here that's funny that I cna't even begin to describe, though I think one of my fav is the tampon cleaning... http://www.rabbitlaser.com/DemoVideo/HX40A.mpg
[16:56:48] <alex_joni> whoa.. 300MB
[16:57:07] <Jymmm> and 30 minutes too
[16:57:13] <Jymmm> it's worth it though
[16:57:43] <alex_joni> going slow unfort...
[16:57:52] <Jymmm> stream it
[17:02:09] <alex_joni> too much effort
[17:02:24] <Jymmm> your loss
[17:02:41] <alex_joni> I'll be waiting for the download and watch it then
[17:04:14] <Jymmm> k
[18:05:00] <alex_joni> Jymmm: nice music
[18:15:24] <alex_joni> great stuf.. hold the door safety mechanism with your hand while you shoot the laser
[18:16:01] <anonimasu> lol
[18:16:06] <anonimasu> great idiocy
[18:16:31] <alex_joni> seriously.. they keep adjusting the lasers with the hands in there
[18:17:01] <Jymmm> Here are the rest... http://www.rabbitlaser.com/DemoVideo/
[18:17:18] <Jymmm> alex_joni Yeah, my gf said the music was nice too
[18:21:15] <alex_joni> hmm.. did you guys ever see/use ENUM?
[18:21:39] <alex_joni> this ENUM: http://en.wikipedia.org/wiki/Telephone_Number_Mapping
[18:22:36] <Jymmm> Yeah, just what I want... spamming my VoIP *AND* POTS line
[18:25:01] <alex_joni> don't forget your email
[18:25:11] <alex_joni> which is now in DNS..
[18:25:17] <Jymmm> Nah, I got that covered =)
[18:32:01] <Jymmm> After I chuck up a tool, I use a flashlight to zero the Z axis to top of material (more accurate that a piece of paper) by holding the flahslight so that it creates the longest shadow, then slowly jog the tool till it touches down to the material (i.e. no light shining between the tool tip and the material). This works pretty well (for me at least), but I'd like to be able to automate this if at all possible, any ideas?
[18:34:36] <ds3> why would you say this is more accurate then the paper method?
[18:34:46] <Jymmm> it is, for me.
[18:35:01] <ds3> is there some reasoning behind it?
[18:35:14] <Jymmm> lets leave it at that, I'd rather not go thru that debate right now
[18:35:22] <ds3> ok
[18:35:35] <alex_joni> Jymmm: for automation I'd use a tool length measuring thingie
[18:35:42] <alex_joni> one of those contacts
[18:35:46] <Jymmm> alex_joni example?
[18:35:48] <alex_joni> where you do a probe move down
[18:36:09] <Jymmm> oh no, I havne't been able to find one with costing an arm AND leg
[18:36:23] <Jymmm> I think the optical way is pretty damn good.
[18:36:24] <alex_joni> http://www.vibrafree.com/image/Products/touch-probe.jpg
[18:36:40] <ds3> seems like everyone has their own favorite methods
[18:37:02] <ds3> but I suppose you could use a CCD (cheap camera)
[18:38:20] <ds3> or maybe even a photo transistor
[18:38:31] <Jymmm> When I used to work microelectronics, we had a microscope that (no kidding) had a TI adding machien connected to it. You focused on the substrate, zeroed out the adding machien. then moved it to the thinkfilm you were trying to measure the thickness of and refocus/ the diff would be in microns of thickness.
[18:38:56] <alex_joni> Jymmm: I would try using a simple pushbutton first
[18:39:06] <alex_joni> and check how accurate that is
[18:39:08] <Jymmm> It was simplistic, and accurate.
[18:39:09] <ds3> that sounds like a special purpose optical comparator
[18:39:27] <Jymmm> A $39 Texis Instruments adding machine?!
[18:39:43] <ds3> the focusing mechanism on the microscope
[18:39:52] <tomp> hal-gEDA parports http://imagebin.org/8048 one is cfg'd 'in' the other cfgd 'out' the only change is in pins 2 thru 9
[18:40:20] <ds3> wonder if you could do the same with a servo setup and oneof those QX3+'s
[18:40:23] <Jymmm> alex_joni I contacted CHERRY (microswitches), they said you wouldn't get good repeatable accuracy tdue to the way theose switches work.
[18:40:48] <tomp> NUM should be replaced
[18:41:49] <alex_joni> tomp: params still missing :)
[18:42:59] <tomp> params cant be wired, but you're right, i'm leaving those for now, havent decided if params are visualized as pins or not
[18:43:20] <alex_joni> not as pins.. but they need some visualisation
[18:44:48] <tomp> as pins they can have an attribute 'value', while if not pins.. well thats outside of the available tool :(
[18:47:03] <alex_joni> Jymmm: what kind of precision are you aiming for?
[18:47:44] <tomp> originally i used pins and a constant component for every parameter ( wired a constant into a larger component, so it was visual & had names, and was editable )
[18:48:29] <bytecolor> does EMC have a lathe auto rough/finish cycle?
[18:49:11] <cradek> no the only lathe-specific cycle it has is for threading
[18:49:15] <cradek> (so far)
[18:49:36] <ds3> Jymmm: someone showed me a switch that will work; except the price was about $300/switch.... don't recall who makes it but it does exists
[18:50:34] <cradek> bytecolor: it wouldn't be hard to add - like usual coming up with a decent spec is the hard part
[18:50:41] <bytecolor> ok, thanks cradek, me and Danial were discussing lathe cycles in #cam
[18:51:08] <bytecolor> nod, you could model it after Fanuc
[18:51:25] <ds3> cradek: do you have a template for writing up a spec?
[18:51:35] <cradek> no
[18:51:44] <ds3> got a sample spec that i can model it off of?
[18:51:47] <cradek> by spec I only mean what gcodes to use, how the parameters work, etc
[18:51:53] <ds3> Okay
[18:52:12] <cradek> if you'd like to write something we could discuss, that would be great, the best place would be a wiki page
[18:52:16] <ds3> cradek: if I do the writing, can you help me with the wiki part when I am done?
[18:52:47] <cradek> for the wiki see the BasicSteps page, you can paste your text into a new page wholesale and worry about formatting later or not at all :-)
[18:53:02] <ds3> okay
[18:53:08] <cradek> BasicSteps mostly tells you the password to edit the wiki
[18:53:19] <ds3> one of these nights I will get sick of doing what I should be and spew it out
[18:53:32] <cradek> great
[18:53:34] <ds3> all you need is the G code, parameters, and description of how it works, right?
[18:53:45] <ds3> no need for pseudo code on the strategy for it?
[18:54:18] <cradek> nah just plain english language description of what it should do and how the gcode works
[18:54:27] <ds3> the strategy part is the thing that prevents me from writing up a perl preprocessor
[18:54:49] <bytecolor> G71 horizontal rough, G72 vertical rough, G73 copy turn (offset the profile), G70 finish
[18:54:49] <ds3> okay, I'll see if I can find my examples and maybe photographs if the output
[18:55:14] <ds3> bytecolor: would I be irritating other people if I describe the Haas version? (slightly different but not overly so)
[18:55:27] <bytecolor> havent used those in a while, I'd have to look up the spec
[18:55:41] <bytecolor> ds3, no that would be fine, it's close enough to Fanuc
[18:56:02] <ds3> I am willing to kick it off as I really want the G70/G71 codes implemented =)
[18:56:14] <bytecolor> they are handy :)
[18:56:19] <ds3> yes
[18:56:35] <bytecolor> probably not easy to implement well, though
[18:56:42] <ds3> but then I also want the lathe style tool changes
[18:56:50] <ds3> *nod*
[18:57:31] <tomp> :) there's so much data to be dragged around now, i got an effect similar to the handwheel wind-up , the mass of wires and components has gotten 'springy'
[18:57:42] <cradek> what's a lathe style tool change
[18:58:06] <Jymmm> alex_joni .0005
[18:58:14] <Jymmm> "
[18:58:42] <alex_joni> 0.0127 mm ?
