**** BEGIN LOGGING AT Thu Jan 17 03:00:01 2013
Jan 17 03:34:36 <prpplague> mranostay: can't believe you didn't +1 the cowboid with purple hat
Jan 17 03:34:47 <coldsoup|work> toneeee: or rather, those networks don't need to be switching in public. How indecent
Jan 17 03:36:22 <mranostay> i commented didn't i?
Jan 17 03:36:36 <XorA> prpplague: rockin!!!!#
Jan 17 03:37:49 <prpplague> mranostay: not that i saw...
Jan 17 03:38:01 <toneeee> lol
Jan 17 03:38:08 <prpplague> mranostay: https://plus.google.com/101339419642360856354/posts/TeuGnKNPTWL
Jan 17 03:41:01 <mranostay> drinking a brew and getting my tax paperwork ready..
Jan 17 03:41:31 <mranostay> prpplague: you will love CA taxes :)
Jan 17 03:42:29 <prpplague> mranostay: hehe yea those are words i often use in the same sentence "love" and "taxes"
Jan 17 03:42:44 <prpplague> mranostay: the whole idea of moving to CA makes me want to vomit
Jan 17 03:43:07 <mranostay> prpplague: you'll get over it :)
Jan 17 03:44:53 <prpplague> mranostay: hehe i doubt it, but i have to be realistic about work
Jan 17 03:45:44 <XorA> prpplague: CA is known to cause cancer :-D
Jan 17 03:45:59 <prpplague> XorA: hehe
Jan 17 03:46:00 <coldsoup|work> but only to the state of CA
Jan 17 03:46:46 <prpplague> hehe
Jan 17 03:47:26 <prpplague> argh
Jan 17 03:47:33 * prpplague has way too much work to do tomorrow
Jan 17 03:50:34 <prpplague> now a move to colorado might be interesting...
Jan 17 03:51:03 * prpplague has never lived anywhere that had regular snow falls
Jan 17 03:52:06 <prpplague> hmm
Jan 17 03:52:26 <prpplague> XorA: think a hdmi debug device would sell for $100 ?
Jan 17 03:52:52 <XorA> prpplague: no idea, I never done HDMI stuff
Jan 17 04:02:37 <mrpackethead_> engineers are the worst peopel to set prices
Jan 17 04:02:56 <mrpackethead_> products should not be priced based on what they cost
Jan 17 04:03:06 <mrpackethead_> they shoudl be based on teh value that you can get for them.
Jan 17 04:08:24 <mranostay> prpplague: just need to come up with our pet rock idea :)
Jan 17 04:09:31 <prpplague> mrpackethead_: hehe i don;t set the prices for our stuff....
Jan 17 04:09:34 <prpplague> mranostay: hehe indeed
Jan 17 04:19:07 <mranostay> prpplague: or in our field the rpi idea
Jan 17 04:22:57 <prpplague> mranostay: well for me the flyswatter2 is doing pretty darn good
Jan 17 04:23:11 <prpplague> mranostay: we have about 10 other tools we are looking at producing
Jan 17 04:23:42 <mranostay> need consultants? :)
Jan 17 04:24:07 <prpplague> mranostay: this HDMI debugger that i am whipping up for robclark might be something that people would be interested in
Jan 17 04:24:10 <prpplague> mranostay: hehe
Jan 17 04:24:16 <prpplague> mranostay: i'll keep it in mind hehe
Jan 17 04:24:39 * mranostay curses panto!
Jan 17 04:25:49 <mranostay> wtf does mean "alias tbclk already exists"..
Jan 17 04:28:09 <mranostay> emeb_mac: fpga dude
Jan 17 04:28:25 <emeb_mac> mranostay: how goes it?
Jan 17 04:29:07 <emeb_mac> this evening's fiddling around: http://www.youtube.com/watch?v=7DP4uyz-n7E
Jan 17 04:29:08 <mranostay> it goes
Jan 17 04:29:36 <mranostay> with the logic level converte hack right?
Jan 17 04:29:51 <emeb_mac> yep
Jan 17 04:30:07 <emeb_mac> I get a pretty decent framerate with that hack
Jan 17 04:30:19 <emeb_mac> about 100fps on the spinny cube
Jan 17 04:30:37 <emeb_mac> can run the spi clock > 18MHz.
Jan 17 04:31:31 <prpplague> nice
Jan 17 04:33:04 <emeb_mac> probably ought to try hooking that display up to my beaglebone at some point just for teh lulz
Jan 17 04:44:17 <mrpackethead_> your gun fight escalated today i see
Jan 17 04:44:59 <emeb_mac> see y'all at the OK Corral.
Jan 17 04:45:09 <emeb_mac> it's a shorter drive for me. :)
Jan 17 05:04:58 <mranostay> emeb_mac: you sure that is a good idea?:)
Jan 17 05:05:25 <emeb_mac> mranostay: prolly not - place is a tourist trap.
Jan 17 05:07:26 <mranostay> probably shoot a tourist?
Jan 17 05:19:50 <mranostay> hi sakoman
Jan 17 05:32:16 <mrpackethead_> whole days work
Jan 17 05:32:35 <mrpackethead_> i've finished a front end audio section
Jan 17 05:33:03 <mrpackethead_> new cape design.
Jan 17 05:33:24 <mrpackethead_> pro audio line level input, and two line level outputs
Jan 17 05:33:40 <mrpackethead_> four bidirectional DMX input/outputs
Jan 17 05:36:36 <mranostay> oh no it is tema
Jan 17 05:37:02 <tema> mranostay, you're on a date with your laptop and irc on
Jan 17 05:37:03 <tema> ?
Jan 17 05:37:17 <KotH> he always is
Jan 17 05:37:27 <KotH> the youth today has no life
Jan 17 05:37:29 <KotH> ;->
Jan 17 05:38:23 <mranostay> need to get this demo working
Jan 17 05:38:36 <mrpackethead_> mranostay: u are on a date with a laptop?
Jan 17 05:38:40 <mrpackethead_> teh laptop is the date?
Jan 17 05:40:04 <emeb_mac> mrpackethead_: what do you consider pro audio line level to mean?
Jan 17 05:40:18 <tema> heh.. mranostay - "need to get demo working".. let's shoot for CES, ESC, Computex or something like that, forget ELC
Jan 17 05:40:19 <mrpackethead_> balanced input
Jan 17 05:40:37 <mranostay> tema: heh at some point
Jan 17 05:41:09 <mrpackethead_> +4 dBu
Jan 17 05:41:37 <emeb_mac> mrpackethead_: cool - I was working on a TRS balanced line-level design a while back using THAT Corp drivers/receivers.
Jan 17 05:42:02 * mranostay is cluebatting panto on his return
Jan 17 05:42:14 <mrpackethead_> im using the TLV320 IC
Jan 17 05:42:28 <mrpackethead_> same one that is on the audio Cape...
Jan 17 05:42:32 <emeb_mac> yeah
Jan 17 05:42:41 <mrpackethead_> its remarkably good
Jan 17 05:42:46 <emeb_mac> but what for the line level interfaces?
Jan 17 05:42:52 <mrpackethead_> same IC
Jan 17 05:43:05 <mranostay> heh bossman wants us to work the holiday.. don't think so :)
Jan 17 05:43:06 <emeb_mac> directly driving line level with the codec?
Jan 17 05:43:13 <mrpackethead_> it has both differential inputs and outputs
Jan 17 05:43:15 <mrpackethead_> yes.
Jan 17 05:43:20 <emeb_mac> hmmm...
Jan 17 05:43:39 <emeb_mac> is it rated for handling the impedances?
Jan 17 05:43:45 <mrpackethead_> yes.
Jan 17 05:43:59 <mrpackethead_> its actually got a Highpower output as well
Jan 17 05:44:07 <mrpackethead_> so you can drive a pair of headphones directly
Jan 17 05:44:32 <mrpackethead_> i was a bit miffed by the cape design,  they stick a 33R resistor in series with it, to pad it down to line level
Jan 17 05:44:58 <mrpackethead_> which was an odd design decison when they had a line level!
Jan 17 05:45:05 <mrpackethead_> its actually got about 10 inputs and 10 outputs
Jan 17 05:45:12 <mrpackethead_> i've got a EVM board for it
Jan 17 05:45:18 <mrpackethead_> and its remarkabley quiet
Jan 17 05:45:20 <emeb_mac> yeah - I took a quick look at it a while back
Jan 17 05:45:33 <mrpackethead_> for my applicaiton, it deos not need to be super studio quality
Jan 17 05:45:36 <emeb_mac> sounds pretty good on the dvi-d + audio cape
Jan 17 05:45:57 <mrpackethead_> the audio cape is the dvi-d cape minus the dvi bits
Jan 17 05:46:09 <emeb_mac> should probably compare the specs on the balanced I/O to those I was designing for back when...
Jan 17 05:46:29 <emeb_mac> ISTR that the THAT Corp parts had some fairly nice features for driving
Jan 17 05:46:33 <mrpackethead_> http://www.ti.com/general/docs/lit/getliterature.tsp?baseLiteratureNumber=slau209&fileType=pdf
Jan 17 05:46:55 <mrpackethead_> for my app, i only need 1 input, but 2 outputs
Jan 17 05:46:57 <mrpackethead_> bizzare
Jan 17 05:46:58 <mrpackethead_> :-)
Jan 17 05:46:59 <mrpackethead_> but true
Jan 17 05:47:29 <mrpackethead_> next job is to get a real time kernal running
Jan 17 05:47:41 <emeb_mac> fun
Jan 17 05:48:10 <mrpackethead_> though indications suggest i have enough horse power not to worry
Jan 17 05:48:37 <mrpackethead_> i thought my smpte Ltc decode would have taken a lot more processig than it does.
Jan 17 05:48:49 <mrpackethead_> it chuggs along at about 2% of the CPU
Jan 17 05:49:04 * XorA gives the angstrom site the evil eye
Jan 17 05:49:22 <mrpackethead_> XorA: I dont' get angstrom
Jan 17 05:49:34 <mrpackethead_> i'm not sure why it is so popular with the trolls that live here
Jan 17 05:49:55 <XorA> because it has an awesome spelling
Jan 17 05:49:57 <emeb_mac> Angstrom is nice - simple, small, does what it says on the tin.
Jan 17 05:50:04 <mrpackethead_> but then i'm a hobbit.
Jan 17 05:50:29 <mrpackethead_> and Hobbits need to avoid trolls, least they get eaten.
Jan 17 05:50:40 * XorA munches on mrpackethead_ 
Jan 17 05:50:44 <XorA> omnomnom
Jan 17 05:50:51 * mranostay backs away
Jan 17 05:52:15 * XorA smiles evilly
Jan 17 05:52:46 <XorA> can I build basic image in 20G these days?
Jan 17 05:53:18 <mrpackethead_> emeb_mac: so its simple and small..
Jan 17 05:53:30 <mrpackethead_> emeb_mac: how does that help me rapidlly develope stuff?
Jan 17 05:53:30 <KotH> XorA: it doesnt say that on the tin ;)
Jan 17 05:53:39 * mranostay checks to see if MOSFET caught fire
Jan 17 05:53:53 <mrpackethead_> i only stuck with debian because its what i am familar with
Jan 17 05:53:57 <mrpackethead_> for no other reason..
Jan 17 05:53:58 <XorA> KotH: normally Ive got a 1TB disk so havent been paying attention to disk usage
Jan 17 05:54:09 <mrpackethead_> and as yet, its not been a problem.
Jan 17 05:54:17 <KotH> mranostay: mosfets do not easily catch fire
Jan 17 05:54:23 <emeb_mac> mrpackethead_: depends on what your preferred dev tools are.
Jan 17 05:54:30 <XorA> nothing wrong with debian, its perfectly acceptable choice
Jan 17 05:54:31 <mrpackethead_> what i'm wanting to find out is what im missing.
Jan 17 05:54:32 <emeb_mac> I like a text editor, make and gcc
Jan 17 05:54:35 <KotH> mranostay: oil filled transformers on the other hand.... ;)
Jan 17 05:55:06 <mrpackethead_> mosfets dont burn well afer the intial 'puff' of smoke.
Jan 17 05:55:18 <mrpackethead_> aka, 1 time smoke machine.
Jan 17 05:55:38 <KotH> juup
Jan 17 05:55:43 <KotH> capacitors are much more fun
Jan 17 05:55:52 <mrpackethead_> im sure if i was to say, hey i'm using ubuntu.. i'd get a beating.
Jan 17 05:55:56 * KotH is currently playing with 20u 2kV capacitors
Jan 17 05:56:15 <XorA> mrpackethead_: well isnt that a valid response?
Jan 17 05:56:17 <KotH> mrpackethead_: using ubuntu is fine until you learn how to spell debian ;)
Jan 17 05:56:54 <mrpackethead_> but my front end of course is my macbook.
Jan 17 05:56:55 <mrpackethead_> :-)
Jan 17 05:57:12 <XorA> mrpackethead_: youll fit into Angstrom perfectly then :-D
Jan 17 05:57:25 <mrpackethead_> I will have a look at it sometime
Jan 17 05:57:30 <mrpackethead_> but right now, time is short
Jan 17 05:57:49 <mrpackethead_> i have to have a working demostration with the client by the end of the month.
Jan 17 05:57:54 <mrpackethead_> :-)
Jan 17 05:58:10 <XorA> mrpackethead_: makes sense to stay with what you know then, no argument from me
Jan 17 05:58:17 <mrpackethead_> the beagle is a better bet for making some money than mining bitcoins
Jan 17 06:01:08 <mrpackethead_> there was a reason why someone sends Angrstrom out with the Beagles though.
Jan 17 06:01:15 <mrpackethead_> and i'm sstill unclear as to what it is.
Jan 17 06:01:26 <XorA> because head Angstrom guy works for beagle manufacturer
Jan 17 06:01:59 <emeb_mac> and Angstrom was the 1st distro for the Beagle - before debian & ubuntu noticed it.
Jan 17 06:02:06 <mrpackethead_> ok ok.
Jan 17 06:02:18 <mrpackethead_> so as much as it was the first on the block.
Jan 17 06:02:43 <XorA> debian arm is pretty old as well, I was running it back in the zaurus days
Jan 17 06:02:53 <mrpackethead_> yes..
Jan 17 06:02:58 <mrpackethead_> i'm runnign squeeze
Jan 17 06:03:01 <mrpackethead_> old..
Jan 17 06:03:08 <mrpackethead_> but its stable
Jan 17 06:03:24 <XorA> got squeeze on my Tranformer Infinity
Jan 17 06:03:31 <mrpackethead_> 3.2.33-psp26 kernel
Jan 17 06:03:41 <XorA> the chrootless debian install is the mutts nuts
Jan 17 06:05:01 <mrpackethead_> never heard of it
Jan 17 06:05:28 <mrpackethead_> XorA: do you have a link?
Jan 17 06:05:40 <XorA> mrpackethead_: http://sven-ola.dyndns.org/repo/debian-kit-en.html
Jan 17 06:06:43 <mrpackethead_> i need to assign an engineer to develop our cloing process
Jan 17 06:06:57 <mrpackethead_> we'll need to build several hundred of the same image
Jan 17 06:07:09 <mrpackethead_> need to be able to do that resonably quickly.
Jan 17 06:07:21 <XorA> cp command doesnt work?
Jan 17 06:08:36 <mrpackethead_> can do dd's
Jan 17 06:08:40 <mrpackethead_> but thats not always great
Jan 17 06:09:40 <calculus> Gareth: pong
Jan 17 06:16:34 <emeb_mac> oi! reading thru the registers on the AIC3106. That's a lot of registers!
Jan 17 06:18:42 <mrpackethead_> yes
Jan 17 06:19:31 <emeb_mac> played with audio on the dvi-d + audio board the other day when I got it.
Jan 17 06:19:51 <emeb_mac> opened up alsamixer and paged thru all the controls. Yikes!
Jan 17 06:20:17 <mrpackethead_> its very capable
Jan 17 06:20:37 <mrpackethead_> but the bbtoys implemenation is slightly sad.
Jan 17 06:21:12 <emeb_mac> oh?
Jan 17 06:21:27 <emeb_mac> aside from not bringing out much of the possible I/O?
Jan 17 06:21:36 <mrpackethead_> they could have exposed a few more in/outputs for very little effort
Jan 17 06:28:16 <mrpackethead_> other than that its fine
Jan 17 06:28:30 <mrpackethead_> its just lost a lot of useful capablity by not doign that
Jan 17 06:28:37 <mrpackethead_> but i guess, it was acutyally designed as a DVI card
Jan 17 06:28:44 <mrpackethead_> and audio was an after thought.
Jan 17 06:29:12 <mrpackethead_> who knows.
Jan 17 07:25:52 <Dunearhp_> Can anyone tell me which is the best optimized recent ubuntu image for the beagleboard xm
Jan 17 07:25:54 <Dunearhp_> ?
Jan 17 07:29:38 <Dunearhp_> I'm currently trying ubuntu-12.04-preinstalled-desktop-armhf+omap.img.gz but is seems a little sluggish to me
Jan 17 07:30:27 <_av500_> http://hackaday.com/2013/01/12/finally-ti-is-producing-simple-cheap-wifi-modules/
Jan 17 07:30:33 <_av500_> wifi all the things
Jan 17 07:32:11 * KotH wifis _av500_ 
Jan 17 07:32:26 <Russ> I think av500 has wifi
Jan 17 07:35:04 <mrpackethead_> is the PRU expert about still
Jan 17 07:35:14 <mrpackethead_> ?
Jan 17 07:35:36 <mrpackethead_> mranostay:
Jan 17 07:35:38 <mrpackethead_> ?
Jan 17 07:35:51 <mrpackethead_> mranostay: pong ping bang crash..
