• [00:01:58] * maelcum (n=quassel@e178181223.adsl.alicedsl.de) Quit (Read error: 110 (Connection timed out))
  • [00:04:14] * maelcum (n=quassel@e178146146.adsl.alicedsl.de) has joined #beagle
  • [00:05:45] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [00:06:12] * dcordes (n=dcordes_@unaffiliated/dcordes) Quit ("Verlassend")
  • [00:15:29] * felipec (n=felipec@a91-153-251-222.elisa-laajakaista.fi) Quit ("Leaving")
  • [00:24:03] * chelli (n=chelli@debian/developer/tschmidt) Quit ("IRC is just multiplayer notepad")
  • [00:25:56] * aleij (n=ad@adsl-84-227-116-11.adslplus.ch) Quit ()
  • [00:30:45] * gbot (i=558a4728@gateway/web/ajax/mibbit.com/x-0b3e5460d0dfed0c) has joined #beagle
  • [00:31:07] <gbot> hi
  • [00:31:38] * gbot is now known as gbot__
  • [00:33:53] * gbot__ (i=558a4728@gateway/web/ajax/mibbit.com/x-0b3e5460d0dfed0c) Quit (Client Quit)
  • [00:57:56] * RobotGuy (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) has joined #Beagle
  • [00:57:56] * zedstar (n=john@fsf/member/zedstar) Quit (Remote closed the connection)
  • [01:13:57] * RobotGuy is now known as ke7yut
  • [01:22:16] * feig1 (n=ejf3@69.205.96.133) Quit (Remote closed the connection)
  • [01:35:05] * feig (n=ejf3@69.205.96.133) has joined #beagle
  • [01:42:58] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) Quit (Read error: 54 (Connection reset by peer))
  • [01:48:07] * fulgas is now known as FuL|OUT
  • [02:05:13] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has left #beagle
  • [02:09:07] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [02:12:10] * ScriptRipper (n=martin@opensuse/member/MartinMohring) Quit (Remote closed the connection)
  • [02:13:45] * blade4 (n=chatzill@modemcable007.76-56-74.mc.videotron.ca) has joined #beagle
  • [02:19:08] * hli (i=chaton@vig91-2-82-232-97-149.fbx.proxad.net) Quit ("ChatZilla 0.9.84 [Firefox 3.0.5/2008120122]")
  • [02:28:51] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [02:31:54] * blade4 (n=chatzill@modemcable007.76-56-74.mc.videotron.ca) Quit (Remote closed the connection)
  • [02:54:31] * ke7yut (n=robotguy@pool-96-225-210-142.ptldor.fios.verizon.net) Quit (Remote closed the connection)
  • [03:03:04] * raster (n=raster@r125-63-180-233.cpe.unwired.net.au) has joined #beagle
  • [03:06:00] <jkridner> I'm editing BeagleBot to talk to us on occasions, such as when the website is updated.
  • [03:07:31] <jkridner> People will have to tell me if it is too much noise.
  • [03:13:50] <raster> /ignore is a useful thing :)
  • [03:14:00] <raster> hey - i had a q...
  • [03:14:44] <raster> who should i poke (talk to) about doing beagleboard "offshoots" - ie somethnig with extra components on it and more trimemd down board?
  • [03:15:02] <ds2> do you have the $$ to even consider it?
  • [03:15:03] <raster> i'm not a hardware guy - beyond the "give me the registers and tell me what they do"
  • [03:15:03] <raster> :)
  • [03:15:11] <raster> depends on th $$ needed
  • [03:15:14] <raster> i do have $$
  • [03:15:30] <raster> and yes - this is a commercial venture - and i have already interested buyers.
  • [03:15:37] <raster> not massive volume
  • [03:15:53] <raster> but lets say the 20k units range so far - and i havent really tried "selling" it yet
  • [03:15:57] <ds2> I can help..... i do software but my formal training is in hardware
  • [03:16:25] <raster> ok first q
  • [03:16:27] <Pavlov> o.O a raster
  • [03:16:32] <ds2> your biggest issue there is 1. funding for assembler (Micro BGAs) and 2. Funding to get the PCB made and possibly having to deal with multiple revs
  • [03:16:57] <raster> whats the ballpark $ needed to lets say.. make a miuch leaner bb with some more components + lcd (of course bom cost based on components matters) but just cost in development?
  • [03:17:07] <raster> talking 20k? 50k? 100k?
  • [03:17:20] <ds2> 20K is probally doable
  • [03:17:43] <raster> so basically to do a first rev of a board and make a few iterations
  • [03:17:47] <ds2> I been racking my head to find a way to make new versions for sub $1K in small quantities
  • [03:17:54] <raster> like make 10 or 20 (making 1 or 20 is probably the same cost really)
  • [03:18:39] <ds2> yeah... 20K shouldbe doable...50K shouldbe more then sufficient unless you are doing something major like adding an ATSC receiver or WiFi chips on the same board
  • [03:19:01] <ds2> what size are you targeting?
  • [03:20:32] <raster> hmm
  • [03:20:34] <raster> ok
  • [03:20:38] <raster> i'm looking at major then
  • [03:21:07] <raster> wifi, bt, 3g, gps,
  • [03:21:16] <raster> and you can then gues the size
  • [03:21:18] <raster> :)
  • [03:21:54] <raster> 3-4" diagonal... so like 4x2.5" or so
  • [03:22:08] <raster> and yeah - kick me - n00b :)
  • [03:22:20] <raster> ie "you have got to be kidding if u can do that"
  • [03:22:20] <raster> :)
  • [03:23:03] <raster> and yes i know 3g modules are pricey
  • [03:23:11] <ds2> WiFi/BT/3G/GPS can be easy if you use module or very painful if you use the raw chip
  • [03:23:11] <raster> the rest is "Reasonable"
  • [03:23:17] <ds2> so you need to clarify that
  • [03:23:37] <raster> i'd go module
  • [03:23:49] <raster> this is a first go
  • [03:23:51] <ds2> I personally wouldn't touch using raw chips as I am not a sufficiently high level analog wizard
  • [03:23:57] <raster> i know thnigs can be refined later
  • [03:24:04] <raster> i wouldnt either
  • [03:24:05] <ds2> Oh that's typically not an issue... and your size is not too painfully small
  • [03:24:10] <raster> 3g raw is pretty much impossible anyway
  • [03:24:20] <raster> basically phone size
  • [03:24:22] <raster> smartphone
  • [03:24:25] <ds2> thought you wanted all that AND something small like 1"x2" ;)
  • [03:24:37] <raster> so it needs to fit behind the lcd
  • [03:24:53] <raster> so its going to have to be compact
  • [03:24:55] <ds2> I'd probally through in an extra 10K or so for regulatory certs
  • [03:25:03] <raster> but you miss connectors - so thats good
  • [03:25:14] <raster> an skip to microsd for media (just 1)
  • [03:25:21] <raster> thts not a problem
  • [03:25:25] <ds2> I am more worried about getting regulatory compliance issues then getting it to work
  • [03:25:28] <raster> basically i have a budget
  • [03:25:47] <ds2> is this intended as a raw board or a final product?
  • [03:26:06] <raster> final product :)
  • [03:26:11] <raster> but you start with a board
  • [03:26:19] <raster> amd "make it work"
  • [03:26:27] <raster> as i said
  • [03:26:33] <raster> already have buyers interested
  • [03:26:50] <raster> so i know the initial unit numbers already just as a starter
  • [03:26:53] <ds2> Ouch... regulatory issues is _MY_ main worry
  • [03:27:03] <raster> and this mostly hinges on software being usable - which is my domain
  • [03:27:19] <raster> the hw i think is just a matter of kicking it off in parallel
  • [03:27:24] <raster> and i'm going to find this out of my own pocket
  • [03:27:31] <raster> so - thus i care about costs
  • [03:27:32] <raster> :(
  • [03:27:33] <ds2> yes, you could do that
  • [03:27:36] <raster> no vc money here
  • [03:27:51] <raster> but i do know that once i can get the flow going - it will pay for itself
  • [03:27:57] <raster> so i'm kick-starting it
  • [03:27:59] <ds2> build a mock up using the beagle + external modules and get coding to work while at the same time building a board to the right dimensions
  • [03:28:10] <raster> yup
  • [03:28:13] <raster> thats kind of the plan
  • [03:28:17] <raster> as such after having done samsung
  • [03:28:19] <raster> pxa
  • [03:28:23] <raster> and omap
  • [03:28:28] <ds2> *nod*
  • [03:28:32] <raster> omap by far and wide has the best powermangement suport
  • [03:28:38] <raster> and best overall support from TI
  • [03:28:46] <raster> for a long time the imagination blob has kept me scpetical
  • [03:28:50] <raster> but thats improving
  • [03:28:52] <raster> and frankly
  • [03:29:04] <ds2> esp. comparing it to the current PXA which for all intents and purposes offers NO support for low volume users
  • [03:29:11] <raster> omap3530 + dsp + cortox + the simd asm extensions should voer 2d gfx just fine
  • [03:29:21] <raster> and i just will ignore the imagination as if it doesnt exist
  • [03:29:22] <raster> :)
  • [03:29:55] <ds2> yep
  • [03:30:01] <raster> after seeing the 3530 actually running in several situations - i'm thoroughly impressed
  • [03:30:13] <raster> and as best i can see.. haven't even started touchign the 3d bit
  • [03:30:26] <ds2> all doable... if you are not a hw person, good luck talking the board fab'ed
  • [03:30:28] <raster> anyway - yes. attach peripherals on existing busses
  • [03:30:33] <raster> eg serial/usb
  • [03:30:39] <raster> code it all up to work
  • [03:30:49] <raster> and "assume" the board weill come through with ti all integrated and shrunk down
  • [03:31:01] <raster> thats my other problem
  • [03:31:09] <raster> gettign it not just desuigned - but fabbed
  • [03:31:21] <raster> ie even prorotypes ned fabbing
  • [03:31:35] <raster> tho low-volume can be done by bespoke board production services for high cost
  • [03:31:41] <raster> but once u talk a few k boards
  • [03:31:46] <raster> then u need real fabs
  • [03:31:54] <raster> and one that can deal with the 3530
  • [03:32:13] <raster> i still need to source battery and get a case design
  • [03:32:24] <raster> so lots of things to "fund"
  • [03:33:10] <raster> so if you're interested...
  • [03:33:37] <Vegar> is this the raster of E?
  • [03:33:48] <raster> could be
  • [03:33:49] <raster> :)
  • [03:34:48] <raster> but i ate him
  • [03:34:53] <raster> so maybe i'm him by proxy
  • [03:34:54] <raster> :)
  • [03:35:13] <Vegar> ok
  • [03:42:00] * ClaudeQC (n=claude@bas1-quebec03-1279637128.dsl.bell.ca) has left #beagle
  • [03:58:36] * JuanG (n=Juan@nat/ti/x-fabd50bb16bfe380) has joined #beagle
  • [04:03:27] * JuanG (n=Juan@nat/ti/x-fabd50bb16bfe380) has left #beagle
  • [04:19:37] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit ("Ex-Chat")
  • [04:27:45] * jsync (n=jess@59.160.172.220) has joined #beagle
  • [04:35:24] <sakoman> raster: you might also consider prototyping with Overo. there will be a 4" LCD/touchscreen available "Real Soon Now"
  • [04:36:00] <sakoman> also a version of the motherboard with BT/wifi/microSD
  • [04:36:25] <raster> overo?
  • [04:36:29] * raster googles
  • [04:37:11] <sakoman> http://gumstix.net/overo
  • [04:37:49] <raster> oooh gumstix!
  • [04:38:07] <raster> i like gumstix
  • [04:38:08] <sakoman> yes, the omap3 version
  • [04:38:20] <raster> didnt know they were doing one
  • [04:38:53] <sakoman> I do the sw for them
  • [04:39:27] <raster> aaah aexellent
  • [04:39:30] <sakoman> struggling a bit with E this evening, but you wouldn't know anything about that I suppose ;-)
  • [04:39:54] <raster> nothing
  • [04:39:57] <raster> i'm a n00b with e
  • [04:40:13] <raster> whats the problem?
  • [04:40:14] <raster> :)
  • [04:40:43] <raster> wow thats a small mobo
  • [04:40:44] <sakoman> in file manager icons don't get drawn
  • [04:40:48] <raster> in the gumstix tradition
  • [04:41:07] <raster> hmmm
  • [04:41:11] <sakoman> if I right click and select any view type they show up
  • [04:41:24] <raster> what os?
  • [04:41:31] <sakoman> for the life of the window, but new windows have the same issue
  • [04:41:42] <sakoman> linux 2.6.28-rc9
  • [04:41:50] <sakoman> on overo :-)
  • [04:42:07] <raster> openemebedded based?
  • [04:42:13] <sakoman> yes
  • [04:42:33] <raster> hmm
  • [04:42:41] <raster> what svnrev of e?
  • [04:42:52] <raster> others have bene complainging of missing icosn - i never saw it
  • [04:42:59] <raster> i just built
  • [04:42:59] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) Quit ()
  • [04:43:00] <raster> EFL_SRCREV ?= "38274"
  • [04:43:09] <raster> (in sane-srcrevs.inc)
  • [04:43:12] <raster> works fine
  • [04:44:08] <sakoman> mine seems to be building 37988
  • [04:44:52] <raster> old
  • [04:44:59] <raster> but then in e land something a few weeks old.. is old
  • [04:44:59] <raster> :)
  • [04:45:02] <sakoman> I'll specify that SRCREV
  • [04:45:08] <sakoman> thanks
  • [04:45:21] <sakoman> let me know if you have any overo questions
  • [04:45:24] <raster> try that
  • [04:45:40] <raster> i assuem your oe setup is "working" so it'll take amyba e20 mins to spin a new fs image
  • [04:45:40] <sakoman> if might help you keep development costs down
  • [04:45:45] <raster> sure
  • [04:45:48] <raster> i might get one too
  • [04:45:57] <raster> i just didnt know gumstix had one coming
  • [04:46:00] <sakoman> correct, I build many times each day :-)
  • [04:46:03] <raster> most gumstix stuff has been pxa
  • [04:46:17] <sakoman> It has been shipping for a couple of months now
  • [04:46:25] <jkridner> raster: looks like you are getting lots of good help from ds2 and sakoman.
  • [04:46:57] <sakoman> I'm working on the sw for some of the follow on products on the roadmap - i.e. wifi/bt/lcd/ts
  • [04:47:11] <jkridner> I'm trying to create a page that summarizes some of the Beagle off-shoots (http://beagleboard.org/breeds), but really do need to do more to summarize what services there are to make your own hardware as well.
  • [04:47:14] <raster> jkridner: yup.
  • [04:47:43] <jkridner> you can get the beagle design materials from http://beagleboard.org/hardware/design, but for just 20k units, I think something like the Overo makes a lot of sense.
  • [04:47:48] <raster> sakoman: aaah so still "in progress"
  • [04:48:12] <sakoman> well, hw in hand and sw is working
  • [04:48:25] <jkridner> working with someone like ds2 for the kind of expansion board might also be good.
  • [04:48:26] <raster> jkridner: chances are it will go up from there
  • [04:48:31] <sakoman> so it is a matter of it making it through the production ramp
  • [04:48:36] <raster> i guesstimate theres a market for 50k in a year
  • [04:48:39] <jkridner> both have been great beagleboard.org community participants.
  • [04:48:44] <raster> but i have almsot definite sales of 20k right now
  • [04:49:02] <raster> cool
  • [04:49:13] <raster> my aim for now is somethng that works
  • [04:49:42] <raster> the problem with the gumstix thing is possibly that it will have connectors et.c that are not wanted/needed
  • [04:49:52] <raster> eg DVI
  • [04:50:09] <raster> and in a phone formfactor you want to nuke any connectors u dont need
  • [04:50:19] <sakoman> the motherboard doesn't have any extra connectors
  • [04:50:36] <raster> so i already "assumed" there would need to be a "custom board" anyway - but it can start its life off as an offshoot
  • [04:50:43] <sakoman> the hdmi, etc are on the daughtercard - that would be custom in your case
  • [04:50:57] <raster> oooh thats what i was looking at
  • [04:50:58] <raster> :)
  • [04:51:07] <raster> hmm
  • [04:51:11] <sakoman> all the hard stuff is done on the motherboard
  • [04:51:18] <raster> whats the overo go for in wuantity?
  • [04:51:24] <raster> no detaisl - ballpark
  • [04:51:27] <raster> details
  • [04:51:29] <sakoman> you can do a simple 2/4 layer daughtercard
  • [04:51:42] <raster> i am sure gumstix do the "well negotiate with us - if u have large orders of a few k units or more"
  • [04:51:55] <sakoman> I'm not a gumstix employee, just a sw contractor
  • [04:51:59] <raster> yup
  • [04:52:04] <sakoman> but one moment, I think they have an oem page
  • [04:52:22] <raster> my webbrowser si hyper-slow
  • [04:52:27] <raster> i'm stuc on 32kbit
  • [04:52:35] <raster> thansk to my isp being slow-as-molasses
  • [04:52:41] <raster> in reconnecting me after my move
  • [04:53:11] <raster> a daughterboard makes sense
  • [04:53:21] <raster> depends on the interconnect
  • [04:53:26] <raster> oi assume the spi/sdio etc. is all broken out there?
  • [04:53:31] <raster> i2c
  • [04:53:33] <raster> blah blah...
  • [04:53:45] <raster> hmm
  • [04:53:48] <sakoman> yes, all that stuff is on the 70 pin expansion connectors
  • [04:53:56] <sakoman> 1.8v logic
  • [04:54:07] <sakoman> you can do all the things you mention quite easily
  • [04:54:22] <raster> whats the 3530 x 3503...