[18:58:52] <ds3> T0101 instead of T01 M06; G43 H01
[18:58:53] <Jymmm> ds3: I have a switch that does .003mm accuracy
[18:59:12] <Jymmm> alex_joni nfc, inches baby, inches
[18:59:18] <ds3> Jymmm: but does it repeat to the same place everytime? accuracy and repeatability are 2 different things
[18:59:34] <Jymmm> ds3 haven't tested it yet
[18:59:43] <alex_joni> repeatability is alwayworse then accuracy
[19:00:11] <Jymmm> Well, anything mechanical is gonna be icky, that's whay I was thinking optical
[19:00:16] <alex_joni> Jymmm: .003mm is way better than .0005"
[19:00:34] <Jymmm> alex_joni I just remembered, it's sloppy.
[19:00:42] <ds3> make an interferometer with a laseer and use that ;)
[19:00:53] <ds3> laser
[19:00:59] <alex_joni> just make a U shaped laser barier
[19:01:23] <alex_joni> laser pointer, and optotransister as receiver
[19:01:42] <Jymmm> to measure what, intensity of light?
[19:02:28] <alex_joni> Jymmm: no.. just move down till you brake teh ray
[19:03:10] <Jymmm> I'm dealing with a tool tip of .005"
[19:03:26] <Jymmm> flat
[19:03:35] <alex_joni> I assume you have homeswitches
[19:03:41] <Jymmm> nope
[19:04:48] <Jymmm> I sorta like the CCD thing... could measure the qty of light, no?
[19:13:06] <Jymmm> Mitutoyo 547-258 http://www.mitutoyo.com/TerminalMerchandisingGroup.aspx?group=1137 $372 USD
[19:14:46] <Jymmm> I wonder if you could rip apart an optical mouse
[19:15:04] <Jymmm> they have a CCD camera in them
[19:15:29] <Jymmm> ds3 gave up on making your H bridges?
[19:15:40] <jmkasunich> mechanical is gonna be more accurate than optical
[19:15:48] <Jymmm> how so?
[19:16:14] <jmkasunich> think about it....
[19:16:32] <jmkasunich> how big is the beam of a laser pointer? 1/16"? huge!
[19:17:00] <jmkasunich> how fine can you focus a camera? can you see a human hair reliably? that is 0.004" or so
[19:17:03] <Jymmm> jmkasunich you mean in the sense of "breaking the beam" ?
[19:17:09] <Skullworks_PGAB> there is a writeup somewhere out on the net of some guy who hacked up a mouse and used the wheels for ecoders
[19:17:22] <jmkasunich> Skullworks_PGAB: that's a mechanical mouse
[19:17:31] <Skullworks_PGAB> lots of crazy glue to tack and hot glue to hold
[19:17:41] <Skullworks_PGAB> exactly
[19:17:47] <jmkasunich> besides, mice have a resolution of a few hundred counts per inch
[19:18:00] <jmkasunich> 0.005 to maybe 0.002"
[19:18:28] <jmkasunich> and thats with the mechanical advantage of the axle
[19:18:50] <jmkasunich> (small shaft driving large slotted wheel - a small movement at the shaft surface = a large one at the slots
[19:19:44] <jmkasunich> Jymmm: this is for setting tool length, right?
[19:20:00] <Jymmm> basically
[19:20:03] <jmkasunich> how about a dial indicator with a flat tip, mounted upside down (plunger facing up)
[19:20:52] <Skullworks_PGAB> There is a write up on making one of those on the IH website
[19:20:55] <jmkasunich> the simple approach would be to jog down on it, then do 0.0001 incremental jogs up and down till it reads zero, then touch off at that point
[19:21:15] <jmkasunich> the fancy way would be to put an optical interruptor such that the needle breaks the beam
[19:21:29] <ds3> Jymmm: eh? never had a need for H bridges
[19:21:41] <jmkasunich> again with the mechanical advantage - 0.001" of tool movement = 1/16" or more of needle movement
[19:22:00] <ds3> hacking a CD player head might be better then a mouse
[19:22:05] <Jymmm> jmkasunich I see two thing here.... 1) I need to know where the material height is and 2) know where the tool tip is. Not sure if you saw what I described above, so here it is again:
[19:22:16] <anonimasu> hi
[19:22:29] <Jymmm> After I chuck up a tool, I use a flashlight to zero the Z axis to top of material (more accurate that a piece of paper) by holding the flahslight so that it creates the longest shadow, then slowly jog the tool till it touches down to the material (i.e. no light shining between the tool tip and the material). This works pretty well (for me at least), but I'd like to be able to automate this if at all possible, any ideas?
[19:22:40] <Skullworks_PGAB> Well I still need H-Bridges!
[19:23:07] <anonimasu> hm, I find papers pretty reliable..
[19:23:28] <Jymmm> lets leave it at that, I'd rather not go thru that debate right now
[19:23:42] <alex_joni> Jymmm: the dial indicator sounds like a good idea
[19:24:05] <alex_joni> dial indicator + optical sensor on the needle
[19:24:15] <Jymmm> alex_joni The problem is I still don't know where the tool tip is in relation to TOM (top of material)
[19:24:18] <anonimasu> why not a microswitch?
[19:24:25] <Jymmm> anonimasu scroll up
[19:24:39] <ds3> Jymmm: what kind of material?
[19:24:40] <Jymmm> alex_joni I contacted CHERRY (microswitches), they said you wouldn't get good repeatable accuracy tdue to the way theose switches work.
[19:24:46] <Jymmm> anonimasu: I contacted CHERRY (microswitches), they said you wouldn't get good repeatable accuracy tdue to the way theose switches work.
[19:24:54] <Skullworks_PGAB> http://www.industrialhobbies.com/howto/tips/touch_probe.htm
[19:25:04] <anonimasu> Jymmm: Well, call baumer..
[19:25:07] <Jymmm> ds3 plastics
[19:25:07] <ds3> surface textures will be significant
[19:25:07] <anonimasu> about prox switches..
[19:25:21] <anonimasu> 1um, repeatability..
[19:25:26] <ds3> Jymmm: transparent or opaque?
[19:25:40] <Jymmm> anonimasu sensor is what?
[19:25:46] <Jymmm> ds3 yes
[19:25:48] <anonimasu> they have switches that accurate too..
[19:25:53] <anonimasu> www.baumerelectric.de
[19:25:57] <Jymmm> anonimasu under $1000 ?
[19:26:16] <anonimasu> Jymmm: yeah..
[19:26:19] <ds3> transparent would be tricky
[19:26:22] <anonimasu> I can bet they are under 1000$..
[19:26:27] <alex_joni> Jymmm: you can get one for about 300$
[19:26:37] <anonimasu> I dont have a price but I have a 400page catalog ;)
[19:26:52] <alex_joni> http://www.cadcamcadcam.com/index.asp?PageAction=VIEWPROD&ProdID=120\
[19:26:53] <anonimasu> they have lots of diffferent ones with diff precision..
[19:26:58] <Jymmm> alex_joni and what kind of sensor is it?
[19:27:01] <alex_joni> without the \ at the end
[19:27:13] <anonimasu> the ones I talked at was inductive, or microwitches..
[19:27:36] <Jymmm> so again, I only get the position of the tool or the material, but not both.
