• [00:00:07] <ds2> 128M is plenty
  • [00:00:18] <ds2> stop writing bloated code :P
  • [00:00:23] <mru> indeed
  • [00:00:42] <mru> at work the 128MB boxes feel quite roomy
  • [00:00:44] <Ravi> I assume there's no way a Java-written server software of any kind can run on this?
  • [00:00:51] <Ravi> in the Jabber space that is
  • [00:00:55] <mru> java? are you mad?
  • [00:01:01] <Ravi> mru: yes :)
  • [00:01:02] <ds2> java can run
  • [00:01:11] <ds2> but then I don't recommend anyone use java, EVER
  • [00:01:11] <mru> well, crawl more likely...
  • [00:01:12] <Ravi> ds2: I assumed as much
  • [00:01:28] <Ravi> ok... I have alternatives which are written in lisp
  • [00:01:40] <mru> lisp is at least fun
  • [00:01:46] <mru> even if it's a bit slow sometimes
  • [00:01:57] <Ravi> C
  • [00:01:58] <mru> java is both painful and slow
  • [00:02:06] <mru> assembler is both FUN and FAST
  • [00:02:23] <Ravi> if I can get that C version to compile, it probably would be good enough
  • [00:03:36] <mru> C is ok
  • [00:03:57] <Ravi> I assume berkeley db is ok to run...
  • [00:04:47] <Ravi> what do most of you guys use to get programs up and running on this board?
  • [00:05:03] <Ravi> do you take the angst of installing an actual OS on this thing?
  • [00:05:10] <Ravi> write from scratch in ASM?
  • [00:05:47] <mru> linux here
  • [00:05:53] <Ravi> mru: what flavour?
  • [00:05:56] <mru> gentoo
  • [00:06:02] <mru> cross-compiled of course
  • [00:06:12] <Ravi> mru: cool, and you run your programs on top of that?
  • [00:06:16] * calculus cheers for mru
  • [00:06:39] <Ravi> python seems to have good support from angstrom
  • [00:06:45] <Ravi> anyone try angstrom?
  • [00:07:01] * mru just finished a 100-line assembler function, phew
  • [00:07:12] <mru> lots of people run angstrom
  • [00:07:18] <Ravi> so you run that on top of gentoo?
  • [00:07:25] <mru> or ??ngstr??m as it's properly written
  • [00:07:40] <Ravi> not angstrom, I mean your assembler
  • [00:07:49] <mru> yes
  • [00:07:52] <mru> it's part of ffmpeg
  • [00:08:33] <Ravi> mru: I feel bad for you - I hear drivers for the video unit on this thing haven't been fully released?
  • [00:08:47] <mru> that's the 3d accelerator
  • [00:08:52] <mru> I don't need that
  • [00:08:58] <Ravi> mru: ah, ok
  • [00:09:08] <Ravi> mru: doesn't any of possible post-processing steps use that?
  • [00:09:36] <mru> nobody knows what's inside the powervr
  • [00:09:42] <mru> maybe it could be used for video processing
  • [00:09:49] <asterick> The OMAP3 doesn't have an MCU, does it?
  • [00:10:00] <mru> mcu?
  • [00:10:02] <Ravi> mru: I would think so - when are they gonna release this stuff? never?
  • [00:10:06] <asterick> memory control unit.
  • [00:10:17] <mru> it has an mmu, if that's what you mean
  • [00:10:19] <asterick> used for mapping logical memory addresses to physical ones.
  • [00:10:19] <Ravi> asterick: mmu
  • [00:10:25] <asterick> ahh, sweet.
  • [00:10:31] <Ravi> asterick: of course it does
  • [00:10:32] <mru> think of it as a small pc
  • [00:11:09] <Ravi> amazing power footprint - how much power consumption do you guys see @ peaks?
  • [00:11:11] <mru> it's basically your high-end pc of 10 years ago
  • [00:11:22] <mru> I haven't measured the power
  • [00:11:28] <mru> it should be a few watts at most
  • [00:11:41] <Ravi> mru: so it's like less than a bulb.. amazing
  • [00:11:42] <asterick> mru: I just wasn't sure if it had a mmu. Lots of embedded chipsets forgo them, even though they are essential for real platform security.
  • [00:12:07] <mru> that's why I said think of it as a pc, not an embedded processor
  • [00:12:17] <asterick> I've been pretty impressed with it's capibilities so far.
  • [00:13:11] <asterick> I was especially happy to see it uses the PowerVR. deferred rendering makes me happy.