Jan 17 07:37:24 <mrpackethead_> anyone else?
Jan 17 07:58:41 <cykon> I got a nice C program running on my beagleboard, and now I tried some C++ -- yet when I try to include iostream -- it gives me this error "/usr/local/angstrom/arm/lib/gcc/arm-angstrom-linux-gnueabi/4.3.3/../../../../arm-angstrom-linux-gnueabi/include/c++/4.3.3/iostream:77: undefined reference to `std::ios_base::Init::Init()'"
Jan 17 07:59:09 <cykon> Is the compiled iostream just... not working? or is it perhaps something on my end
Jan 17 09:12:17 <mrpackethead_> anyone got an idea of how to listen to a serial data stream (250kps) using the PRU?
Jan 17 09:13:44 <Russ> headphones?
Jan 17 09:14:22 <mrpackethead_> my ears crap out at about 15k
Jan 17 09:14:24 <mrpackethead_> :-(
Jan 17 09:14:43 <mrpackethead_> shoudl be doable
Jan 17 09:17:42 <KotH> listen?
Jan 17 09:17:45 <KotH> what do you mean?
Jan 17 09:17:52 <mrpackethead_> receive
Jan 17 09:17:59 <mrpackethead_> sorry bad words
Jan 17 09:18:03 <KotH> ah..
Jan 17 09:18:12 <KotH> doesnt the PRU have an uart?
Jan 17 09:18:12 <mrpackethead_> easy to bitbang out the serial streams
Jan 17 09:18:29 <mrpackethead_> only good to 115200
Jan 17 09:18:31 <KotH> simlarly easy to bitreceive in the serial streams
Jan 17 09:18:45 <KotH> how fast does the PRU run?
Jan 17 09:18:52 <mrpackethead_> 200Mhz
Jan 17 09:18:58 <KotH> o_0
Jan 17 09:19:11 <KotH> taht should be easily doable
Jan 17 09:19:19 <mrpackethead_> i woudl ahve thought.
Jan 17 09:19:25 <mrpackethead_> the serial isnt quite straight forward
Jan 17 09:19:26 <mrpackethead_> its DMX
Jan 17 09:19:30 <KotH> ?
Jan 17 09:19:33 <KotH> what's dmx?
Jan 17 09:19:34 <mrpackethead_> it has some oddball breaks in it
Jan 17 09:19:46 <mrpackethead_> its a protocol for controlling lighting
Jan 17 09:20:04 <KotH> ah
Jan 17 09:20:11 <KotH> hmm..
Jan 17 09:20:15 <Russ> can't you just sample at an appropriate frequency?
Jan 17 09:20:23 <mrpackethead_> yes.
Jan 17 09:20:32 <Russ> dmx512, that on PRU, that would have been a good elc talk
Jan 17 09:20:41 <av500> yep
Jan 17 09:20:52 <mrpackethead_> the outbound works easily
Jan 17 09:21:01 <mrpackethead_> i just want to do inbound as well
Jan 17 09:21:02 <av500> mrpackethead_: you have that?
Jan 17 09:21:07 <av500> outbound?=
Jan 17 09:21:15 <mrpackethead_> yes, i can transmit it
Jan 17 09:21:38 <mrpackethead_> one of the beaglecape contest projects inspired me
Jan 17 09:21:47 <KotH> mrpackethead_: sample with about 5-10times the frequency, estimate the midpoint of the bits, use that sample,
Jan 17 09:22:15 <KotH> shouldnt be too hard
Jan 17 09:22:50 <KotH> i dont know whether you've a timer or anything like that on the PRU, but even if not, just counting cycles should do the job
Jan 17 09:23:05 <mrpackethead_> its very deterministic for timing
Jan 17 09:23:10 <mrpackethead_> thats what the PRU is good at
Jan 17 09:23:25 <Russ> https://en.wikipedia.org/wiki/File:Annotated_trace_of_DMX-512_signal.png
Jan 17 09:24:33 <Russ> also don't forget you have two prus
Jan 17 09:24:38 <mrpackethead_> yes
Jan 17 09:24:50 <mrpackethead_> plan was to use one for TX
Jan 17 09:24:52 <mrpackethead_> and one for RX
Jan 17 09:25:11 <mrpackethead_> the memory mapping makes things easy on the other side
Jan 17 09:26:31 <mrpackethead_> this one is better http://www.erwinrol.com/dmx512/
Jan 17 09:27:58 <Russ> with rs232 style serial, you do you timing based on the start bit
Jan 17 09:28:22 <Russ> so long as neither side stays too far from 250khz, you shouldn't have a problem over the course of 8 bits
Jan 17 09:29:31 <mrpackethead_> i think you're right.
Jan 17 09:30:42 <Russ> you can schedule your sampling of bits based on the arrival of the start bit and the bit period
Jan 17 09:31:45 <mrpackethead_> yup, i think we will be ok.
Jan 17 09:31:53 <Russ> with PRU 800 cycles per bit, you should be ok :), even if you only use one PRU
Jan 17 09:32:03 <Russ> er, 800 PRU cycles
Jan 17 09:32:18 <mrpackethead_> and you get 1 instruciton per cycle
Jan 17 10:31:05 <panto> hello trolls
Jan 17 10:32:11 <av500> hello troll
Jan 17 10:37:39 <panto> now I'm hurt
Jan 17 10:38:49 <av500> good
Jan 17 10:39:36 <panto> somehow I will survive having this black spot on my soul
Jan 17 10:39:40 <panto> it won't be easy
Jan 17 10:44:14 <dm8tbr> *burp*
Jan 17 10:47:21 <av500> CC3000 sample ordered
Jan 17 10:48:07 <dm8tbr> nice :)
Jan 17 10:48:33 <dm8tbr> what cpu is in that thing? I didn't follow the specs yet. M3?
Jan 17 10:48:46 <av500> no idea
Jan 17 10:48:50 <av500> its an SPI device
Jan 17 10:48:54 <av500> you still need a host
Jan 17 10:49:04 <av500> I would prefer it to be an MCU too
Jan 17 10:49:08 <av500> with a few GPIOs
Jan 17 10:49:15 <av500> as is you need to add an MSP430 or so
Jan 17 10:49:53 <av500> ti has a trick to configure wifi on it: http://processors.wiki.ti.com/index.php/CC3000_Smart_Config
Jan 17 10:50:17 <av500> I think it transmits crpyted essid and pwd in wifi broadcasts
Jan 17 10:50:23 <av500> and the cc3000 picks that up
Jan 17 10:50:46 <av500> not sure what kind of a broadcast an android/ios app can send
Jan 17 10:51:10 <av500> I think I have some slides on that somewhere
Jan 17 10:51:30 <dm8tbr> interesting
Jan 17 10:51:36 <dm8tbr> but might also be error prone
Jan 17 10:52:06 <av500> https://itunes.apple.com/us/app/ti-wifi-smartconfig/id580969322?mt=8
Jan 17 10:52:19 <av500> dm8tbr: it forces the pax to renter his wifi stufdf
Jan 17 10:52:23 <av500> re-enter
Jan 17 10:52:48 <av500> which means to have no idea about the ESSID and to turn over the router to read the stock password :)
Jan 17 10:53:10 <dm8tbr> yeah :)
Jan 17 10:53:28 <dm8tbr> well it would be a security issue if random apps could read your network credentials...
Jan 17 10:54:57 <av500> yes
Jan 17 10:55:06 <av500> user has to enter it manually
Jan 17 10:56:20 <dm8tbr> nevertheless nice stuff if you want to offload the complete network and ip stack
Jan 17 10:57:04 <av500> yes
Jan 17 11:06:12 <av500> ah: http://processors.wiki.ti.com/index.php/File:Cc3000_ftc_process.PNG
Jan 17 11:06:29 <av500> but how can an android "app" request to connect to an AP?
Jan 17 11:07:45 <av500> http://processors.wiki.ti.com/index.php/File:Cc3000_ssid_fields.PNG
Jan 17 11:07:49 <av500> hmm, 32bytes max
Jan 17 11:35:41 <dm8tbr> heh, it will do wep64 but not wep128 in that mode
Jan 17 11:54:33 <mdp> mmm...the internet of things
Jan 17 11:55:10 <mdp> av500, a further demonstration of the TI focus "elsewhere"
Jan 17 11:55:45 <panto> I am always wary when I see piddling micros connect to the 'internets'
Jan 17 11:55:54 <panto> it's a wild wild world out there
Jan 17 11:58:07 <ynezz> not much stack for overflows anyway :p
Jan 17 12:02:14 <ynezz> mdp: you mean your network keys "elsewhere"?
Jan 17 12:02:55 <ynezz> as a showcase it's nice, but real world usability - none?
Jan 17 12:03:12 <mdp> panto, speaking of wild wild west: http://26-26-54.hardwarebug.org/123
Jan 17 12:03:43 <panto> mdp, as long as it's a soldering gun
Jan 17 12:03:47 <mdp> ynezz: I didn't say it was any good ;)
Jan 17 12:29:00 <KotH> mdp: what do you mean with "elsewhere"?
Jan 17 12:29:45 <mdp> KotH: just a comment about no more parts focus on handsets/tablets
Jan 17 12:30:01 <mdp> the rule also applies to wireless stuff
Jan 17 12:30:14 <KotH> mdp: ah..
Jan 17 12:30:22 <KotH> mdp: juup...
Jan 17 12:30:38 <KotH> mdp: i asked myself what they are going to do with the whole chipcon stuff they still own
Jan 17 12:31:11 <mdp> most of it applies "elsewhere"
Jan 17 12:31:36 <KotH> juup
Jan 17 12:31:52 <KotH> but there was also a big suppart that did wlan kind of stuff
Jan 17 12:31:59 <mdp> but stuff like wilink8 is purpose built for a handset...kinda useless now
Jan 17 12:32:00 <KotH> subpart*
Jan 17 12:32:08 <KotH> juup
Jan 17 12:32:26 <KotH> which reminds me... i always wanted to try to do something with the cc430
Jan 17 12:32:52 <mdp> we had some moron ask why we weren't using wilink8 in a board that needed to be an NFC reader ;)
Jan 17 12:32:54 <mdp> duh
Jan 17 12:33:55 <KotH> it would do the job ;)
Jan 17 12:34:27 <mdp> except for not having NFC reader capability ;)
Jan 17 12:34:49 <mdp> they forgot it's only the client protocols...purpose-built
Jan 17 12:35:03 <mdp> it was a "checkbox feature" person...
Jan 17 12:35:39 <mdp> cargo cult technologist ;)
Jan 17 12:35:44 <KotH> hehe
Jan 17 12:36:40 <KotH> btw: is nfc the same mess as rfid?
Jan 17 12:37:31 <dm8tbr> nfc essentially _is_ RFID, just with some parameters hardcoded and additional protocols on top
Jan 17 12:38:03 <ynezz> mess in RFID?
Jan 17 12:38:09 <mdp> KotH: what dm8tbr said
Jan 17 12:38:15 <mdp> a superset
Jan 17 12:39:23 <ynezz> there're ISO standards, there's no room for mess :p
Jan 17 12:39:49 <mdp> hehe
Jan 17 12:39:55 <ynezz> then GS1/EPC/NFC standards on top of it
Jan 17 12:39:57 <ynezz> then ...
Jan 17 12:40:03 <KotH> so, it's a total mess, not just a huge mess
Jan 17 12:40:25 <ynezz> where's mess, there's always market :)
Jan 17 12:40:47 <dm8tbr> matroska-style
Jan 17 12:41:08 <ynezz> gangman-style I would say
Jan 17 12:42:22 <bradfa> EPC's not a mess
Jan 17 12:42:34 <bradfa> NFC is though :)
Jan 17 12:42:44 <ynezz> oh really?
Jan 17 12:43:15 <ynezz> how's that
Jan 17 12:43:28 <bradfa> ynezz, which statement?
Jan 17 12:43:31 <bradfa> both?
Jan 17 12:43:33 <bradfa> :)
Jan 17 12:43:52 <bradfa> if you want to see a mess, go look at ISO 18000-7
Jan 17 12:44:20 <ynezz> -ENOMONEY
Jan 17 12:44:24 <bradfa> the spec is a mess (it's missing about 100 pages of detail) and the org writing it had a coup last year
Jan 17 12:44:44 <bradfa> then they added a patent clause to their example code
Jan 17 12:44:52 <ynezz> geniuses :D
Jan 17 12:44:59 <bradfa> yes
Jan 17 12:45:11 <KotH> bradfa: a coup?
Jan 17 12:45:33 <ynezz> = want more money
Jan 17 12:45:42 <bradfa> KotH, overthrew the former leadership
Jan 17 12:45:53 <bradfa> probably in a farily peaceful way, I imagine, though
Jan 17 12:46:03 <bradfa> threw out the old leaders and put new ones in
Jan 17 12:46:15 <KotH> peacefull? like beheading?
Jan 17 12:46:17 <bradfa> and then kicked all the members out of the area where the specs are on the website
Jan 17 12:46:26 <bradfa> KotH, somewhat like that
Jan 17 12:46:43 <KotH> sounds like fun
Jan 17 12:46:48 <bradfa> KotH, oh yes
Jan 17 12:46:50 <KotH> and how come that did happen?
Jan 17 12:47:10 <bradfa> KotH, cause they've been building the spec for years and no one's adopting it (or paying them money, I presume)
Jan 17 12:47:14 <bradfa> plus the spec's not done
Jan 17 12:47:22 <bradfa> and 6lowpan's about to eat their lunch
Jan 17 12:47:31 <bradfa> ieee and ietf know what they are doing
Jan 17 12:47:42 <KotH> then, why would anyone even bother to stage a coup?
Jan 17 12:47:51 <bradfa> KotH, why would anyone want to get an MBA?
Jan 17 12:48:03 <KotH> to earn more money for doing less
Jan 17 12:48:18 <bradfa> KotH, well, some MBA types think they can save it, I guess
Jan 17 12:48:26 <bradfa> the engineers don't seem to think so
Jan 17 12:49:15 <KotH> well.. considering that most of these protocols are first implemented by smal companies... the one that has their specs online for free will usually win
Jan 17 12:49:25 <bradfa> KotH, yes
Jan 17 12:49:25 <KotH> apropos...
Jan 17 12:49:41 <KotH> does anyone have a copy of the scsi specs from before they closed the acces to the documents?
Jan 17 12:49:46 <bradfa> KotH, 802.15.4 cost me like $15 for all the parts, ietf specs are free
Jan 17 12:50:00 * KotH doenst pay for ieee standards
Jan 17 12:50:00 <bradfa> KotH, scsi is closed now?
Jan 17 12:50:10 <KotH> has been for 3 or 4 years
Jan 17 12:50:23 <bradfa> KotH, the core one is free but some of the amendments aren't free yet for the 802 series
Jan 17 12:50:29 <bradfa> Ethernet's mostly free
Jan 17 12:50:31 <KotH> the announcement read like "because we became important, we now charge you to get access"
Jan 17 12:50:53 <bradfa> I don't mind a few dollars, it's the ISO ones that each part costs $100 that I mind
Jan 17 12:50:54 <KotH> bradfa: i still dont pay for ieee standards ;)
Jan 17 12:51:17 * KotH likes gramatical ambiguity ;)
Jan 17 13:08:00 <KotH> does anyone know when distributors like arrow open?
Jan 17 13:08:35 <KotH> it's already 6am there, and still nobody picking up the phone
Jan 17 13:09:51 <av500> phone?
Jan 17 13:09:57 <av500> are you like "old"?
Jan 17 13:10:03 <KotH> yes
Jan 17 13:10:29 <av500> nice, my new ip cam has ssh
Jan 17 13:10:35 <av500> Linux AirCam 2.6.28 #1 PREEMPT Wed Sep 5 15:41:51 EEST 2012 armv5tel GNU/Linux
Jan 17 13:10:59 <aholler_> you don't use beagles for that?
Jan 17 13:11:09 <av500> not outside
Jan 17 13:11:22 <KotH> new? 2.6.28?
Jan 17 13:11:28 <KotH> that's even older than me
Jan 17 13:11:29 <av500> yes, newe
Jan 17 13:11:31 <mdp> av500, no telnet?
Jan 17 13:11:36 <av500> mdp: that too
Jan 17 13:11:40 <av500> just disabled :)
Jan 17 13:11:40 <mdp> rock on
Jan 17 13:11:44 <mdp> bah
Jan 17 13:11:53 <mdp> you and your fancy new ssh technology
Jan 17 13:11:55 <av500> http://www.ubnt.com/airvision#airCamHardware
Jan 17 13:12:05 <av500> ~100€ for a 720p outdoor ipcam
Jan 17 13:12:10 <av500> including POE injector
Jan 17 13:12:39 <KotH> trying to spy on your girls while playing outside?
Jan 17 13:12:43 <mdp> nice
Jan 17 13:12:52 <av500> KotH: no
Jan 17 13:13:16 <KotH> trying to spy on your neighbors while "playing" outside?
Jan 17 13:13:24 <av500> too cold for outside
Jan 17 13:13:26 <av500> :)
Jan 17 13:13:41 <KotH> there are all kind of fetishes ;)
Jan 17 13:14:06 <mdp> impressive..3 pack of those is 250USD shipped on ebay
Jan 17 13:14:10 * KotH just hopes this isnt one of these family friendly channels
Jan 17 13:15:05 <av500> mdp: yes
Jan 17 13:15:47 <mdp> KotH: creeper card time? ;)
Jan 17 13:16:05 <KotH> mdp: i hope not
Jan 17 13:16:11 * KotH isnt creepy...most of the time
Jan 17 13:16:17 <KotH> mdp: but i've been banned for less ^^'
Jan 17 13:17:55 <mdp> av500, thanks for being my lazyweb...researching this stuff has been on my todo list for some time. ;)
Jan 17 13:18:18 <av500> you are welcome
Jan 17 13:18:19 <mdp> I'm using some old axis indoor cameras but needed a couple outdoor ones..with more capability
Jan 17 13:18:28 <mdp> and telnet!