[19:27:50] <anonimasu> Jymmm: so make yout thing a set height..
[19:28:04] <anonimasu> you know how much is between the base and when the switch triggers..
[19:28:12] <anonimasu> or buy a electric tool height setter for 300$
[19:28:50] <Jymmm> I like my flashlight idea, I just want to automate it.
[19:28:54] <anonimasu> you place it on material and when it gives a signal your 10mm from the top..
[19:28:57] <alex_joni> Jymmm: just place the tool setter on the workpiece
[19:29:05] <anonimasu> of your material..
[19:29:42] <anonimasu> Jymmm: write a app that monitors your camera..
[19:29:52] <anonimasu> and triggers when the amount of bright is less then X
[19:29:57] <ds3> Jymmm: a laser pointer would be better, it is hard to detect a flashlight disappearing given all the side spillage
[19:29:58] <Jymmm> for what, brightness?
[19:30:05] <anonimasu> yeah
[19:30:34] <alex_joni> I doubt you can automate the flashlight/laser approach
[19:30:49] <ds3> with a camera, you might be able to
[19:31:00] <ds3> take an image before the light source is enabled
[19:31:02] <Jymmm> that where I was thinking an optical laser mouse that already ahs the camera and light source in it
[19:31:03] <anonimasu> yeah, stll what happens with different flutes?
[19:31:04] <ds3> take one with it enabled
[19:31:10] <anonimasu> there arent cameras in them
[19:31:16] <anonimasu> they have reading forks..
[19:31:17] <Jymmm> yes there are
[19:31:17] <jmkasunich> gotta automate the positioning of the light and the camera too (unless you permanently mount them - then you have dust and maybe clearance issues)
[19:31:26] <ds3> anonimasu: actually, some of them do have cameras in them!
[19:31:29] <anonimasu> really?
[19:31:35] <Jymmm> yep
[19:31:39] <jmkasunich> optical mice are not the answer
[19:31:47] <jmkasunich> yes, they do have little CCD cameras in them
[19:32:04] <jmkasunich> but they are fixed focus, a few mm ahead of the lens
[19:32:06] <anonimasu> I still think that kind of setup is a hack..
[19:32:22] <jmkasunich> and they don't give straight video output, they have built in processing that looks for movement
[19:32:38] <anonimasu> probably it works much better when doing manually..
[19:32:47] <anonimasu> Jymmm: how about a touchprobe?
[19:32:51] <ds3> jmkasunich: actually, they do give you video out
[19:33:03] <jmkasunich> show me
[19:33:12] <ds3> link coming
[19:33:14] <jmkasunich> (like - URL of a datasheet)
[19:33:21] <Jymmm> jmkasunich gimme a few to find the link with pictures
[19:33:21] <ds3> http://www.spritesmods.com/?art=mouseeye
[19:33:32] <ds3> they point to the datasheet
[19:33:40] <ds3> there is a debug port that will return the pixels seen
[19:34:05] <ds3> this is the datasheet if you want to go direct
[19:34:07] <ds3> http://cp.literature.agilent.com/litweb/pdf/5988-9774EN.pdf
[19:34:16] <jmkasunich> I see
[19:34:26] <jmkasunich> and I stand corrected - at least some mice can give you an image
[19:34:31] <jmkasunich> a whole 18x18 pixels
[19:34:34] <ds3> *nod*
[19:34:39] <Jymmm> there's one that has the lens built in too
[19:34:43] <Jymmm> and 64x64
[19:34:43] <jmkasunich> and judging by the pics, each pixel is large
[19:34:48] <ds3> a CD player optical assembly has a sensor that returns 4x4 info
[19:35:21] <jmkasunich> I still think this is silly
[19:35:51] <Skullworks_PGAB> 2nd
[19:35:59] <anonimasu> 3rd
[19:36:16] <ds3> if someone wants rope to hang themselves, I don't object ;)
[19:36:46] <Jymmm> Here ya go..... http://www.mstarmetro.net/~rlowens/OpticalMouse/
[19:37:16] <ds3> Hmmm
[19:37:34] <ds3> prehaps a scanning laser disk + phototransistor w/miss pulse detector
[19:37:48] <ds3> when it gets blocked, it will cut out a few pulses and generate a stop
[19:37:51] <Jymmm> ds3 there is a shutter on the cameras
[19:38:11] <ds3> shutter won't help
[19:38:16] <Jymmm> k
[19:38:19] <ds3> I am thinking of modulating it pretty fast
[19:38:47] <ds3> so you will get a constant stream of pulses but the slightiest blockage should cut out a pulse or two, just enough to flag things
[19:38:57] <alex_joni> 4th
[19:39:07] <jmkasunich> have you guys ever heard of diffraction?
[19:39:19] <Jymmm> http://www.chipmunk.nl/ProMouse/part9.html
[19:39:37] <alex_joni> ds3: but you don't need to draw a picture how they should tie the knot around their neck..
[19:39:41] <ds3> Hmmm single edge diffractionHmmmmm
[19:40:37] <Jymmm> http://www.business-sites.philips.com/lasersensors/about/article-15031.html
[19:42:15] <ds3> jmkasunich: that might not work with transparent stock
[19:42:19] <Skullworks_PGAB> I use these
[19:42:22] <Skullworks_PGAB> http://www.use-enco.com/CGI/INSRIT?PARTPG=INSRAR2&PMAKA=636-7044&PMPXNO=951637
[19:43:09] <jmkasunich> ds3: you don't understand me... I was just pointing out that diffraction is just one of the many reasons why doing this optically is silly - transparent stock is another one
[19:43:40] <Jymmm> Skullworks_PGAB doesnt that require electrical conductivity?
[19:43:45] <jmkasunich> I'm going to shut up now, and let you guys have your fun
[19:43:52] <Skullworks_PGAB> yes
[19:43:58] <ds3> jmaksunich: Ohhhhh. misunderstood you; thought you were suggesting looking at the single edge diffraction (beam movement w/a monochromatic source)
[19:44:01] <Jymmm> Skullworks_PGAB I'm doing plastics
[19:44:17] <anonimasu> there are other kinds.
[19:44:24] <Skullworks_PGAB> so fof plastics I set it on a sheet of al foil
[19:44:30] <Jymmm> ds3 Most of the time, the plastic does have a paper film on it.
[19:44:34] <anonimasu> http://www.use-enco.com/CGI/INSRIT?PMAKA=308-0391&PMPXNO=12093229&PARTPG=INSRHI
[19:44:57] <Jymmm> anonimasu that one conducts too
[19:45:00] <ds3> Jymmm: you must use better suppliers then I do.... most of the time I have a clear film on them ;)
[19:45:12] <anonimasu> sigh..
[19:45:19] <Jymmm> ds3 you are using extruted plastic, I'm not.
[19:45:27] <anonimasu> Jymmm: there are others..
[19:45:29] <anonimasu> with switche..
[19:45:48] <ds3> Jymmm: is that the difference? don't know enough to tell
[19:46:09] <Jymmm> ds3 Actually yeah... extrutuded is blue plastic film.
[19:46:22] <Jymmm> sometimes white depending on mfg
[19:46:36] <anonimasu> http://www.centroidcnc.com/images/tt1_toolset200.jpg
[19:46:39] <ds3> cool, learned something
[19:47:01] <anonimasu> http://img.alibaba.com/photo/51257691/Z_Axis_Setter.summ.jpg
[19:47:07] <jmkasunich> that toolsetter would work - just sit it on top of the plastic
[19:47:08] <anonimasu> like the one on the bottom
[19:47:22] <Jymmm> Maybe I'll run this by Les, he likes this stuff
[19:47:55] <anonimasu> Jymmm: I think les will point out the same thing as jmk just did..