  • [00:13:24] <Ravi> but that stuff isn't complete asterick
  • [00:13:37] <asterick> software will be complete eventually. ;)
  • [00:13:44] <Ravi> I want it now :)
  • [00:13:45] <mru> software is never complete
  • [00:13:57] <asterick> mru: 'functional' work for ya? :D
  • [00:14:13] <Ravi> I would love to see full 3d out of this thing
  • [00:14:36] <Ravi> can be handy for live visualizations based on data using DVI out
  • [00:14:39] <mru> we've had 3d demos running at 720p resolution
  • [00:15:05] <Ravi> mru: how do you guys pull that off if the manufacturer of the video unit hasn't released the drivers?
  • [00:15:12] <Ravi> mru: for the 3d stuff at least
  • [00:15:36] <mru> beta drivers are available under nda
  • [00:15:44] <mru> at least to the likes of TI
  • [00:16:43] <Ravi> mru: :( assume it doesn't get out of TI though, right?
  • [00:17:28] <mru> we have them at work
  • [00:17:41] <Ravi> mru: where do you work?
  • [00:17:44] <mru> not TI
  • [00:17:53] <Ravi> mru: ah, ok
  • [00:18:30] <Ravi> anyone know how the performance for python is?
  • [00:18:58] <mru> poor, but that's the case wherever you run it
  • [00:18:58] <Ravi> I see that Angstrom has plenty of packages available: http://www.angstrom-distribution.org/repo/?action=section&section=devel%2Fpython
  • [00:19:31] <Ravi> mru: I'm not so concerned about speed as I am about hitting swap
  • [00:19:39] <Ravi> how is the memory consumption looking for Python?
  • [00:20:00] * mru likes dogs better than snakes
  • [00:20:04] <mru> snakes eat dogs
  • [00:20:22] <Ravi> :)
  • [00:21:12] <asterick> Ravi: are we talking relative performance or on the OMAP specifically
  • [00:21:24] <Ravi> asterick: OMAP
  • [00:21:28] <asterick> Ah, no idea. xD
  • [00:21:37] <Ravi> I know about the relative performance of course
  • [00:21:41] <asterick> I'll be happy when someone ports Mono. ;)
  • [00:21:52] * mru runs and hides
  • [00:22:01] <asterick> from mono?
  • [00:22:22] <asterick> don't hide from the mono, it loves you
  • [00:23:04] <asterick> Holy crap it builds on ARM. <3
  • [00:23:23] <asterick> nvm then
  • [00:24:22] <mru> but can it do 8 multiplications per clock cycle?
  • [00:24:41] <Ravi> asterick: it's already on this board: http://www.angstrom-distribution.org/repo/?action=section&section=devel%2Fmono
  • [00:25:01] <asterick> huzzah!
  • [00:25:15] <Ravi> mru: with a good enough optimizer, a compiler can be pretty damn good mru
  • [00:25:29] <mru> I'll believe that when I see it
  • [00:25:57] <Ravi> mru: then again, do we really need it in places where you're serving data out through the network though mru?
  • [00:26:12] <mru> it depends on where the bottleneck is
  • [00:26:29] <mru> for the stuff I'm doing, every cycle counts
  • [00:26:35] <Ravi> mru: exactly, in most embedded scenarios - it does count
  • [00:26:49] <mru> I'm talking about HD video here
  • [00:27:13] <Ravi> mru: oh, ok - this board can do 720p right?
  • [00:27:22] <Ravi> mru: no 1080p possible I bet?
  • [00:27:34] <asterick> Ravi: pixelclock isn't fast enough
  • [00:27:38] <mru> 1080p is definitely pushing it
  • [00:27:47] <mru> the pixel clock is enough for 1080p30
  • [00:27:52] <asterick> true.
  • [00:28:34] * rsalveti (n=salveti@189.70.119.44) Quit (Read error: 104 (Connection reset by peer))
  • [00:29:51] <Ravi> is there any kind of resource/guide out there that gives an intro to development on this board?