  • [04:54:30] <raster> cool
  • [04:54:43] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [04:54:47] <sakoman> there will be both options, 3503 doesn't have 3d/dsp
  • [04:54:47] <raster> ok
  • [04:54:51] <sakoman> just bare omap
  • [04:54:52] <raster> retail single-unit $149
  • [04:54:56] <raster> hmm
  • [04:54:59] <raster> also 600mhz
  • [04:55:09] <raster> 3530 v 3503... what vanished?
  • [04:55:15] <raster> AAAAH
  • [04:55:17] <raster> no dsp or 3d
  • [04:55:19] <raster> cool
  • [04:55:26] <raster> dsp would be nice
  • [04:55:34] <raster> 3d i can live without atm thansk to it being closed
  • [04:55:46] <raster> but as such i see this as a drop-in replacement anyway
  • [04:55:52] <sakoman> seems that volume discounts are available if you buy more than 120 units
  • [04:55:55] <raster> so drop in a 3530 if u MUSt have dsp and 3d
  • [04:55:57] <raster> for example
  • [04:56:03] <raster> sure
  • [04:56:10] <sakoman> yes, they are drop in, same PCB
  • [04:56:14] <raster> and if u buy 10k units.. you probably negotiate special prices
  • [04:56:18] <raster> :)
  • [04:57:23] <raster> aah 800x480 4.3" samsung lcd
  • [04:58:03] <raster> 21% > 3k
  • [04:58:06] <sakoman> I believe there will be several lcd options
  • [04:58:09] <raster> (for mb's)
  • [04:58:44] <raster> so hmm
  • [04:58:56] <raster> lcd not really the one i wanted - that i see
  • [04:59:03] <raster> some svhs-style connectors there
  • [04:59:58] <sakoman> what size/resolution are you looking for?
  • [05:00:27] <sakoman> I know they are discussing a couple of other lcd options to offer
  • [05:00:37] <raster> simple
  • [05:00:41] <raster> cellphone
  • [05:00:41] <raster> :)
  • [05:00:44] <sakoman> you really could choose anything since you are doing a custom daughter card
  • [05:00:45] <raster> ie really damn small
  • [05:00:49] * jkridner|work (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) has joined #beagle
  • [05:01:00] * jkridner|work (n=jason@c-76-31-18-64.hsd1.tx.comcast.net) has left #beagle
  • [05:01:02] <ds2> for 800x480, prehaps the one from the N810 or N800 could be uused
  • [05:01:03] <raster> but i guess gaughterboard would do the lcd connector
  • [05:01:16] <sakoman> correct, really it could be anything
  • [05:01:20] <ds2> it is known to work with the omapfb and the timings are already in the kernel
  • [05:01:24] <raster> was considering
  • [05:01:36] <raster> but - i thnk it's a tad too big
  • [05:01:36] <raster> if that is used
  • [05:01:38] <raster> u'd have to have an incredibly flush case
  • [05:01:41] <raster> otherwise
  • [05:01:54] <raster> ie the case would be the lcd dimensions exactly
  • [05:01:55] <ds2> what is the target resolution?
  • [05:02:07] <ds2> 320x320 tolerable?
  • [05:02:08] <raster> plus a little above/below for mic/speaker
  • [05:02:13] <raster> hmm nah
  • [05:02:22] <raster> was hopine 272x480 @ 3" minimum
  • [05:02:34] <ds2> so a smaller PSP display essentially
  • [05:02:34] <raster> i have looked into one of those before from AOU
  • [05:02:45] <raster> 3x0640 would be nice
  • [05:02:50] <raster> nokia produced somethnig with that
  • [05:02:54] <raster> 3.5" i thnik
  • [05:02:58] <raster> dont know details
  • [05:03:10] <raster> that'd be as high res as i think is sensible
  • [05:03:13] <raster> and in terms of size
  • [05:03:15] <raster> 320x480 is ok
  • [05:03:22] <raster> but too "iphone"
  • [05:03:31] <sakoman> you should easily be able to fit what you need on a footprint the size of the lcd
  • [05:03:39] <raster> yup
  • [05:03:43] <raster> thats the main thing
  • [05:03:46] <ds2> nicething about 320x320 is that is what's on teh Treo so buying a stack of replacement treo screens....
  • [05:03:48] <raster> lcd is going to heavily define the footprint
  • [05:04:01] <ds2> sakoman: the killer might be Z dimenion
  • [05:04:02] <raster> ds2: true
  • [05:04:05] <raster> i have a treo
  • [05:04:06] <raster> its nice
  • [05:04:11] <raster> but... not nice enough
  • [05:04:33] <ds2> those HiRose gumstix connectors take up almost as much Z as a board :/
  • [05:05:17] <raster> errr
  • [05:05:23] <raster> AUO i mean
  • [05:05:30] <raster> they have a really slim 272x480
  • [05:05:31] <raster> 3"
  • [05:05:35] <raster> looked very nice
  • [05:06:00] <raster> ds2: true - the conectors are a problem
  • [05:06:01] <raster> BUT
  • [05:06:19] <raster> u can do ti differently and do a sandwich-style
  • [05:06:28] <raster> and then make up for the space with a different battery
  • [05:06:40] <raster> so instead of board backing the hole lcd
  • [05:06:43] <sakoman> ds2: you are thinking of the previous generation
  • [05:06:46] <raster> it uses a portion only
  • [05:06:49] <raster> battery fills the rest
  • [05:06:55] <sakoman> overo uses (2) x 70-pin AVX 5602-14 connectors
  • [05:07:25] * ds2 hides from his guests more and goes to look for CAD drawings of the AVX5602-14
  • [05:07:48] <raster> hehehe
  • [05:07:54] <raster> boring guests?
  • [05:08:46] * rupeshgujare (n=Administ@59.160.172.220) has joined #beagle
  • [05:08:56] <ds2> parents ;)
  • [05:09:07] <raster> AAAAH
  • [05:09:13] <raster> xmas
  • [05:09:13] <raster> family
  • [05:09:29] <ds2> AVX5602 has a stacking height of 1.5mm... that's about 0.060 which is the thickness of a common PCB
  • [05:09:29] <raster> hide.
  • [05:09:50] <raster> http://auo.com/
  • [05:09:54] <raster> that have some nice lcd's
  • [05:11:30] <sakoman> ds2: better than the old 2mm and 3mm spacings :-)
  • [05:12:27] <ds2> sakoman: well... fair enough
  • [05:12:46] <sakoman> I don't think it would be a big issue in a cell phone design. A good mechanical engineer could tuck it in there
  • [05:13:36] <ds2> unless you want something like what motorola did with their thinest phones ;)
  • [05:14:27] <sakoman> when you are only making 20K it is really tough to justify the development expense that Apple, Nokia, and Moto can
  • [05:14:48] <sakoman> unless unit cost isn't a big issue
  • [05:14:52] <ds2> raster: nice... going to risk a capacitive touch screen design? :D
  • [05:15:09] <raster> well starting at 20k
  • [05:15:20] <raster> but yes - dont have the mopney for that kind of expense
  • [05:15:23] <raster> but at the same time
  • [05:15:33] <raster> cant go producing something half-arsed
  • [05:15:38] <raster> ds2: i'm tempted
  • [05:15:52] <raster> synaptics have been hocking their wares around
  • [05:16:00] <raster> but i havent dug in much
  • [05:16:25] <raster> its going to be murder on the design tho separating touch surface and screen
  • [05:16:25] <ds2> the other issue with these thin connectors is how vibration resistant... you can screw it together but dealing 0-80's in a close quarters is not exactly what you want for cheap mfg
  • [05:16:35] <raster> ie the mechanical
  • [05:16:49] <raster> as u need to connect now bits of the outer caseing electrically
  • [05:17:04] <ds2> yep... the Atmel Q touch stuff has been tempting
  • [05:17:06] <raster> yeah
  • [05:17:07] <ds2> no
  • [05:17:09] <raster> thats the other problem
  • [05:17:15] <raster> making it small is possible
  • [05:17:25] <raster> but then u need tomake it sanely manufacturable at a sane cost
  • [05:17:28] <ds2> you can have a solid block of plastic on top... but you do need the indium oxide film made though
  • [05:17:55] <raster> but then that film needs to be connected
  • [05:18:10] <ds2> it is internal justlike a resistive screen
  • [05:19:58] <raster> ummm but the resistive ts is glued on the lcd
  • [05:20:08] <raster> and the lcd ribbon connector is what waires it up
  • [05:20:15] <raster> in this case its not attached to the lcd anymore
  • [05:20:19] <raster> so need some more connectors
  • [05:21:01] <sakoman> do you have a good mechanical guy?
  • [05:24:49] <raster> right now... that's me
  • [05:24:50] <raster> :)
  • [05:24:54] <raster> so "no"
  • [05:24:54] <raster> :)
  • [05:25:13] <sakoman> understood! they are worth their weight in gold :-)
  • [05:27:03] <sakoman> I started my career doing calculators for HP and ended the "big company" phase of my life 25 years later doing iPods so I've seen the benefits of good iD/ME types
  • [05:27:25] <raster> sure - i dont doubt u
  • [05:27:28] <raster> problem is budget
  • [05:27:56] <raster> at least initial budget
  • [05:28:27] <raster> right now i am hoping to find a way to do this for "pocket change"
  • [05:28:46] <raster> (relative to what the industry normally spends on such things)
  • [05:28:56] <raster> basically because i dont have an existing product pipeline to feed it
  • [05:29:09] <sakoman> a good mechanical type might save you several iterations and hence $
  • [05:29:16] <raster> yes
  • [05:29:26] <raster> tho depends how u make it
  • [05:29:32] <raster> eg moulds - absolutely
  • [05:29:40] <raster> cost of a mold is insane
  • [05:29:50] <raster> cnc things...
  • [05:29:51] <sakoman> can't tell you how many times I've seen seen PCB spins required for silly mechanical issues
  • [05:29:57] <raster> mich easier
  • [05:30:02] <raster> higher per unit cost tho
  • [05:30:08] <raster> ooh pcb
  • [05:30:11] <raster> thats another
  • [05:30:27] <raster> as such i want to delay real manufacturing as much as feasible
  • [05:30:32] <raster> ie design everythign virtually
  • [05:30:36] <raster> if possible
  • [05:30:55] <raster> that means software starts before hw is there using desktops+x86 (already done/underway)
  • [05:31:02] <raster> then using almost equivalent hw
  • [05:31:20] <raster> eg beagleboard or overo
  • [05:31:23] <sakoman> that's the way the big guys do it too :-)
  • [05:31:45] <raster> + stuff jerry-rigged on with stickytape and wires
  • [05:32:30] <raster> yeah
  • [05:32:35] <sakoman> I licensed sw to cell phone companies for a few years and I've seen lots of phone protos that looked like that ;-)
  • [05:32:39] <raster> but they also have 892 levesl of beaurocracy
  • [05:32:45] <sakoman> no kidding!
  • [05:32:49] <raster> and a lot of cruft to get in the way
  • [05:33:01] <sakoman> try to get license agreements done with those guys!
  • [05:33:12] <raster> hehee i know
  • [05:33:17] <raster> i contract for one of them
  • [05:33:24] <raster> admittedly its their r&d dept
  • [05:33:41] <raster> so its a step removed from the production line
  • [05:34:02] <raster> and i dont see the hw bits - just software
  • [05:34:17] <raster> what i'm doing is entirely orthogonal
  • [05:34:32] <raster> except that they are building their r&d next gen software ont he same stuff i'm going to use anyway
  • [05:36:39] <sakoman> sounds like a fun project
  • [05:37:02] * soman (n=somnath@61.16.248.242) has joined #beagle
  • [05:37:05] <raster> well i'm hoping i can fund it to get it started
  • [05:37:13] <raster> been saving for a while for this
  • [05:41:03] <sakoman> I'm happy to offer free advice :-)
  • [05:45:43] <raster> sure :) most appreciated
  • [05:46:02] <raster> just worting out my hw spec list
  • [05:46:52] <sakoman> I've been involved in doing consumer devices for a lot of years & if I can help you avoid some of the mistakes I made it would make me happy
  • [05:47:14] <raster> sure
  • [05:48:33] <raster> well here's the goal - omap3530 based "smartphone" (yes - i know every man and his dog is doing it but i've had this on my burner and radar for about 8 years now and been following everythng in the market i can - and i'm finally in a position where i know just enough to get started and not enough to do it all myself, and have enouhg money to maybe do it.. and customers willing to buy if execution happens)
  • [05:48:50] <raster> so its kind of - the perfect storm is here and its time to finally live the dream
  • [05:49:02] <raster> open phone
  • [05:49:22] <raster> ie all compnents have drivers that are open (or the component is simply ignored - eg the 3d core)
  • [05:49:47] <raster> and yes i know android si all the rage now with those who are hip
  • [05:49:56] <raster> or those that dont care abotu linux just say "buy an iphone"
  • [05:50:00] <raster> but there is a market
  • [05:50:19] <raster> and that is something after 8 years of making sure there is a market.. i know exists
  • [05:50:21] <raster> :)
  • [05:50:28] <raster> other than that give the users what they want
  • [05:50:47] <raster> that is 3g, wifi, bt, gps, fast soc, decent ram and flash, but expandable
  • [05:50:50] <raster> decent screen
  • [05:51:10] <raster> mechanically i'd love to give them a qwerty kbd - but frankly.. dont see that happening without divine intervention
  • [05:51:22] <raster> (ie mechanically thats now beyond my experience)
  • [05:51:54] <raster> after much kerfulling about on many soc's ompa 35xx seems about the best
  • [05:51:57] <sakoman> yes, big tooling bucks for that
  • [05:52:20] <sakoman> definitely agree on the omap3
  • [05:52:37] <raster> yup
  • [05:52:44] <raster> so chances are it'd going to be mostly touchscreen
  • [05:52:51] <raster> if thats the case capacitive is probably the go
  • [05:53:07] <raster> frankly no stylus to worry about shovelling into the case
  • [05:53:17] <raster> and using it with your fingers is frankly useful
  • [05:53:42] <sakoman> yes, stylus for a phone is just bad UI
  • [05:54:53] <raster> well in the past its been a necessity
  • [05:54:58] <raster> but now its not
  • [05:56:25] <raster> ds2: btw - accel - $10 is their 1 unit pricing
  • [05:56:36] <raster> but buy a few k sand itd down to $6-7
  • [06:01:48] <raster> you are right - mechanical is a problem
  • [06:01:56] <raster> but just production will be too
  • [06:01:59] <raster> assembly
  • [06:02:10] <raster> its a chain of problems
  • [06:05:16] <sakoman> yeah, finding a good contract assembly house is a challenge
  • [06:06:05] <sakoman> one good reason to start with an overo module -- someone else already paid for the OMAP/POP/WIFI/BT learning curve :-)
  • [06:06:41] <raster> sure
  • [06:06:48] <raster> tho missing dsp is a little unhappiness
  • [06:07:08] <sakoman> you can get 3530 just by asking
  • [06:07:16] <sakoman> I have a couple on my desk right now
  • [06:07:39] <raster> 3525 might be the go there
  • [06:07:41] <sakoman> For volume orders you just have to ask
  • [06:07:46] <raster> 3525?
  • [06:07:54] <sakoman> any of the choices should work
  • [06:07:56] <raster> (same as 3530 minus 3d)
  • [06:08:00] <raster> ooh ok cool
  • [06:08:02] <sakoman> all pin compatible
  • [06:08:08] <raster> so u can do that
  • [06:08:09] <sakoman> just a load option
  • [06:08:16] <raster> i have no idea how flexible gumstix are
  • [06:08:38] <sakoman> since they aren't a retail SKU yet, there might be a minimum order
  • [06:09:04] <raster> that is not going to be the problem
  • [06:09:04] <raster> :)
  • [06:10:13] <raster> so a custom daughterboard then osunds sane
  • [06:10:19] <sakoman> definitely
  • [06:10:20] <raster> u dont have any lcd optiosn that seem sane tho
  • [06:10:47] <sakoman> not introduced yet, but on the roadmap (and on my desk running e17)
  • [06:11:14] <sakoman> LCD on daughter card is a trivial task
  • [06:11:31] <raster> yeah
  • [06:11:38] <raster> the problem is more the lcd itself
  • [06:11:43] <raster> and the ts
  • [06:11:48] <sakoman> Gumstix publishes their daughter card schematics so it is an easy cut and paste
  • [06:12:17] <sakoman> Touchscreen for the Samsung is resistive
  • [06:12:41] <sakoman> it is an SPI controller, capacitive would use a different controller
  • [06:13:03] <sakoman> just look for one that has an i2c or SPI interface
  • [06:13:08] <raster> no wifi/bt/gps yet either :)
  • [06:13:14] <sakoman> (and an existing linus driver :-)
  • [06:13:33] <raster> yeah - the samsung would be a no-brainer
  • [06:13:42] <sakoman> wifi/bt are on the current motherboard layout
  • [06:13:44] <raster> and for a raw prototype "pieced together" would be good
  • [06:13:56] <raster> oh wait
  • [06:13:59] <raster> wifi/bt on the mb?
  • [06:14:00] <sakoman> working through yield issues before it goes on the price list
  • [06:14:16] <raster> aaah ok
  • [06:14:17] <sakoman> yes, there are protos floating around
  • [06:14:22] <raster> combined wifi+bt chip?
  • [06:14:24] <sakoman> yes
  • [06:14:34] <raster> usb?
  • [06:14:34] <sakoman> sw is working for both
  • [06:14:48] <raster> working ... or "working"
  • [06:14:49] <sakoman> yes OTG is working, ehci port is limping
  • [06:15:04] <raster> ie "it boots" vs "iit handles everything we throw at it" :)
  • [06:15:08] <sakoman> wifi is libertas_sdio
  • [06:15:34] <sakoman> BT is serial and supported by standard bluez
  • [06:15:41] <raster> aaah ok
  • [06:15:52] <sakoman> so more "handles everything I've thrown at it"
  • [06:16:08] <raster> cool
  • [06:16:14] <raster> i know hwo things work in software
  • [06:16:26] <raster> "it boots" often == "ship it"
  • [06:16:27] <raster> :)
  • [06:16:41] <raster> so the driver is open... right?