[19:48:00] <ds3> I personally am happy with using a dowel pin, but that's me
[19:48:32] <Jymmm> anonimasu Maybe so, but I really like the optical idea/concept.
[19:48:49] <anonimasu> Jymmm: you are blind to what problems it poses.
[19:48:58] <Skullworks_PGAB> anonimasu - I use the other one - not the Fowler model because the contact is spring loaded so your micro EM won't break if you over travel Z.
[19:49:15] <anonimasu> yep
[19:49:19] <anonimasu> I dont have one :p
[19:49:29] <Jymmm> anonimasu You are more than welcome to "show me the light"
[19:50:14] <anonimasu> Jymmm: you cant show anything to a blind person, sorry.
[19:50:25] <anonimasu> ^_^
[19:52:27] <anonimasu> have anyone investigated custom panels?
[19:52:32] <anonimasu> with buttons and stuff?
[19:52:32] <Skullworks_PGAB> (BTW it also works for setting Z on lathe tooling.)
[19:52:42] <ds3> custom panels for what, anonimasu?
[19:52:47] <Skullworks_PGAB> yep
[19:52:50] <anonimasu> a cnc machine...
[19:53:21] <Skullworks_PGAB> hoping that 8255 PCI card would work for the I/O
[19:53:40] <ds3> I'd think finding a suitable display is a bigger issue then a panel?
[19:53:47] <anonimasu> why?
[19:53:52] <Skullworks_PGAB> Mode switch would be 3 bits
[19:53:56] <anonimasu> yeah
[19:53:59] <ds3> $$$ reasons
[19:54:08] <anonimasu> there are loads of cheap tft's..
[19:54:12] <alex_joni> simple LCD should do the trick for a panel
[19:54:18] <ds3> anonimasu: what's cheap?
[19:54:20] <anonimasu> and 10" ones too
[19:54:27] <ds3> and where?!!!!!!
[19:54:33] <anonimasu> cheap is less then 200$£
[19:54:34] <anonimasu> err 200$
[19:54:47] <Skullworks_PGAB> I got a NEC 15" LCD cheap because the base was broken...
[19:54:53] <ds3> that costs more then my PC ;)
[19:54:55] <anonimasu> if ebay ftw!
[19:55:02] <Jymmm> all I hear is you can't/shouldn't do it, but I don't hear WHY?
[19:55:21] <Skullworks_PGAB> ebay will eat you alive on shipping for some stuff
[19:55:36] <jmkasunich> Jymmm: you are being told why as well, but you aren't listening
[19:55:43] <Jymmm> I saw it done before, and it worked good, so I don't understand why it's a bad idea?
[19:55:55] <Jymmm> jmkasunich then explain it differently please.
[19:55:57] <jmkasunich> saw what done before?
[19:56:09] <jmkasunich> optical mice as toolsetters - somehow I doubt it
[19:56:21] <anonimasu> a 7" tft costs 55eur..
[19:56:24] <jmkasunich> flashlights into cameras as toolsetters - ditto
[19:56:25] <ds3> I'd be fun to implement as a computer vision exercise
[19:56:26] <Jymmm> I said optical
[19:56:26] <anonimasu> 79$
[19:56:38] <ds3> anomimasu: some place other then ebay?
[19:56:48] <anonimasu> any computer store..
[19:56:59] <jmkasunich> Jymmm: what kind of optical?
[19:57:03] <anonimasu> I dont know about the us, but secondhand monitors are free/cheap..
[19:57:04] <anonimasu> almost..
[19:57:11] <ds3> don't see sub $100 much around here or for that matter sub 15" monitors
[19:57:18] <Skullworks_PGAB> 7" is too small and likely uses composite vid not DVI or VGA
[19:57:44] <ds3> 7" would be nice if it can do VGA resz
[19:57:44] <jmkasunich> 15" LCDs should be available for $100 new if you shop around
[19:57:47] <anonimasu> the one with vga is double the price..
[19:57:58] <anonimasu> so, where do you get panels?
[19:57:58] <jmkasunich> if you need smaller, that makes it tough
[19:58:01] <anonimasu> with buttons and stuff?
[19:58:02] <anonimasu> :D
[19:58:06] <jmkasunich> oh, buttons...
[19:58:10] <anonimasu> yeah..
[19:58:15] <ds3> jmkasunich: yep
[19:58:15] <jmkasunich> you make the panel, and buy the buttons
[19:58:29] <Skullworks_PGAB> have to roll your own for now...
[19:58:55] <anonimasu> I want thoose square black buttons ;)
[19:59:33] <Skullworks_PGAB> only panels I seen so far are very limited function - for Mach using USB
[19:59:43] <Jymmm> jmkasunich whatever kind you are saying "won't work" I guess. I never gave specifics.
[19:59:43] <ds3> making the panel is, IMO, the easy part
[20:00:11] <ds3> the buttons are pretty straight forward; machine a pusher; and make a PCB for the microswitches
[20:00:13] <jmkasunich> Jymmm: anything based on breaking a beam can only be as accurate as the beam is wide
[20:00:16] <anonimasu> pusher?
[20:00:21] <jmkasunich> and making a beam of light that is only 0.001" wide is not easy
[20:00:22] <Jymmm> jmkasunich I never siad breaking a beam
[20:00:27] <jmkasunich> diffraction will spread it out
[20:00:32] <jmkasunich> I didn't finish!
[20:00:34] <ds3> yeah a thing that moves in and out to actuate the switch
[20:00:53] <ds3> if you disassemble most modern electronics, it is waht you will find
[20:01:01] <anonimasu> ds3: I'm not talking about that kind of thing..
[20:01:01] <anonimasu> HMI stuff..
[20:01:03] <jmkasunich> anything based on looking at something with a camera will only be as accurate as the pixel size of the camera, plus error do to focus, etc
[20:01:12] <Jymmm> was it alex_joni or anonimasu that found that software that uses a camera and light ( and a ruler too) to make a 3d object scanner???
[20:01:15] <ds3> HMI?
[20:01:20] <Skullworks_PGAB> * Skullworks_PGAB uses a laser to do engraving - and its .003-.004" dia.
[20:01:22] <anonimasu> human machine interface..
[20:01:30] <anonimasu> panel, for entering data..
[20:01:41] <ds3> anonimasu: you can make a matrix of those switches
[20:01:55] <anonimasu> ds3: well, the issue is more to make the actual panel..
[20:02:04] <ds3> Jymmm: the resolution of those is worse then using a paper to set it
[20:02:52] <Skullworks_PGAB> I like the idea of using the 8255 because it can output status to LED's for each button
[20:03:25] <Skullworks_PGAB> seems a waste though
[20:03:47] <Skullworks_PGAB> using all that I/O just for panal functions
[20:04:00] <anonimasu> 8255?
[20:04:00] <ds3> they have chips to do matrix scanning and encoding
[20:04:11] <jmkasunich> IO is cheap, if you don't need to isolate it, etc
[20:04:28] <jmkasunich> its the isolation, drivers, cables, and breakouts that cost $
[20:04:32] <anonimasu> yeah
[20:04:47] <SWPadnos> a $3 (or less) AVR can do 256 switches and 256 lights in scanned matrices
[20:04:46] <jmkasunich> for buttons and LEDs you don't need most of that
[20:05:06] <SWPadnos> and send it to the PC via serial (or maybe USB)
[20:05:25] <Jymmm> http://www.sjbaker.org/projects/scanner/
[20:06:51] <anonimasu> http://www.elfa.se/images/highres/h1623.jpg
[20:07:20] <anonimasu> that's something like what I want
[20:08:37] <jmkasunich> Jymmm: tell me, how accurate do you think that scanner is? 0.1", 0.01", or 0.001"?