Jan 17 13:18:32 <av500> :)
Jan 17 13:18:49 <av500> now I need to run ethernet up my air intake to the roof
Jan 17 13:19:22 <av500> luckily the patch panel is in the same room where the air axchanger is
Jan 17 13:19:24 * mdp imagines new camera names..."creepercam1" ...
Jan 17 13:19:55 <aholler_> I don't like do get monitored too, so it that doesn't match
Jan 17 13:20:15 <KotH> btw: i got my boss credit card... does anyone want to have some fancy dinner tonight? ;)
Jan 17 13:20:30 * mdp books flight
Jan 17 13:20:57 <av500> KotH: buy 2 tickets to ELC
Jan 17 13:21:29 <mdp> heh
Jan 17 13:21:37 <brainuee> is there anyway to boot beagleboard without micro sd card, because, board is not reading file from any sd card
Jan 17 13:21:52 <brainuee> can i boot using pendrive?
Jan 17 13:22:26 <KotH> have you read documentation?
Jan 17 13:22:31 <aholler> no, sounds like you've killed it with that fraction of a second
Jan 17 13:22:49 <aholler> electrons are fast
Jan 17 13:23:05 <KotH> fractions of a second?
Jan 17 13:23:19 * KotH didnt know that fractured seconds could damage electrons
Jan 17 13:23:42 <av500> brainuee: no, not using pendrive
Jan 17 13:23:48 <av500> it boots fine from SD
Jan 17 13:24:08 <brainuee> can i use pendrive for root partion?
Jan 17 13:24:15 <brainuee> if yes, then how?
Jan 17 13:24:51 <brainuee> because i can perhaps boot by sending u-boot file from laptop to board through y-modem
Jan 17 13:25:08 <KotH> brainuee: do you know linux?
Jan 17 13:25:15 <brainuee> no :(
Jan 17 13:25:24 <KotH> thought so
Jan 17 13:25:32 <KotH> go, get a good book on linux
Jan 17 13:25:55 <brainuee> i have stuck, and dont have time to finish project, so plz help me
Jan 17 13:26:03 <aholler> brainuee: try if you can and checkout how to proceed afterwards
Jan 17 13:26:11 <KotH> brainuee: i cost 500CHF/hour
Jan 17 13:26:17 <KidBeta> CHF?
Jan 17 13:26:33 <KotH> the only stable currency left on this planet.. besides chcolate
Jan 17 13:26:36 <brainuee> http://pastebin.com/FHvcz2y9
Jan 17 13:26:54 <jonand> KotH: 500CHF/h to idle on IRC? :)
Jan 17 13:27:01 <KotH> jonand: ofc
Jan 17 13:27:05 <av500> brainuee: fix your SD card
Jan 17 13:27:10 <brainuee> my problem is above, so i concluded that board is not reading anything from sd card, so thought to use pendrive
Jan 17 13:27:49 <av500> it is reading from sd
Jan 17 13:27:57 <av500> where else would the boot loader come from?
Jan 17 13:28:04 <brainuee> then waht is problem?
Jan 17 13:28:10 <av500> no idea
Jan 17 13:28:14 <av500> make a new card
Jan 17 13:28:26 <av500> ** Unable to use mmc 0:1 for fatload **
Jan 17 13:28:28 <av500> ** Bad partition 2 **
Jan 17 13:28:34 <av500> your card might be bad/corrupted
Jan 17 13:28:41 <brainuee> i had back up of original card , so i ran other card, then that also showed same result
Jan 17 13:28:50 <brainuee> how can both card get corrupted at same time
Jan 17 13:28:51 <brainuee> ?
Jan 17 13:28:59 <KotH> wrong type of card
Jan 17 13:29:44 <KotH> try yet another card, different brand, different size
Jan 17 13:30:05 <av500> try a more recent MLO/uboot
Jan 17 13:31:04 <av500> try http://circuitco.com/support/index.php?title=BeagleBoard-xM#BeagleBoard-xM_Rev_C.2FC1.2FC2_Image
Jan 17 13:32:06 <brainuee> i have copy of original card, should not it work on my board in all case?  what is the purpose or reason for getting more recent MLO?
Jan 17 13:33:29 <brainuee> if sd card is booting, then can't i use pendrive for root file system?
Jan 17 13:36:35 <av500> yes
Jan 17 13:36:44 <av500> but you need to load the kernel from sd first
Jan 17 13:37:06 <aholler> no
Jan 17 13:38:55 <aholler> u-boot can load the kernel from usb (ehci) too
Jan 17 13:39:39 <aholler> but as he seems to be unable to build/verify/check partions on a sd-card, it's unlikely he can do that with an usb-stick ;)
Jan 17 13:40:08 <av500> yes
Jan 17 13:40:18 <av500> I'd still go with a know good image first
Jan 17 13:40:23 <av500> known*
Jan 17 13:51:41 <av500> https://fosdem.org/2013/schedule/track/embedded_and_mobile/
Jan 17 13:51:45 <av500> https://fosdem.org/2013/schedule/track/operating_systems/
Jan 17 13:54:25 <mdp> ok, OSHW bagpipes is reason alone to be there
Jan 17 13:55:16 <av500> http://wiki.baserock.org
Jan 17 13:55:23 <av500> gee, yet another build system
Jan 17 13:55:25 <aholler> mdp: if you get a paid trip, fosdem is worth the time
Jan 17 13:55:39 <mdp> yeah, some friends have been there and recommend it
Jan 17 13:55:51 <mdp> I just don't have time to go to everything I'd like to
Jan 17 13:55:53 <mdp> ;)
Jan 17 13:56:29 <KotH> mdp: just do it like av500 :)
Jan 17 14:03:45 <aholler> mdp: just be sure to have something else than just t-shirts in your luggage ;)
Jan 17 14:16:43 <bradfa> device tree, here I come! :)
Jan 17 14:18:42 <mdp> bradfa, you finally admit that it will fix your problems....that's the first step.
Jan 17 14:20:39 <av500> lol: Barkley Socket API
Jan 17 14:20:46 <av500> I prefer Barney Sockets
Jan 17 14:21:28 <aholler> firewalled?
Jan 17 14:21:46 <mdp> mmm Purple Sockets
Jan 17 14:23:55 <jackmitchell> With the 3.8/3.7 beaglebone kernel from Github does anyone else experience slow ssh connections. It takes around 5 seconds for me to log in over SSH
Jan 17 14:24:08 <jackmitchell> I'm trying to determine if it is a userland or kernel issue...
Jan 17 14:24:25 <bradfa> mdp, not only device tree is solving all my problems, it's going to do so on 3.8-rc3!
Jan 17 14:24:30 <bradfa> mainline, baby!
Jan 17 14:24:56 <bradfa> built! first try! (we'll see if it actually runs next :)
Jan 17 14:24:57 <aholler> jackmitchell: try -4
Jan 17 14:25:00 * mdp sounds like a mgr he knows ;)
Jan 17 14:25:24 <mdp> er, bradfa, you sound like somebody at work here ;)
Jan 17 14:25:52 <bradfa> mdp, doesn't everyone at your work embrace mainline and device tree? :)
Jan 17 14:26:00 <bradfa> I thought mainline was the new company line
Jan 17 14:26:02 <bradfa> no?
Jan 17 14:26:25 <mdp> bradfa, sure
Jan 17 14:26:37 <bradfa> u-boot 2013.01, linux-3.8-rc3, what's next for me?  dropping debian for some cross built thing? :)
Jan 17 14:26:52 <jackmitchell> aholler: ssh 192.168.0.167 -l root -4
Jan 17 14:26:53 <mdp> the particular exclamation of "mainline, baby!!!" in excitable glee is what is humorous
Jan 17 14:26:56 <jackmitchell> aholler: still the same
Jan 17 14:27:01 <mdp> as if it's some "new thing" ;)
Jan 17 14:27:22 <Crofton|work> must beat the mainline drum constantly to overcome the inertia toward a million git repos
Jan 17 14:27:38 <Crofton|work> particualrly created by silicon vendors
Jan 17 14:27:52 <bradfa> Crofton|work, better only 1 million than 2 million :)
Jan 17 14:27:54 <mdp> bradfa, it's sometimes funny to step back and see how exciting it is to the ARM world
Jan 17 14:27:59 <jackmitchell> bradfa: I'm having a lot of success on mainline + devicetree
Jan 17 14:28:03 <aholler> Crofton|work: too late, the interwebs won't forget the million gits
Jan 17 14:28:05 <jackmitchell> bradfa: who would have guessed!
Jan 17 14:28:06 <bradfa> jackmitchell, hope to join you soon!
Jan 17 14:28:40 <mdp> bradfa, the list of gaps shrinks daily...
Jan 17 14:28:48 <bradfa> the big question for me, soon, is SPI performance on am335x on mainline, if that works well now I might just drop evil vendor tree
Jan 17 14:28:50 <jackmitchell> I've got a feeling my SSH issues may be related to logging...
Jan 17 14:28:51 <jackmitchell> Dec 11 13:11:17 beaglebone auth.info sshd[488]: Accepted none for root from 192.168.0.76 port 39019 ssh2
Jan 17 14:28:51 <jackmitchell> Dec 11 13:11:17 beaglebone auth.info sshd[492]: lastlog_openseek: Couldn't stat /var/log/lastlog: No such file or directory
Jan 17 14:28:51 <jackmitchell> Dec 11 13:11:17 beaglebone auth.info sshd[492]: lastlog_openseek: Couldn't stat /var/log/lastlog: No such file or directory
Jan 17 14:28:51 <jackmitchell> Dec 11 13:20:10 beaglebone auth.info sshd[488]: Received disconnect from 192.168.0.76: 11: disconnected by user
Jan 17 14:28:51 <jackmitchell> Dec 11 13:20:18 beaglebone auth.info sshd[494]: Accepted none for root from 192.168.0.76 port 39122 ssh2
Jan 17 14:28:51 <jackmitchell> Dec 11 13:20:18 beaglebone auth.info sshd[498]: lastlog_openseek: Couldn't stat /var/log/lastlog: No such file or directory
Jan 17 14:28:51 <jackmitchell> Dec 11 13:20:18 beaglebone auth.info sshd[498]: lastlog_openseek: Couldn't stat /var/log/lastlog: No such file or directory
Jan 17 14:29:08 <mdp> bradfa, rephrase question ;)
Jan 17 14:29:20 <jackmitchell> bradfa: SPI is my main use case for mainline, it works a lot better than the 3.2.vendor.tree.makes.me.suck
Jan 17 14:29:21 <bradfa> mdp, does SPI work well (with DMA) on am335x on mainline?
Jan 17 14:29:25 <mdp> yes
Jan 17 14:29:29 <mdp> for a long time now
Jan 17 14:29:36 * bradfa hugs mdp and dances
Jan 17 14:29:40 <mdp> well, using koen's evil vendor tree based on mainline
Jan 17 14:29:49 <koen> yes!
Jan 17 14:30:09 * bradfa hugs koen, too
Jan 17 14:30:21 * bradfa throws beer bottles up and dances like mranostay
Jan 17 14:30:31 <mdp> he has the latest v5 dmaengine/mmc/spi patches applied so we are "aligned"
Jan 17 14:30:32 <mdp> ;)
Jan 17 14:30:39 <bradfa> maybe this device tree really will solve my problems :)
Jan 17 14:30:47 <mdp> it will
Jan 17 14:30:54 * Crofton|work will beat koen at fosdem
Jan 17 14:31:12 * av500 will watch
Jan 17 14:31:14 <Crofton|work> although as a representative of a use of someone elses silicon
Jan 17 14:31:18 <Crofton|work> I understand his pain
Jan 17 14:31:29 <bradfa> fosdem rumble?
Jan 17 14:31:33 <av500> beating involves pain, no?
Jan 17 14:31:38 <mdp> koen, do you have the audio patches applied to 3.8?
Jan 17 14:31:44 * mdp is too lazy to look
Jan 17 14:32:10 <bradfa> jackmitchell, good to hear you're success with SPI
Jan 17 14:32:17 <Crofton|work> I totally undertstand Wind River wanting to use mainline + patches, not base from evil vendor trees now
Jan 17 14:32:23 <bradfa> gimme a few hours and I might have more successes here, I'll report back
Jan 17 14:32:27 <koen> mdp: yes
Jan 17 14:32:34 <mdp> koen, groovy!
Jan 17 14:32:37 <bradfa> Crofton|work, Wind River doing something sane?
Jan 17 14:32:37 <koen> av500: curse you for that .gif!
Jan 17 14:32:54 <Crofton|work> snae and insane all at the same time
Jan 17 14:33:13 <Crofton|work> when working with crazy upstream, you end up maintaining the patches
Jan 17 14:33:19 <mdp> koen, when will you switch our new and improved vendor tree? ;)
Jan 17 14:33:35 <mdp> Crofton|work: +1
Jan 17 14:33:36 <koen> mdp: likely never
Jan 17 14:33:44 <mdp> koen, ahh, c'mon ;)
Jan 17 14:33:51 <koen> mdp: "everything in master, no topic branches" is a recipe for disaster
Jan 17 14:34:13 <aholler> rebase patches once every 3 month shouldn't be that hard for proper supported hw.
Jan 17 14:34:14 <Crofton|work> koen, +1
Jan 17 14:34:38 <mdp> koen, they're not sure how it's going to be yet..that process thing is a proposal looking for comments.
Jan 17 14:34:54 <av500> koen: you are welcome :)
Jan 17 14:35:26 <av500> aholler: lol
Jan 17 14:35:46 <mdp> koen, I haven't fully reviewed, but my understanding is that there will be a new rebase for every two rcS of all our pending stuff
Jan 17 14:35:53 <koen> mdp: right now it looks like "this tree is on a different server, so it's better!!!!!one!"
Jan 17 14:35:59 <mdp> heh, s/rcS/rcs/
Jan 17 14:36:09 <av500> rc.d
Jan 17 14:36:39 <koen> mdp: I still can't (easily) extract a minimal patchset with that merge hell
Jan 17 14:36:59 <av500> guys, let me tell you how to do it:
Jan 17 14:36:59 <koen> mdp: it's exactly like a PSP tree, but with western names as committers
Jan 17 14:37:06 <av500> 1) finish the linux kernel!
Jan 17 14:37:08 <koen> av500: tarball on gopher!
Jan 17 14:37:09 <mdp> koen, yeah, I think they need a couple different trees
Jan 17 14:37:14 <av500> 2) add support for the BBone
Jan 17 14:37:17 <av500> 3) done
Jan 17 14:37:38 <jackmitchell> 4) ????
Jan 17 14:37:40 <jackmitchell> 5) profit
Jan 17 14:37:44 <koen> av500: the kernel is finished with 3.8
Jan 17 14:38:08 <av500> finally
Jan 17 14:38:11 <mdp> koen, somebody did recently mention that our needs for a -next style tree and then what is a useful stable tree are different
Jan 17 14:38:37 <koen> mdp: sudden outbreak of common sense?
Jan 17 14:38:40 <mdp> koen, Tartarus is supposed to get some direct input to this so it would help if you talk to him
Jan 17 14:38:51 * Tartarus runs away
Jan 17 14:38:55 <Tartarus> i mean, hm?
Jan 17 14:39:06 * koen offers Tartarus an americano
Jan 17 14:39:20 <mdp> Tartarus, he's a customer ;) can't hurt to see what would be useful
Jan 17 14:39:28 <av500> koen: 3 words on baserock
Jan 17 14:39:57 <mdp> koen, in theory, the person making that tree out to be able to provide something to save you time in culling all those patches for your product tree ;)
Jan 17 14:40:07 <Tartarus> koen: so, you're pretty insistent upon a break-downable tree, rather than a git URI to work off of, yes?
Jan 17 14:40:15 <mdp> considering the pain we've all gone through grabbing crap from all the mystery TII branches
Jan 17 14:40:56 <Tartarus> koen: have you seen the v3.8-rc3 based semi-proof of concept tree?
Jan 17 14:41:02 <mdp> koen, my only need from them is a -next-style tree to catch early conflicts...as upstream -next entry happens way too late to be useful.
Jan 17 14:42:32 <koen> av500: they love drugs
Jan 17 14:42:36 <KotH> mdp: use svn, then you can have best of both words: tons of branches and one single tree ;)
Jan 17 14:42:52 <koen> Tartarus: yeah, the git(ourious).ti.com
Jan 17 14:42:56 <mdp> KotH: not my department...don't care ;)
Jan 17 14:43:07 <mdp> my union contract says upstream ;)
Jan 17 14:43:08 <koen> Tartarus: it has complex merges, I don't like it
Jan 17 14:43:10 <Tartarus> koen: The one Dan finished yesterday I mean?
Jan 17 14:43:26 <aholler> suggesting svn is just dumb
Jan 17 14:43:30 <mdp> koen, it's -next
Jan 17 14:43:31 <av500> koen: that bad?
Jan 17 14:43:37 <koen> [remote "ti-linux"] url = git://gitorious.ti.com/ti-linux-kernel/ti-linux-kernel.git
Jan 17 14:43:42 <av500> aholler: no, it's a clever troll
Jan 17 14:43:43 <koen> Tartarus: I pulled that yesterday
Jan 17 14:43:56 <panto> huh, I need to get my crap in there too?
Jan 17 14:44:08 <mdp> panto, you were added to the thread, buddy
Jan 17 14:44:14 <Tartarus> koen: I assume Dan also published the what went into it file
Jan 17 14:44:15 * av500 is confused
Jan 17 14:44:16 <aholler> av500: some people do believe using svn solves something
Jan 17 14:44:19 <panto> yes, I am on the hook I guess
Jan 17 14:44:21 <Tartarus> or no, did that remain internal only?