  • [06:17:06] <sakoman> yes, and the libertas firmware is re-distributable
  • [06:17:15] <raster> but still a blob?
  • [06:17:21] <sakoman> yes, still a blob
  • [06:17:26] <raster> ok
  • [06:17:34] <raster> the tru open fanatics will be unhappy
  • [06:17:44] <raster> but thats like 2% of them
  • [06:17:46] <sakoman> good luck finding wifi that doesn't have similar or worse issues
  • [06:18:05] <raster> the other 98% go "well if tis a blob firmware i have to upload to card - ok. as long as it doesnt run in my kernel space"
  • [06:18:35] <raster> atheros is pretty much the only other option i know of
  • [06:19:26] <sakoman> wifi is a big issue for anyone who isn't a major manufacturer
  • [06:19:36] <raster> yup
  • [06:19:44] <sakoman> module/chip vendors just want to deal with the big guys
  • [06:19:51] <raster> antennae built into chip
  • [06:19:54] <raster> or need external?
  • [06:19:59] <sakoman> external
  • [06:20:06] <raster> ok nothing funky then
  • [06:20:07] <raster> :)
  • [06:20:52] * soman (n=somnath@61.16.248.242) Quit (Read error: 60 (Operation timed out))
  • [06:20:58] <raster> my main problem is going to be this 3g goop
  • [06:22:08] <sakoman> see picture at the bottom of the page I linked to for pictures of the UFL connectors for BT and WIFI
  • [06:22:08] * mik18 (n=asdf@76.85.180.121) Quit (Read error: 54 (Connection reset by peer))
  • [06:22:08] * mike18 (n=asdf@76.85.180.121) has joined #beagle
  • [06:22:32] <raster> component docs?
  • [06:22:53] <sakoman> which components?
  • [06:24:24] <raster> oh the picture - the one with dvi/dhmi, usb etc. etc.
  • [06:24:27] <raster> ?
  • [06:24:47] <ds2> UFL blows chunks ;)
  • [06:24:48] <raster> aah i see
  • [06:24:50] <raster> that one
  • [06:25:08] <raster> heheh
  • [06:25:09] <ds2> and this still an understatement
  • [06:25:09] <sakoman> ds2: agree, but options are limited
  • [06:25:14] <raster> well - connectors are a problem
  • [06:25:17] <raster> point fo failure
  • [06:26:26] <sakoman> hey, if you have the bucks to go full custom that would give you the cleanest design
  • [06:26:53] <raster> well the q is
  • [06:26:57] <sakoman> otherwise it's trade-off time :-)
  • [06:26:59] <raster> since so much is going on the board
  • [06:27:19] <raster> is it more expensive to make a packed daughterboard
  • [06:27:34] <raster> and then risk the point of failure etc. etc. and complicate assembly
  • [06:27:49] <raster> or simplify by having a single board
  • [06:27:54] <raster> and yes - all tradeoff
  • [06:28:09] <raster> the overo still isnt too cheap even in high volume
  • [06:28:25] <raster> as such there's still margin in there :)
  • [06:28:32] <raster> the q is... is it that much
  • [06:28:36] <raster> and is it worht the pain
  • [06:28:40] * emeb_mac (n=ericb@ip72-223-90-212.ph.ph.cox.net) Quit ()
  • [06:28:45] <raster> (of doin it all custom)
  • [06:28:53] <raster> i'm not qualified to know that
  • [06:29:06] <raster> all i know is that there is a call to be made
  • [06:32:26] <sakoman> yeah, then there are the time to market considerations
  • [06:32:53] <raster> yup
  • [06:32:59] <sakoman> it is not trivial to get a board design working and into production
  • [06:33:11] <raster> though i know those and there isn't a big hurry
  • [06:33:19] <raster> but trie
  • [06:33:21] <raster> true
  • [06:33:29] <raster> tho software will probably lag
  • [06:33:32] <sakoman> witness the fact that TI is still working on a rev C and Gumstix is still working on the wifi/bt version
  • [06:33:34] <raster> so i'm not rushing the hw
  • [06:35:00] <sakoman> It's never been completely clear to me at what point the cpu module approach begins to not make sense
  • [06:35:33] <sakoman> certainly for 100's to single digit thousands it makes sense
  • [06:36:02] <sakoman> then there is a big grey area where time to market and unit cost sensitivities make it a really complex decision
  • [06:36:17] <raster> yeah
  • [06:36:22] <raster> sucky for me i'm in there
  • [06:36:43] <raster> i'd be most happy to be able to be in the 100,000's or millions
  • [06:36:47] <sakoman> yup. the hw business sucks ;-)
  • [06:36:48] <raster> but.. that aint happening
  • [06:36:50] <raster> :)
  • [06:36:55] <raster> software is nice
  • [06:36:59] <raster> u create somethnig
  • [06:37:01] <raster> it takes time
  • [06:37:04] <raster> thats it
  • [06:37:13] <raster> anyone with a brain can do it with minimal setup
  • [06:37:13] <ds2> sakoman: I would be so much happier if the Overo had a series of pads instead of a connector
  • [06:37:37] <raster> sakoman: pads - would that be possible instead of a connector?
  • [06:37:46] <raster> if there are large enough orders?
  • [06:37:58] <ds2> not with their current board
  • [06:38:06] <sakoman> you'd have to ask, I have no idea
  • [06:38:13] <raster> ok - fair enough
  • [06:38:23] <raster> mind u
  • [06:38:31] <raster> a connector does have positives
  • [06:38:35] <raster> for the nerds of this world
  • [06:38:39] <raster> it measn the phone si upgradable
  • [06:38:41] <ds2> sakoman: actually, if people want a rapid system... the Zoom is always there ;)
  • [06:38:44] <raster> or side-gradable
  • [06:38:49] <sakoman> if you ask for a custom, why not just do a custom yourself?
  • [06:38:54] <raster> eg
  • [06:38:57] <raster> if 3g is too pricey
  • [06:39:03] * ds2 ducks to prepare for the flying tomatoes
  • [06:39:03] <raster> "ok- buy the 2g version"
  • [06:39:08] <raster> and it just has a 2g daughterboard
  • [06:39:27] <sakoman> right, flexibility does have its positives
  • [06:39:36] <raster> sakoman: well if the soc is drop-in changable...
  • [06:39:48] <raster> :)
  • [06:39:56] <ds2> connectors are expensive as a con for having connectors
  • [06:39:57] <sakoman> ds2: I think it is great for TI that there are so many options available
  • [06:40:21] <ds2> sakoman: sorry, I forgot the </sarcasm> there ;)
  • [06:40:40] <ds2> (have to have seen and held the zoom to understand)
  • [06:40:47] <sakoman> I haven't!
  • [06:41:22] <sakoman> ds2: From prior lives I'm used to thinking in millions/tens of millions
  • [06:41:38] <sakoman> So I understand the cost issues
  • [06:41:52] <sakoman> every penny counts at those volumes
  • [06:42:02] * soman (n=somnath@61.16.248.242) has joined #beagle
  • [06:42:04] <ds2> no argument there
  • [06:42:11] <raster> absolutely
  • [06:42:17] <raster> make 10m units
  • [06:42:26] <raster> save $1 each in parts and assembly costs
  • [06:42:32] <raster> thats a big xmas bonus in savings u made
  • [06:42:34] <raster> :)
  • [06:42:38] <ds2> try save 0.01 each in parts
  • [06:42:41] <ds2> ;)
  • [06:42:42] <sakoman> with smaller volumes the development costs become a significant portion of the unit cost
  • [06:43:01] <ds2> the NRE's don't change much with volume
  • [06:43:04] <sakoman> with ipod volumes you can throw vast amounts of engineering $
  • [06:43:33] <sakoman> to save a few pennies
  • [06:44:13] <ds2> but at those volumes, the folks in purchasing can make the engineer's life....
  • [06:44:53] <sakoman> well, I'm sure raster will make the right choices. he's worrying about the right things and it is his own money so he will be careful with it!
  • [06:45:02] <raster> hehehe
  • [06:45:12] <raster> no idea if i will make the right choice...
  • [06:45:14] <raster> i hope so
  • [06:45:26] <raster> but a connector for expanding/changin guts,.. makes sense
  • [06:45:44] <raster> also it falls into the "3g charges royalties.. per unit on the WHOLE unit cost"
  • [06:45:49] <raster> last i checked
  • [06:45:59] <raster> so.. if your unit is cheaper...
  • [06:46:10] <raster> u pay less in royalties
  • [06:46:22] <raster> so if 3g is a plug-on duaghterboard
  • [06:46:28] <raster> (with some other things)
  • [06:46:46] <sakoman> heh, reminds me of discussions on royalties for sw licensed *to* cell phone companies
  • [06:47:02] <sakoman> seems they always wanted it for free!
  • [06:47:08] <sakoman> or nearly so
  • [06:47:17] <raster> of course they do
  • [06:47:20] <raster> everyone wants free
  • [06:47:22] <raster> :)
  • [06:47:39] <sakoman> but the 3g guys manage to extract their pound of flesh
  • [06:47:53] <raster> yeah
  • [06:48:02] <raster> tho the licensing is pretty insane on that
  • [06:48:22] <sakoman> yeah, I'm glad to be out of that business
  • [06:48:53] <raster> just as i step in...
  • [06:49:09] <raster> tho i'm on the hw buying/selling end of 3g
  • [06:49:13] <sakoman> changing of the guard :-)
  • [06:49:19] <raster> hehehe
  • [06:49:27] <raster> i just want an open hackable phone
  • [06:49:33] <raster> as do 1000's of others
  • [06:49:46] <raster> where the maker says "go for it - patch your own kernel. have fun!"
  • [06:49:54] <sakoman> yeah, I'd probably buy one
  • [06:50:00] <raster> "but if u do. dont come screaming to us"
  • [06:50:19] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [06:50:24] <raster> sop many embedded developers who now buy beagle or gumstix would buy a fully working deviuce
  • [06:50:39] <raster> with battery/sceen/phone network/etc. all already done for them
  • [06:50:46] <raster> as its so much less hassle
  • [06:50:48] <raster> and its now portable
  • [06:50:54] <raster> plug in to usb
  • [06:51:00] <raster> reflash via dfu
  • [06:51:09] <raster> minimal hassle embedded
  • [06:53:10] <sakoman> lots of gumstix developers are into robotics and industrial embedded
  • [06:53:31] <sakoman> but many also want 3g
  • [06:53:45] <raster> yeah - so now they'd have something with a debug console (lcd) and phone network connectivity
  • [06:53:54] <raster> and battery built-in
  • [06:54:02] <raster> tho it likely isn not going to be cheap
  • [06:54:14] <raster> realistically its goign to be a pricey toy
  • [06:54:40] <raster> cost alone is going to be like $300 BOM
  • [06:54:42] <ds2> raster: the Zoom meets the letter of that requirement
  • [06:54:44] <raster> it seems
  • [06:54:59] <raster> ds2: is the zoom what i think it is?
  • [06:55:10] <ds2> I shall cease the wise cracks about the zoom for tonight
  • [06:55:46] <ds2> raster: the zoom is a reference board in the form of a phone (comes with BT/WiFi/and a GSM module); the SOM is on a DIMM connector
  • [06:55:58] <ds2> officially it uses the sister 3430 chip
  • [06:56:07] <ds2> LogicPD makes that
  • [06:57:01] * raster checks
  • [06:57:27] <ds2> make sure it is the one with the OMAP;they have at least one other product called the Zoom
  • [06:58:01] <raster> fucking isp
  • [06:58:12] <sakoman> cortex
  • [06:58:16] <raster> wont get dsp up until the 6th of jan
  • [06:58:21] <sakoman> sorry wrong window :-)
  • [06:58:26] <raster> err
  • [06:58:31] <raster> dsl
  • [06:58:32] <ds2> heh
  • [06:59:00] <ds2> sakoman: I think I even know what the question was ;)
  • [06:59:00] <sakoman> multitasking errors :-) maybe time for sleep
  • [06:59:09] <ds2> ad hoc IRC jeopardy ;)
  • [07:01:25] <raster> http://www.logicpd.com/products/devkit/ti/zoom_omap3_development_kit
  • [07:01:32] <raster> they have themselves a 25xx tooo
  • [07:01:35] <raster> it seems
  • [07:01:59] <raster> but just a board
  • [07:02:06] <raster> with lcd etc.
  • [07:02:57] <ds2> http://www.logicpd.com/products/devkit/ti/zoom_mobile_development_kit
  • [07:03:00] <ds2> that one
  • [07:03:10] <sakoman> raster: btw, the version bump on e did indeed fix the file manager icon issue. thanks!
  • [07:03:14] <raster> yeah found that too
  • [07:03:18] <raster> and man its ugly
  • [07:03:30] <ds2> raster: there is a reason I was making the wise cracks ;)
  • [07:03:56] <raster> hehehehe
  • [07:04:01] <raster> i see now
  • [07:04:14] <raster> man... that had better never leave the lab
  • [07:04:22] <sakoman> and you were complaining about the overo connectors!
  • [07:04:31] <raster> sakoman: cool! :) problem solved
  • [07:04:31] <raster> :)
  • [07:04:54] <sakoman> raster: now I can move on the next one!
  • [07:05:18] <sakoman> a never ending supply
  • [07:07:05] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Read error: 110 (Connection timed out))
  • [07:07:14] <raster> sorry to have caused you one of them :(
  • [07:07:26] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [07:07:41] <raster> ds2: actually only 2g
  • [07:08:01] <ds2> raster: but most of that stuff is on a SOM so in theory you could also use that as a board
  • [07:08:03] <raster> and missing a bunch of stuff like gps, accelerometers etc.
  • [07:08:38] <raster> hehehe
  • [07:08:41] <raster> well.. yeah
  • [07:08:52] <raster> a custom board does have bonuses of being.. slimmer
  • [07:09:05] <raster> it depends on the 3g really
  • [07:09:28] <ds2> _IMO_, none of these existing boards are a good fit for low volume production uses...
  • [07:11:36] <raster> in the end.. whatever makes it worka nd be the lower cost when we are talking 10-50k units
  • [07:13:32] <raster> a do agree tha a single board is going to make it slimmer
  • [07:13:50] <raster> and frankly - a very high bar has been set there by competitors whoa re not open
  • [07:14:59] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [07:15:37] <ds2> I am just thinking if you could also create a moderate volume module as a side effect
  • [07:15:40] * dirk2 (n=dirk@p5B04201B.dip0.t-ipconnect.de) has joined #beagle
  • [07:15:55] <raster> ?
  • [07:16:20] <ds2> a module that can be used for otherprojects that is more suited for moderate volume then the existing boards
  • [07:16:45] <raster> ooh i asee
  • [07:16:53] <raster> yes - so produce same board
  • [07:16:56] <raster> remove some parts
  • [07:17:03] <raster> sell as a baseboard for other projects
  • [07:17:08] <sakoman> if you are going to the expense of custom why not tailor it to be precisely what you want?
  • [07:17:13] <raster> eg done sell lcd/battery/case
  • [07:17:23] <sakoman> the CPU module business is a lousy business
  • [07:17:27] <raster> sakoman: that was the original plan
  • [07:17:37] <ds2> thinking a module that wouldbe solderable instead of connectors
  • [07:17:43] <raster> as i've been looking for an oem who has what i want
  • [07:17:50] <raster> and so far.. come up empty handed
  • [07:18:00] <raster> i'd happily go for an oem device with everythin done
  • [07:18:03] <raster> and a case too
  • [07:18:18] <raster> just have the logo changed and res-ell with new software
  • [07:18:29] <raster> but that isnt going to happen after much hunting
  • [07:18:34] <raster> so i'm now going to the other plan
  • [07:19:15] <raster> ds2: really tho.. just producing a singel pcb is cheapest and easiest
  • [07:19:40] <sakoman> raster: another quick e question if you don't mind - I noticed that OE packages like evince and gimp don't have an icon displayed in the apps menu
  • [07:19:43] <ds2> raster: right, just saying, maybe the pcb can alsobe sold as a module
  • [07:19:53] <sakoman> do you know what the issue is?
  • [07:20:58] <raster> sakoman: no icon - but an entry?
  • [07:21:06] <sakoman> correct
  • [07:21:13] <raster> ds2: possible
  • [07:21:27] <raster> though i;'d say volume is minimal comapred to the whole kit and kaboodle
  • [07:21:32] * favor (n=user@221.10.25.6) has joined #beagle
  • [07:21:38] <raster> as i said - i have barely scratched the volume surface
  • [07:21:54] <raster> i suspect reality will be more like 50k/yr
  • [07:21:57] <raster> or more
  • [07:22:10] <favor> hi. how to build gstreamer to beagle board using OE?
  • [07:22:15] <raster> thats what my market analysis has said for a long time
  • [07:22:28] <raster> the only reason i can do this is i know i have at least 1 10k order sittign ready to go
  • [07:22:33] <raster> that means if the margins are right
  • [07:22:45] <raster> the single order from a telco cusotmer can cover intiial production costs
  • [07:23:11] <raster> the rest then is just keeping the pipe flowing
  • [07:23:19] <ds2> raster: just more ways ofgenerating revenue
  • [07:23:30] <raster> and eventually when enough has been stached - work on version 2
  • [07:23:31] <raster> :)
  • [07:23:38] <raster> ds2: sure :)
  • [07:24:05] <ds2> raster: if you don't mind answering, what UI do you have in mind?