[20:09:14] <anonimasu> jmkasunich: you might know, can you do serial communication in realtime somehow?
[20:09:27] <jmkasunich> I'm sure its possible
[20:09:35] <Jymmm> jmkasunich You keep asking me questions, but you havne't explained why this is a bad idea in a way I can understand.
[20:09:37] <jmkasunich> you could always write code that talks directly to the uart
[20:09:44] <jmkasunich> Jymmm: tell me, how accurate do you think that scanner is? 0.1", 0.01", or 0.001"?
[20:10:01] <jmkasunich> that concept is probably accurate to 0.050" at best
[20:10:08] <jmkasunich> that is 50 times worse than what you are looking for
[20:10:14] <anonimasu> jmkasunich: looking on how to get a panel to talk to emc easily
[20:10:20] <jmkasunich> so you are fooling yourself thinking about it
[20:10:26] <Jymmm> jmkasunich You keep asking me questions, but you havne't explained why this is a bad idea in a way I can understand.
[20:10:41] <Jymmm> that was just one concept
[20:11:15] <jmkasunich> if you can't tell just by looking at that "one concept" that it will NEVER be accurate enough to do what you need, then I CAN'T explain it to you
[20:11:16] <alex_joni> Jymmm: I worked together with some people who do PC-vision based inspection
[20:11:28] <alex_joni> camera based part inspection
[20:11:35] <alex_joni> they were very happy to do .5mm
[20:11:50] <alex_joni> and that was using cameras worth 3000$
[20:11:58] <anonimasu> I have some fork sensors they do 0.2mm object detection..
[20:12:06] <jmkasunich> anonimasu: what do you use for I/O now? steps out a parport, or some servo card, or...
[20:12:13] <anonimasu> jmkasunich: I use a usc..
[20:12:22] <jmkasunich> ok
[20:12:28] <jmkasunich> hmm, thats epp bus
[20:12:31] <anonimasu> im all open for ideas..
[20:13:03] <anonimasu> I have a plc running via serial, for I/O and toolchanging and stuff
[20:13:29] <jmkasunich> non RT serial, right?
[20:13:36] <anonimasu> no
[20:13:40] <anonimasu> that's why im interested
[20:13:46] <jmkasunich> no?
[20:13:50] <jmkasunich> you mean it is RT?
[20:13:55] <anonimasu> no, it's not
[20:13:58] <jmkasunich> ok ;-)
[20:13:59] <anonimasu> I wish it were
[20:14:29] <anonimasu> I'm running a python program to handle the communication today
[20:14:43] <jmkasunich> what you want is some kind of I/O concentrator - lots of bits of IO going in/out thru a single port
[20:14:48] <anonimasu> yeah
[20:14:58] <anonimasu> ideally I'd use a can card and a HMI module..
[20:14:59] <anonimasu> :)
[20:15:06] <anonimasu> they are pretty cheap..
[20:15:08] <anonimasu> the modules..
[20:15:16] <jmkasunich> go for it - write a HAL driver for the can card
[20:15:21] <jmkasunich> ;P
[20:15:30] <anonimasu> the io cards are about 285$
[20:15:38] <anonimasu> but with lots of io, and IP-something
[20:15:47] <anonimasu> err io modules..
[20:16:25] <Skullworks_PGAB> http://www.futurlec.com/PCI8255.shtml
[20:16:41] <anonimasu> sweet.
[20:16:53] <anonimasu> I'd buy one if I could write a driver for it..
[20:16:55] <Skullworks_PGAB> thats the 8255 board I was thinking of
[20:17:07] <jmkasunich> jepler has at least partly written the driver already
[20:17:13] <anonimasu> thats perfect..
[20:17:14] <Skullworks_PGAB> Jeff is playing with one now
[20:17:36] <jmkasunich> if your computer is close enough to the control panel that you can run a bundle of wires - thats perfect
[20:17:40] <anonimasu> yep
[20:17:45] <anonimasu> thats what I were thinking
[20:18:14] <Skullworks_PGAB> use a LVD SCSI cable
[20:18:16] <jmkasunich> for my panel (eventually) I want something that will crunch down the wire count, because the computer will be in an oil-tite enclosure a few feet from the panel
[20:18:25] <anonimasu> well can ;)
[20:18:34] <Skullworks_PGAB> 68 wires
[20:18:46] <anonimasu> I'd love to run that to my plc :)
[20:18:56] <anonimasu> but I dont think there are cheap cards to do it
[20:19:11] <jmkasunich> I'm thinking about a shift register based thing - I'm using a 5i20 - if I dedicate 3-4 of its pins to I/O, I can get 32 ins and 32 outs at the panel with only a few wires
[20:19:25] <anonimasu> yeah
[20:20:00] <jmkasunich> but that relies on the logic in the 5i20 to do the serial stuff
[20:20:18] <alex_joni> jmkasunich: or bitbang it in HAL
[20:20:25] <alex_joni> it's still fast enough
[20:20:25] <jmkasunich> yeah
[20:20:52] <jmkasunich> true - if you bitbang in a 20KHz base thread, you can read 32 buttons every 2mS with time left over
[20:21:12] <anonimasu> :)
[20:22:22] <anonimasu> a kvaser pcican card is 450$
[20:22:33] <jmkasunich> ouch
[20:22:42] <jmkasunich> that buys 2 USCs or 2 5i20s
[20:22:50] <Skullworks_PGAB> my bleeding wallet...
[20:22:56] <anonimasu> I wonder if you can do can in software,
[20:23:03] <anonimasu> or to a pic, and then bitbang to hal.
[20:23:08] <jmkasunich> youcandocan
[20:23:43] <anonimasu> 500kbit/s is the speed i run can at.. ot 615kbit/s
[20:24:29] <alex_joni> jmkasunich: seen the work Rugludallur did?
[20:24:36] <jmkasunich> the THC? yeah
[20:24:41] <alex_joni> great stuff
[20:24:48] <jmkasunich> yep
[20:25:51] <anonimasu> im looking at can because it's pretty much standard for pheriperals and stuff..
[20:27:09] <anonimasu> http://www.c-a-n.com/canpci.html
[20:27:11] <anonimasu> hm
[20:27:12] <anonimasu> even cheaper..
[20:28:43] <alex_joni> bbl
[20:44:42] <Jymmm> panel, as in control panel with lots of buttons?
[20:45:23] <Skullworks_PGAB> yes - buttons switches, bells & whistles
[20:45:26] <Jymmm> http://www.ultimarc.com/ipac1.html
[20:45:56] <Jymmm> that's what I plan on using one of these years
[20:45:59] <ds3> so you must want the steam powered santa sleigh ;)
[20:46:23] <ds3> sleigh bells + steam whistle is stock
[20:46:32] <Skullworks_PGAB> Jymmm thats fine for buttons but useless for switches
[20:47:13] <Jymmm> why?
[20:47:23] <Skullworks_PGAB> and all feedback would have to be via GUI on screen
[20:47:26] <Jymmm> you can't hold down a button?
[20:47:50] <Skullworks_PGAB> ever get a stuck key on your keyboard?
[20:47:57] <Jymmm> nope
[20:48:01] <Jymmm> never
[20:48:10] <Jymmm> same kybd for the last 15 years too
[20:49:02] <ds3> be fun to get a jog key stuck ;)
[20:49:16] <Skullworks_PGAB> well - the way the I/O works any key held down would block other key signals
[20:49:28] <anonimasu> what happ
[20:49:33] <anonimasu> happens when you jog xy ;)
[20:49:34] <anonimasu> or xz ;)
[20:49:37] <Jymmm> like the SHIFT key?