Jan 17 14:44:30 * av500 thought there is no git tree, just an md5sum and patches
Jan 17 14:44:35 <av500> not there are gits everywhere
Jan 17 14:44:38 <av500> now*
Jan 17 14:44:40 <mdp> panto, I specifically told them they fscked up by not having you on the named tree owner list.
Jan 17 14:44:56 <bradfa> gitorious.ti.com says "System notice: Do Not Use - Under Test", I like it :)
Jan 17 14:44:57 <mdp> panto, since dt overlays is required for proper functionality...
Jan 17 14:45:14 <mdp> bradfa, they're not expecting us to start telling people about this
Jan 17 14:45:23 <panto> mdp, now I'm confused
Jan 17 14:45:32 <panto> (it is a common affliction)
Jan 17 14:45:46 <mdp> bradfa, it's because IT has been setting up this git.ti.com thing for several years now ;)
Jan 17 14:45:58 <mdp> panto, you have doubt?
Jan 17 14:46:00 <bradfa> mdp, ah, now I understand
Jan 17 14:46:23 <bradfa> I created an account
Jan 17 14:46:29 <panto> ok, so how do I go adding a topic branch to that tree?
Jan 17 14:46:30 <koen> Tartarus: I'm a huge fan of linear history in a vendor tree, so IMO the 'release' branches should be a clean rebase of every patchset against mainline
Jan 17 14:46:42 <aholler> +1
Jan 17 14:47:01 <mdp> "those above" do not like the way patches are just flung all over patchwork, they require an official TI linux tree with some rules
Jan 17 14:47:12 <Tartarus> So you want an un-kernel.org tree, koen?
Jan 17 14:47:15 <mdp> bradfa, just fwiw...why this is appearing now
Jan 17 14:47:16 <Tartarus> By which I mean
Jan 17 14:47:29 <bradfa> mdp?
Jan 17 14:47:30 <Tartarus> When maintainer X pushes things up the chain, it's not vs latest at the time
Jan 17 14:47:32 <koen> Tartarus: I like how greg-kh manages linux-stable
Jan 17 14:47:44 <bradfa> +1 for linux-stable type setup
Jan 17 14:47:46 <koen> Tartarus: patch queue in git, result is a number of clean branches
Jan 17 14:47:59 <koen> Tartarus: bleeding edge people can pull linux or even -next
Jan 17 14:48:12 <koen> linus*
Jan 17 14:48:25 <Tartarus> hmm
Jan 17 14:48:28 <mdp> bradfa, the trees you don't see is the linux-stable stuff...following the same process as greg's.
Jan 17 14:48:42 <koen> Tartarus: so the current master branch in ti-linux is more like -next: nice to test, unfit for use in products
Jan 17 14:48:46 <Tartarus> So linux-stable like, but with the "already in master" rule modified a bit
Jan 17 14:48:47 <bradfa> mdp, I can't see the forest, you expect me to see trees?
Jan 17 14:48:55 <mdp> bradfa, http://processors.wiki.ti.com/index.php/Linux_Kernel_Integration
Jan 17 14:49:33 <mdp> IMHO, they have a lot of things right...needs refined
Jan 17 14:49:51 <Tartarus> koen: So, kinda-sort, if the current master was rebase -q's on top of v3.8-rc3, and format-patch -o'd into another tree as well?
Jan 17 14:49:53 <bradfa> mdp, I'll take your word for it till I get device tree to solve my problems this morning :)
Jan 17 14:49:53 <mdp> to be sure it's useful
Jan 17 14:50:06 * bradfa goes and tries to boot, then read
Jan 17 14:50:09 <Tartarus> So ti-linux.git and ti-linux-queue.git
Jan 17 14:50:10 <mdp> bradfa, as always, I have doubt
Jan 17 14:50:16 <Tartarus> and some name coordination between branches of the two?
Jan 17 14:50:24 <koen> Tartarus: it'd be happy with a clean rebase already, seperate patches would be icing on the cake
Jan 17 14:50:47 <mdp> bradfa, I know, in practice, integration trees are the source of failure as they are abused.
Jan 17 14:51:00 <koen> Tartarus: I had customers at TI that required us to list each and every commit that wasn't in mainline to validation reasons
Jan 17 14:51:04 <Tartarus> koen: and you won't use git am to apply the series, or will expect git am failing things to have been fixed?
Jan 17 14:51:19 <koen> Tartarus: PSP couldn't tell us what was mainline and what was their work
Jan 17 14:51:20 <Crofton|work> Some poor bastards dissected the Xilinx zynq tree into sane patches
Jan 17 14:51:22 <Crofton|work> http://git.yoctoproject.org/cgit/cgit.cgi/meta-zynq/tree/recipes-kernel/linux/files
Jan 17 14:51:25 <Crofton|work> to make that
Jan 17 14:52:05 <mdp> Crofton|work: fairly concise set
Jan 17 14:52:22 <koen> Tartarus: personally, I'd use 'git am' or 'git am -3' on an old version of git to apply the patches, git cherry-picking a range is also fine
Jan 17 14:52:31 <Crofton|work> yeah
Jan 17 14:52:44 <Crofton|work> it would appear the deltas ar not that bad
Jan 17 14:52:46 <mdp> Crofton|work: I noticed that our integration team's first attempt ended up with 133 patches on top of mainline for both am33xx and omap.
Jan 17 14:52:47 <Tartarus> Just thinking of a problem we've had already with *someone*
Jan 17 14:52:49 <Crofton|work> or this is incomplete :)
Jan 17 14:52:51 <Tartarus> And non-am'able patches
Jan 17 14:52:52 <Tartarus> But, ok
Jan 17 14:53:03 <Tartarus> You'd be OK doing a little whacking of silly things
Jan 17 14:53:23 <Crofton|work> but if you start moving patches upstream, the situation gets better
Jan 17 14:53:34 <Crofton|work> mdp, how many of those patches would go upstream easily?
Jan 17 14:53:36 <mdp> Crofton|work: but that includes a crazy lot more of driver functionality patches that are in-flight on the lists, so not surprising.
Jan 17 14:53:49 <mdp> Crofton|work: 120+
Jan 17 14:53:53 * Tartarus does a quick apt-get source
Jan 17 14:54:29 <mdp> Crofton|work: in this case, I can honestly say that almost everything in there is stuff that's already in late stages of review
Jan 17 14:54:40 <mdp> or is in -next
Jan 17 14:54:44 <Crofton|work> mdp +1
Jan 17 14:54:51 <Crofton|work> the hard part is getting started
Jan 17 14:55:00 <Crofton|work> and changing mindset
Jan 17 14:55:16 <aholler> they thats the hardest part
Jan 17 14:55:45 <mdp> I contributed 25 or so alone to that list, 6 of which come from -next, and 18 or so waiting to be taken.
Jan 17 14:56:07 <mdp> the waiting is the hardest part, I hear
Jan 17 14:57:02 <koen> Tartarus: so, what I usually do as a test
Jan 17 14:57:03 <mdp> koen, so the verdict is "not useful"?
Jan 17 14:57:13 <koen> Tartarus: git cherry -v linus/master..vendor/branch
Jan 17 14:58:07 <koen> ehm
Jan 17 14:58:18 <koen> git cherry -v vendor/branch linus/master
Jan 17 14:59:00 * koen git cherry -v origin/master ti-linux/TI-Linux-3.8-rc3
Jan 17 15:00:43 <koen> Tartarus: the current ti-linux tree will list some patches twice or more
Jan 17 15:01:12 <Tartarus> fetching
Jan 17 15:01:12 <koen> wait
Jan 17 15:01:17 <koen> I'm blind
Jan 17 15:01:32 <koen> [koen@Angstrom-F16-vm-rpm kernel]$ git cherry -v origin/master ti-linux/TI-Linux-3.8-rc3 | grep noise
Jan 17 15:01:35 <koen> + 2d37eb202f3364317d44764c39bb7a4cd85491a8 crypto: omap-sham - Remove unnecessary pr_info noise
Jan 17 15:01:39 <koen> + 4a7bb9e430bda60ba5ce009998a56ecab90c20b6 crypto: omap-aes - Remmove unnecessary pr_info noise
Jan 17 15:01:42 <koen> that looks like a duplicate, but might not be
Jan 17 15:02:00 <mdp> koen, please note that the current tree has some mistakes too. the merger picked up some incorrect commits as this was the "first time"
Jan 17 15:02:02 <Tartarus> Yeah, bet it's not
Jan 17 15:02:50 <mdp> those two are legitimate commits above
Jan 17 15:03:52 <mdp> those are straight from mark's crypto sham and crypto aes series that are posted
Jan 17 15:04:43 <koen> the stuff below the merge commits looks sane
Jan 17 15:05:07 <jsabeaudry> mdp, how come I can't devmem2 McSPI registers? (I can devmem2 gpmc registers)
Jan 17 15:28:43 <jsabeaudry> join ##kernel
Jan 17 15:28:46 <jsabeaudry> oops
Jan 17 15:31:24 <mdp> join #bootloader + #kernel + #userspace
Jan 17 15:31:40 <mdp> jsabeaudry: nfc
Jan 17 15:32:14 <mdp> the only thing that shouldn't be able to be accessed via mmap is register files requiring privileged access
Jan 17 15:32:22 <mdp> e.g. CM
Jan 17 15:32:34 <jsabeaudry> what does nfc mean?
Jan 17 15:32:49 <mru> near-field comm?
Jan 17 15:32:50 <agmlego> Near-Field Communications.
Jan 17 15:33:01 <agmlego> RFID is a subset of NFC.
Jan 17 15:33:53 <jsabeaudry> not sure how that relates to devmem2 and mcspi
Jan 17 15:34:28 <mru> nfc module attached over spi?
Jan 17 15:34:42 <mdp> jsabeaudry: no [fine] clue
Jan 17 15:35:13 <mru> mdp: why'd you have to spoil a perfectly good troll?
Jan 17 15:36:18 <mdp> mru, I had to take pity on jsabeaudry ;)
Jan 17 15:37:10 <jsabeaudry> Yes, give a break to the imbecile here who can't make SPI work using DT, I deserve it
Jan 17 15:37:50 <mdp> what's the relationship of devmem2 to using McSPI w/ DT, btw?
Jan 17 15:38:49 <agmlego> "It's Complicated"
Jan 17 15:39:12 <jsabeaudry> mdp, Well, I attempted spidev using DT, and it does not work, so I want to check if the spi config registers are set properly
Jan 17 15:40:01 <mdp> jsabeaudry: hrm, would you believe it works for me? ;)
Jan 17 15:40:31 <jsabeaudry> I'm using "ti,pindir-d0-out-d1-in = <1>;" and I figured it probably did not get a huge amount of testing
Jan 17 15:40:37 <jsabeaudry> mdp, it probably does yes :)
Jan 17 15:41:02 <mdp> entirely possible
Jan 17 15:46:43 <jsabeaudry> When people refer to mainline, is it beagleboard/kernel @ 3.8 ?
Jan 17 15:48:16 <mdp> it depends on the person
Jan 17 15:48:49 <jsabeaudry> the person is bradfa
Jan 17 15:49:02 * bradfa scrolls back
Jan 17 15:49:04 <mdp> most people use it ambiguously, hence the problem
Jan 17 15:49:49 <jsabeaudry> There has been dancing about SPI and I would like to take part
Jan 17 15:49:49 <mdp> http://kernel.org has the canonical definition of what mainline is
Jan 17 15:50:05 <av500> mdp: ok, there is a drawback on the aircam
Jan 17 15:50:09 <av500> it can only stream
Jan 17 15:50:14 <koen> mdp: native compilation?
Jan 17 15:50:19 <av500> no jpg upload and such stuff
Jan 17 15:50:27 <av500> not even a full res jpg out of it
Jan 17 15:50:29 <av500> :(
Jan 17 15:50:31 <mdp> koen, I was careful to avoid the proper name ;)
Jan 17 15:50:31 <bradfa> jsabeaudry, when I say "mainline" I mean Linus tree
Jan 17 15:50:52 * mdp backslaps bradfa
Jan 17 15:51:04 <bradfa> jsabeaudry, when I say "evil vendor" I mean anything that TI puts out which isn't hosted on kernel.org
Jan 17 15:51:21 <bradfa> jsabeaudry, when I say 'linux-omap' I mean the linux-omap tree hosted on kernel.org
Jan 17 15:51:21 <koen> or what I put in the beagleboard tree
Jan 17 15:51:26 * koen is an evil vendor now as well
Jan 17 15:51:33 <mdp> when I say "evil vendor" I include the tree CCO puts out too ;)
Jan 17 15:51:33 <bradfa> koen's just koen to me :)
Jan 17 15:51:46 <panto> koen==evil
Jan 17 15:51:48 * bradfa has avoided koen's tree so far with great success
Jan 17 15:51:52 <mdp> "CCO puts out"
Jan 17 15:52:16 <panto> "CCO doesn't lift the toilet seat"
Jan 17 15:52:31 <jsabeaudry> bradfa, on which hardware?
Jan 17 15:53:48 <bradfa> jsabeaudry, I don't understand the question
Jan 17 15:53:58 <jsabeaudry> bradfa, also, when you refer to koen's tree is it koenkooi/linux or beagleboard/kernel?
Jan 17 15:54:04 <bradfa> jsabeaudry, either for koen
Jan 17 15:54:15 <bradfa> jsabeaudry, I don't use trees from koen or beagleboard group
Jan 17 15:54:24 <bradfa> jsabeaudry, I've been using PSP tree from TI
Jan 17 15:54:33 <bradfa> jsabeaudry, now trying to get kernel.org linus tree to boot
Jan 17 15:54:45 <jsabeaudry> bradfa, on bboard or bbone?
Jan 17 15:54:47 <bradfa> jsabeaudry, but I keep getting interrupted (both on irc and irl)
Jan 17 15:54:56 <bradfa> bone
Jan 17 15:54:56 <mdp> bradfa, that makes your tree an "evil customer tree"
Jan 17 15:55:07 <bradfa> mdp, yes, I'd agree with that :)
Jan 17 15:55:16 <bradfa> stupid evil customers, keep asking questions!
Jan 17 15:56:17 <mdp> bradfa, there's solid agreement that if we could get rid of the customers things would run more smoothly.
Jan 17 15:56:29 <jsabeaudry> PSP tree from TI, here's another one
Jan 17 15:56:46 <jsabeaudry> so confusing
Jan 17 15:56:59 <bradfa> jsabeaudry, it's not confusing once you understand :)
Jan 17 16:01:17 <jsabeaudry> I need some kind of matrix, a column for each tree and a row for each peripheral, feature, etc.
Jan 17 16:01:38 <bradfa> jsabeaudry, mainline -> all the features (real soon now)
Jan 17 16:01:58 <bradfa> jsabeaudry, if you just want something that works, use either koen's tree or PSP tree
Jan 17 16:02:03 <bradfa> on 3.2
Jan 17 16:02:19 <jsabeaudry> all the features but can you use GPMC on mainline
Jan 17 16:02:32 <bradfa> jsabeaudry, I don't know the answer to that
Jan 17 16:06:00 <mdp> jsabeaudry, is your spi pinmux set correctly?
Jan 17 16:06:32 <mdp> if you don't have both d0/d1 set with receiver enabled and you use the new property to flip MOSI/MISO...it will fail
Jan 17 16:08:59 <jsabeaudry> mdp, Oh both must have receiver enabled! That is certainly part of the problem
Jan 17 16:09:15 <jsabeaudry> mdp, Thanks for the info
Jan 17 16:09:39 <mdp> discussion with jackmitchell actually triggered that, thx to him
Jan 17 16:09:42 <mdp> np
Jan 17 16:09:52 <jsabeaudry> Is there still a way to change pinmux from debugfs in 3.8 or did it go away with omap_mux
Jan 17 16:10:04 <mdp> it's gone
Jan 17 16:10:43 <mdp> at some point, somebody will get annoyed and address that
Jan 17 16:10:56 <mdp> I was only annoyed enough to add read support to start
Jan 17 16:11:38 <jsabeaudry> mdp, Oh well thanks for that, i've been less'ing pins a lot to debug my DT
Jan 17 16:12:49 <mdp> too many top priorities atm, I might look at that soon though
Jan 17 16:13:14 <bradfa> well this isn't good -> "internal compiler error: Segmentation fault"
Jan 17 16:14:31 <bradfa> I shall not go to device tree today (it seems, people want me to make product work, or something...)
Jan 17 16:16:06 <mdp> bradfa, you promised! hell hath no fury like a DT scorned.
Jan 17 16:16:29 <bradfa> mdp, well, giving it one more shot, then going to eat food, then fix product problems for trade shows, or something...
Jan 17 16:17:21 * bradfa stares at snow waiting for build...
Jan 17 16:17:25 * bradfa needs more CPU
Jan 17 16:18:57 * mdp hugs his cost-effective amd 8 core
Jan 17 16:19:40 <panto> get a room!
Jan 17 16:20:00 <mdp> I am alone in the room with it already
Jan 17 16:20:10 * panto hands mdp his creeper card 
Jan 17 16:21:16 <mdp> heh
Jan 17 16:21:35 * mdp wonders if Vinod is on holiday this week
Jan 17 16:22:45 * av500 wishes we could discuss badly soldered serial cables like in ye olde days
Jan 17 16:22:53 <av500> all that tree talk makes me dizzy
Jan 17 16:23:09 <mdp> we could go back to shrubbery talk
Jan 17 16:24:32 <mdp> I'll review my new .cn-direct pl2303/ttl breakout boards when they arrive
Jan 17 16:24:48 <mdp> we can talk cold solder joints
Jan 17 16:25:10 <mdp> I can just see the cargo ship steaming my direction now
Jan 17 16:29:52 <aholler> no ti-chip for those breakout boards?