  • [07:24:09] <sakoman> raster: looks like packages that have an icon in /usr/share/pixmaps get displayed, those with icons in /usr/share/icons do not
  • [07:24:20] <raster> in fact i'd be more than willing to look at people doign custom short-run cases
  • [07:24:23] <raster> eg 100units
  • [07:24:31] <raster> they do the case and do some custom thing
  • [07:24:34] <raster> for their market
  • [07:24:37] <raster> u just sell them the guts
  • [07:25:07] <raster> sakoman: oh sorry - um.. /usr/share/icons should work
  • [07:25:08] <raster> BUT
  • [07:25:18] <raster> the .deskotp would need to specify the path relative
  • [07:25:20] <raster> ie
  • [07:25:22] <raster> if its in a subdir
  • [07:25:26] <raster> dir/icon.png
  • [07:25:28] <raster> OR
  • [07:25:36] <raster> it'd need to be in the icon theme ebing used
  • [07:25:54] <raster> value "icon_theme" string: "openmoko-standard";
  • [07:26:02] <raster> is the default icon theme set in the illume profile
  • [07:26:07] <ds2> small volume custom cases are easy
  • [07:26:07] <raster> so if the icon is not in that icon theme
  • [07:26:15] <raster> or not in dir/file.png
  • [07:26:19] <raster> or file.png
  • [07:26:21] <raster> it wont be found
  • [07:26:39] <raster> ds2: yup - so someone wants to make 10 custom ones for xmas presents to people
  • [07:26:40] <raster> go fo rit
  • [07:26:48] <ds2> people just need to realize time on a CNC machine is not that cheap
  • [07:26:54] <raster> Sept: there is a gui to select icon theme
  • [07:27:00] <raster> it just may not be that nice on your screen.
  • [07:27:03] <raster> depending...
  • [07:27:12] <ds2> if they want a $5case, forget it
  • [07:27:20] <raster> oh
  • [07:27:22] <raster> of course not
  • [07:27:30] <raster> i know that this project if anything is a premium thing
  • [07:27:36] <raster> u cant compete with the $200 hpone guys
  • [07:27:39] <raster> who sell 1million
  • [07:27:53] <raster> even at 50k its a "niceh" device
  • [07:27:57] <raster> so if its going to be niche
  • [07:28:02] <ds2> I was playing with the Tmobile G1 at the store and thatthing feels flimsy
  • [07:28:11] <raster> i have no problem goign to ectra costs/lengths to make it somethign special
  • [07:28:16] <raster> ie the alumunium casing
  • [07:28:19] <raster> etc.
  • [07:28:27] <raster> so somethnig specizal that sets it apart
  • [07:28:48] <raster> make nag and olfsen stand up and take notice
  • [07:28:50] <raster> err
  • [07:28:51] <raster> bang
  • [07:28:57] <raster> and olfsen
  • [07:28:58] <raster> :)
  • [07:29:27] <sakoman> raster: for these apps there are multiple icons under /usr/share/icons/hicolor/16x16, 22x22, etc
  • [07:29:41] <sakoman> which path should I specify?
  • [07:29:49] <sakoman> er, size wise
  • [07:29:53] <raster> sakoman: oh dear. looks like u cant change icon theme
  • [07:29:57] <raster> its controleld by the filemanager
  • [07:30:07] <raster> and... fm conf panel not there in illume profile
  • [07:30:18] <raster> ph u shouldnt have to specify path
  • [07:30:27] <raster> thats part of icon theme handlign to use the right path and icon size
  • [07:32:03] <raster> hmm ok icon theme selector useless
  • [07:32:10] <raster> no icon theme data installed
  • [07:32:17] <raster> well xdg-compliant stuff
  • [07:32:20] <khasim> sakoman: what is the cost of Overo with motherboard + wifi
  • [07:32:23] <sakoman> perhaps I could just hack the recipe to create a 48x48 entry in /usr/share/pixmaps
  • [07:32:39] <raster> ds2: yes. g1 feels flimsy - especially the screen on that flip-out arm
  • [07:32:50] <sakoman> khasim: I don't know
  • [07:32:54] <raster> sakoman: enerally i'd throw a default icon in there anyway
  • [07:32:55] <khasim> :)
  • [07:33:10] <sakoman> I don't think that gumstix has announced pricing yet
  • [07:33:34] <sakoman> so "I don't know" is the safest answer :-)
  • [07:33:56] <sakoman> khasim: have you gotten wifi working in the office on your overo?
  • [07:34:32] <khasim> sakoman: the office network was not open, I have board with me now at home
  • [07:34:39] <ds2> what WiFi module is on the overo?
  • [07:34:42] <khasim> sakoman: will check this out today
  • [07:35:02] <sakoman> raster: OK, I'll update the gimp, evince recipes to do that
  • [07:35:13] <khasim> sakoman: when I show board overo and beagle, many are interested in overo :)
  • [07:35:27] <sakoman> khasim: it is really tiny isn't it?
  • [07:36:05] <sakoman> khasim: you could use iwconfig to set wifi up for secure networks
  • [07:36:15] <khasim> sakoman: its really good, I was under an impression that total cost is 149 but then on gumstix site it says only the processor module to be 149$
  • [07:36:47] <khasim> sakoman: ok
  • [07:36:48] <sakoman> khasim: right. TI set the bar low on pricing with beagle
  • [07:37:11] <sakoman> very difficult to compete with that
  • [07:37:13] <khasim> sakoman: but the design is really good as others can buy only the processor module
  • [07:37:39] <sakoman> yeah, they serve different markets
  • [07:37:50] <raster> ds2: of as for gui.. frankly- whatever users want - but i was goign to ship X11+e
  • [07:38:08] <khasim> sakoman: I was trying to compare LogiPDs SOM module v/s overo. not much info on SOM module available and on avnet I saw it to be 350$
  • [07:38:13] <sakoman> some overlap, but overo is realy for low volume OEMs
  • [07:38:28] <sakoman> (compared to beagle)
  • [07:38:40] <sakoman> more direct competition with LogicPD
  • [07:39:13] <sakoman> raster: overo ships with x11+e in nand
  • [07:39:15] <khasim> more than competition I was looking at different ranges and feature set,
  • [07:39:26] <khasim> I think overo is good enough for any kind of products
  • [07:39:46] <khasim> with wifi on processor board, it gives good flexibility to desing the base board
  • [07:40:02] <raster> ds2: but really - id want to boot to x+e and as part of "First run" offer a choice to download and flasg another image if u want
  • [07:40:04] <raster> eg pure android
  • [07:40:07] <raster> or qte
  • [07:40:08] <khasim> but wifi is not available in 149 module
  • [07:40:08] <raster> if u like
  • [07:40:21] <sakoman> khasim: yeah, all the hard stuff is on the motherboard, daughter cards can be cheap 2/4 layer boards
  • [07:40:28] <raster> but provide a "Brwoser" that lists images, a bunch of screenshots and some info on them as wel as some "Support" matrix
  • [07:40:36] <raster> ie what the image supports etc.
  • [07:40:39] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Connection timed out)
  • [07:40:55] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [07:40:58] <raster> sakoman: thats awesome. :) conquoring the wolrd processor by processor :)
  • [07:41:03] <khasim> sakoman: I am demonstrating overo in India
  • [07:41:27] <sakoman> khasim: hopefully wifi/bt will be out soon. there were yield issues that delayed intro
  • [07:41:30] <raster> sakoman: if u had a tamer product in mind where i could live with just bt and wifi - overio might be perfect
  • [07:41:32] <khasim> sakoman: the overo is getting more interest due to the expandability and low cost
  • [07:41:36] <raster> err overo
  • [07:41:43] <sakoman> turns out to be a component vendor issue
  • [07:41:44] <raster> eg
  • [07:41:45] <raster> chumby
  • [07:41:54] <raster> stick an overo + lcd in a case
  • [07:42:00] <raster> (and add software)
  • [07:42:01] <khasim> sakoman: but the module will be priced high is it?
  • [07:42:43] <ds2> raster: I was just wondering you were making yet-another-android phone
  • [07:43:03] <sakoman> khasim: I don't think it will be too bad compared to LogicPD :-)
  • [07:43:09] <raster> ds2: oh hell no
  • [07:43:14] <raster> ds2: every man and his dog is
  • [07:43:18] <raster> and a lot of them are not "open"
  • [07:43:22] <raster> they're semi-open
  • [07:43:26] <raster> or mostly closed
  • [07:43:31] <ds2> hahaha
  • [07:43:39] <raster> also i really dont gel with this whole
  • [07:43:53] <raster> "lets force develoeprs to use our language and our vm and our api's to develop"
  • [07:44:01] <raster> what happened to re-cycling desktop developers
  • [07:44:10] <raster> who already know the whole java stack
  • [07:44:14] <raster> and include a full jvm
  • [07:44:17] <raster> or the gtk devs
  • [07:44:21] <raster> or qt/kde devs
  • [07:44:29] <raster> and the list goes on
  • [07:44:57] <raster> imho trying to avoid having x is just puting a blocker on a whole dev community waiting to release themselves onto toys
  • [07:45:02] <khasim> sakoman: the other wifi modules were around 40 $ from gumstix, so I am assuming the OMAP processor board with Wifi will be some where around 149 + 40 = 189
  • [07:45:03] <ds2> Speaking of desktop, I wonder if QEMU can run fast enough on the OMAP3 to run say... win98
  • [07:45:13] <raster> produce the toy they really want with the openness they want
  • [07:46:41] <sakoman> khasim: guess we'll have to wait and see where gumstix sets the price
  • [07:46:53] <khasim> sakoman: ok
  • [07:47:13] <raster> ds2: possibly
  • [07:47:19] <raster> i was toying with another idea
  • [07:47:23] <raster> wine + qemu
  • [07:47:36] <raster> much like scratchbox uses qemu in reverse
  • [07:47:45] <ds2> *nod*
  • [07:47:47] <raster> ie interpret the app code
  • [07:47:54] <raster> but have the wine code native
  • [07:47:59] <raster> better perf than pure qemu
  • [07:48:01] <raster> b ut trickier
  • [07:48:04] <ds2> an ARM hand held with ability to run legacy x86 code would be neat
  • [07:49:22] <sakoman> khasim: the new wifi module is much more highly integrated than the old one, which will likely help on price
  • [07:49:33] <raster> ds2: face it
  • [07:49:38] <raster> windows is legacy :)
  • [07:49:39] <raster> :)
  • [07:49:53] <khasim> sakoman: thats good
  • [07:49:56] <ds2> along with x86
  • [07:50:24] <khasim> sakoman: trying to push overo in India, let's see how it will execute :)
  • [07:50:25] <sakoman> khasim: though it also includes BT which may hurt price :-)
  • [07:50:38] <raster> heheh
  • [07:50:51] <sakoman> khasim: let me know if there's anything I can do to help
  • [07:51:13] <khasim> sakoman: sure
  • [07:51:20] <sakoman> with raster's help there will be a new more polished image out in the next day or so :-)
  • [07:51:37] <raster> but in the end everyone will bring all these hacks if you make the platform open and go
  • [07:51:45] <raster> "do it your way, not our way"
  • [07:51:49] <sakoman> I'm fixing a lot of the annying little things that were wrong
  • [07:52:25] <sakoman> missing icons, broken time zone support, . . .
  • [07:54:00] <sakoman> well, time for sleep. midnight here.
  • [07:54:02] <raster> its amazing how many little thnigs break in images
  • [07:54:12] <sakoman> thanks for the help raster!
  • [07:54:19] <sakoman> I really appreciate it
  • [07:54:25] <raster> no problems
  • [07:54:32] <raster> email the e-devle lists any time
  • [07:54:36] <raster> lots of people ready to help
  • [07:54:38] <raster> also on #e
  • [07:54:45] <sakoman> I'll do that!
  • [07:55:17] <sakoman> good night all
  • [07:55:31] <raster> nite!
  • [07:55:33] <raster> and thanks :)
  • [07:55:33] <khasim> is there a easy way to know the glibc version with latest code sourcey toolchain
  • [07:57:59] <AV500> raster: you still there?
  • [07:58:18] <raster> yup
  • [08:00:08] <khasim> found here http://www.codesourcery.com/gnu_toolchains/arm
  • [08:00:50] * mrc3 (n=ddiaz@189.157.115.100) has joined #beagle
  • [08:16:55] * LesterBurnham (n=luca@85.17.165.5) Quit (Read error: 110 (Connection timed out))
  • [08:42:03] * ssvb (n=ssvb___@a88-114-221-132.elisa-laajakaista.fi) Quit ("Leaving")
  • [08:43:28] <koen> good morning all
  • [08:53:53] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) Quit ("Leaving")
  • [08:54:15] <geckosenator> I"m learning how to write gcc lisp
  • [09:01:07] <tomba> I remember once seeing a rootfs build system that had kernel menuconfig-like configuration system. anyone know what the name of this is?
  • [09:02:44] <tomba> so what I'm interested in is building a minimal rootfs with just the basic stuff. I have my own script system that builds busybox and some other components, but I was wondering if there was a better ready made system somewhere
  • [09:03:28] <koen> tomba: try http://amethyst.openembedded.net/~koen/narcissus/ :)
  • [09:04:46] * guillaum1 (n=gl@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [09:05:14] <tomba> koen: interesting =) but that has a bit too high level stuff for me. I just want busybox, some libs like libpng etc.
  • [09:05:17] * guillaum1 (n=GZBA4143@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) has joined #beagle
  • [09:06:04] <tomba> hmm I wonder if I could use debian's arm packages for that...
  • [09:06:31] <koen> debian doesn't have anything above armv4t
  • [09:06:51] <tomba> that's ok, I don't need optimized stuff. better if it works on both omap2 and 3
  • [09:06:52] <koen> if you *only* want busybox and a few libs then buildroot would do
  • [09:07:20] <koen> beyond busybox a tool like matrix or openembedded would be better
  • [09:09:04] <koen> gregoiregentil: pong
  • [09:09:20] * dirk2 (n=dirk@p5B04201B.dip0.t-ipconnect.de) has left #beagle
  • [09:14:18] <tomba> koen: thanks, buildroot was the one I used earlier. let's see if it's enough. but I guess I should try openembedded at some point also, just to see what's it about =)
  • [09:17:10] <zuh> matrix+sb2 would be a nicely lightweight solution, but not sure how well it works currently...
  • [09:21:46] <gregoiregentil> Koen: Ping-pong...
  • [09:22:08] <gregoiregentil> Koen: I have a few questions/reports for you
  • [09:22:21] * aleij (n=ad@adsl-62-167-83-179.adslplus.ch) has joined #beagle
  • [09:23:03] <gregoiregentil> Koen: First, a suggestion: http://pastebin.com/d29b5a283
  • [09:24:17] * ldesnogu (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) has joined #beagle
  • [09:24:23] <tomba> zuh: what's matrix? quick googling didn't award me
  • [09:24:39] <kulve> tomba: yeah, "matrix" is a bad name because you really can't google it..
  • [09:25:07] <kulve> tomba: sandbox.movial.com
  • [09:25:43] <kulve> tomba: actually linux.onarm.com tells better how the matrix is (can be) used
  • [09:26:46] <zuh> tomba: In short, i
  • [09:26:54] <kulve> that is short
  • [09:27:00] <gregoiregentil> Koen: Another one (due to latest of ffmpeg in Angstrom) http://pastebin.com/dd5077d4, http://pastebin.com/d6eb3d8bb
  • [09:27:10] <zuh> t's what oe, jhbuild and other similar tools are too :)
  • [09:28:07] <zuh> but with it's own emphasis and conventions (yes, one tends to think if creating N build systems are really worth it ;)
  • [09:28:22] <zuh> oe meaning bitbake here I guess
  • [09:29:21] <gregoiregentil> Koen: thirdly, I was trying to compile clutter and you added a dependency to libgles-omap3 (I guess that it's a wrapper to pull the header and the binaries lib), which makes sense but this recipe is not available anywhere... Any hint to find this recipe?
  • [09:33:05] * JayFoxRox (n=opera@a89-182-135-8.net-htp.de) has joined #beagle
  • [09:41:32] * chelli (n=chelli@p54B86E43.dip.t-dialin.net) has joined #beagle
  • [09:56:27] * fille_ (n=fille@a83-245-252-47.elisa-laajakaista.fi) has joined #beagle
  • [09:57:08] * fille_ is now known as philippe
  • [09:59:27] <gregoiregentil> Koen: fourthly, I'm trying to compile the kernel with CONFIG_PVR=y and I get this strong warning: http://pastebin.com/d7f236bad. Should I be worried?
  • [10:00:53] * JayFoxRox (n=opera@a89-182-135-8.net-htp.de) Quit (Read error: 110 (Connection timed out))
  • [10:03:40] * garren (n=chatzill@mail.dm.co.za) has joined #beagle
  • [10:03:54] <gregoiregentil> Koen: fifthly, I'm also trying to add abiword-plugins, which compiles fine but generates an error during rootfs "log.do_rootfs.691: * Cannot find package abiword-plugins.". In the ipk folder, there are a bunch of abiword-plugins*.ipk but no abiword-plugins. Have you tried to run this recipe?
  • [10:04:26] * abitos (n=nixgibts@p5B2E6809.dip.t-dialin.net) has joined #beagle
  • [10:06:44] * guillaum11 (n=GZBA4143@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) has joined #beagle
  • [10:07:30] * guillaum1 (n=GZBA4143@AMontsouris-153-1-20-180.w86-212.abo.wanadoo.fr) Quit (Nick collision from services.)
  • [10:07:36] * guillaum11 is now known as guillaum1
  • [10:09:20] * JayFoxRox (n=opera@a89-183-18-130.net-htp.de) has joined #beagle
  • [10:10:01] * abitos (n=nixgibts@p5B2E6809.dip.t-dialin.net) Quit (Nick collision from services.)
  • [10:10:17] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) has joined #beagle
  • [10:13:45] * ScriptRipper (n=martin@85.233.34.53.dynamic.cablesurf.de) has joined #beagle
  • [10:29:01] * DJWills (i=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) Quit ("Manny: It's my scythe. I like to keep it next to where my heart used to be.")