[20:50:12] <Skullworks_PGAB> shift only modifys the ASCII value
[20:50:38] <ds3> think he is refering to matrix scanning where if a key gets stuck, you loose the row + column
[20:50:45] <anonimasu> I guess I should use a avr to decode keys, and then bitbang to hal..
[20:50:56] <Jymmm> Does not use a matrix - no ghost keys.
[20:50:58] <Skullworks_PGAB> that is handled in the kyb processor -
[20:51:06] <Jymmm> * Does not use a scanning method which causes a variable delay. Each input has it's own dedicated connection into the on-board CPU .
[20:51:06] <Jymmm> * Fast running interrupt-driven software gives much better response than a standard keyboard controller. Key debounce uses a state method for each key. No delays between keys depressed at the same time. (essential for fighting games).
[20:51:06] <Jymmm> * 2 operation modes. Either uses standard MAME keys or you can program your own character set.
[20:51:06] <Jymmm> * Shift functions. Holding "Start1" and pressing other buttons sends a range of codes for MAME functions such as "escape", "Coin 1", "tab", "enter". This means no extra buttons are needed on the cabinet. In programmable mode any input can be the shift key and all keys can be programmed with a shifted code.
[20:51:23] <Jymmm> really sorry about that
[20:51:23] <ds3> hmmmm
[20:53:30] <Skullworks_PGAB> I looked at that unit a year ago - and its fine if you don't want switches
[20:54:53] <Jymmm> I know what SPI is, but SPC? http://www.mitutoyo.com/TerminalMerchandisingGroup.aspx?group=1137
[20:55:11] <Skullworks_PGAB> but I want to make a "normal" CNC panal - including status lights... which that kyb interface can't do
[20:55:35] <Jymmm> Skullworks_PGAB, well why didn't you say so?! =)
[20:56:12] <Skullworks_PGAB> SPC = Statistical Process Control
[20:56:46] <Jymmm> propritary?
[20:57:05] <Skullworks_PGAB> the SPC I/O cable hooks into a PC for logging actual part measurement
[20:57:19] <Skullworks_PGAB> WE have it at work
[20:57:37] <Jymmm> serial port? special software?
[20:57:38] <Skullworks_PGAB> PITA - it feeds an ACESS data base
[20:57:58] <Skullworks_PGAB> only a serial port driver
[20:58:07] <Jymmm> hmmmm,
[20:58:36] <Jymmm> heck my cheapy dmm has datalogger sw with it
[20:58:36] <Skullworks_PGAB> M$ ACCESS - crashes atleast twice daily
[20:59:18] <Jymmm> Skullworks_PGAB ahve you ever just poked the serial port and see what you get? maybe with a terminal?
[20:59:24] <Skullworks_PGAB> worse cause they try to use Novell with Client for M$ networking - a real FRANKEN NET
[20:59:33] <anonimasu> oh
[21:00:42] <Skullworks_PGAB> The whole IT dept is clueless. (How they got hired I don't know.)
[21:02:25] <Jymmm> NetBios FTW!
[21:03:49] <Skullworks_PGAB> RE: I-PAC - You could just chop up a bad kyb and wire up the I/O from that - the only added bonus the I-pac hase is its programmable
[21:04:15] <Skullworks_PGAB> but then in linux you can map keys anyway...
[21:11:44] <anonimasu> well, it's too small..
[21:12:01] <anonimasu> and it's ultimately not nearly what a panel is..
[21:12:02] <anonimasu> more like a pendant..,
[21:44:38] <robin_sz> alex_joni, meep?
[22:07:38] <robin_sz> whoever wrote youtubes search algo needs a damn good kicking
[22:08:23] <Skullworks_PGAB> its intended to take you off on tangents
[22:09:49] <robin_sz> you think?
[22:10:31] <robin_sz> ok, so I seached for "black sabbath" ... becuase I fancied listening to a bit of rock ... sure enough it returns about 30 different tracks ...
[22:10:31] <cradek> http://www.youtube.com/watch?v=vEtM1q6gm9Q
[22:11:14] <robin_sz> so, I click on the first one I fancy listening to.. happens to be "heaven and hell" ... and what do I get inthe "related" bar along with the video?
[22:11:32] <robin_sz> about 30 copies of the same song ...
[22:11:43] <robin_sz> different uploaders, but the same thing
[22:12:06] <robin_sz> so to see anyting differetn, I have to go back a page, and then forward to the next thing
[22:12:50] <robin_sz> so yeah, normally, when its something obscure like CNC mill, it does lead you off at tangents .... and you cna cruise around to places you woudlnt otherwise go
[22:13:07] <robin_sz> but when its something popular, well, it just saturates
[22:13:35] <Skullworks_PGAB> then there are all the "reply" comment vids
[22:15:21] <eric_u> too bad it's not considered fair use to have Les Rita Mitsouko vids on youtube
[22:16:31] <Skullworks_PGAB> that passed right over my head - Les Rita who?
[22:17:15] <eric_u> actually, under the record companies' current idea about digital content, none of the rock vids are legal
[22:17:59] <eric_u> someday, they will be out of business, and artists will all be promoting themselves on youtube
[22:18:18] <eric_u> just like the ones that don't have contracts do today
[22:19:18] <Skullworks_PGAB> RIAA takes all the money - artists don't get crap
[22:19:52] <eric_u> that's true, the lawsuits of downloaders give the money to RIAA
[22:20:08] <eric_u> the settlement does not go to the artist who's song was downloaded
[22:20:14] <eric_u> that artist could still sue you
[22:20:39] <eric_u> don't really understand why riaa has standing to sue, it's racketeering
[22:21:12] <Skullworks_PGAB> well IIRC the BEASTY BOYS as a group were paid a total of $10,000 for "License to ill" yet it sold multi-million copies
[22:21:37] <Skullworks_PGAB> thats $10K divided amoung the whole group
[22:22:24] <eric_u> better than a lot of groups who end up owing money
[22:35:09] <eric_u> http://www.youtube.com/watch?v=BggXhzUhZ94
[22:37:45] <alex_joni> robin_sz: still there?
[22:42:01] <Rugludallur> Any comments/suggestions for this PyVCP panel -> http://imagebin.org/8051
[22:43:38] <alex_joni> Rugludallur: besides nice?
[22:44:04] <Skullworks_PGAB> yeah very nice - looks like the bases are covered
[22:44:19] <Rugludallur> alex_joni: more like "wouldn't a dudahbutton be better for that"
[22:44:20] <alex_joni> weell.. I would align the texts left (not center)
[22:44:40] <Rugludallur> alex_joni: hmm alignment works ?
[22:44:48] <alex_joni> Rugludallur: hmm.. no idea
[22:44:51] <alex_joni> :P
[22:45:29] <Rugludallur> alex_joni: cradek might know if he is around
[22:45:36] <Skullworks_PGAB> does that sidebar resize or fixed?
[22:45:39] <alex_joni> http://wiki.linuxcnc.org/uploads/axis_pyvcp.png
[22:45:48] <alex_joni> those seem leftaligned
[22:47:08] <Rugludallur> alex_joni: hmm I'm using the source for those as a reference ,, trying to find a way now
[22:47:59] <alex_joni> Rugludallur: don't waste time on that.. it's silly
[22:50:21] <Rugludallur> alex_joni: if jepler/cradek drop by i'll ask but I think that change requires Anchor tag which I suspect is not in head yet
[22:50:30] <robin_sz> alex_joni, I am
[22:50:45] <cradek> I don't know anything about vcp ...