Jan 17 16:31:39 <mdp> a usb-serial converter? no
Jan 17 16:32:10 <mdp> otherwise you would see a TI one on all the ref brds that provide usb only debug support
Jan 17 16:33:46 <mdp> since I'm not longer permitted to have a dsub9 on boards, my #1 desire is to see a ftdi header on everything
Jan 17 16:33:56 <mdp> at least one future board gets that right ;)
Jan 17 16:34:00 <av500> mdp: there is TIUSB
Jan 17 16:34:15 <av500> er TUSB3410
Jan 17 16:34:20 <mdp> av500, I should have said, "cost effective" ;)
Jan 17 16:34:23 <av500> I have no idea why TI uses FTDI
Jan 17 16:34:31 <av500> when there is TUSB
Jan 17 16:35:10 <aholler> yes, the tiusb is used on the $5launchpad
Jan 17 16:35:25 <av500> aholler: no
Jan 17 16:35:34 <av500> lunchpack uses a 2nd MSP430
Jan 17 16:35:39 <av500> no?
Jan 17 16:35:51 <mdp> av500, legacy...ftdi is fundamental to the xds100v1/v2 definition and then wrapped into CCS
Jan 17 16:35:59 <av500> http://www.flickr.com/photos/av500/4941371504/
Jan 17 16:36:21 <aholler> av500: actually I'm not sure, but it's ti-only ;)
Jan 17 16:36:50 <av500> yes, no second source for MSP430
Jan 17 16:36:56 <av500> the .mil guys cannot use it....
Jan 17 16:36:58 <mdp> heh
Jan 17 16:37:07 <av500> what if north korea bombs texas....
Jan 17 16:37:22 * Tartarus sets his mentor ID badge picture to his TI.com gravatar
Jan 17 16:37:57 <mdp> av500, they are targeting Colorado..I saw this movie...
Jan 17 16:38:02 * av500 wonders if trolling e2e is any fun?
Jan 17 16:39:09 <aholler> av500: I think the second msp is for the emulation but usb is done by tiusb on the launchpad
Jan 17 16:39:55 <av500> ah yes
Jan 17 16:39:57 <av500> you are right
Jan 17 16:40:12 <aholler> it loads it's fw from an eeprom
Jan 17 16:40:28 <av500> forgot there are 3 chips
Jan 17 16:40:30 <av500> sorry
Jan 17 16:40:40 <aholler> http://www.flickr.com/photos/av500/4940706569/in/photostream/
Jan 17 16:40:54 <av500> jaja
Jan 17 16:43:29 <aholler> I remember having used tp1 to get 5v ;)
Jan 17 16:45:30 <aholler> but it's now old tech, no rgb-led ;)
Jan 17 16:47:48 <bradfa> mdp, ftdi headers are OK, I guess, but I still miss real RS-232C levels
Jan 17 16:47:50 * bradfa sheds tear
Jan 17 16:48:11 <mdp> bradfa, I find it fixable if they just give me the header
Jan 17 16:48:28 <bradfa> mdp, yes, I have sparkfun adapter on "that new board"
Jan 17 16:48:47 <mdp> bradfa, I have an old calao one on mine ;)
Jan 17 16:48:54 <bradfa> now if only I could find time to play with it...
Jan 17 16:48:58 <mdp> only the hated EVMSK confounds me now
Jan 17 16:49:29 <mdp> bradfa, panto rewrote the incorrect cpsw interrupt handling so networking works now.
Jan 17 16:49:41 <mdp> finally, I can switch "production" work to it ;)
Jan 17 16:50:56 * bradfa watered the device tree and sparks came out...
Jan 17 16:51:15 * bradfa does a dance again with mranostay!
Jan 17 16:51:17 <bradfa> it lives!
Jan 17 16:51:41 <panto> don't dance with mranostay
Jan 17 16:51:44 <panto> it's dangerous
Jan 17 16:51:52 <aholler> how to build a cloud
Jan 17 16:52:46 * mdp becomes entranced at "cloud"
Jan 17 16:52:57 <mdp> like watching an av500 animated gif
Jan 17 16:52:59 <bradfa> aholler, 1. get water 2. ??? 3. profit?
Jan 17 16:53:32 <bradfa> yey! https://gist.github.com/4557465
Jan 17 16:53:35 <aholler> water + sparks
Jan 17 16:53:43 <av500> bradfa: at work we ordered a 20pack of 1.8V FTDI cables
Jan 17 16:53:54 <av500> and I have a 5pack of sparkun 3.3V ones
Jan 17 16:54:04 <bradfa> av500, you're packing some heat there, eh?
Jan 17 16:54:24 * bradfa imagines av500 wearing a trenchcoat full of FTDI cables
Jan 17 16:54:45 <av500> mranostay: creeper card this guy please
Jan 17 16:54:53 <mdp> heh
Jan 17 16:55:01 * mdp notifies the HRbot
Jan 17 16:55:04 * bradfa accepts the honor of creepercard
Jan 17 16:55:29 <av500> mdp: easy: avconv -rtsp_transport tcp -i rtsp://192.168.1.135/live/ch00_0 -frames 1 snapshot.jpg
Jan 17 16:55:41 <av500> pull vs push, but I can live with that
Jan 17 16:55:59 <mdp> av500, nice
Jan 17 16:56:12 <jsabeaudry> mdp, I've been double checking everything, and still nothing on SPI1, would you be so kind as to take a look at my 20 lines of DT : http://pastebin.ca/2303957
Jan 17 16:56:44 <aholler> av500: it's linux ;)
Jan 17 16:57:05 <mdp> jsabeaudry: spi1_d1 needs to be muxed to 0x33
Jan 17 16:57:28 <mdp> aholler: are you sure it's not ubuntu?
Jan 17 16:57:28 <jsabeaudry> I'm not switching them anymore though
Jan 17 16:57:43 <aholler> tizeniced
Jan 17 16:57:46 <jsabeaudry> I'll change it anyway
Jan 17 16:57:53 <mdp> jsabeaudry: yeah, missed the commenting
Jan 17 16:58:01 <mdp> it was too obvious for me
Jan 17 17:00:00 <aholler> av500: maybe just installing a cron job on the thingy itself is possible
Jan 17 17:00:28 <av500> aholler: yes, but I need to grab the actual image
Jan 17 17:00:39 <av500> I was hoping for a way to get the sensor data
Jan 17 17:00:51 <av500> I need to poke around in the binaries...
Jan 17 17:01:05 <aholler> evil binary blobs
Jan 17 17:02:48 <av500> oops, I killed the streamer process
Jan 17 17:02:55 <av500> no more rtsp :)
Jan 17 17:03:21 <aholler> isn't it v4l2?
Jan 17 17:04:01 <aholler> just use opencv ;)
Jan 17 17:04:04 <av500> ah
Jan 17 17:04:09 <av500> that causes it to reboot
Jan 17 17:04:12 <av500> some watchdog
Jan 17 17:07:01 <mdp> jsabeaudry: I did a quick external loopback test here using spidev_test.c
Jan 17 17:07:33 <mdp> 3.8-rc3 + my dmaengine v5 stuffs (since that was handy)
Jan 17 17:08:05 <mdp> in summary, it works for me with the same configuration
Jan 17 17:08:28 <mdp> though loopback doesn't test any of the sclk/cs functionality.
Jan 17 17:12:34 <jsabeaudry> could you paste your dt?
Jan 17 17:12:51 <jsabeaudry> perhaps it's the spidev that is broken
Jan 17 17:14:41 <mdp> I also checked a modified version (yes, must send upstream rsn) that sends much more data and overcomes the dma threshold to make sure that's working
Jan 17 17:14:50 <mdp> yeah, sec
Jan 17 17:15:15 <jsabeaudry> oh, with the same pinmux is it for spidev1.0 or spidev 2.0 ?
Jan 17 17:16:10 <mdp> spidev1.0
Jan 17 17:16:53 <jsabeaudry> ah that's the issue
Jan 17 17:17:01 <jsabeaudry> those pins were spidev2.0 in 3.2
Jan 17 17:17:12 <av500> devmem 0x99000038 32 0xF81FE2A3
Jan 17 17:17:14 <av500> devmem 0x9900003C 32 0xFBFF80C4
Jan 17 17:17:17 <av500> aholler: thats in /init :)
Jan 17 17:17:28 <av500> membanging
Jan 17 17:17:33 <aholler> omg
Jan 17 17:17:53 <aholler> they've attented this channel
Jan 17 17:18:44 <aholler> s/t/d
Jan 17 17:18:51 <mdp> jsabeaudry:  https://gist.github.com/4557648
Jan 17 17:18:53 <av500>  /init is a script
Jan 17 17:19:02 <av500> that does a lot of mount and mknod
Jan 17 17:19:07 <av500> and checks the reset button :)
Jan 17 17:19:30 <aholler> I meant that they learned how to use devmem in #beagle ;)
Jan 17 17:19:35 <mdp> jsabeaudry: ok, yeah, they used fortran numbering in 3.2
Jan 17 17:19:36 <av500> yes
Jan 17 17:19:47 <av500> RST_FILE="/proc/gpio/reset_btn"
Jan 17 17:19:54 <mdp> jsabeaudry: be warned that other peripherals still use fortran numbering in mainline
Jan 17 17:20:00 <mdp> more crap to fix upstream
Jan 17 17:20:18 <mdp> e.g. gpio
Jan 17 17:20:20 <jsabeaudry> mdp, how come I have /dev/spidev1.0 and /dev/spidev2.0 yet no /dev/spidev0.0 ?
Jan 17 17:20:39 <mdp> spi1 -> spidev1.0
Jan 17 17:21:03 <av500> $error_msg = dict_translate("warn_third_party_firmware|WARNING: Uploaded firmware is third-party, make sure you're familiar with recovery procedure!");
Jan 17 17:21:05 <av500> hmm
Jan 17 17:21:11 <av500> so I can flash homebrew?
Jan 17 17:21:48 <jsabeaudry> mdp, oh wow here we are, back to inverting pins :), thanks for your help!
Jan 17 17:22:37 <aholler> av500: likely some tftp-stuff at bootup
Jan 17 17:23:54 <av500> Linux version 2.6.28 (buildd@builder) (gcc version 4.4.0 (Faraday C/C++ Compiler Release 20100325) ) #1 PREEMPT Wed Sep 5 15:41:51 EEST 2012
Jan 17 17:23:59 <av500> Faraday?
Jan 17 17:24:13 <av500> ah Machine: Faraday GM8126
Jan 17 17:24:44 <av500> console handover: boot [uart0] -> real [ttyS0]
Jan 17 17:25:40 <aholler> http://openipcam.com/
Jan 17 17:27:53 <mdp> jsabeaudry: btw, the mcspi driver also does this fortran numbering issue internally.
Jan 17 17:28:10 <mdp> so whoever shows up first is 1
Jan 17 17:28:50 <mdp> so with two instances enabled, spi0-> linux bus 1 and spi1-> linux bus 2
Jan 17 17:29:01 <mdp> if you only have spi1 enabled it will show as linux bus 1
Jan 17 17:29:06 <mdp> awesome? yes!
Jan 17 17:29:50 <mdp> this is to weed out people and preserve jobs in the embedded linux space
Jan 17 17:32:15 <bradfa> mdp, mmc will do this too
Jan 17 17:32:15 <jsabeaudry> mdp, brilliant!
Jan 17 17:32:30 <bradfa> which gets fun if u-boot sees 2 mmc devices but Linux doesn't
Jan 17 17:32:53 <mdp> bradfa, yep..it's a thing from OMAP(tm) where the IPs are fortran numbered..so the assumptions were hardcoded
Jan 17 17:33:17 <bradfa> no one ever said TI was consistent :)
Jan 17 17:33:25 <bradfa> just that they make calculators
Jan 17 17:33:26 <mdp> and also legacy from pre-OF where you had all instances turned on, all the time
Jan 17 17:33:41 <mdp> bradfa, you should read the thread Benoit and I had about this ;)
Jan 17 17:33:47 <bradfa> mdp link?
Jan 17 17:33:54 <mdp> lots of Sitara blame here ;)
Jan 17 17:33:55 <bradfa> or which ml?
Jan 17 17:34:00 <mdp> linux-omap
Jan 17 17:34:32 <mdp> https://patchwork.kernel.org/patch/1433001/
Jan 17 17:35:21 <mdp> there's a subtle jab in there about wtf they documented stuff differently in Sitara TRMs
Jan 17 17:35:22 <mdp> :(
Jan 17 17:37:07 <bradfa> oh fun
Jan 17 17:37:35 <bradfa> I already got to see the uart1 really is uart0 with my quick boot of device tree
Jan 17 17:38:00 <bradfa> will have to keep this in mind for my further dt work
Jan 17 17:38:03 <aholler> 01
Jan 17 17:38:15 <mru> bradfa: http://26-26-54.hardwarebug.org/3
Jan 17 17:39:14 <bradfa> mru, yes
Jan 17 17:39:41 <mru> hmm, the universe is consipiring... my credit card statement has a total of £1999.98
Jan 17 17:39:55 <mru> that looks intentional
Jan 17 17:40:11 <bradfa> mru, if only it was 999.99 you could save quite a lot :)
Jan 17 17:40:12 <Crofton|work> not much for disappearing for weeks
Jan 17 17:40:25 <Crofton|work> how far off the grid where you?
Jan 17 17:41:12 <mru> bradfa: save how?
Jan 17 17:41:43 <bradfa> if statement was 999.99 less, it'd be 999.99, thus you "save" money, no?
Jan 17 17:41:59 <bradfa> would require you to have bought less
Jan 17 17:42:00 <mru> if the amount was lower, I'd have spent less, yes
Jan 17 17:48:03 <bradfa> eh, stupid question, but should I be expecting mmc to have/need device tree data in order to be useful?
Jan 17 17:49:13 <panto> bradfa, in a DT kernel yes
Jan 17 17:49:20 <bradfa> panto, thanks, that's what I figured
Jan 17 17:49:41 <bradfa> panto, am I blind or do the dts and dtsi files in 3.8-rc3 for am335x not have mmc yet?
Jan 17 17:49:54 <panto> mainline?
Jan 17 17:49:56 <bradfa> yes
Jan 17 17:50:01 <panto> of course they don't
Jan 17 17:50:06 <bradfa> well, of course not
Jan 17 17:50:10 <panto> :)
Jan 17 17:50:29 <panto> use the evil vendor *cough*koen*cough* tree
Jan 17 17:50:40 <bradfa> ja ja ja, ok, I'll break down and use evil koen treee
Jan 17 17:51:25 <bradfa> panto, thanks!
Jan 17 17:51:58 * panto watches the big board
Jan 17 17:52:04 <mdp> bradfa, that's correct...still collecting acks...and besides..you need dmaengine support for that driver to work too
Jan 17 17:52:22 <bradfa> mdp, so what you're telling me is that mainline != for bradfa
Jan 17 17:52:32 <mdp> you only said you needed spi
Jan 17 17:52:37 <bradfa> true
Jan 17 17:52:40 <bradfa> touche
Jan 17 17:52:58 * bradfa goes to ml to find ackless patches
Jan 17 17:52:59 <mdp> I even granted you a uart driver for free
Jan 17 17:53:21 * bradfa hugs PSP kernel, "don't worry, I still love you!"
Jan 17 17:53:52 <mdp> https://lkml.org/lkml/2013/1/15/569
Jan 17 17:54:21 <bradfa> mdp just wants to prove he does work, sending links to lkml :)
Jan 17 17:55:06 <mdp> do note all the dependencies...either posted or in -next you need to collect as well
Jan 17 17:55:15 <bradfa> mdp, yes, noted
Jan 17 17:55:58 <mdp> heh, yeah, real proof, I suspose
Jan 17 17:56:16 <bradfa> mdp, I'll try your tree, see how that goes
Jan 17 17:56:20 <bradfa> thanks!
Jan 17 17:56:29 * bradfa suggests bigboard gets +1 mdp and +1 panto
Jan 17 17:56:53 <mdp> there's some testing commits in there and such...but the dependencies are all there in a concise set against -rc3
Jan 17 17:57:03 <mdp> if you don't want everything in an evil koen tree
Jan 17 17:57:45 <bradfa> mdp, I'll try your less-evil tree first :)
Jan 17 17:57:48 <mdp> bradfa, I  just noticed...the v4 branch cited...it was not updated in the cover letter to v5
Jan 17 17:58:07 <bradfa> mdp, ya, I noticed but was easy to find v5 branch for me
Jan 17 17:58:12 <bradfa> so no worries
Jan 17 17:58:16 <bradfa> fetching now
Jan 17 17:58:27 <mdp> bradfa, TIers can't figure that out..I should have known you could
Jan 17 17:58:42 <bradfa> mdp, that's why I'm not a TIer :) I know how github works
Jan 17 17:58:56 <mdp> true story, too. but I can't really tell it...
Jan 17 17:59:01 <mdp> :)
Jan 17 17:59:36 <bradfa> mdp, is github just not showing the linux repo files now?  they just show up for a second then all the files dispear in the web site
Jan 17 17:59:39 <bradfa> odd
Jan 17 17:59:54 <bradfa> firefox is also mad since apparently github uses flash?
Jan 17 17:59:56 <bradfa> wtf
Jan 17 18:00:02 <bradfa> maybe I don't know how to use github...