  • [10:39:09] * favor (n=user@221.10.25.6) Quit ("Leaving")
  • [10:39:42] <garren> when does the beagleboard contest officially start?
  • [10:43:43] <JayFoxRox> whats that? some coding competition? using the beagleboards power most or what?
  • [10:44:47] <garren> yeah check out: http://elinux.org/BeagleBoard/contest
  • [10:44:53] <garren> I think they are still planning it
  • [10:46:33] <JayFoxRox> so, is this only low level stuff or also the higher level stuff?
  • [10:47:00] <garren> high level stuff as well I think
  • [10:47:00] <JayFoxRox> because I wrote some 3D stuff for PSP and PC before and I m preparing a demo for the 09 demopartys right now - might make a good entry :P
  • [10:47:20] <JayFoxRox> (seeing how I was targeting pandora anyway)
  • [10:50:48] <JayFoxRox> but 5th jan is really close. not much time - there is xmas etc which will hold me back, and I still didn't get the SGX drivers
  • [10:51:21] <garren> yeah I'm away for christmas so also wont have much time, but your 3D stuff will probably work nicely
  • [10:54:15] <ldesnogu> except 3D drivers are not there yet
  • [10:54:48] <JayFoxRox> yeh - well the kernel stuff was compiled by koen, but I only send a mail about the userspace stuff to TI yesterday
  • [10:55:02] <JayFoxRox> and seeing how some people claim it takes a few weeks - I guess I m doomed
  • [11:13:15] * TehUni (n=cory@c-24-30-35-55.hsd1.ga.comcast.net) Quit (Read error: 104 (Connection reset by peer))
  • [11:13:35] * TehUni (n=cory@c-24-30-35-55.hsd1.ga.comcast.net) has joined #beagle
  • [11:18:15] <ldesnogu> ScriptRipper, please provide more information for your qemu issue so that I can check what's going on :)
  • [11:18:32] <ScriptRipper> Hi,
  • [11:18:52] <ScriptRipper> ldesnogu: of course.
  • [11:19:41] <ScriptRipper> I register QEMU for ARM with a binfmt handler...
  • [11:20:59] <ldesnogu> oh
  • [11:30:10] * simon42 (n=simon@89.238.65.10) Quit ("leaving")
  • [11:31:23] * dirk2 (n=dirk@p5B0410D0.dip0.t-ipconnect.de) has joined #beagle
  • [11:31:31] <tomba> well, I don't seem to be able to get buildroot working. sigh. cross compiling is always such a pain.
  • [11:34:11] <kulve> tomba: what are you cross compiling?
  • [11:34:32] <tomba> trying to build a minimal rootfs with selected components
  • [11:34:58] <tomba> I have my own scripts that do it, but I though there must be more extensive setups out there
  • [11:35:13] <kulve> if you are not afraid of SB, try the linux.onarm.com :)
  • [11:35:26] <garren> else OE is cool :-)
  • [11:35:54] <tomba> I wasn't able to use OE either, it used cvs to check out something and nokia's firewalls don't let anything out =)
  • [11:36:40] <garren> there is a guide on using a firewall in the wiki... I think...
  • [11:36:42] <kulve> l.o.c uses git by default but with little tweaking you can set it to use http..
  • [11:37:46] <dirk2> Anybody from EMBINUX here? ping
  • [11:42:16] <dirk2> garren: Contest: What's about adding your project to http://elinux.org/BeagleBoard/contest#Contest_projects (and beagleboard.org) and we will see how things develop then... You would be the first one, I think starting already can't be wrong ;)
  • [11:43:13] <garren> To early I still need to do some investigation and I'm going to be away for the rest of December so will only start looking at it around the 1st of Jan so I'm not sure if I have enough time
  • [11:43:18] <rupeshgujare> dirk: yes
  • [11:43:39] <dirk2> rupeshgujare: You are from EMBINUX?
  • [11:43:45] <rupeshgujare> yup
  • [11:44:11] <dirk2> I don't like the changes http://elinux.org/index.php?title=BeagleBoard&diff=8637&oldid=8613
  • [11:44:45] <dirk2> *adding* info is fine, but not removing other info and replace it by "use EMBINUX"
  • [11:45:50] <rupeshgujare> actually we were trying to edit Android section. and we didn't edited any other section
  • [11:47:42] <rupeshgujare> dirk2: as previous info was a bit outdated, and only deleted section which was not relevant.
  • [11:48:16] <dirk2> rupeshgujare: Yes, that's fine. But please *add* your Android information and don't remove information others have already put there before you.
  • [11:48:49] <dirk2> rupeshgujare: You decide that the deleted stuff was not relevant? What if others like it?
  • [11:49:30] * Openfree (n=df@222.65.143.118) has joined #beagle
  • [11:49:39] <dirk2> rupeshgujare: If I got it right, you removed e.g. link to https://omapzoom.org/gf/project/omapandroid/wiki/ . You think this is not relevant?
  • [11:51:18] <rupeshgujare> dirk2:all right. i thought it was for omap zoom architecture, and for beagle we updated info..
  • [11:51:38] <rupeshgujare> all rite.. i apologise.. i will restore it back..
  • [11:51:54] <dirk2> rupeshgujare: And I think the link to Android OMAP wiki page is gone, too?
  • [11:52:29] <rupeshgujare> i will restore as it was, and append EMBINUX stuff.
  • [11:53:02] <dirk2> rupeshgujare: Hmm, maybe we can do it better than just restore the old stuff?
  • [11:53:07] <dirk2> rupeshgujare: E.g.:
  • [11:53:40] <rupeshgujare> dirk2:if you can suggest, what are stuff u ll like see, i can modify accordingly
  • [11:54:24] <dirk2> rupeshgujare: Stay with the first paragraph, this intro is nice
  • [11:55:14] <dirk2> rupeshgujare: Then have two or three sub-sections: E.g.
  • [11:55:18] <dirk2> * Zoom
  • [11:55:31] <dirk2> * EBINUX
  • [11:55:38] <dirk2> * Android on OMAP wiki
  • [11:56:29] <dirk2> In Zoom section and Android on OMAP wiki section you restore the old info, in EMBINUX section you put your new stuff
  • [11:57:54] <rupeshgujare> dirk2:ok.. let me do it.
  • [11:58:58] <dirk2> rupeshgujare: E.g: http://elinux.org/BeagleBoard#Android
  • [12:01:07] <dirk2> rupeshgujare: And in EMBINUX section maybe you like to add a link to video. To put it into video section is fine, but I'm not sure everybody will find it there. So maybe an additional link in Android EMBINUX section might help ;)
  • [12:02:05] * raster (n=raster@r125-63-180-233.cpe.unwired.net.au) Quit ("Gettin' stinky!")
  • [12:07:49] * FuL|OUT is now known as fulgas
  • [12:13:42] <JayFoxRox> how do I change the angstrom default filemanager to gpe-filemanager?
  • [12:22:20] <rupeshgujare> dirk2: please check now. you can edit "OMAPZOOM" and "Android on OMAP wiki" if you wish to add/remove anything.
  • [12:22:39] <garren> JayFoxRox: should be somewhere in the OE recipes
  • [12:25:15] * rsalveti (n=salveti@189.70.38.175) Quit (Remote closed the connection)
  • [12:28:37] <JayFoxRox> garren: might be a stupid question, but, what are the OE recipes?
  • [12:31:29] <garren> JayFoxRox: are you using the beagleboard demo image supplied or did you build it yourself?
  • [12:31:44] <JayFoxRox> the demo image
  • [12:31:59] <dirk2> rupeshgujare: Looks quite better :) Thanks!
  • [12:32:09] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Read error: 110 (Connection timed out))
  • [12:32:31] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [12:33:21] <rupeshgujare> dirk2: welcome, and sorry for inconvience
  • [12:33:24] <garren> ok, not really sure then, you might be able to install it and then in Xorg.conf set it but not really sure...
  • [12:34:34] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) Quit (Nick collision from services.)
  • [12:34:50] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) has joined #beagle
  • [12:42:04] * and-ri (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) has joined #beagle
  • [12:49:33] <and-ri> hello. i have trouble with omapfbplay ...i using the kernel 2.6.27-omap1 with a older Angstrom fs where omapfbplay with the original kernel 2.6.26-omap1 usually work. I receive an error after starting omapfbplay:"FBIOPAN_DISPLAY: Invalid argument"....i think this is a kernel issue but i have no idea whether i have to install a patch or only set a option in the kernel.....has anybody an idea to fix it?
  • [12:55:51] * JuanG (n=Juan@nat/ti/x-ffd1c1b6ed681ac2) has joined #beagle
  • [12:56:16] * JuanG (n=Juan@nat/ti/x-ffd1c1b6ed681ac2) has left #beagle
  • [13:01:19] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
  • [13:05:10] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [13:10:21] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) has joined #beagle
  • [13:10:48] * soman (n=somnath@61.16.248.242) Quit (Remote closed the connection)
  • [13:20:54] <and-ri> Koen, are you there? i have a question regarding the kernel 2.6.27-rc4-omap1. which i get from the angstrom website....the problem is that ALSA seems to be disabled
  • [13:24:07] * cbrake_away is now known as cbrake
  • [13:32:18] <jkridner> good morning all
  • [13:39:49] <rupeshgujare> jkridner:good morning
  • [13:39:50] <and-ri> he;-)
  • [13:41:01] <Crofton|work> gm
  • [13:46:20] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) Quit (Nick collision from services.)
  • [13:46:38] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) has joined #beagle
  • [13:52:35] * garren (n=chatzill@mail.dm.co.za) Quit (Read error: 110 (Connection timed out))
  • [14:04:01] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
  • [14:13:46] * garren (n=chatzill@mail.dm.co.za) has joined #beagle
  • [14:14:54] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [14:18:48] <and-ri> where can i find the libc6-dev? cause angstrom.org has a dead link http://www.angstrom-distribution.org/repo/?pkgname=libc6-dev
  • [14:18:49] * rsalveti (n=salveti@200.184.118.130) has joined #beagle
  • [14:20:47] <koen> and-ri: have you tried entering 'libc6-dev' into the searchbox on that page?
  • [14:23:48] * dirk2 (n=dirk@p5B0410D0.dip0.t-ipconnect.de) has left #beagle
  • [14:25:59] <jkridner> bad timing on my part to miss dirk2.
  • [14:26:11] <jkridner> can the judges start to add votes to the contest criteria?
  • [14:26:17] * atin_ (n=atin@c-71-232-7-71.hsd1.ma.comcast.net) has joined #beagle
  • [14:28:32] <jkridner> what is the alternative format besides a 'dd' image? rootfs tarball with uImage under /boot ?
  • [14:29:15] <jkridner> if such a tarball were used, then we can automate creation of the 'dd' image, as long as no bootargs or u-boot changes were required.
  • [14:29:33] <jkridner> if there are u-boot or bootargs changes, then those would need to be supplied in some regular way as well.
  • [14:30:14] <jkridner> also, where do contestants distribute their images? Do we expect them to provide hosting? I think that is our only option on this time-frame.
  • [14:39:06] * and-ri (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) Quit (Read error: 110 (Connection timed out))
  • [14:42:28] * jsync (n=jess@59.160.172.220) Quit (Remote closed the connection)
  • [14:45:36] * JuanG (n=Juan@nat/ti/x-919612ed478cb295) has joined #beagle
  • [14:45:43] * JuanG (n=Juan@nat/ti/x-919612ed478cb295) has left #beagle
  • [14:46:57] * and-ri (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) has joined #beagle
  • [14:47:22] <and-ri> damm always this stupid timeout error here....
  • [14:47:25] * maelcum_ (n=quassel@e178155234.adsl.alicedsl.de) has joined #beagle
  • [14:49:33] * maelcum (n=quassel@e178146146.adsl.alicedsl.de) Quit (Read error: 110 (Connection timed out))
  • [14:50:06] <and-ri> koen: i have tried the libc6-dev in the search bar...but on the mainpage it gives no resolution and in the package browser he gives only the deadlink back http://www.angstrom-distribution.org/repo/?pkgsearch=libc6-dev&arch=
  • [14:54:28] * jonnor (n=jon@ti0016a380-1975.bb.online.no) has joined #beagle
  • [14:55:31] * emeb (n=ericb@ip72-223-90-212.ph.ph.cox.net) has joined #beagle
  • [15:01:37] <and-ri> i still have the trouble with omapfbplay : omapfbplay:"FBIOPAN_DISPLAY: Invalid argument"...i would be appreciative about any suggestion
  • [15:02:33] * rupeshgujare (n=Administ@59.160.172.220) has left #beagle
  • [15:04:50] * Beagle3 (n=Beagle3@vtelinet-216-66-97-215.vermontel.net) has joined #beagle
  • [15:07:06] <maelcum_> has somebody found an easy way to get documentation for the omap cpu? i'd really like to have a detailed instruction set & cpu features reference...
  • [15:07:43] <maelcum_> there are enough 30 page whitepapers, i don't need another one of those
  • [15:11:49] <Beagle3> Good Morning. I just booted angstrom but I can't seem to get a mouse or keyboard working. The screen from DVI shows what looks like a touch screen keybord.
  • [15:11:57] <Beagle3> Any ideas?
  • [15:12:12] <and-ri> maelcum_: may it could help you out http://www.ti.com/lit/gpn/omap3530
  • [15:12:29] <and-ri> http://focus.ti.com/general/docs/gencontent.tsp?contentId=36915
  • [15:13:38] <maelcum_> hm, i should have made myself clearer: i need an instruction set reference.
  • [15:13:51] <and-ri> Beagle3: may you should try to compile your own kernel with kybourd support
  • [15:14:15] <and-ri> *keyboard
  • [15:14:52] <TehUni> Beagle: most likely you have the wrong usb cable
  • [15:14:56] <and-ri> maelcum_: you mean a list for assembler commands?
  • [15:14:59] <TehUni> unless you searched far and wide for the one you've got
  • [15:15:11] <TehUni> Beagle3 even.
  • [15:15:18] <maelcum_> and-ri: yes
  • [15:15:31] <and-ri> hmm .. sorry no idea
  • [15:16:00] <Beagle3> I am using standard cables. Power to the board is comming from the USB hub.
  • [15:16:00] <maelcum_> and-ri: this is not your fault, it seems to be arm inc's. they just don't give that information away to anyone...
  • [15:16:38] <and-ri> but linux have, i iguess
  • [15:16:45] <TehUni> Beagle3: you have to use a usbB to usbA cable, one that's hard to find. not just a standard cable that comes with most electronics
  • [15:17:56] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [15:18:02] <and-ri> look at the linux kernel....you will found some hardware specific asm files , maybe you have a look at this to find out what you need
  • [15:18:04] * NishanthMenon (n=nmenon@nat/ti/x-f344fc59b10c59ba) has joined #beagle
  • [15:18:05] <Beagle3> How can I tell the difference?? Also if I am feeding power to the board isint that the correct cable?
  • [15:18:31] <maelcum_> fwiw, i've applied for an account at arm's website and fortunately i work for a software company that i could enter there. but it's silly to require so much for the docs.
  • [15:18:41] <TehUni> Beagle3: you have to power the board through the power connecter (barrel) if you want to use usb also
  • [15:23:28] <and-ri> maelcum_: well, i took a look at the linux kernel and unfortunately i didn't found any arm specific asm file
  • [15:23:37] <maelcum_> Beagle3: one of the easiest ways to get the correct adapter cable is to buy it on ebay.
  • [15:23:46] <maelcum_> and-ri: arch/arm
  • [15:24:11] <and-ri> but no ASM
  • [15:24:19] <maelcum_> *.S, there are some
  • [15:24:52] <and-ri> asm is usually the assembler file extension i guess
  • [15:24:53] <jonnor> maelcum_: google for Arm Cortex A-8 instruction set
  • [15:25:08] <jonnor> *ARM Cortex-A8
  • [15:25:16] <maelcum_> jonnor: does that find more than the quick reference card?
  • [15:26:38] <Beagle3> Thanks
  • [15:27:22] <maelcum_> and-ri: in the unix world it is mostly .S
  • [15:27:31] <maelcum_> or .s
  • [15:27:53] <maelcum_> that might come from a time when it meant "source" :)
  • [15:28:47] <jonnor> maelcum_: http://infocenter.arm.com/help/index.jsp?topic=/com.arm.doc.set.cortex/index.html
  • [15:29:52] * Beagle3 (n=Beagle3@vtelinet-216-66-97-215.vermontel.net) Quit ()
  • [15:29:54] <maelcum_> jonnor: the documentation i'm looking for is explicitly only available to registered users there.
  • [15:30:32] * Beagle5 (n=Beagle5@vtelinet-216-66-97-215.vermontel.net) has joined #beagle
  • [15:30:51] <maelcum_> i'm not too stupid to find it - i was rather asking for loopholes or tricks to get it without the paperwork... sorry, should have made that more clear.
  • [15:31:18] <maelcum_> sometimes a link is available in some corner of the website because it was forgotten, etc.
  • [15:31:33] <maelcum_> or TI has what arm doesn't have, etc
  • [15:32:21] <jonnor> you could try to find the info for the 3 supported instruction sets seperatly
  • [15:32:28] <jonnor> might turn up something useful
  • [15:35:35] <maelcum_> don't think so... for example, for the neon instruction i've found exactly nothing so far.
  • [15:35:43] <maelcum_> *instruction set
  • [15:35:59] <maelcum_> anyway, i'll just hope that i get an account soonish
  • [15:39:19] <and-ri> maelcum_: well, with your hint, i found some sources ;-) but it is indeed stupid to writ a whole program only with the commands which linux use....nobody knows what else would be possible....but if you write for Omap there must be an instruction set outsite to optimize for DSP
  • [15:42:11] * Openfree (n=df@222.65.143.118) Quit (Remote closed the connection)
  • [15:43:47] <ldesnogu> maelcum_, "i was rather asking for loopholes or tricks to get it without the paperwork" <-- I hope you won't find such trick :)
  • [15:44:28] <ldesnogu> because that'd mean someone broke ARM's license
  • [15:44:34] <maelcum_> ldesnogu: in contrast to what the suits may think, i think it would be good for arm overall...