[22:50:50] <robin_sz> alex_joni, have you run that HAL ui on a plasma yet?
[22:50:52] <cradek> are you saying there are things in 2.1 but not trunk?
[22:51:29] <alex_joni> robin_sz: not yet
[22:51:40] <alex_joni> cradek: no, it
[22:51:46] <robin_sz> thought about tech tables (ie cutting data) yet?
[22:51:48] <alex_joni> cradek: no, it's not available at all I bet
[22:52:00] <alex_joni> robin_sz: what about it?
[22:52:39] <robin_sz> thought about how to implement a store of say, feedrates, pierce times, arc voltage in EMC for each material
[22:53:41] <robin_sz> rather like a tool table on a mill
[22:54:00] <Rugludallur> cradek: There are things which are on jepler's web page (in the sample configs) which are not in head yet
[22:54:24] <cradek> interesting - I didn't think jepler did any of vcp
[22:54:37] <cradek> shows what I know
[22:55:04] <Rugludallur> cradek: actually might just show what I know, just a sec to verify
[22:55:56] <Rugludallur> cradek: my mistake, I mixed up stuff on linuxcnc.org and jeplers page :P
[22:55:58] <alex_joni> Rugludallur: bet that was awallin
[22:56:11] <alex_joni> he was the one working on pyvcp
[22:56:31] <Rugludallur> alex_joni: IC, thx
[22:59:51] <alex_joni> although I haven't seen him around lately
[23:00:47] <Rugludallur> alex_joni: me neither, but I think it might be exam/finals time for him, I seem to recall him attending the University of Helsinki
[23:01:46] <Skullworks_PGAB> * Skullworks_PGAB wishes he could afford to go back to college again.
[23:02:38] <alex_joni> Skullworks_PGAB: over here it's free :P
[23:02:46] <alex_joni> at least the first one..
[23:03:11] <Rugludallur> Skullworks_PGAB: it's also free over here, all education is free
[23:03:39] <Skullworks_PGAB> meanwhile I have a mortgage to pay etc.
[23:03:51] <Rugludallur> Skullworks_PGAB: and you can get a 3% intrest rate loan to pay for the cost of living
[23:04:26] <Skullworks_PGAB> that is tempting...
[23:04:36] <Rugludallur> Skullworks_PGAB: but then again we pay 38% income tax :P
[23:05:20] <Skullworks_PGAB> I'm not that far behind... between state / fed and local taxes
[23:06:26] <Skullworks_PGAB> besides - I was searching colleges here... they are going down hill badly
[23:07:03] <Skullworks_PGAB> Unless I want a degree in Cultural Arts...:(
[23:07:44] <Rugludallur> Skullworks_PGAB: :P
[23:10:59] <eric_u> Skullworks, do you have a degree?
[23:11:15] <Skullworks_PGAB> * Skullworks_PGAB has thought about Finland... and going back to working as a gunsmith.
[23:11:28] <Skullworks_PGAB> Welding metalurgy
[23:11:34] <Skullworks_PGAB> AS
[23:11:42] <Skullworks_PGAB> degree
[23:11:44] <jmkasunich> Skullworks_PGAB: finland is gun-friendly?
[23:11:53] <Skullworks_PGAB> very
[23:11:58] <jmkasunich> cool
[23:12:59] <eric_u> who posted the video of the gun blowing up the other night?
[23:13:13] <Skullworks_PGAB> shooting sports are still part of high school intramural sports competition
[23:14:00] <jmkasunich> could you imagine the wailing and whining if they did that here?
[23:14:38] <eric_u> I don't see that anyone would want to participate
[23:14:48] <Skullworks_PGAB> yeah - they kicked out all the jr ROTC programs and most the regular ROTC programs too
[23:15:06] <jmkasunich> eric_u: why do you say that?
[23:15:17] <Skullworks_PGAB> well it depends on where
[23:15:39] <eric_u> it's a chicken and egg situation
[23:15:59] <eric_u> plus, where would they put the firing range in the average town?
[23:16:19] <jmkasunich> .22 indoor - 50 feet long is plenty
[23:16:36] <Skullworks_PGAB> thats just it - in heavy urban areas there is no places anymore
[23:16:46] <eric_u> the way they build schools around here, that would be millions of bucks
[23:16:49] <Skullworks_PGAB> except indoor facilities
[23:17:36] <Skullworks_PGAB> and those they are shutting down via EPA due to trace airbourn lead particiles
[23:17:53] <jmkasunich> arrg
[23:17:52] <jtr> detention hall could be downrange - solve a lot of discipline problems :-D
[23:17:53] <eric_u> of course, you can't bring a fingernail clipper to school, so that's pretty much out anyway
[23:18:01] <jmkasunich> yeah
[23:18:48] <Skullworks_PGAB> yep - almost everything in my electronics tool box would be VERBOTEN
[23:19:04] <eric_u> I saw a cop show the other night
[23:19:21] <Skullworks_PGAB> I dred flying now
[23:19:35] <eric_u> he is going through a guys backpack, pulls out a batch of electronics tools and says "burglary tools"
[23:19:45] <Skullworks_PGAB> IIRC I'll have to check my laptop at the gate?
[23:20:02] <eric_u> I said holy crap, better not be pulled over in my car
[23:20:18] <jtr> Re: laptop - not here; just flew with two of them.
[23:20:28] <Skullworks_PGAB> cool
[23:20:38] <eric_u> I hope they never search my basement
[23:20:43] <Rugludallur> jtr: was that domestic ?
[23:20:53] <Skullworks_PGAB> they have free 802.11b at the local airport
[23:21:06] <Skullworks_PGAB> good to kill time while waiting
[23:21:18] <eric_u> Mr. Unterhausen, why do you have 10 different crowbars?
[23:21:18] <robin_sz> unusual
[23:21:41] <eric_u> most airports are charging big bucks and don't let you plug in your laptop
[23:21:41] <jtr> Yes, domestic. Had a small pouch with electronics tools as well - nothing over 7 inches long. No knives.
[23:22:00] <robin_sz> the ultimate burglary tools is:
[23:22:09] <eric_u> credit card
[23:22:15] <Skullworks_PGAB> they freaked over a #2 phillips
[23:22:18] <robin_sz> 24v lithium battery powered angle grinder
[23:22:25] <robin_sz> 1mm cutting disc
[23:22:31] <Rugludallur> robin_sz: true :D
[23:22:34] <robin_sz> seen those?
[23:22:38] <eric_u> should have gone with the #1 phillips
[23:22:44] <robin_sz> the 1mm thick discs?
[23:22:52] <jmkasunich> Skullworks_PGAB: I once found myself in an airport line with a single socket in my pocket- 3/8 drive, torx bit
[23:22:59] <Rugludallur> robin_sz: it's the only disks I use
[23:23:02] <jmkasunich> they wouldn't let me keep it
[23:23:13] <jmkasunich> security theater, make a show for the sheeple
[23:23:41] <jmkasunich> meanwhile, a flight attendent recently _accidentally_ carried a gun thru security and wasn't detected - not even trying to conceal it
[23:23:47] <robin_sz> Rugludallur, we use em a lot, amazing how quick they go through hardened steel
[23:23:56] <Rugludallur> robin_sz: for cutting my motto is the thinner the disk the better as long as they don't fly apart
[23:24:04] <robin_sz> yeah
[23:24:11] <Jymmm> Skullworks_PGAB you do gunsmithing?