Jan 17 18:02:10 <mdp> works for me
Jan 17 18:02:35 <bradfa> they're really there, I think, just the web interface is barfing
Jan 17 18:02:37 <mdp> the only thing that bombs consistently on github for me is the Branches tab...50% of time I get webserver errors
Jan 17 18:02:50 <mdp> so I only use the branch: foo dropdown
Jan 17 18:02:52 <bradfa> mdp, yeah, and this week I've had all sorts of github issues
Jan 17 18:02:58 <bradfa> ponies lots
Jan 17 18:03:05 <mdp> not getting what we paid for
Jan 17 18:03:56 <bradfa> mdp, heh, $100 for us
Jan 17 18:03:59 <bradfa> per month
Jan 17 18:04:19 <mdp> ok, well, maybe I'm getting more than my money's worth here ;)
Jan 17 18:04:43 <bradfa> mdp, recommended config I should use from your tree?
Jan 17 18:05:07 <mdp> that omap2plus_defconfig in a later commit (not included in the posting) is known good
Jan 17 18:05:16 <bradfa> k
Jan 17 18:05:26 <bradfa> I had been using omap2plus, I'll continue
Jan 17 18:05:31 <bradfa> thanks again!
Jan 17 18:05:56 <mdp> my point with this branch is to ease testing since people tend not to be able to come up with working .configs , dts etc. for their specific applications
Jan 17 18:06:31 <mdp> there's even a throwaway gpio dma event driver example in there I needed since wl127x has no DT support
Jan 17 18:07:14 <mdp> perhaps some day someone will need a gpio dma event example ;)
Jan 17 18:08:22 <bradfa> dma bit banging?
Jan 17 18:08:53 <mdp> gpio banks can be dma event sources to start a transfer
Jan 17 18:09:14 <mdp> short version is that it was a way to test the am33xx crossbar support
Jan 17 18:09:22 <mdp> I'm a little crazy about testing
Jan 17 18:09:25 <bradfa> mdp, what, no uart2 in your dts?
Jan 17 18:09:34 <bradfa> :)
Jan 17 18:09:51 <mdp> there's a commit for UART4 in the branch ;)
Jan 17 18:10:04 <bradfa> yeah yeah
Jan 17 18:10:27 <bradfa> thankfully I can just use "uart1" real soon now :)
Jan 17 18:11:00 <mdp> I switched to UART4 on my bones as I have no use for gpmc for most work atm
Jan 17 18:12:13 <mdp> will be a non-issue with the new board
Jan 17 18:14:36 <bradfa> hmm, u-boot's ext2load no longer works within if statements...
Jan 17 18:14:51 <bradfa> or I'm an idiot
Jan 17 18:14:54 * bradfa is an idiot
Jan 17 18:20:12 <bradfa> mdp, success! mmc working, thanks!
Jan 17 18:20:41 <bradfa> 'dpkg --configure -a'ing now
Jan 17 18:20:46 <mdp> now, don't break anything that cause me to post a new version ;)
Jan 17 18:20:50 <bradfa> mdp, :)
Jan 17 18:21:33 <mdp> fwiw, anybody that's been using koen's tree for the last 3 kernel revs has been using some version of this..so hopefully nothing shocking falls out now ;)
Jan 17 18:23:13 <bradfa> https://gist.github.com/4558225
Jan 17 18:23:32 <jkridner>  anyone going to SCaLE?
Jan 17 18:24:30 <mdp> bradfa, mainline-ish! :P
Jan 17 18:24:43 <bradfa> mdp, close enough! :)
Jan 17 18:25:22 <mdp> grab am180x if you want to run an unaltered mainline tree :)
Jan 17 18:28:41 <mranostay> hey gentleman and jkridner
Jan 17 18:28:49 <mranostay> *gentlemen
Jan 17 18:28:58 <jkridner> gm mranostay
Jan 17 18:29:00 * panto kicks mranostay 
Jan 17 18:30:23 <bradfa> mdp, I might do that for upcoming possible project
Jan 17 18:30:52 <mranostay> ouch
Jan 17 18:31:05 <mdp> cool, also will work well on an EV3 brick when it comes out ;)
Jan 17 18:31:06 <jkridner> koen: you or anyone looking at pulling panto's patches into http://github.com/beagleboard/kernel ?
Jan 17 18:31:30 <jkridner> er, on 3.7 branch.
Jan 17 18:33:41 <bradfa> mdp, cost target on project I want to do is retail under $75 (my number)
Jan 17 18:33:53 <bradfa> doesn't need to be high performance
Jan 17 18:34:06 <bradfa> we'll see if it goes anywhere
Jan 17 18:34:42 <mdp> cool
Jan 17 18:35:05 <mdp> kickstarter retirement project?
Jan 17 18:35:14 <bradfa> mdp, sadly no
Jan 17 18:35:30 <bradfa> for "the company"
Jan 17 18:35:47 <mdp> I've already picked Belize once I run off with everybody's money.
Jan 17 18:35:51 <mdp> ;)
Jan 17 18:36:24 <mranostay> mdp: no island?
Jan 17 18:36:32 <mdp> that's played out!
Jan 17 18:36:35 <mdp> too cliche
Jan 17 18:51:02 <bradfa> mdp, any reason I shouldn't expect to get panics if user space is trying to use uarts that aren't setup properly? (ie: not all uarts userspace is trying to use have device tree data)
Jan 17 18:51:26 <bradfa> correction, not panics
Jan 17 18:51:41 <bradfa> circular locking dependencies
Jan 17 18:52:01 <aholler> mdp: is the ev3 a one-chip solution or does the cpu has .little (besides pru)?
Jan 17 18:58:42 <aholler> maybe nda
Jan 17 19:09:18 <mdp> aholler, just am1808
Jan 17 19:09:39 <mdp> bradfa: not sure
Jan 17 19:10:18 <bradfa> mdp, no worries
Jan 17 19:10:30 * bradfa gets back to other things
Jan 17 19:10:51 <XorA> identify amnin666
Jan 17 19:11:03 <XorA> ooops
Jan 17 19:11:08 <XorA> tralala
Jan 17 19:11:33 <dm8tbr> someone is chosing a new password tonight :)
Jan 17 19:12:02 <dm8tbr> quick, before someone ghosts you ;)
Jan 17 19:16:40 <mrpackethead_> has anyone thought about buidling a floating battery power supply to sit across the Beaglbones power input, to protect againt power failure?
Jan 17 19:17:01 <jkridner> apparently you have.
Jan 17 19:17:10 <mrpackethead_> anyone other than I
Jan 17 19:17:17 <mrpackethead_> and maybe done more thinking than i
Jan 17 19:22:18 <mrpackethead_> mranostay: are you working?
Jan 17 19:22:21 <_av500_> mrpackethead_:
Jan 17 19:22:24 <_av500_> oops
Jan 17 19:22:31 <_av500_> mrpackethead_: any of the battery packs
Jan 17 19:22:41 <_av500_> that charge from usb and deliver usb
Jan 17 19:23:02 <_av500_> they should work "inline"
Jan 17 19:23:17 <mrpackethead_> i dont' power up via usb
Jan 17 19:23:24 <mrpackethead_> but..
Jan 17 19:23:41 <mrpackethead_> yes, interesting concept.
Jan 17 19:24:01 <mrpackethead_> not sure what woudl happen if you tryed charging it on the usb port
Jan 17 19:24:06 <mranostay> i suppose that is what i'm doing
Jan 17 19:24:41 <_av500_> mrpackethead_: usb delivers 5v
Jan 17 19:24:46 <_av500_> feed that to the dcin
Jan 17 19:25:10 <_av500_> but these mobile phone battery packs mostly output on usb
Jan 17 19:25:55 <mrpackethead_> mranostay:  just a quick question for you..  is it fesible to expect to be able to receive four aysnc serial streams on the PRU and map that to the CPUs' main memory.. Data is being recevied at 250kbps
Jan 17 19:26:22 <mrpackethead_> easy to bit bang and TX it..
Jan 17 19:26:55 <mranostay> well that is wrong way to do it
Jan 17 19:27:27 <mrpackethead_> mranostay: ?
Jan 17 19:27:35 <mranostay> it is non-determistic once you touch outside the PRU
Jan 17 19:27:55 <mrpackethead_> yes.
Jan 17 19:28:02 <mranostay> you'd write that stuff to PRU data memory
Jan 17 19:30:33 <mrpackethead_> what you are saying is that its non deterministic to write to the CPU's memory from teh PRU?
Jan 17 19:30:42 <mrpackethead_> so, write to PRU memeory.
Jan 17 19:30:52 <mrpackethead_> and and then have the CPU read the PRU's memor?
Jan 17 19:31:23 <mranostay> yeah
Jan 17 19:31:29 <mrpackethead_> yeah.. ok
Jan 17 19:31:32 <mrpackethead_> follow
Jan 17 19:31:36 <mrpackethead_> ha'dnt quite got that far
Jan 17 19:31:45 <mrpackethead_> more a mind map of how i'm doig this.
Jan 17 19:31:49 <mranostay> there are demos in the pasm repo
Jan 17 19:32:04 <mrpackethead_> bit banging out the TX works fine
Jan 17 19:32:17 <mrpackethead_> but i want to receive as well
Jan 17 19:32:51 <mrpackethead_> i woudl ahve used the PRU Uarts but they are not fast enough it seems
Jan 17 19:46:46 <KotH> for the usian gun fanatics around here: http://sphotos-f.ak.fbcdn.net/hphotos-ak-ash4/408563_10152453731710515_950145928_n.jpg
Jan 17 19:47:21 <KotH> (and yes, you can see things like this in switzerland... or people shopping with a full automatic ausault rifle on their back)
Jan 17 19:47:23 <mrpackethead_> perhaps the crime rate has nothing to do with guns
Jan 17 19:47:30 <Crofton|work> heh
Jan 17 19:47:36 <KotH> mrpackethead_: pah!
Jan 17 19:47:37 <Crofton|work> I doubt crime follows guns
Jan 17 19:47:44 <Crofton|work> but they change the mix
Jan 17 19:47:45 <KotH> dont confuse me with facts!
Jan 17 19:48:03 <mdp> crime is stalker
Jan 17 19:48:12 <mdp> it needs a creeper card, badly
Jan 17 19:48:12 <_av500_> in fact, the "violence" rate is similar
Jan 17 19:48:25 <_av500_> its just that in the US, violence is more deadly
Jan 17 19:48:27 <KotH> these girls are probably 16 or so and taking one of the many shooting classes around in switzerland, that train you to use an assault rifle
Jan 17 19:48:45 <mrpackethead_> taking ont eh swiss in a fight woudl be insane
Jan 17 19:48:46 <mrpackethead_> :-)
Jan 17 19:48:56 <mrpackethead_> they have a fleet of armed paragliders
Jan 17 19:48:57 <mrpackethead_> :-)
Jan 17 19:49:07 <_av500_> http://www.slate.com/articles/health_and_science/science/2013/01/gun_violence_summit_at_johns_hopkins_researchers_present_data_and_analyses.single.html
Jan 17 19:49:08 <KotH> mrpackethead_: as an friend of mine put it: neutral, but heavily armed :)
Jan 17 19:49:21 <mrpackethead_> love switzerland
Jan 17 19:49:21 <_av500_> ...“We’re not more violent, but when we are violent, we kill.”
Jan 17 19:49:33 <mrpackethead_> have spent quite a ot of time there
Jan 17 19:49:45 <mrpackethead_> paragliding in the Lauterbrunnen valley
Jan 17 19:56:45 * mranostay gets beer bottles and guns
Jan 17 19:57:54 <emeb> drinking & plinking?
Jan 17 20:00:34 <mranostay> how else am i going to have targets?
Jan 17 20:03:12 <mrpackethead_> mranostay: i'm a target, shoot me
Jan 17 20:05:19 * mranostay switchs to airsoft
Jan 17 20:05:29 <mranostay> which really isn't that soft btw :)
Jan 17 20:06:21 <koen> jkridner: no, everyone ignores panto and his patches
Jan 17 20:06:58 <Russ> _av500_, sigh, yes, if only the magazine was limited to 10 rounds, then the kindergardeners could have fought back during the reload
Jan 17 20:06:58 <panto> koen, :(
Jan 17 20:07:06 <emeb> here's something fun to do w/ airsoft: buy the glow-in-the-dark pellets. mount UV LED inside autofire airsoft gun. take outside at night. fire into air for WWII tracer effect.
Jan 17 20:08:06 <jkridner> koen: sarcasm aside, it seems trying it on 3.7 is going to be a bit of a pain.
Jan 17 20:09:27 <koen> jkridner: I'm tired of people asking me to pull in patches that are already present
Jan 17 20:09:53 <koen> so everyone asking about patches gets the sarcasm treatment
Jan 17 20:10:18 <jkridner> I'm quite confident this one isn't there, since it is only a couple hours old.
Jan 17 20:10:37 <emeb> maybe patches need "sell-by" date.
Jan 17 20:11:16 <koen> jkridner: the irq storm is already in 3.8, but I haven't pushed it to github yet
Jan 17 20:11:34 <jkridner> Is touchscreen the only significant regression?
Jan 17 20:11:36 <koen> I'll see if it applies to 3.7 as well
Jan 17 20:11:42 <jkridner> it doesn't.
Jan 17 20:11:43 <koen> ts and adc, yes
Jan 17 20:11:52 <jkridner> readme says ADC works.
Jan 17 20:12:11 <koen> yeah, but no sysfs
Jan 17 20:12:24 <jkridner> oh. :(
Jan 17 20:12:34 <mranostay> jkridner: going to ELC? (not sure if i asked)
Jan 17 20:12:39 <jkridner> yes
Jan 17 20:12:50 <mrpackethead_> last thign to bolt down.. battery backup
Jan 17 20:12:53 <mrpackethead_> ugh.
Jan 17 20:12:54 <mranostay> oh good we get to troll^H^H^H^H talk to you :)
Jan 17 20:13:19 <mranostay> jkridner: booth this year?
Jan 17 20:13:32 <jkridner> don't think so.
Jan 17 20:13:44 <jkridner> tutorial room is the plan
Jan 17 20:13:50 <koen> jkridner: conflicts in 3.7 seem easy enough to resolve
Jan 17 20:13:52 <jkridner> prpplague is looking at details.
Jan 17 20:13:54 <mranostay> are exact steps available there?
Jan 17 20:14:26 <jkridner> koen: some of the structures changed, so it isn't 100% trivial to resolve.
Jan 17 20:15:23 <ka6sox> morning
Jan 17 20:15:47 <koen> jkridner: btw
Jan 17 20:15:50 <koen> jkridner: http://dominion.thruhere.net/koen/angstrom/beagleboard/Angstrom-systemd-GNOME-image-eglibc-ipk-v2012.12-beagleboard-2013.01.17.img.xz
Jan 17 20:16:09 <jkridner> I'm still writing 16 to a card!
Jan 17 20:16:11 <koen> jkridner: I did that for a customer today, I'll see if I can add more stuff to it so we can update the xM uSD card
Jan 17 20:16:25 <koen> jkridner: that's for board, not bone :)
Jan 17 20:16:26 <jkridner> oh, BeagleBaord!
Jan 17 20:16:37 <jkridner> nice.
Jan 17 20:16:41 <koen> still kernel 3.2
Jan 17 20:16:53 <koen> but a lot more recent userspace :)
Jan 17 20:18:07 <jkridner> with rcn-ee making lots of use of 3.7/3.8, I'd expect a move to be relatively easy. it will be nice to unify the kernel at some point. :( anyway, nice to see a new userspace build!
Jan 17 20:19:39 <koen> I'd like to move to 3.8, but that wasn't what the customer asked for :)
Jan 17 20:20:40 <koen> jkridner: like I said in the email, if we can get musb working in 3.8 we should ship with that
Jan 17 20:20:53 <jkridner> I see 3.8 has a new config option: MULTI_RNDIS...
Jan 17 20:20:58 <jkridner> I hope that'll help.
Jan 17 20:21:03 <jkridner> I'm going to try that out next.
Jan 17 20:21:10 <koen> PG2.0 fixes everything, you know
Jan 17 20:21:15 <koen> haven't you heard?
Jan 17 20:21:30 <koen> that combined with the fact that linux will be finished with 3.8 nothing can go wrong!
Jan 17 20:21:31 <jkridner> I really, really need an image with g_multi.
Jan 17 20:21:52 <koen> the .16 image should have all modules present
Jan 17 20:21:58 <jkridner> well, musb on PG2.0 worked fine for me on 3.7.2
Jan 17 20:22:12 <koen> of course usb is completely borken on 3.8
Jan 17 20:22:25 <jkridner> 2.2GB written to my SD card and waiting.
Jan 17 20:23:09 <jkridner> any specific issues to follow-up on?
Jan 17 20:23:49 <panto> jkridner, btw, any docs of the emmc part?
Jan 17 20:25:54 <_av500_> Russ: the more interesting part of the article are the final remarks
Jan 17 20:27:52 <Russ> they are only talking about a single data point, so it isn't statistically significant
Jan 17 20:28:30 <jkridner> panto: is https://www.micron.com/parts/nand-flash/managed-nand/mtfc2gmvea-0m-wt not sufficient?
Jan 17 20:29:12 <Russ> plus, I'm willing to bet that most firearm homicides are of someone who is known to the person, uses much less than 10 rounds, and doesn't matter what kind of gun it is, much less weapon
Jan 17 20:29:12 <panto> needs sign-in
Jan 17 20:29:35 <panto> anyway, time to bed
Jan 17 20:29:36 <panto> cya tomorrow
Jan 17 20:30:34 <jkridner> k, I'll complete and let you know what it looks like. :)
Jan 17 20:31:18 <panto> thx, time to go :)
Jan 17 20:34:23 <_av500_> Russ: I know, if you'd ban guns, mass shootings would be done by expert knife throwers...