  • [15:44:50] <ldesnogu> I think the same, but it's not the case :)
  • [15:45:28] <maelcum_> does everybody who asks nicely get an account or is that somewhat difficult as well?
  • [15:45:54] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) Quit (Read error: 110 (Connection timed out))
  • [15:46:19] <ldesnogu> when did you ask for that?
  • [15:46:45] <maelcum_> yesterday, so it's going to take a couple more days until i hear back, i guess.
  • [15:47:03] <ldesnogu> I'm afraid it will take longer than that due to the period...
  • [15:47:10] * Beagle3 (n=Beagle3@137.205.147.151) has joined #beagle
  • [15:47:44] <Beagle3> hi
  • [15:48:08] <maelcum_> yeah, obviously... everybody on holidays. i just wonder if there are strict criteria like "you need to work for a company who is going to buy a license/has a license"
  • [15:48:46] <Beagle3> i have a problem on reload the boot to the nand flash
  • [15:48:52] * dirk2 (n=dirk@p5B043348.dip0.t-ipconnect.de) has joined #beagle
  • [15:49:05] * JuanG_ (n=Juan@nat/ti/x-080c419b53a151d4) has joined #beagle
  • [15:49:30] <ldesnogu> maelcum_, there's no such requirement IIRC
  • [15:49:38] * JuanG_ (n=Juan@nat/ti/x-080c419b53a151d4) has left #beagle
  • [15:50:09] <maelcum_> okay, good. thanks. i'll shut up and wait for an answer for now.
  • [15:50:22] <ldesnogu> but you'd better have a good reason to use it and then you'll obviously will have some restrictions on what you can do with the document (such as no copying to anyone)
  • [15:50:55] <Beagle3> when i boot from the MMC it said..Starting on with MMC Reading boot sector ECC failed , page 0x00080100
  • [15:51:03] <maelcum_> the reason is that i'm compiling and maybe developing software for it *shrug*
  • [15:51:06] <Beagle3> anyone come across it?
  • [15:52:14] <maelcum_> Beagle3: sounds like an incompatible memory card(?)
  • [15:52:37] <Beagle3> it worked b4 with the same card
  • [15:52:47] <ldesnogu> Beagle3, I had such errors; they disappeared after I flashed X-loader and u-boot
  • [15:52:55] <maelcum_> ha...
  • [15:53:10] <Beagle3> Today it crushed suddenly...then i cant reload/erase the NAND
  • [15:53:24] <ldesnogu> hum my problem was different from yours
  • [15:53:33] <ldesnogu> it happened before reaching u-boot prompt
  • [15:53:41] <ldesnogu> I never let the card boot on its own
  • [15:53:44] * erbo (n=erik@c-7b7de455.115-16-64736c14.cust.bredbandsbolaget.se) has joined #beagle
  • [15:53:53] <and-ri> is there a package available to install JAMVM on ARMv7 in one step?
  • [15:54:28] * Beagle5 (n=Beagle5@vtelinet-216-66-97-215.vermontel.net) Quit ()
  • [15:54:55] <Beagle3> it appears 1.Texas Instruments X-loader 1.41 2.Starting on with MMC 3.Reading boot sector 4.ECC Failed, page 0x00080100
  • [15:55:04] * JimDonova1 (n=root@c-66-30-221-249.hsd1.ma.comcast.net) Quit (Read error: 104 (Connection reset by peer))
  • [15:55:33] <Beagle3> i have tried 2 different SD cards
  • [15:56:00] <Beagle3> i still cant manage to flash the boot...
  • [15:56:26] <Beagle3> any suggestoins?
  • [15:57:10] <and-ri> flash the boot?
  • [15:57:29] <and-ri> have you setup the environemt?
  • [15:57:35] <Beagle3> yes
  • [15:57:48] <JayFoxRox> did you check the partition size?
  • [15:57:53] <Beagle3> i have managed to install the Angstrom linux
  • [15:58:27] <Beagle3> but it suddenly crashed today...
  • [15:58:29] * nemequ (n=nemequ@ip68-111-215-155.sd.sd.cox.net) has joined #beagle
  • [15:58:56] <maelcum_> what's it like to run angstroem compared to debian? package management, available packages, reliability...
  • [15:59:00] <Beagle3> and the terminal shows those errors
  • [15:59:47] <maelcum_> http://elinux.org/BeagleBoardRecovery
  • [16:00:03] <Beagle3> actually a company called Grumstix has the same design as Beagleboard..
  • [16:00:15] <Beagle3> they use the same TI chip
  • [16:00:45] <ldesnogu> Beagle3, are you sure? last time I checked they were using omap3503 and only had plans for 3530
  • [16:00:46] * JuanG (n=Juan@nat/ti/x-1bb78b2bdbb6f0d8) has joined #beagle
  • [16:00:54] <Beagle3> i have followed and tried every single instruction i could find on goodle
  • [16:01:01] * JuanG (n=Juan@nat/ti/x-1bb78b2bdbb6f0d8) has left #beagle
  • [16:01:26] <Beagle3> but i still cant manage to replace the error u-boot
  • [16:02:51] <Beagle3> i have tried the USB load too... but my machine cant detect the beagleboard...i dont know the reason
  • [16:03:15] <Beagle3> the TI omap3 usbload
  • [16:05:38] <Beagle3> guys any ideas?
  • [16:08:01] <atin_> can you boot off the sd card?
  • [16:08:18] <Beagle3> maelcum the performance of Angstrom is slightly better comparing to debian
  • [16:08:53] <Beagle3> i can boot from the MMC with errors..ECC Failed page 0x00081000
  • [16:09:33] <atin_> hmm, I haven't seen that ... maybe one of the others here
  • [16:10:22] <Beagle3> if i can erase the NAND.. then i think i can solve the problem
  • [16:10:49] <atin_> I had managed to brick mine too the first day I had it, but using the MLO_recovery image brought it back
  • [16:11:57] <Beagle3> i have tried ... i still cant bring it back
  • [16:12:35] <atin_> have you tried building the images yourself?
  • [16:12:51] <Beagle3> no
  • [16:13:07] <atin_> what were the steps you've taken till now?
  • [16:13:24] <Beagle3> i just download all the files i need from http://code.google.com/p/beagleboard/downloads/list
  • [16:14:26] <Beagle3> i assume if i boot from the MMC all the files inside the NAND will be replaced
  • [16:14:35] <atin_> no
  • [16:14:53] <Beagle3> then how can i replace them?
  • [16:15:05] <atin_> you have to overwrite nand unless you are using one of the auto-replace images I saw somewhere
  • [16:15:28] <atin_> the autoflash image will do it for you
  • [16:15:42] <atin_> but otherwise you have to do it by hand.
  • [16:15:53] <atin_> let me see where the instructions were, I jsut had them up
  • [16:16:05] <atin_> http://www.sakoman.net/omap3/flash%20procedure.txt
  • [16:16:06] <Beagle3> yes..i have followed the steps on the manual...
  • [16:16:23] <atin_> tho maybe the latest version needs nandecc as one word
  • [16:16:47] <Beagle3> i cant type any commands on the terminal
  • [16:17:01] <atin_> first boot off mmc
  • [16:17:11] <atin_> and do the commands from that prompt
  • [16:17:32] <atin_> but make sure those are the right commands first - I think I used those, but now I don't remember!
  • [16:17:36] <Beagle3> with no commands prompt on my terminal..i cant do anything
  • [16:17:51] <Beagle3> the only thing i can see
  • [16:17:53] <Beagle3> is
  • [16:18:08] <Beagle3> Texas Instruments X-loader 1.41
  • [16:18:13] <Beagle3> Starting on with MMC
  • [16:18:19] <Beagle3> Reading boot sector
  • [16:18:25] <atin_> http://elinux.org/BeagleBoardNAND
  • [16:18:31] <atin_> this page has stuff on it too
  • [16:18:32] <Beagle3> ECC Failed, page 0x00080100
  • [16:18:38] <Beagle3> f7 de 37
  • [16:18:44] <Beagle3> ee ff 77
  • [16:18:47] <atin_> ok, then your image is bad on the mmc too
  • [16:18:56] <atin_> or your card is bad
  • [16:19:12] <Beagle3> i have format my SD many times
  • [16:19:21] <atin_> you have to copy the files in order too - MLO is the first file to copy to the partition
  • [16:19:24] <Beagle3> and i have tried a new card too
  • [16:19:32] <Beagle3> yes i did that
  • [16:20:12] <Beagle3> MLO , x-load.bin.ift, flash-uboot, u-boot, rd-ext2, uImage
  • [16:20:31] <Beagle3> that is wat i loaded on my SD card
  • [16:20:58] <atin_> hmm, sec
  • [16:21:06] <Beagle3> it worked the first day
  • [16:21:27] <Beagle3> this is the 2nd week...it no longer works
  • [16:21:40] * garren_ (n=garren@dsl-244-208-104.telkomadsl.co.za) has joined #beagle
  • [16:22:18] <Beagle3> i think the only option i have is to load uboot.bin via USB
  • [16:22:26] <dirk2> Beagle3: Can you try a recent U-Boot?
  • [16:22:27] <atin_> is the MLO you have MLO_revb?
  • [16:22:51] <atin_> there we go - dirk can help you with u-boot :)
  • [16:23:06] <Beagle3> i have tried MLO_revb also tried MLO_restore
  • [16:23:25] <dirk2> Beagle3: sakoman should have one, I have to look for a link
  • [16:23:42] <dirk2> Beagle3: Or send me mail, I will attach one then
  • [16:24:29] <jkridner> Beagle3: what extensions did you use for those file? did you name them appropriately?
  • [16:24:41] <Beagle3> yes
  • [16:24:47] <Beagle3> all double checked
  • [16:25:04] <Beagle3> i have done all these the 1st day i received the board
  • [16:25:22] <dirk2> Beagle3: I remember an bug we had with U-Boot NAND, maybe there are some 'old' U-Boots around
  • [16:25:25] <jkridner> per http://code.google.com/p/beagleboard/wiki/BeagleBoardDiagnostics ?
  • [16:25:42] <jkridner> so, u-boot.bin_autoflash was renamed to u-boot.bin?
  • [16:25:50] <Beagle3> Today is my 2nd week having it... it suddenly crashed this morning...then i cant manage to recover it
  • [16:26:07] <atin_> there are - I managed to brick mine the first day I had it too - I ended up building myself and using that.
  • [16:28:19] <dirk2> Beagle3: Can you please try the U-Boots at http://www.sakoman.net/feeds/omap3/glibc/images/beagleboard/
  • [16:28:52] <atin_> koen: ping?
  • [16:29:08] <erbo> I'm trying to get started building Angstrom using OE for the beagleboard.. But when I try to bitbake nano as a first try it fails on shasum-native.bb on do_compile, bit the do_compile log is all empty. Ideas?
  • [16:29:13] <Beagle3> ok im trying
  • [16:29:28] <atin_> koen: pidkin is failing to build? do_fetch failed - you just changed the .bb fiel
  • [16:29:30] <atin_> file
  • [16:30:20] <atin_> err, pidgin
  • [16:30:57] <atin_> weird, working now.
  • [16:31:04] <atin_> maybe the site was hung or something
  • [16:32:08] <atin_> if I want to add uinput to the modules being built for linux-omap, do I just modify the defconfig?
  • [16:32:24] <atin_> seems to work, but wondering if that is the way to do it in openembedded.
  • [16:32:46] * garren_ (n=garren@dsl-244-208-104.telkomadsl.co.za) Quit ("Ex-Chat")
  • [16:33:28] <atin_> erbo: did you set up the openembedded environment as per the instructions? are you sure ;)
  • [16:34:35] * tharvey (n=tharvey@adsl-76-205-222-173.dsl.snlo01.sbcglobal.net) has joined #beagle
  • [16:34:46] <atin_> http://elinux.org/BeagleBoardAndOpenEmbeddedGit
  • [16:34:49] <erbo> atin_: I'm tripple checking that right now, but so far it seems to be right.. any ideas what could be wrong? otherwise I'll just keep on checking some more
  • [16:35:11] <atin_> no - my first guess would be configuration if this is your very first build
  • [16:36:23] <atin_> did it first build gcc etc?
  • [16:36:35] <atin_> or was that the first do_compile it tried?
  • [16:37:48] <atin_> (btw, I am heartily tired of openembedded already. its great if all you're doing is building the packages already there, and not modifying any sources). Otherwise I am starting to prefer handbuilding everything using the cross-compiler that was created by openembedded.
  • [16:38:59] <atin_> I hacked one of the run.do_compile scripts to create my own file for setting up the compiler paths etc and I just call make as usual on whatevr I am trying to build now.
  • [16:39:21] <atin_> now and then I have to modify the Makefile for a package if it somehow, but usually it just works.
  • [16:39:23] <erbo> atin_: got it working after a tip in #oe, apperently there's an issue using ccache in arch linux
  • [16:42:38] <atin_> huh. I seem to be able to use it. or at least, my CC= is set to ccache arm-angstrom-linux-gnueabi-gcc -march=armv7-a -mtune=cortex-a8 -mfpu=neon -mfloat-abi=softfp
  • [16:43:18] <atin_> basically I ran openembedded once successfully, then took one of the run.do_compile scripts and edited it
  • [16:43:29] <atin_> to create my own shell script that sets up the environment
  • [16:43:42] <atin_> so I can now just call make inside my own copies of sources
  • [16:43:55] <atin_> was just able to compile CWiid for the beagle that way
  • [16:44:13] <atin_> now my nintendo wiimote can be used as a mouse on the beagle
  • [16:44:22] <atin_> painfully, but it can
  • [16:44:29] <and-ri> lol
  • [16:44:30] <guillaum1> :-)
  • [16:44:31] <and-ri> cool
  • [16:45:02] * and-ri (n=ubuntu@pd95b4498.dip0.t-ipconnect.de) has left #beagle
  • [16:45:03] <erbo> i wrote some recipes at work for OE, the first one took quite a while before I started to understand how it worked
  • [16:45:20] <erbo> atin_: nice :)
  • [16:45:30] <Beagle3> anybody has done something on loading u-boot via USB
  • [16:45:41] <atin_> Beagle3 - yes, TI
  • [16:45:49] <atin_> let me get you the link that dirk gave me
  • [16:46:05] <atin_> http://lists.denx.de/pipermail/u-boot/2008-December/044896.html
  • [16:46:32] <atin_> follow the thread. the stuff isn't available yet, I was going to work on it but since they're already on it, I'm going to let them ;)
  • [16:47:00] <Beagle3> wait..let me check
  • [16:47:10] <atin_> http://lists.denx.de/pipermail/u-boot/2008-December/044905.html
  • [16:47:19] * DJWillis (n=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) has joined #beagle
  • [16:47:23] <atin_> they did it for davinci first
  • [16:47:41] <atin_> oh wait, you mean somethign else ;)
  • [16:47:52] <atin_> you mean loading u-boot via usb.
  • [16:48:20] <Beagle3> yesyes
  • [16:48:22] <atin_> sorry - that's not what that is about, that's about getting usb devices visible within u-boot.
  • [16:48:46] <atin_> there is a link on elinux about booting off usb
  • [16:48:51] <atin_> requires some magic
  • [16:49:00] <atin_> I've never attempted it.
  • [16:49:32] <Beagle3> i have tried TIomap3_usbload...but it cant detect the beagleboard
  • [16:49:40] <dirk2> http://code.google.com/p/omap-u-boot-utils/ http://nishanthmenon.blogspot.com/ http://groups.google.com/group/beagleboard/browse_thread/thread/c5bfb1b8ed528b52#
  • [16:50:17] <Beagle3> i have tired it with my ubuntu machine
  • [16:50:25] <Beagle3> not working
  • [16:50:28] <atin_> you're in some other purgatory than usual it seems.
  • [16:50:54] <atin_> did you try the images from sakoman?
  • [16:51:01] <Beagle3> i dont understand wat im experiencing now
  • [16:51:15] <Beagle3> im trying sakoman images
  • [16:51:58] <Beagle3> im trying to get the board boot from the MMC
  • [16:52:49] <atin_> yes, get that going first. the images from sakoman should at least get you that.
  • [16:53:05] <atin_> nand is for later :)
  • [16:53:31] <Beagle3> but it is quite tough to get it boot from MMC
  • [16:53:48] <atin_> tough how? I do it all the time ..
  • [16:53:58] <atin_> since I keep mucking in u-boot..
  • [16:54:16] <Beagle3> u hold the user button?
  • [16:54:25] <Beagle3> and keep the reset on and off
  • [16:54:27] <Beagle3> ?
  • [16:54:43] <atin_> I just powercycle the darn thing
  • [16:55:09] <atin_> while holding the reset button, then let go the button
  • [16:55:30] * LesterBurnham (n=luca@adsl-99-190-249-129.dsl.irvnca.sbcglobal.net) has joined #beagle
  • [16:56:26] <Beagle3> holding thereset button...then let it go?
  • [16:56:34] <atin_> yes.
  • [16:56:57] <Beagle3> doesnt work
  • [16:56:59] <atin_> I pull the power cable, then press the reset button, plug it back in, then let the button go
  • [16:57:36] <Beagle3> not working now
  • [16:57:46] <Beagle3> it is still booting from Nand
  • [16:58:13] <Beagle3> any other way
  • [16:58:54] <atin_> hmm - let me see. maybe that's not what I am doing anymore - haven't booted since friday ;)
  • [16:59:26] <Beagle3> my screen...showing Timed out in wait_for_bb:I2C_STAT=1000
  • [16:59:32] <Beagle3> Reading boot sector
  • [16:59:36] <Beagle3> not there
  • [16:59:41] <Beagle3> Botting from Nand
  • [16:59:55] <Beagle3> ECC Failed , page 0x00080000
  • [17:00:05] <Beagle3> eb f7 fb fe....