[23:24:14] <Rugludallur> robin_sz: you can probably cut through a deadbolt in sub 3 sec for most doors
[23:24:25] <Skullworks_PGAB> worse - many years ago I arrived in LA and while unpacking found a box of 9mmx19 in my camera bag...
[23:24:25] <robin_sz> not had one fly apart yet, probably used 500 or so of them
[23:24:56] <eric_u> when I worked in a bike shop people would ask us to cut off their locks but then they would look at us like we were crooks when we lopped it off with the boltcutters
[23:25:02] <jmkasunich> many years ago as in pre 2001?
[23:25:06] <Rugludallur> robin_sz: I have had a couple break apart at 10k rpm +
[23:25:15] <Skullworks_PGAB> I wonder what would have happened if they had found it
[23:25:22] <Skullworks_PGAB> yeah like 96
[23:25:35] <jmkasunich> probably less than would happen now
[23:25:41] <robin_sz> just check it in as hold luggage
[23:25:46] <Skullworks_PGAB> of that I'm sure
[23:26:00] <eric_u> in 79 I carried my toolbox on a domestic flight
[23:26:06] <robin_sz> so long as its CIP made ammo and not handloads, you can check it in
[23:26:09] <eric_u> no problem at the local airport, they didn't like it in chicago
[23:26:12] <cradek> I haven't been on a plane since Nov '01 - I hope to avoid it for a few more years
[23:26:30] <jmkasunich> I traveled (by plane) 14 times for work last year
[23:26:35] <Skullworks_PGAB> BTW I was detained for that #2 phillips on a Greyhound bus
[23:26:37] <jmkasunich> thats above average - usually its about 6-8
[23:26:37] <eric_u> I haven't either, taken a batch of people to the airport though
[23:26:58] <eric_u> I've heard the busses can be worse
[23:27:00] <Skullworks_PGAB> I was going 1 way to pick up a Mustang GT I had bought
[23:27:02] <robin_sz> weird
[23:27:16] <eric_u> they don't really have an established security protocol
[23:27:19] <jmkasunich> do they have metal detectors for busses? how'd they find it?
[23:27:20] <robin_sz> when it gets that bad, the terrorists HAVE won
[23:27:31] <cradek> I took amtrak once and had no problems
[23:27:33] <eric_u> the terrorists have won a long time ago
[23:27:43] <Skullworks_PGAB> yeah well they waited until 440 miles down the road before running anyone thru security
[23:27:50] <eric_u> they got us to lose our heads, they won
[23:28:21] <jmkasunich> lose our heads?
[23:28:25] <cradek> eric_u: I have my fingers crossed for some sanity after '08
[23:28:34] <robin_sz> unlikely
[23:28:42] <Jymmm> cradek that an election year?
[23:28:54] <Skullworks_PGAB> then at the next stop - a truck stop - was a complete store - I bought a Buck knife and another screwdriver - and took pics to send the bus company.
[23:28:57] <robin_sz> election?
[23:29:07] <cradek> new president
[23:29:23] <robin_sz> oh, you elect them now?
[23:29:22] <Jymmm> cradek As long as the patriot acts goes bye bye
[23:29:23] <cradek> people are getting sick of the shit - it took years but they are
[23:29:39] <Jymmm> should have never been
[23:29:59] <cradek> robin_sz: every 8 years we get a new one - hard to say what the process is other than that
[23:30:06] <Skullworks_PGAB> yeah they belt the law into a pretzel with the patriot act
[23:30:11] <Skullworks_PGAB> bent
[23:30:14] <jmkasunich> patriot act: bad.... pretending we are not at war (and have been for many years): also bad
[23:30:24] <cradek> jmkasunich: we're not at war
[23:30:34] <cradek> jmkasunich: congress declares war in this country
[23:30:44] <robin_sz> undeclsared war is still war
[23:30:46] <jmkasunich> the war was declared on us, not by us
[23:30:57] <robin_sz> quite
[23:31:04] <Jymmm> Like I said, the Patriot Act should have never existed.... Since when in the last oh say 50+ years have we ever needed a "law" to allow us to go kick some butt.
[23:31:13] <Skullworks_PGAB> what was last offical war - Korea?
[23:31:19] <robin_sz> nah
[23:31:30] <Jymmm> Desert Storm
[23:31:36] <Jymmm> or was that not a "war"
[23:31:37] <jmkasunich> 1972, 1979, 1983, ..... 2001, 2005...
[23:31:53] <cradek> (I think it was korea)
[23:32:04] <cradek> certainly not vietnam I or II
[23:32:24] <Skullworks_PGAB> * Skullworks_PGAB needs to reboot - major java curruption issues.
[23:32:24] <robin_sz> desert storm was going fine, but the endign was mesed up by politics
[23:32:31] <Jymmm> no war in the 90's?! who was president?
[23:32:37] <robin_sz> carter?
[23:32:44] <cradek> clinton mostly
[23:32:52] <robin_sz> ah yeah
[23:32:55] <Jymmm> BRING BACK CLINTON!!! Make Love not War
[23:32:56] <Skullworks_PGAB> carter was late 70's
[23:33:13] <robin_sz> right
[23:33:23] <Jymmm> Can he re-run?
[23:33:25] <Skullworks_PGAB> * Skullworks_PGAB rebooting
[23:33:26] <Rugludallur> Jymmm: which clinton ,,, ,,,,
[23:33:30] <Jymmm> Bill
[23:33:38] <jmkasunich> she is running
[23:33:40] <Jymmm> Not PResident Hilary
[23:34:03] <robin_sz> will she be wearing THE dress on the campaign trail?
[23:34:10] <Jymmm> Bill "I got a velcro zipper" Clinton should re-run
[23:34:44] <Jymmm> Things were pretty good in the early 90's
[23:34:50] <Jymmm> mid
[23:34:56] <robin_sz> I see we justmanaged to create a whole lott trouble for ourselves in Iran ...
[23:35:32] <robin_sz> "I wonder what the brits will do about this nuclear thing ... I know, go grab some sailors .. see what happens"
[23:35:39] <jmkasunich> don't kid yourself - the war was going on then too: 1993 WTC, 1996 Khobar Towers, 1998 Embassies in Africa. 2000 USS Cole
[23:35:44] <robin_sz> "nothing, apparently"
[23:35:52] <jmkasunich> yep...
[23:35:53] <Jymmm> jmkasunich ah
[23:36:21] <Jymmm> jmkasunich I reject your reality, and substitute my own!
[23:36:28] <robin_sz> there are people over here calling for us to decomission Trident
[23:36:28] <jmkasunich> after 2001 we woke up, for a little while... .then back to sleep
[23:36:42] <Jymmm> trident missle?
[23:36:45] <cradek> and then we invaded an irrelevant country
[23:36:49] <robin_sz> we should have "decomissioned"one over tehran
[23:37:11] <robin_sz> cradek, but they do have oil
[23:38:31] <tomp> jmkasunich: comments on gEDA-Hal components? whats needed? http://imagebin.org/8048
[23:39:04] <robin_sz> at the end of the day, the supply of oil is crucial. It feeds our economies, we'd all be poor and cold without it. I mean really poor. this internet thing woudl be a way too expensive luxury for us
[23:39:28] <jmkasunich> I would not have the pins labeled "parport.N.pin-08-in.... just use pin-08-in, the parport.N applies to the part as a whole (makes it less crowded)
[23:40:39] <robin_sz> I'm pretty certain I don't want to be that poor and cold, so yeah, it is worthwhile to make sure it keeps rolling in. I suspect thats not a popular view, but i suspect its closer to the truth than many will admit
[23:51:59] <alex_joni> good night all
[23:52:05] <jmkasunich> night alex
[23:55:13] <tomp> gnite alex