Jan 17 20:34:45 <_av500_> or drowners with 10 buckets of water
Jan 17 20:34:54 <_av500_> :)
Jan 17 20:36:25 <jsabeaudry> mass throwings and mass drownings... can't wait for the hollywood videos calling for banning of hands and water
Jan 17 20:36:29 <wmat> _av500_: or by the really, really boring
Jan 17 20:36:52 <_av500_> mass boredom kills class of 20
Jan 17 20:36:57 <_av500_> that has happened :)
Jan 17 20:37:03 <_av500_> we barely escaped
Jan 17 20:37:35 <jsabeaudry> whew class of '20 didn't realize you were that old av500
Jan 17 20:37:40 <_av500_> and here I agree, only a good teacher can stop a bad one
Jan 17 20:37:50 <mranostay> jsabeaudry: +1
Jan 17 20:40:23 * wmat drops dead
Jan 17 20:42:46 <mdp> Friday?!?
Jan 17 20:46:38 <prpplague> wmat: wth?
Jan 17 20:47:45 <mdp> I think he dropped dead from boredom
Jan 17 20:48:03 <mdp> _av500_'s discussion killed him
Jan 17 20:48:58 <mranostay> can i have his stuff?
Jan 17 20:49:18 <mdp> is there a garbage collector that runs on this channel?
Jan 17 20:49:53 <mdp> mranostay, leave the wiki pages...those are for the betterment of humanity
Jan 17 21:12:11 <mrpackethead_> trolls, your opinion please.. other than (a) hardeing the filesystem (b) using very good SD cards what else coudl be done to harden a beagle so its more capable of sustaining 'unschedule' power cuts without curruption.
Jan 17 21:12:31 <mru> mount fs read-only
Jan 17 21:14:09 <_av500_> +1
Jan 17 21:14:40 <_av500_> read only rootf
Jan 17 21:14:42 <_av500_> rootfs
Jan 17 21:15:17 <_av500_> or some kind of battery backup
Jan 17 21:15:30 <_av500_> that at least lets you unmount cleanly
Jan 17 21:15:38 <_av500_> for the volatile data
Jan 17 21:15:45 <mrpackethead_> any other ideas?
Jan 17 21:15:56 <_av500_> move to europe, much more stable power
Jan 17 21:15:59 <mrpackethead_> lol.
Jan 17 21:16:11 <mrpackethead_> teh AC power here is fine.
Jan 17 21:16:19 <_av500_> boot from ethernet, keep filesystem in RAM
Jan 17 21:16:24 <_av500_> save money on sdcard
Jan 17 21:16:53 <mrpackethead_> the ethernet boot is nice
Jan 17 21:16:53 <mru> does the device need any non-volatile storage?
Jan 17 21:16:59 <mrpackethead_> but won't work for this app
Jan 17 21:17:09 <mrpackethead_> as its mostly never connected to the outside world
Jan 17 21:17:22 <mrpackethead_> and sticking another server to provide bootp or similar is'nt sensible.
Jan 17 21:17:32 <mrpackethead_> you just move the problem.
Jan 17 21:17:35 <_av500_> have a fallback solution with 2 filesystems
Jan 17 21:17:44 <_av500_> and something to decide when to fallback
Jan 17 21:17:57 <_av500_> now, that could break, so have a fallback for that too
Jan 17 21:19:31 <mrpackethead_> there comes a point at which you cna't protect everything
Jan 17 21:19:40 <mrpackethead_> i just need to make a "resonable" effort
Jan 17 21:19:44 <_av500_> yes
Jan 17 21:20:00 <mrpackethead_> for this app, i'll also have sd-cards swapped out as peridoic maintaince
Jan 17 21:20:15 <mrpackethead_> they are not that expensive in teh big scheme fo thigns
Jan 17 21:20:43 <mrpackethead_> not when yu have up to 15,000 people watching a show at a time
Jan 17 21:24:06 <XorA> ka6sox: yo!
Jan 17 21:24:07 <jkridner> koen: ugh. I got a kernel panic when I tried to boot the image, so I mounted it over USB and altered the symlink to ...bonelt-3.8. Now, I'm getting the damn 'CCCCC'. :(
Jan 17 21:24:44 <mrpackethead_> http://www.mini-box.com/picoUPS-120-12V-DC-micro-UPS-battery-backup?sc=8&category=980   <-- something like this might be good
Jan 17 21:24:55 <_av500_> mrpackethead_: show?
Jan 17 21:24:56 <mrpackethead_> this ones only good up to 18V though
Jan 17 21:24:59 <_av500_> can we wathc too=
Jan 17 21:25:01 <_av500_> too?
Jan 17 21:25:12 <mrpackethead_> yes, you can.. everybodys welcome
Jan 17 21:25:13 <mru> _av500_: you like watching?
Jan 17 21:25:18 <mrpackethead_> you might have already been to see it
Jan 17 21:26:45 <koen> jkridner: cccc would mean MLO is gone as well
Jan 17 21:27:13 <koen> jkridner: if it's the cec_write thing, remove /lib/firmware/*hdmi*.dtbo
Jan 17 21:27:36 <koen> jkridner: my board crashes as well, I need to try the A1 and get a serial cable
Jan 17 21:37:32 <jkridner> yeah, MLO blew up somehow, but I don't know how. I mounted the disk and can see it fine.
Jan 17 21:37:33 <NishanthMenon> koen, quick question - OE boots on any 3.8rcs private branch to UI?
Jan 17 21:37:39 <NishanthMenon> koen, on regular bone
Jan 17 21:37:43 <koen> OE doesn't boot
Jan 17 21:37:46 <koen> it's a buildsys
Jan 17 21:38:25 <koen> but the angstrom build I have boots to GUI with an LCD3 cape on a white bone with 3.8rc3
Jan 17 21:38:39 <NishanthMenon> neat
Jan 17 21:38:43 <NishanthMenon> koen, on your tree?
Jan 17 21:39:05 <koen> https://github.com/beagleboard/kernel/tree/3.8
Jan 17 21:39:11 <mdp> koen, my angstrom boots to fbcon on st7735 cape with a different 3.8-rc3 ;)
Jan 17 21:39:14 <NishanthMenon> koen, awesome
Jan 17 21:39:15 <NishanthMenon> thanks
Jan 17 21:39:51 <mdp> koen, I'm going to call fbcon a GUI today
Jan 17 21:40:04 <koen> fbcon is a way of life
Jan 17 21:40:39 <mdp> tragedy
Jan 17 21:40:45 <mdp> asoc is a huge DT gap
Jan 17 21:43:02 <mrpackethead_> bit of latteral thinking
Jan 17 21:43:05 <mrpackethead_> i've found my solution!
Jan 17 21:51:21 <_av500_> ?
Jan 17 21:51:29 <mrpackethead_> http://datasheet.octopart.com/DR-RDN20-Mean-Well-datasheet-10900782.pdf
Jan 17 21:51:36 <mrpackethead_> its easy.
Jan 17 21:51:43 <mrpackethead_> just an extra battery
Jan 17 21:51:44 <mrpackethead_> :-)
Jan 17 21:53:28 <_av500_> basically, this thing is 2 diodes
Jan 17 22:01:27 <mranostay> wb davest
Jan 17 22:02:22 <bradfa> mrpackethead_, the fall off at 70C is quite drastic :)
Jan 17 22:02:39 <mrpackethead_> yes.
Jan 17 22:03:01 <mrpackethead_> its a stupid graph
Jan 17 22:07:49 <mrpackethead_> av500: sometimes two diodes is all you need.
Jan 17 22:14:31 <ka6sox> morning XorA
Jan 17 22:16:09 <XorA> ka6sox: will be busy tonight :-(
Jan 17 22:16:31 <ka6sox> ya, same here
Jan 17 22:16:46 <XorA> ka6sox: did you get my email though?
Jan 17 22:16:58 <ka6sox> yes, came thru fine!
Jan 17 22:17:15 <XorA> ka6sox: sweet!
Jan 17 22:17:38 * XorA never quite trusts couple of years old records :-D
Jan 17 22:17:50 <ka6sox> thanks...
Jan 17 22:26:17 <_av500_> so, mail are coming through again?
Jan 17 22:26:20 <_av500_> with all the snow....
Jan 17 22:26:25 <_av500_> a marvell
Jan 17 22:26:28 <_av500_> -l
Jan 17 22:37:42 <wmat> snowing here too
Jan 17 22:37:58 <wmat> finally
Jan 17 22:38:13 <mru> not here, thankfully
Jan 17 22:38:48 <mranostay> mru: that would cut into rain time
Jan 17 22:40:05 <wmat> i hate seeing green grass in January in Canada, it's just not right
Jan 17 22:40:39 <mru> in england the grass is always green
Jan 17 22:40:45 <_av500_> mru: are you the IT professional from southhampton that visited my linkedin recently?
Jan 17 22:40:59 <wmat> I thought it was more a shade of brown, perhaps tan
Jan 17 22:41:06 <_av500_> mauve
Jan 17 22:41:17 <mrpackethead_> the grass should be brown here right nwo
Jan 17 22:41:20 <mrpackethead_> but its very green
Jan 17 22:41:26 <mrpackethead_> had a very wet summer
Jan 17 22:41:28 <mrpackethead_> and warm
Jan 17 22:41:47 <mru> wmat: most of it stays the same colour the whole year
Jan 17 22:41:53 <XorA> _av500_: there is snow, I see bright sunshine
Jan 17 22:42:22 <wmat> mru: island living eh
Jan 17 22:49:43 <mdp> wmat, we're a bit confused here with the green grass showing, it's dry, and a little sunny today
Jan 17 22:50:17 <mdp> normally we should see the sun until late march ;)
Jan 17 22:58:47 <mrpackethead_> sorry for all the questiosn today..
Jan 17 22:59:10 <mrpackethead_> has anyone seen or done anting with drycontacts and their begales?
Jan 17 23:00:32 <Russ> whoza whatzer fer?
Jan 17 23:01:28 <mranostay> relays?
Jan 17 23:01:47 <Russ> mranostay, I'm not sure what he's asking about, but this is a family channel
Jan 17 23:02:13 <Russ> so keep your begales to yourself
Jan 17 23:02:47 <mranostay> heh
Jan 17 23:02:53 <mrpackethead_> not sure what your on about at all
Jan 17 23:02:54 <mrpackethead_> :-)
Jan 17 23:06:06 <_av500_> mranostay: 3pm is awfully early for being drunk
Jan 17 23:06:26 <_av500_> dont you have patches to rebase?
Jan 17 23:06:42 <mranostay> cute
Jan 17 23:06:47 <mdp> omg
Jan 17 23:06:55 * mdp takes his son away from the screen
Jan 17 23:07:04 <mdp> I thought this was a safe environment
Jan 17 23:07:17 <Russ> I believe you are thinking of #rpi
Jan 17 23:07:23 <mdp> lol
Jan 17 23:07:39 <mdp> what the heck do you mean?
Jan 17 23:07:57 <Russ> try talking about anting your drycontacts in #rpi, and see what happens
Jan 17 23:08:00 <mdp> oops, another language violation on #rpi
Jan 17 23:08:11 <mdp> salty language, indeed
Jan 17 23:08:24 <Russ> TIL anting is actually a thing
Jan 17 23:09:06 * Russ tracks down a youtube video of a black drongo anting
Jan 17 23:09:37 <mranostay> Russ: weird isn't it?
Jan 17 23:09:38 * mdp covers his eyes
Jan 17 23:09:53 <Russ> hmm...this video has classical music, I'm not sure appropriate that is
Jan 17 23:10:15 <mranostay> 70's slow jazz better?
Jan 17 23:11:00 <Russ> you know you've found youtube gold when the comments include 'No, the "moustache" means it's a male.'
Jan 17 23:11:31 * mranostay takes notes
Jan 17 23:48:02 <alan_o> CareBear\: Best pull request yet: https://github.com/signal11/hidapi/pull/96
Jan 17 23:58:29 <aholler> hmm, where can I find that irq-storm commit?
Jan 18 00:01:13 <mrpackethead_> does anyone know if the schematics for the BeBoPr are online someone?
Jan 18 00:03:14 <Russ> it says 'schematics no'
Jan 18 00:03:38 <mrpackethead_> sorry, what says schematics no?
Jan 18 00:03:52 <Russ> http://circuitco.com/support/index.php?title=BeBoPr_Cape
Jan 18 00:04:20 <mrpackethead_> deos too.
Jan 18 00:04:21 <aholler> found it, wrong storm ;)
Jan 18 00:04:27 <mrpackethead_> did'nt see that
Jan 18 00:08:45 <mranostay> email flood to koen + panto
Jan 18 00:13:19 * mranostay finally setup gmail + git send-email
Jan 18 01:13:10 <toneeee> has anyone built the cloud9-image lately ? (is there a daily autobuild?)
Jan 18 01:15:25 <toneeee> oh I just saw angstrom-devel for today; I have similar issue on "| No package 'libudev' found"
Jan 18 01:16:04 <toneeee> I did not have the "preferred version 189 of systemd not available (for item udev/systemd)"
Jan 18 01:28:47 <NickTompkins> Hello al
Jan 18 01:28:48 <NickTompkins> l
Jan 18 01:29:25 <NickTompkins> I recently came across an idea to use the beagle bone as a part of a process at work
Jan 18 01:30:09 <NickTompkins> I have a question if beaglebone can be used as a MCU practically
Jan 18 01:30:25 <NickTompkins> The idea is to take a scan from a barcode scanner
Jan 18 01:30:39 <NickTompkins> send that data to a host that has a database
Jan 18 01:31:00 <NickTompkins> recieve data from a host and then send a hex signal to some pins
Jan 18 01:31:04 <NickTompkins> using Wifi
Jan 18 01:31:15 <NickTompkins> the host will just be a basic pc
Jan 18 01:31:43 <NickTompkins> and there might be like maybe 50 of these systems doing this
Jan 18 01:32:04 <NickTompkins> my question is if that can be done without a GUI for the system
Jan 18 01:32:14 <aholler> the beaglebone runs linux
Jan 18 01:33:06 <NickTompkins> ok.., so that means that it cannot be done? because it runs linux?
Jan 18 01:34:12 <NickTompkins> like I know it might be dumbing down alot of what the system is capable of but I would have no need for a GUI for this program
Jan 18 01:34:21 <toneeee> well seems pretty do-able /because/ it has linux.
Jan 18 01:34:37 <toneeee> (I'm going afk , back in a bit)
Jan 18 01:34:54 <NickTompkins> I would guess that there would need to be an app that will just setup on startup
Jan 18 01:35:07 <NickTompkins> so it could just function in that app enviroment
Jan 18 01:35:45 <NickTompkins> and by the way thank you for this room and to listen to this question
Jan 18 01:36:48 <NickTompkins> basically the idea would be this..... input from the USB is a Scanner the other usb is a WiFi jack......
Jan 18 01:37:01 <NickTompkins> takes logic in sends logic to the server
Jan 18 01:37:12 <XorA> NickTompkins: basically if you can do it with a linux desktop it can be done witht he bone
Jan 18 01:37:39 <aholler> no need for a desktop, a linux server can do that too ;)
Jan 18 01:37:42 <XorA> NickTompkins: as long as there isnt some speed critical stage that requires an uber CPU
Jan 18 01:37:43 <NickTompkins> the server will send the logic some bits that it will be sent to an existing MCU
Jan 18 01:38:03 <prpplague> NickTompkins: yea the bone can be used as a price checker
Jan 18 01:38:13 * mranostay zzzzzzzzzzz's
Jan 18 01:38:24 <NickTompkins> basically your right its a basic price checker
Jan 18 01:38:48 <NickTompkins> but its from um fragial of an application than that
Jan 18 01:39:02 <NickTompkins> more* of
Jan 18 01:39:08 <aholler> you will need to learn linux
Jan 18 01:39:15 <aholler> +about
Jan 18 01:39:37 <NickTompkins> last question and I will stop this spamming session
Jan 18 01:39:48 <NickTompkins> you can remote into the BB correct?
Jan 18 01:39:53 <XorA> yes
Jan 18 01:40:04 <NickTompkins> and now Im a very happy man
Jan 18 01:40:09 <XorA> you can do anything you can do with an x86 linux system
Jan 18 01:40:13 <NickTompkins> you guys rock
Jan 18 01:40:14 <NickTompkins> !
Jan 18 01:40:17 <XorA> so prototype on a PC then transfer
Jan 18 01:40:23 <aholler> learning linux is best done on a pc. you can just use a live-cd
Jan 18 01:40:39 <aholler> or usb-stick
Jan 18 01:40:42 <NickTompkins> I will make a partition
Jan 18 01:40:50 <NickTompkins> oh just boot form a stick lol
Jan 18 01:40:52 <NickTompkins> ok
Jan 18 01:41:00 <prpplague> or just buy a price checker, hehe
Jan 18 01:41:18 <XorA> you price today sir is Purple$
Jan 18 01:41:32 <NickTompkins> this is kinda like a large manufacuring enviroment
Jan 18 01:41:50 <prpplague> NickTompkins: hehe doesn't make much difference
Jan 18 01:42:02 <NickTompkins> and im doing an internship as a Lean Sigma Engineer and im not.... Lean or Sigma ....
Jan 18 01:42:07 <XorA> I hope this isnt going to make the poor box counter dude redundant
Jan 18 01:42:09 <NickTompkins> but im an EE student
Jan 18 01:42:46 <XorA> what is a Small Sum Engineer?
Jan 18 01:42:50 <NickTompkins> so I found out that they where looking for systems that might take up too 10,000 for this basic design
Jan 18 01:42:57 <prpplague> NickTompkins: have fun, hehe
Jan 18 01:43:00 <XorA> not a term I ever heard
Jan 18 01:43:10 <NickTompkins> lol
Jan 18 01:43:13 <prpplague> 10000 what?