  • [17:00:36] <Beagle3> this is weird
  • [17:01:04] <adj_> Beagle3: could you pastebin the whole boot sequence log?
  • [17:02:01] <dirk2> Beagle3: Yes, please
  • [17:02:53] <Beagle3> Texas Instruments X-Loader 1.4.2 (Aug 8 2008 - 16:59:05) timed out in wait_for_bb: I2C_STAT=1000 timed out in wait_for_bb: I2C_STAT=1000 Reading boot sector not there Booting from nand ECC Failed, page 0x00080000 eb f7 fb fe c7 dd ff cc e7 ef f7 ff e5 cc 77 ff cf df bf fe fb fd 7b b7 ff fc 33 ff ff ce ff 3f f3 bd b3 ff ed ef 3f ce fd ee 7f da f3 cf 7f f6 77 df bf fe ff 5 d bb f1 df ce bb eb df ec 33 7f fb 4e 5b fd d3 fd 3a ff ff ee bb 7f ff df 3e 7c
  • [17:03:10] <atin_> http://pastebin.com
  • [17:03:32] <atin_> its very hard to read stuff like that here.
  • [17:04:37] <Beagle3> i havepaste it
  • [17:04:37] <atin_> weird, timed out wait_for_bb? i2c_stat timed out?
  • [17:04:47] <atin_> link?
  • [17:05:23] <atin_> click send after you paste it, then paste the url here ..
  • [17:05:53] <Beagle3> http://pastebin.com/m5b04a6eb
  • [17:06:55] <atin_> how does one force it to go to MMC anyway? holding user button isn't doing it for me either!
  • [17:07:02] <Beagle3> it is weird..that is the reason why i wanna erase the NAND via host machine
  • [17:08:21] <dirk2> atin_: Press user button, then switch/power on, then release user button
  • [17:08:33] <Beagle3> the manual said... Press and hold the user button while presing and relasing REST button
  • [17:09:00] <dirk2> Beagle3: Yes, user reset or power cycle
  • [17:09:13] <dirk2> s/user/use
  • [17:09:13] <atin_> does the u-boot stuff say it is booting from MMC if it does?
  • [17:09:34] <Beagle3> let me try it again
  • [17:09:37] <Beagle3> hold on
  • [17:10:18] <dirk2> atin_: Yes: http://pastebin.com/
  • [17:11:00] <atin_> hmm, this is what I get when I hold user-button, power board on, and then let go the button
  • [17:11:02] <atin_> com
  • [17:11:08] <atin_> http://pastebin.com/d1e725f6b
  • [17:11:30] <atin_> dirk2, you missed the code on the pastebin url :)
  • [17:12:06] <dirk2> atin_: Old X-Loader?
  • [17:12:19] <atin_> possible, I am building everything right now
  • [17:12:38] <atin_> will use what I get out of the build and burn that and hopefully not brick everything ;)
  • [17:12:54] <dirk2> atin_: sounds like you know how to unbrick ;)
  • [17:13:19] <Beagle3> let me have the files...if it ubbrick
  • [17:13:58] <Beagle3> i have been spending more than 6hrs on solving this prob
  • [17:14:02] <dirk2> atin_: Before building, you might try sakoman's x-loader and U-Boot images
  • [17:14:03] <atin_> I didn't put it in my wiki like an idiot, so it'll most probably entail re-figuring out how I did it last time :)
  • [17:14:28] <atin_> yeah, might as well do that while I wait
  • [17:14:59] * LesterBurnham (n=luca@adsl-99-190-249-129.dsl.irvnca.sbcglobal.net) Quit (Read error: 110 (Connection timed out))
  • [17:15:51] <dirk2> bbl, hope I find success in the logs when back ;)
  • [17:16:53] * ZeZu (n=null@c-98-227-56-17.hsd1.in.comcast.net) has joined #beagle
  • [17:17:01] * dirk2 (n=dirk@p5B043348.dip0.t-ipconnect.de) has left #beagle
  • [17:18:17] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [17:19:03] * JoeyBorn (n=jborn@dsl017-022-252.chi1.dsl.speakeasy.net) has joined #beagle
  • [17:20:40] * richardw (n=richardw@nat/ti/x-48988265f92055a6) has joined #beagle
  • [17:23:08] * eFfeM1 (n=Frans@195-241-226-180.ip.telfort.nl) has joined #beagle
  • [17:23:19] <Beagle3> http://pastebin.com/m7dd14a20
  • [17:23:32] <Beagle3> that is wat i get
  • [17:26:59] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [17:30:13] * uwe_ (n=uwe_@dslb-088-064-094-252.pools.arcor-ip.net) Quit (Read error: 60 (Operation timed out))
  • [17:30:39] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) Quit (Read error: 60 (Operation timed out))
  • [17:31:26] <Beagle3> atin?
  • [17:33:56] * uwe2 (n=uwe@dslb-088-064-075-087.pools.arcor-ip.net) has joined #beagle
  • [17:35:23] * eFfeM1 is now known as eFfeM
  • [17:36:35] <atin_> hi, yes - sec
  • [17:36:43] <Beagle3> damn it... it just wont read
  • [17:37:14] <Beagle3> the terminal shows......Texas Instruments Xloader 1.41
  • [17:37:17] <atin_> it is definitely trying MMC if I am reading the header correctly (Starting on with MMC)
  • [17:37:20] * mrc3__ (n=mrc3@nat/ti/x-b22b94a1d4e54e76) Quit (Remote closed the connection)
  • [17:37:25] <Beagle3> Starting on with MMC
  • [17:37:32] <Beagle3> Reeading boot sector
  • [17:37:38] <atin_> but the reading bootsector is failing - are you use you formated the card correctly?
  • [17:37:38] <Beagle3> then stopped
  • [17:38:06] <Beagle3> yes formatted FAT32
  • [17:38:10] <atin_> http://code.google.com/p/beagleboard/wiki/LinuxBootDiskFormat
  • [17:38:42] <Beagle3> i can try it with another card
  • [17:38:44] <Beagle3> wait
  • [17:40:09] * rupeshgujar2 (n=Administ@220.226.78.141) has joined #beagle
  • [17:41:00] <Beagle3> it is still stay the same
  • [17:41:15] <Beagle3> reading boot sector..then stopped
  • [17:41:28] <Beagle3> i think i need to bin it
  • [17:42:30] * jonnor (n=jon@ti0016a380-1975.bb.online.no) has left #beagle
  • [17:42:53] <Beagle3> it wont read the boot sector from the mmc
  • [17:43:28] <Beagle3> is there another way to erase the nand
  • [17:43:51] <atin_> dunno what to tell you - you say you are creating the mmc correctly?
  • [17:47:45] * uwe__ (n=uwe@dslb-088-064-094-252.pools.arcor-ip.net) Quit (Read error: 110 (Connection timed out))
  • [17:48:02] * alecrim (n=alecrim@189.2.128.130) has joined #beagle
  • [17:49:17] <Beagle3> yes
  • [17:49:27] <Beagle3> formatted to FAT32
  • [17:50:52] <atin_> I dunno, the message seems to say it can't read your boot sector, or the u-boot itself.
  • [17:50:57] <Beagle3> and storing files, MLO, x-load.bin.ift, flash-uboot, u-boot, rd-ext2 and uImage
  • [17:51:30] <Beagle3> i did the same the 1st day i received the board
  • [17:51:34] <Beagle3> for testing
  • [17:51:38] <Beagle3> it worked
  • [17:55:49] <Beagle3> i dont know how to fix it
  • [17:56:48] <atin_> I am trying out sakoman's latest MLO, x-load, u-boot
  • [17:56:51] <atin_> let me see what I get
  • [17:57:27] <atin_> (I was distracted because I had to get vnc working on my ubuntu box since I moved the monitor to the beagle)
  • [17:57:43] <Beagle3> ok
  • [18:05:41] * rupeshgujar2 (n=Administ@220.226.78.141) has left #beagle
  • [18:09:02] <atin_> sorry, taking a few - have to get my card to have all the right stuff, and I had to reboot my ubuntu machine too to get vnc going
  • [18:09:23] <Beagle3> mine is still stuck
  • [18:09:34] <Beagle3> stuck at Reading boot sector
  • [18:10:06] <Beagle3> with the latest uboot..it shows...REading boot sector not there loading u-boot.bin from nand
  • [18:10:24] <Beagle3> i think it is something wrong with my board
  • [18:12:22] <jkridner> I just found out about http://www.2009dsp.org/training.html by looking at the website analytics.
  • [18:12:34] <jkridner> seems like I should have known about BeagleBoard training.
  • [18:12:49] <jkridner> strange that people don't push for this stuff to be promoted on the BeagleBoard.org site.
  • [18:13:18] <jkridner> I don't bite. I don't even make people contact me to put content up there.
  • [18:13:33] <prpplague> jkridner: example of a disconnect between different areas of business
  • [18:14:07] * Beagle3 (n=Beagle3@137.205.147.151) Quit (Remote closed the connection)
  • [18:14:11] <jkridner> well, I thought that BeagleBoard.org would be a way to bring that together, but apparently it is still too difficult for people to update.
  • [18:15:24] <prpplague> jkridner: one of the first things we found at TinCanTools was that alot of people are completely clueless about usage of open community items like wikis and forums
  • [18:17:08] * uwe_ (n=uwe_@dslb-088-064-075-087.pools.arcor-ip.net) has joined #beagle
  • [18:20:27] * valhalla (n=valhalla@81-174-23-116.dynamic.ngi.it) has joined #beagle
  • [18:21:17] * Crofton (n=balister@pool-71-171-26-175.nwrknj.east.verizon.net) has joined #beagle
  • [18:23:36] <ds2> openid isn't that open either ;)
  • [18:24:37] <ds2> the log archives seems to be well used
  • [18:26:12] * robtow (n=rtow@64.62.142.114) has joined #beagle
  • [18:30:30] * guillaum1 (n=GZBA4143@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [18:32:13] * guillaum1 (n=gl@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) has joined #beagle
  • [18:38:44] * alecrim (n=alecrim@189.2.128.130) Quit (Remote closed the connection)
  • [18:41:08] <atin_> dang it! I bricked my bb now.
  • [18:41:29] <atin_> and its very hosed ... just 40T
  • [18:41:33] <atin_> and nothing at all.
  • [18:41:59] <tomba> you can't really brick it, as you can always boot omap via serial or usb
  • [18:42:01] <atin_> nand was erased, and so I'm trying to boot off mmc.
  • [18:42:11] * dcordes (n=dcordes@unaffiliated/dcordes) has joined #beagle
  • [18:42:17] <atin_> but pressing the user button doesn't seem to try the mmc
  • [18:48:59] <sakoman> atin_: are you positive you have a good mmc image? the boot loader is very picky
  • [18:49:13] <atin_> can someone tell me the exact steps to make the bb not go to nand but go to the mmc instead?
  • [18:49:40] <atin_> I had used your images :) - but I can try again. :)
  • [18:49:54] <sakoman> I mean card formatting
  • [18:51:15] <atin_> yeah - that's the one I've been using. But now I'm going to do everything from scratch to make sure!
  • [18:51:18] <sakoman> if partitioning, file copy order, etc are not right you will see the symptom you are seeing
  • [18:51:33] * dcordes is now known as dcordes__
  • [18:52:33] <sakoman> atin_: the overo instructions will work for beagle too: http://www.gumstix.net/Overo/view/Overo/Creating-a-bootable-microSD-card/115.html
  • [18:52:52] <atin_> ok, let me make certain now.
  • [18:54:14] * dcordes__ is now known as dcordes
  • [18:56:26] <JayFoxRox> if you have don't have a linux machine the HP storage tool should also work for formating on windows, thats easier and faster in that case
  • [18:56:51] <JayFoxRox> however, once you want angstrom you will need the second partition which windows won't do for you ;)
  • [18:57:13] <atin_> I run only linux :)
  • [18:58:19] * zedstar (n=john@fsf/member/zedstar) has joined #beagle
  • [18:58:51] * blade45 (n=chatzill@modemcable007.76-56-74.mc.videotron.ca) has joined #beagle
  • [18:59:11] <blade45> did someone play with arm implementation of openMax DL
  • [18:59:13] <blade45> ?
  • [19:01:41] <JayFoxRox> [bitch]how can TI be so slow with the SGX driver? I mailed them yesterday, I would have liked an reply today. I m used to 3 hours waiting time with other companies.. I m a paying customer, they should show some respect[/bitch]
  • [19:02:28] <blade45> JayFoxRox: the SGX driver is ready ?
  • [19:02:41] <JayFoxRox> http://groups.google.com/group/beagleboard/browse_thread/thread/44a0c0f5354afaab/7a9a5c78f6a8068f?lnk=gst&q=SGX#7a9a5c78f6a8068f
  • [19:03:08] <JayFoxRox> seems so, koen has a package ready on his space. but you have to request the userspace tools to start them
  • [19:03:23] <blade45> i see
  • [19:03:39] <blade45> what userspace tools are we talking again ?
  • [19:03:58] <JayFoxRox> I was let down enough because I thought they were ready when I ordered, it was used in "promo." videos. I was really confused when I heard that they were not ready yet
  • [19:04:15] <JayFoxRox> blade45: ?
  • [19:04:38] <blade45> JayFoxRox: you said some usertools space tool are needed to start them
  • [19:04:59] <blade45> what king of userspace tool ?
  • [19:05:03] <JayFoxRox> yeh, check the link I just mentioned
  • [19:05:25] * beewoolie (n=beewooli@66.147.194.238) has joined #beagle
  • [19:05:49] <blade45> i read it , i just dont understand the role of those userspace tools...
  • [19:05:55] <blade45> "to init the driver "??
  • [19:06:07] <JayFoxRox> I assume the driver is in the kernel but you have to activate it
  • [19:06:32] <JayFoxRox> otherwise the chip won't do anything. which means that only having the right kernel gets you nowhere
  • [19:07:05] * JimDonova1 (n=angstrom@c-66-30-221-249.hsd1.ma.comcast.net) has joined #beagle
  • [19:07:19] <blade45> but since the driver ssource code are available
  • [19:07:25] <blade45> cant we just recompile them ?
  • [19:07:51] <JayFoxRox> according to the mailinglist entry only the kernel driver is open source - the userspace tools aren'T
  • [19:08:03] <JayFoxRox> you have to request them manually
  • [19:09:02] <blade45> JayFoxRox: my point is (and i am not a specialist) since the source are open , it should be pretty easy to compile a version that dont neeed the userspace tools to run
  • [19:10:11] <JayFoxRox> well... true :P I m not an expert or anything (more the opposite) so I m not sure, i can only say what I have heard so far.
  • [19:11:01] * JimDonova1 (n=angstrom@c-66-30-221-249.hsd1.ma.comcast.net) has left #beagle
  • [19:11:02] <JayFoxRox> but maybe the kernel driver only forwards something from the userspace tools which activates the chip - that would make sense. So even if the kernel is open, you don't know which data you have to pass (until you debug it)
  • [19:12:58] <blade45> JayFoxRox: i see
  • [19:13:08] <blade45> do they plan on free the userspace tool code also ?
  • [19:14:12] <atin_> sakoman: ping?
  • [19:14:29] <JayFoxRox> I have no idea :/ . From what I have heard on the pandora forums the hardware distributor has to provide the drivers and not ImgTec. I think its TIs job to share them. But the pandora team said that they will publish the hardware SGX ready at some point. So I would assume that we will find out soon
  • [19:14:30] <atin_> sakoman: I finally built the card all over again using your latest images
  • [19:14:31] <sakoman> atin_: pong
  • [19:14:33] <dcordes> koen: how's the ti projector?
  • [19:14:52] <atin_> and it booted - gives a *** Warning - bad CRC or NAND, using default env
  • [19:14:56] <atin_> message, but then goes on to boot
  • [19:14:59] <atin_> what is that warning?
  • [19:15:09] <blade45> JayFoxRox: thanks for the info
  • [19:15:10] <koen> dcordes: somewhere between utrect and here, I suspect DHL will drop it off tomorrow
  • [19:15:15] <atin_> is that because my nand is erased right now?
  • [19:15:18] <sakoman> perfect, that is what it should do. your nand env partition is empty
  • [19:15:33] <sakoman> just do a saveenv and you won't get that anymore
  • [19:15:37] <atin_> great, can I just write the u-boot image sitting on my mmc into nand now?
  • [19:15:49] <atin_> the one that booted?
  • [19:16:44] <sakoman> if you want to. if you will be running off mmc, there is no real reason to do so
  • [19:17:33] <sakoman> if you decide to do so, I recommend using a method like this: http://www.gumstix.net/Overo/view/Overo/Writing-images-to-onboard-nand/115.html
  • [19:17:41] <JayFoxRox> blade45: don't quote me on that :P
  • [19:17:59] * guillaum1 (n=gl@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [19:18:03] <Crofton|work> anyone know what would be a good voltage input range if you were making something to run from batteries
  • [19:18:09] <atin_> ok, great.
  • [19:18:12] <blade45> JayFoxRox: lol
  • [19:18:13] <Crofton|work> high end 6 votls from small lead-acid
  • [19:18:17] <atin_> Beagle3: ping?
  • [19:18:23] <Crofton|work> what would low input be>
  • [19:18:24] * guillaum1 (n=Guillaum@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) has joined #beagle
  • [19:18:24] <Crofton|work> ?
  • [19:18:31] <blade45> i am still wayting for rev c boards
  • [19:18:43] <atin_> I was able to get my mmc to boot - you really have to be quite particular about how you make the card.
  • [19:18:56] <Crofton|work> NiMH 4*1.2 = 4.8 -> 4-6 volts?