Jan 18 01:43:24 <mrpackethead_> bitcoins
Jan 18 01:43:29 <NickTompkins> its 10,000 for that same thing I just asked about
Jan 18 01:43:29 <prpplague> hehe
Jan 18 01:43:36 <NickTompkins> for design
Jan 18 01:43:41 <NickTompkins> just design
Jan 18 01:43:43 <prpplague> NickTompkins: they are looking in the wrong spot then
Jan 18 01:43:48 <NickTompkins> yes
Jan 18 01:43:55 <prpplague> NickTompkins: http://www.amltd.com/product.asp?pid=kdt750
Jan 18 01:44:18 <mrpackethead_> who's an EE and needs a job?
Jan 18 01:44:23 <prpplague> hehe
Jan 18 01:45:12 <mrpackethead_> im deisgning a new supercape
Jan 18 01:45:23 <mrpackethead_> and i and debating if i should but on some user leds
Jan 18 01:45:28 <NickTompkins> almost would work but they need a hand held scanner
Jan 18 01:45:31 <XorA> oh right, its QA
Jan 18 01:46:02 <prpplague> NickTompkins: http://www.amltd.com/product.asp?pid=m7225
Jan 18 01:46:14 <mrpackethead_> i suppose i better
Jan 18 01:46:15 <mrpackethead_> :-)
Jan 18 01:46:20 <mrpackethead_> since you won't see the beagle
Jan 18 01:47:02 <prpplague> NickTompkins: or even cheaper model: http://www.amltd.com/product.asp?pid=m7220
Jan 18 01:47:17 <mrpackethead_> theres another ever so slightly annoying beagle feature
Jan 18 01:47:35 <mrpackethead_> theres no easy access to the ethernet indicator LED's
Jan 18 01:47:41 <mrpackethead_> so you can put them somwhere you can see them
Jan 18 01:47:58 <aholler> just use iptables and a trigger
Jan 18 01:48:16 <NickTompkins> but is that cheaper than a Beagle Bone a usb scanner and a Wifi adapter
Jan 18 01:48:43 <aholler> mrpackethead_:  simple rule and you can use any led you want to visualize traffic
Jan 18 01:48:46 <NickTompkins> plus it needs to send out packets to pins for hex logic to light up LEDs on a MCU
Jan 18 01:49:23 <mrpackethead_> aholler.. that is an interesting concept
Jan 18 01:49:34 <mrpackethead_> very latteral thinking..
Jan 18 01:49:41 <mrpackethead_> +2 points.
Jan 18 01:50:16 <mrpackethead_> though doesn't help me with connectivity
Jan 18 01:50:21 <mrpackethead_> just tells me if theres activity
Jan 18 01:50:26 <NickTompkins> I dont think those scanner solutions have the flexiblity they need to control the 54 leds they need to light up
Jan 18 01:50:42 <prpplague> NickTompkins: it is if you factor in getting everything working, plus battery support, and having all of the software already there for you...
Jan 18 01:50:51 <aholler> mrpackethead_: define connectivity
Jan 18 01:51:09 <prpplague> NickTompkins: hehe they do
Jan 18 01:51:10 <mrpackethead_> ethernet cable plugged in, and also connected to a switch..
Jan 18 01:51:36 <prpplague> what is the purpose of the 54 leds?
Jan 18 01:51:46 <NickTompkins> they can send data to pins for another mcu?
Jan 18 01:51:55 <mrpackethead_> 54 leds. must be a christmas tree.
Jan 18 01:51:56 <mrpackethead_> :-)
Jan 18 01:52:03 <NickTompkins> too another *
Jan 18 01:52:15 <prpplague> NickTompkins: uh what?
Jan 18 01:52:27 <NickTompkins> its a manufacturing quality tool actually
Jan 18 01:52:43 <aholler> mrpackethead_: http://dpaste.de/5iicy/
Jan 18 01:53:03 <NickTompkins> so its like say you built a deathstar out of legos
Jan 18 01:53:30 <NickTompkins> and a part of the completion is flaging certion sections
Jan 18 01:53:53 <NickTompkins> but your workers are not putting flags on or in the right places
Jan 18 01:54:17 <NickTompkins> and this can result into the buyer rejecting the project everytime
Jan 18 01:54:31 <mrpackethead_> aholler: thats traffic.
Jan 18 01:54:47 <aholler> mrpackethead_: yes, the hard part
Jan 18 01:54:47 <NickTompkins> so you make LEDS to where the flags need to go on the deathstar lego table
Jan 18 01:55:12 <NickTompkins> and you will now have barcodes on your falgs
Jan 18 01:55:30 <NickTompkins> so you can force them to place every flag in the right position
Jan 18 01:55:30 <prpplague> NickTompkins: hehe you you need the KDT750 with no internal barcode scanner, a handheld usb scanner, and the i/o expansion module
Jan 18 01:55:40 <toneeee> is the buyer The Empire ?
Jan 18 01:55:40 <NickTompkins> or it cannot move down the line
Jan 18 01:55:43 <prpplague> NickTompkins: basically the same as what you are asking for less than $400USD
Jan 18 01:55:48 <mrpackethead_> aholler: it is still a clver idea.
Jan 18 01:56:18 <aholler> thanks
Jan 18 01:56:32 <mrpackethead_> and with a little bit of extension to the idea
Jan 18 01:56:43 <mrpackethead_> if we had a periodic ping going.
Jan 18 01:57:07 <aholler> I'll forward it to the author of the led-trigger-iptables-module ;)
Jan 18 01:57:08 <mrpackethead_> you probably woudl get the desired result.
Jan 18 01:57:48 <aholler> there are enough ifup/ifdown triggers available
Jan 18 01:58:05 <NickTompkins>  hehe you you need the KDT750 with no internal barcode scanner, a handheld usb scanner, and the i/o expansion module or could this be done just as well with a beagle bone though
Jan 18 01:58:59 <NickTompkins> the idea that it might be possible to do a system for under $150 per unit
Jan 18 01:59:06 <prpplague> NickTompkins: oh sure if you know how to do it..... hehe just remember that it probably will end  up costing a lot more than what you think....
Jan 18 01:59:42 <NickTompkins> might make it where there can be software engineers setup to make it happen for far less
Jan 18 02:00:07 <prpplague> hehe, good luck
Jan 18 02:00:08 <prpplague> hehe
Jan 18 02:00:34 <NickTompkins> lol I will bring your idea to the next meeting
Jan 18 02:00:35 <NickTompkins> lol
Jan 18 02:00:40 <aholler> per unit price doesn't say much without the number of units
Jan 18 02:01:08 <NickTompkins> 35-250
Jan 18 02:01:14 <NickTompkins> depends on sucess
Jan 18 02:01:22 <prpplague> nor does it factor in all the time involved or the ability to replace a unit when it fails
Jan 18 02:01:37 <aholler> and qa
Jan 18 02:01:58 <NickTompkins> your correct in that having a company that will just do an RMA is very nice
Jan 18 02:02:01 <XorA> heh, only QA would fail to QA their QA devices :-D
Jan 18 02:02:18 <aholler> don't forget to calculate what happens when the system fails.
Jan 18 02:02:18 <NickTompkins> and they would
Jan 18 02:02:39 <NickTompkins> a built solution would be better
Jan 18 02:02:56 <NickTompkins> yes
Jan 18 02:03:01 <aholler> 10.000 isn't that much to design an idiot proof solution
Jan 18 02:03:46 <NickTompkins> well it seemed like alot for just the scanner and the server lookup to send a hex packet to an MCU
Jan 18 02:03:48 <aholler> s/./,/
Jan 18 02:03:58 <aholler> it isn't
Jan 18 02:04:18 <NickTompkins> yea I will be less critical
Jan 18 02:04:40 <NickTompkins> perhaps this is just life in the real engineering world
Jan 18 02:04:49 <NickTompkins> im just not seeing it
Jan 18 02:04:50 <NickTompkins> lol
Jan 18 02:05:00 <aholler> it's the ee-problem, they only see hw
Jan 18 02:05:09 <toneeee> Yeah my brother-in-law ran a company that made those scanners, Nick, it's not cheap to handle the whole life cycle of the units.
Jan 18 02:05:11 <prpplague> hehe indeed
Jan 18 02:05:58 <toneeee> note past-tense "ran." ;-)
Jan 18 02:06:02 <NickTompkins> yea this might be a better intership then I expected and I might get alot of lessons out of it that I wont make in a career
Jan 18 02:06:20 <mranostay> NickTompkins: it only gets worse from here :)
Jan 18 02:06:24 <NickTompkins> HA
Jan 18 02:07:05 <NickTompkins> to me though it seemed like it would almost be cheaper just to put a freaking PC at the station
Jan 18 02:07:44 <aholler> doesn't make a differece
Jan 18 02:14:10 <mrpackethead_> anyone got a preference for method when setting up a sinple LED indicator off the beagle
Jan 18 02:14:31 <mrpackethead_> die'snt seem samrt to try to directly drive them
Jan 18 02:14:39 <mrpackethead_> small transistor?
Jan 18 02:15:04 <mranostay> why do you need a transistor?
Jan 18 02:15:25 <mrpackethead_> green led
Jan 18 02:15:36 <mranostay> so? what is the voltage drop on it?
Jan 18 02:15:45 <mrpackethead_> 3.0v nomimal
Jan 18 02:15:55 <mrpackethead_> could be higher.
Jan 18 02:16:01 <mrpackethead_> ouput is just 3.3V
Jan 18 02:16:51 <mrpackethead_> data sheet time.
Jan 18 02:18:31 <mrpackethead_> did i tell you, that you are all mighty fine looking trolls
Jan 18 02:19:35 <prpplague> 3 nomimal is pretty high for a generic led
Jan 18 02:20:36 <emeb_mac> generally 2.8v for a blue LED, more like 1.5 for a red one.
Jan 18 02:21:17 <prpplague> mranostay: http://www.jamesglanville.com/wordpress/wp-content/uploads/2013/01/2013-01-09-18.58.19.jpg
Jan 18 02:21:47 <emeb_mac> nice!
Jan 18 02:21:49 <mranostay> very modern art
Jan 18 02:21:53 <mrpackethead_> prpplague: yeah, i needed to look at the datasheet
Jan 18 02:22:05 <emeb_mac> love those circuits that include the HV boost.
Jan 18 02:22:38 <mrpackethead_> that is truley scary!
Jan 18 02:22:50 <prpplague> mrpackethead_: you generall can find a nice led that can operate around 10mA, using the correct current limiting resistor, it is safe to drive the LED directly from the GPIO
Jan 18 02:23:21 <mrpackethead_> 5mA is mpre than enough gnerally
Jan 18 02:23:34 <mranostay> there you go
Jan 18 02:23:54 <mrpackethead_> theres a lot of "mis-information" out there
Jan 18 02:23:57 <prpplague> mranostay: are you saying that the LED is in 5mA draw or that you expect to find one that is 5mA
Jan 18 02:24:34 <mrpackethead_> i'm saying that i'll cetianly find a LED thats sutible and running it at 5mA will be enough for indicaiton
Jan 18 02:25:11 <prpplague> yea 5mA is harder to find, 10mA is a good choice for both cost and availibility
Jan 18 02:25:28 <mrpackethead_> i'll probably find a led that is rated at 20mA
Jan 18 02:25:37 <mrpackethead_> but i'll only drive it to 5mA
Jan 18 02:25:57 <mrpackethead_> im just tyring to find the max currnet per port in the datasheets.
Jan 18 02:26:36 <mrpackethead_> reducing part count is very helpful!
Jan 18 02:26:59 <prpplague> mrpackethead_: generally speaking you can count on most ARM SoC's to have about 20mA per gpio
Jan 18 02:27:17 <prpplague> with some SoCs having configurable drive strengths
Jan 18 02:28:30 <mranostay> prpplague: i assume that was at mrpackethead_ :)
Jan 18 02:28:52 <prpplague> mranostay: yes indeed, tab completion failure
Jan 18 02:29:09 <mranostay> prpplague: the gpio outputs are 20mA right?
Jan 18 02:29:53 <prpplague> mranostay: i'd have to double check the datasheet for the am335x stuff, but in general you can use a rule of 20mA sink current
Jan 18 02:29:57 <prpplague> mranostay: max
Jan 18 02:30:48 <mranostay> yeah  i meant max
Jan 18 02:30:55 <mrpackethead_> its not burnig down at 10mA!
Jan 18 02:31:09 <prpplague> the omap4 has mostly 20mA gpios , but it does have about 15 or 16 that are maxed at 10mA
Jan 18 02:31:15 <aholler> it burns up
Jan 18 02:32:46 <prpplague> hmm
Jan 18 02:33:02 <mrpackethead_> on teh beaglebone itself, the Four USR Leds are buffered
Jan 18 02:33:15 <prpplague> datasheet seems to indicate 6mA for most signals and 4mA for a bunch of the high speed lines
Jan 18 02:33:24 <toneeee> I saw 18mA and lots of 10mA and some 5mA pins
Jan 18 02:33:35 <toneeee> (3358)
Jan 18 02:33:40 <toneeee> sorry 3358
Jan 18 02:34:26 <prpplague> yea i am looking at 3352 atm, grabbing 3358 now
Jan 18 02:34:45 <mrpackethead_> 3359
Jan 18 02:35:01 <mrpackethead_> http://www.ti.com/lit/ug/spruh73g/spruh73g.pdf
Jan 18 02:35:21 <prpplague> mrpackethead_: just stepping through several of the model line to get an idea of the differences
Jan 18 02:36:52 <mrpackethead_> which documentat are you reading
Jan 18 02:37:16 <prpplague> http://www.ti.com/lit/ds/symlink/am3359.pdf
Jan 18 02:38:13 <mrpackethead_> m assuming Circuitco spent the money on the transisitors for a rason
Jan 18 02:40:38 <prpplague> mranostay: yea usually for safety, but yea it appears that from the datasheet it looks as if it is about 6mA for an average gpio
Jan 18 02:41:02 <prpplague> i am surprised at the low value
Jan 18 02:41:28 <mrpackethead_> which section did you find the rating for the AM3359
Jan 18 02:44:44 <prpplague> table 2-7
Jan 18 02:44:50 <prpplague> buffer strength
Jan 18 02:45:17 <mrpackethead_> ahh
Jan 18 02:45:22 <mrpackethead_> i was looking in teh wrong place
Jan 18 02:46:05 <toneeee> ahhh me too.
Jan 18 02:46:13 <toneeee> gj sir prplague!
Jan 18 02:46:28 <mdp> a faq
Jan 18 02:46:51 * mdp updates prpplague's tickets on the big board
Jan 18 02:47:22 <prpplague> jeeze
Jan 18 02:47:36 <mrpackethead_> a 0.03 led
Jan 18 02:47:41 <mrpackethead_> and i've spent an hour thinking about it
Jan 18 02:48:13 <mrpackethead_> 6mA it is
Jan 18 02:48:29 <mrpackethead_> though i'm driving one at 9.5mA right now
Jan 18 02:48:33 <mrpackethead_> and its doing ok.
Jan 18 02:48:36 <mrpackethead_> but
Jan 18 02:49:12 <mrpackethead_> so, mranostay in reply to your question.. why do you need a transistor
Jan 18 02:49:24 <mrpackethead_> beucase 6mA is'nt that much.
Jan 18 02:49:56 <prpplague> mrpackethead_: other options you could always use an external buffer or some other shifting device
Jan 18 02:50:04 <mrpackethead_> yes
Jan 18 02:50:11 <prpplague> mrpackethead_: all depends on how many leds and what you want to use them for
Jan 18 02:50:23 <mrpackethead_> lowering the part count is always good.
Jan 18 02:50:35 <aholler> use a cpld
Jan 18 02:50:40 <prpplague> aholler: hehe
Jan 18 02:51:07 <mrpackethead_> the clc in some of the micrchip 8bit PIC's is very handy
Jan 18 02:51:09 <aholler> if just the sw wouldn't be that heavy and ugly to install
Jan 18 02:51:12 <mrpackethead_> very low cost
Jan 18 02:51:13 <prpplague> mrpackethead_: how many leds are you going to have?
Jan 18 02:51:18 <mrpackethead_> 4
Jan 18 02:51:59 <mrpackethead_> User0, 1
Jan 18 02:52:20 <mrpackethead_> a gpio to blink when ethernet's passing packets ( very nice trick )
Jan 18 02:52:29 <mrpackethead_> and another one, for a dont' know what
Jan 18 02:52:34 <mrpackethead_> but it will be handy to have it there
Jan 18 02:53:26 <prpplague> http://www.digikey.com/product-detail/en/DMN2005DLP4K-7/DMN2005DLP4KDICT-ND/1774234
Jan 18 02:54:05 <mrpackethead_> yes a small fet
Jan 18 02:54:19 <prpplague> mrpackethead_: those are a good choice
Jan 18 02:54:43 <prpplague> mrpackethead_: small package size and dual
Jan 18 02:55:30 <mrpackethead_> yes..
Jan 18 02:55:41 <mrpackethead_> using a fet liek that saves me a resitor
Jan 18 02:55:50 <mrpackethead_> over using an NPN transistor
Jan 18 02:56:26 <prpplague> you;d still want to put a resistor in there...
Jan 18 02:56:35 <mrpackethead_> yes,
Jan 18 02:56:47 <mrpackethead_> you still need 1 resistor
Jan 18 02:56:48 <mrpackethead_> not 2
Jan 18 02:56:56 <prpplague> yea
Jan 18 02:57:37 <prpplague> plus the dual package saves you about $0.15 on assembly
Jan 18 02:57:45 <mrpackethead_> yes
Jan 18 02:57:53 <mrpackethead_> every part is abotu 0.10-0.15
Jan 18 02:59:33 * prpplague is very familiar with the costs of production
**** ENDING LOGGING AT Fri Jan 18 02:59:59 2013