  • [19:19:05] <Crofton|work> what about Li ION
  • [19:19:32] <blade45> someone played with arm implementation of openMax DL ???
  • [19:19:43] <prpplague> Crofton|work: depends on the voltages you need to generate
  • [19:19:49] <sakoman> Crofton|work: you really need to look at the specs of the batteries you will be using
  • [19:19:58] <Crofton|work> heh
  • [19:20:08] <atin_> sakoman, if I don't write to nand, do I always have to press the user button to make it boot off mmc?
  • [19:20:08] <sakoman> they will tell you what you want to know
  • [19:20:14] <prpplague> Crofton|work: if you need to have +5v on the device, you really should use something slighly higher in voltage
  • [19:20:17] <Crofton|work> to many specifics, I am thinking the "general" case
  • [19:20:20] <sakoman> it varies dramatically with battery technology
  • [19:20:38] <prpplague> Crofton|work: using a buck/boost can get expensive for power consumption
  • [19:21:00] <atin_> I must ... there is nothing else to tell the processor where to go I guess.
  • [19:21:03] <dcordes> koen: ok good luck then. I didn't request one. they wouldn't give me one anyway
  • [19:21:16] <Crofton|work> so basically, there is no magic bullet :)
  • [19:21:17] <prpplague> Crofton|work: generally speaking, you have two ranges, around 1.8 to 5.5v and then you have +6v and up
  • [19:21:20] <sakoman> going down to 4V would not be good for the life of a 6v lead-acid, but would likely be fine with other technologies
  • [19:21:28] <Crofton|work> yeah
  • [19:21:48] <prpplague> Crofton|work: if you look at some of the arduino boards, you can get an idea of the easy ranges to work with
  • [19:21:56] <Crofton|work> I'm not worried about battery life management, just being able to use a battery pack for test runs
  • [19:22:15] <prpplague> Crofton|work: for the beagle?
  • [19:22:27] * Crofton|work has seen pictures of PHD's towing little red wagons around parking lots with laptops and radios in them
  • [19:22:40] <Crofton|work> prpplague, basically, maybe Overo
  • [19:22:40] <sakoman> atin_: write x-load to nand and then you won't need to press user
  • [19:23:07] <Crofton|work> overo exposes GPMC controller
  • [19:23:15] <prpplague> Crofton|work: ahh, a two cell series li-ion would probably be the best bet
  • [19:23:15] <atin_> ok, sounds good.
  • [19:23:24] <sakoman> atin_: my x-load looks for u-boot on mmc first, then nand if there isn't one there
  • [19:23:42] <atin_> the one on your feeds does that?
  • [19:23:43] <Crofton|work> the problem is there would be an fpga involved with its own needs
  • [19:23:49] <atin_> or is that a private version?
  • [19:23:50] <sakoman> atin_: yes
  • [19:23:59] <atin_> cool, I'll write that one then
  • [19:23:59] <sakoman> no the one from my site
  • [19:24:17] <atin_> once gnome finishes init! :)
  • [19:24:24] <prpplague> Crofton|work: what is the total current for the board ?
  • [19:24:37] <Crofton|work> not sure at the moment
  • [19:25:06] <Crofton|work> basically, it sounds like the battery task will be done off board :)
  • [19:25:17] <atin_> then I have to go to radio-shack to get some infrared leds to make the wiimote whiteboard hack work on the beagle
  • [19:25:59] <prpplague> Crofton|work: need to settle on an input voltage, best guess is probably you want to make it easy and provide +5v to the beagle
  • [19:26:27] <prpplague> Crofton|work: just a rough estimate, you could go with something like 1A as a starting point
  • [19:26:43] <Crofton|work> yes
  • [19:27:03] <sakoman> atin_: you should write 4 copies of x-load in case a block or two goes bad
  • [19:27:16] <sakoman> I use a script like this:
  • [19:27:23] <prpplague> Crofton|work: http://www.batteryspace.com/index.asp?PageAction=VIEWPROD&ProdID=4742
  • [19:27:24] <sakoman> http://pastebin.com/m1135bdc0
  • [19:27:32] <prpplague> Crofton|work: that would last you about 2.5 hours
  • [19:27:42] <atin_> cool, ok :)
  • [19:28:25] <prpplague> Crofton|work: there are a number of inexpensive camcorder batteries with basically the same specs
  • [19:33:21] <prpplague> Crofton|work: this one is pricy directly from duracell, but you can find a compatible no-name brand at most of the local electronics shops - http://www.duracelldirect.com/sony_0_oem-battery_camcorder_b-9547.aspx?GCID=S17226x015&KEYWORD=NP-FM50&equiv=(%20NP-FM50%20)&gclid=CMPGprX71JcCFQrAGgod2CXvDg
  • [19:33:32] <prpplague> Crofton|work: model is NP-FM50
  • [19:33:39] <prpplague> Crofton|work: sony camcorder battery
  • [19:34:10] <Crofton|work> yeah, I'm thinking about the general case
  • [19:34:29] <Crofton|work> for example the Beagle uses 5 volts, with a limited range
  • [19:34:31] <prpplague> Crofton|work: then i'd stick with a battery pack similiar to the one from batteryspace.com
  • [19:35:02] <prpplague> Crofton|work: correct, you'll need to use an LDO from the battery to provide the regulated +5v
  • [19:35:06] <Crofton|work> I was hoping it was "easy" to change the power regulators to get more input voltage range
  • [19:35:32] <Crofton|work> but it sounds like that battery adaptation is best doe off board
  • [19:35:34] <prpplague> Crofton|work: i wouldn't bother doing that unless you are very serious about power savings
  • [19:35:43] <prpplague> Crofton|work: yea
  • [19:35:44] <Crofton|work> and really needs to account for the battery chemistry
  • [19:35:52] <prpplague> Crofton|work: yep
  • [19:36:18] <Crofton|work> so no "easy" solution for a general purpose board
  • [19:36:40] <Crofton|work> it seems like so many people end up with 12 volt lead-acid and power inverters feeding wall warts :)
  • [19:36:50] <prpplague> Crofton|work: its pretty easy, actually, iirc spark has a regulator board
  • [19:36:53] * prpplague goes to look
  • [19:39:51] * eFfeM (n=Frans@195-241-226-180.ip.telfort.nl) has left #beagle
  • [19:41:51] <Crofton|work> prpplague, they have some convertors, but not enough current
  • [19:41:52] <gregoiregentil> Koen: Ping?
  • [19:41:58] <koen> gregoiregentil: pong
  • [19:42:30] <gregoiregentil> Koen: I think that this one is not working: http://gitweb.openembedded.net/?p=openembedded.git;a=commitdiff;h=73b85942747e9f560149116b208b11eb9a7951ef
  • [19:43:07] <koen> the commit or is uboot not working?
  • [19:43:25] <Crofton|work> http://www.opensdr.com/files/sym-08/troxel-adroit-vt-20080604.pdf
  • [19:43:26] <prpplague> Crofton|work: http://www.sparkfun.com/commerce/product_info.php?products_id=114
  • [19:43:26] <gregoiregentil> the recipe can't fetch on my system
  • [19:43:44] <Crofton|work> there are a couple of photos that are amusing in this presentation :)
  • [19:43:49] <Crofton|work> sorry about the size ...
  • [19:43:52] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) has joined #beagle
  • [19:43:55] <koen> gregoiregentil: upgrade to bitbake 1.8.12
  • [19:44:45] <gregoiregentil> Koen: I have 1.8.10. I will and then try again
  • [19:44:55] <gregoiregentil> Koen: Thanks for (1), (2), (3)
  • [19:45:05] <gregoiregentil> Koen: what about (5)?
  • [19:45:42] <koen> I can't remember what (5) was, sorry
  • [19:46:11] <gregoiregentil> Koen: fifthly, I'm also trying to add abiword-plugins, which compiles fine but generates an error during rootfs "log.do_rootfs.691: * Cannot find package abiword-plugins.". In the ipk folder, there are a bunch of abiword-plugins*.ipk but no abiword-plugins. Have you tried to run this recipe?
  • [19:46:36] <koen> ah right
  • [19:46:56] <koen> the recipe doesn't generate a 'abiword-plugins' packages, only packages for the plugins
  • [19:47:25] <gregoiregentil> yes. which compiles fine but then fails during rootfs
  • [19:47:46] <koen> 'abiword-plugins' is a recipe name, not a package name
  • [19:48:32] <koen> gregoiregentil: the recipe needs a populate_packages_prepend like this one: http://cgit.openembedded.net/cgit.cgi?url=openembedded/tree/packages/xscreensaver/xscreensaver.inc
  • [19:48:45] <prpplague> Crofton|work: that one should work dandy for you
  • [19:49:32] <Crofton|work> yeah
  • [19:49:53] <Crofton|work> the other approach is use a pico-psu with a 12 volt battery
  • [19:50:25] * Beagle2 (n=Beagle2@91-115-158-248.adsl.highway.telekom.at) has joined #beagle
  • [19:50:29] <Crofton|work> the problem is there are to many solutions
  • [19:51:36] <prpplague> Crofton|work: personally i'd recommed 2-cell li-ion at 7.4v , easily rechargable, good size, and plenty of voltage to step down to a good solid +5v
  • [19:51:45] <prpplague> Crofton|work: just my $0.02
  • [19:51:48] * Beagle8 (n=Beagle8@91-114-255-46.adsl.highway.telekom.at) has joined #beagle
  • [19:54:17] <Beagle2> hi, somebody here, who knows when the first C-Revision of the Beagleboard will be available?
  • [19:55:29] <prpplague> Beagle2: rumor has it will be available at the end of march
  • [19:55:36] <prpplague> Beagle2: or atleast last time i heard
  • [19:58:57] <Beagle2> thank you, this is soon enough for me
  • [20:00:19] <Beagle2> So I guess there is no B7 revision planned
  • [20:02:57] * Beagle2 (n=Beagle2@91-115-158-248.adsl.highway.telekom.at) Quit ()
  • [20:24:20] * valhalla (n=valhalla@81-174-23-116.dynamic.ngi.it) Quit ("Leaving")
  • [20:25:01] * Crofton (n=balister@pool-71-171-26-175.nwrknj.east.verizon.net) Quit (Read error: 113 (No route to host))
  • [20:32:55] * Beagle2 (n=Beagle2@91-114-255-46.adsl.highway.telekom.at) has joined #beagle
  • [20:33:45] * Beagle2 (n=Beagle2@91-114-255-46.adsl.highway.telekom.at) Quit (Client Quit)
  • [20:35:47] * Raja (n=Raja@nat/ti/x-1c55f0cbb0650504) has joined #beagle
  • [20:36:05] * RajaK (n=Raja@nat/ti/x-add26033935537f4) has joined #beagle
  • [20:38:09] * RajaK (n=Raja@nat/ti/x-add26033935537f4) Quit (Client Quit)
  • [20:39:33] * Wowbagger_ (n=wowbagge@d154-20-138-76.bchsia.telus.net) Quit ("Leaving")
  • [20:39:41] * prpplague (n=dave@mail.americanmicrosystems.com) Quit (Nick collision from services.)
  • [20:40:15] * prpplague^2 (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [20:50:58] * Beagle8 (n=Beagle8@91-114-255-46.adsl.highway.telekom.at) Quit (Read error: 110 (Connection timed out))
  • [20:53:52] * Raja (n=Raja@nat/ti/x-1c55f0cbb0650504) Quit ("Leaving")
  • [20:54:01] * Raja (n=Raja@nat/ti/x-8ca7391dd8bbe101) has joined #beagle
  • [20:56:58] * blade45 (n=chatzill@modemcable007.76-56-74.mc.videotron.ca) Quit (Read error: 110 (Connection timed out))
  • [21:22:48] * flo_lap (n=fuchs@f054160083.adsl.alicedsl.de) has joined #beagle
  • [21:23:14] * flo_lap is now known as florian
  • [21:32:56] * DJW|Home (i=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) has joined #beagle
  • [21:32:58] * DJWillis (n=djwillis@82-46-19-72.cable.ubr02.bath.blueyonder.co.uk) Quit (Read error: 104 (Connection reset by peer))
  • [21:33:41] * raster (n=raster@r125-63-180-233.cpe.unwired.net.au) has joined #beagle
  • [21:35:13] * calculus (n=calculus@gentoo/user/calculus) Quit (Read error: 110 (Connection timed out))
  • [21:44:33] * likewise (n=chatzill@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [21:45:39] <likewise> gm all
  • [21:52:06] <florian> hi likewise
  • [21:52:14] <likewise> hi florian
  • [21:52:18] <Crofton|work> gm
  • [21:53:12] <likewise> hi there
  • [21:53:19] <raster> moo
  • [21:56:24] <likewise> dingo's don't do moo, but you do? :-)
  • [21:57:37] <Crofton|work> a dingo ate my baby
  • [21:58:31] <emeb> foetid dingo's kidney
  • [21:58:52] * atin_ (n=atin@c-71-232-7-71.hsd1.ma.comcast.net) has left #beagle
  • [21:59:02] <likewise> first victims disconnect :-)
  • [21:59:19] <likewise> How're your beagles coming along?
  • [22:00:05] <florian> Crofton|work: !
  • [22:00:29] * calculus (n=calculus@gentoo/user/calculus) has joined #beagle
  • [22:43:49] * philippe (n=fille@a83-245-252-47.elisa-laajakaista.fi) Quit (Read error: 60 (Operation timed out))
  • [22:46:36] * Beagle4 (n=Beagle4@cpc1-oxfd6-0-0-cust230.oxfd.cable.ntl.com) has joined #beagle
  • [22:46:53] <Beagle4> hi
  • [22:47:25] <Beagle4> anybody know where can i purchase the beagleboard in uk?
  • [22:48:28] <ghaz``> order it from digikey. afaik they also ship to uk.
  • [22:49:20] <Beagle4> ok
  • [22:54:49] <Beagle4> do u know how much is it in pound?
  • [22:57:09] <ldesnogu> Beagle4, http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail&name=296-23428-ND
  • [22:57:16] <ldesnogu> 108 pounds
  • [22:58:05] <Beagle4> thx
  • [22:59:27] <Beagle4> it is getting more expensive
  • [23:00:21] * abitos (n=nixgibts@p5B2E7C4F.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
  • [23:00:25] <ldesnogu> you mean the pound is crashing down? :)
  • [23:01:22] <ldesnogu> French price has increased slightly too
  • [23:01:40] <ldesnogu> so it's the dollar that has stopped being close to null :)
  • [23:04:42] * JuanG (n=Juan@nat/ti/x-bfdc793d5a7b2af2) has joined #beagle
  • [23:04:48] * JuanG (n=Juan@nat/ti/x-bfdc793d5a7b2af2) has left #beagle
  • [23:05:28] * rsalveti (n=salveti@200.184.118.130) Quit (Remote closed the connection)
  • [23:05:43] <Beagle4> yes
  • [23:06:07] <maelcum_> has anybody else had problems with the screen program in debian lenny on the beagle board?
  • [23:06:07] <Beagle4> i should have bought it during the summer
  • [23:06:41] <maelcum_> the session dies with the funny error message "Suddenly the Dungeon collapses!! - You die..."
  • [23:07:19] * Soopaloo (n=jacques@87-194-189-239.bethere.co.uk) has joined #beagle
  • [23:07:22] <ldesnogu> this looks like some MUD sentence :)
  • [23:07:56] <Soopaloo> does anyone know how to get a ce bsp (demo or otherwise) without having to fill in sales projection of 10000 units/month
  • [23:08:17] <maelcum_> when running a distributed compile of Qt the compiler instances on other machines die sometimes with a segfault and no core file or anything. when i run the compile only on the beagle board the whole screen session dies with the compiler(???)
  • [23:08:41] * raster (n=raster@r125-63-180-233.cpe.unwired.net.au) Quit ("Gettin' stinky!")
  • [23:08:42] <maelcum_> i don't know about the compiler dying because, well, the screen session's dead
  • [23:09:01] * Beagle4 (n=Beagle4@cpc1-oxfd6-0-0-cust230.oxfd.cable.ntl.com) Quit ()
  • [23:09:57] <maelcum_> duh, i should take icecream out of the path if i want core files [i've already removed the ulimit on core files]
  • [23:14:20] <ldesnogu> bbl
  • [23:14:24] * ldesnogu (n=ldesnogu@ven06-2-82-247-86-183.fbx.proxad.net) Quit ()
  • [23:15:19] <vlad_> because I am lazy, anyone have a root fs/kernel image for the android build?
  • [23:20:32] <JayFoxRox> great, still no email by TI :(
  • [23:21:52] * cbrake is now known as cbrake_away
  • [23:28:28] * richardw (n=richardw@nat/ti/x-48988265f92055a6) Quit ("Leaving")
  • [23:29:19] * chelli (n=chelli@debian/developer/tschmidt) Quit ("IRC is just multiplayer notepad")
  • [23:31:07] * dcramer_ (n=davec@dcdsl.ebox.com) has joined #beagle
  • [23:32:14] * dcramer (n=davec@dcdsl.ebox.com) Quit (Read error: 110 (Connection timed out))
  • [23:32:52] * likewise (n=chatzill@82-171-51-231.ip.telfort.nl) Quit ("ChatZilla 0.9.84 [Firefox 3.0.4/2008111317]")
  • [23:33:10] * maelcum_ is now known as maelcum
  • [23:33:15] * NishanthMenon (n=nmenon@nat/ti/x-f344fc59b10c59ba) Quit ("Leaving")
  • [23:37:49] * JayFoxRox (n=opera@a89-183-18-130.net-htp.de) has left #beagle
  • [23:49:56] * guillaum1 (n=Guillaum@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) Quit ("Leaving.")
  • [23:50:12] * guillaum1 (n=gl@AMontsouris-153-1-18-45.w86-212.abo.wanadoo.fr) has joined #beagle