  • [00:30:08] <mru> for general development, it's not that special
  • [00:30:25] * bazbell (n=a0192809@nat/ti/x-adce310f1c61fbfa) Quit (Remote closed the connection)
  • [00:30:37] <Ravi> I'm really looking to do visualizations based on network data
  • [00:31:20] * rsalveti (n=salveti@189.70.119.44) has joined #beagle
  • [00:32:29] <banderson> The previous asm/c/java/python conversation reminds me of this comic http://www.xkcd.com/378/
  • [00:33:31] <mru> oh yes, that one
  • [00:33:49] <NishanthM> banderson: that is my fav xckd comic -> i kinda used that to tell folks about coding standards ;)
  • [00:36:33] <NishanthM> Ravi: for network stuff.. on beagleboard -> you could do either: a) ethernet over usb (RNDIS) b) have a usb2network adapter c)or have a usbmodem.. OR if you are looking for network chip like smc91xx integrated, look at either zoom mdk, evm or sdp
  • [00:38:28] <Ravi> NishanthM: can you elaborate on (a) while I google?
  • [00:39:39] <Ravi> very nice... is RNDIS thing (connecting to networks through ethernet supported on any of the OS's running on beagleboard?
  • [00:39:42] <NishanthM> beagleboard does not have network chip by default. it does have a usbport and can act as a network device to a host pc.. so you can do usbnet network over usb RNDIS gadget.. and u could even do an nfs mount from ur host pc to board
  • [00:40:26] <NishanthM> Ravi: Yes, many folks use RNDIS on beagleboard
  • [00:41:27] <Ravi> NishanthM: oh, I think I'm misunderstanding you - can you connect to other networks (or the internet) through RNDIS?
  • [00:42:22] <NishanthM> RNDIS allows u to plug to the host. now the host can route the packets coming in from the board to the network of it's choice (just like a router).. is usually easy for linux box.. no clue how to do it on windows though..
  • [00:44:07] <Ravi> NishanthM: got it, thanks
  • [00:44:47] <Ravi> most usb hubs just work on this thing right guys?
  • [00:45:35] <NishanthM> it should.. but there have been issues with stability of usb for sometime.. probably others could comment.. i have been off-grid for a while now..
  • [00:46:41] <Ravi> anyone? is the problem power?
  • [00:46:41] <banderson> Ravi: I think all they suggest is that it is powered...(maybe it has to be but don't remember)
  • [00:47:12] <Ravi> banderson: got it, that makes sense though
  • [00:47:38] <banderson> Ravi: I bought the one that is one the linked digikey paged and it seems good.
  • [00:47:52] <Ravi> banderson: it's powered?
  • [00:47:55] <banderson> yes
  • [00:48:32] <banderson> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=DA-70224-ND
  • [00:48:40] <mru> you could network two beagles over usb...
  • [00:48:49] <mru> or several
  • [00:48:54] <mru> a beaglewolf
  • [00:49:16] <banderson> That would be cool!
  • [00:49:31] <mru> koen came up with the name
  • [00:49:35] <banderson> lol
  • [00:50:14] <banderson> Ravi: you might also want to pick up or find one of these http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=AE1474-ND
  • [00:50:42] <Ravi> what is this for?
  • [00:51:03] <banderson> Ravi: makes it easier to connect to Beagle.
  • [00:51:19] <mru> I use a standard A-miniB cable
  • [00:51:28] <Ravi> ok, making a wishlist out of all these pages :)
  • [00:53:05] <Ravi> amazon wishlist
  • [00:54:03] <Ravi> banderson, I see three hubs, anyone of them is fine?
  • [00:55:04] <banderson> Ravi: I am guessing..I bought the one I linked to
  • [00:55:36] <Ravi> hahaha
  • [00:55:47] <Ravi> just reading the beaglewolf comment
  • [00:55:50] <Ravi> sorry :)
  • [00:59:26] * rsalveti_ (n=salveti@189.119.117.6) has joined #beagle
  • [00:59:27] * Olipro_ is now known as Olipro
  • [01:02:24] <Ravi> banderson: do you also suggest the null modem cable?
  • [01:03:02] <Ravi> banderson: I don't have a serial port on any of my computers though
  • [01:03:09] * rsalveti (n=salveti@189.70.119.44) Quit (Read error: 104 (Connection reset by peer))
  • [01:03:40] <Ravi> would you guys suggest usb -> serial to interact with this board guys? or is USB good enough?
  • [01:03:41] * rsalveti (n=salveti@189.70.119.44) has joined #beagle
  • [01:04:10] <mru> you'll need to use the serial console
  • [01:04:55] <Ravi> ok - cool - so the null cable and a usb-> serial should do it?
  • [01:06:35] <Ravi> do I need the null cable?
  • [01:06:58] <mru> yes
  • [01:07:22] <mru> and an IDC10 to DSUB9 cable
  • [01:08:06] <Ravi> ooph, got digikey pages?
  • [01:14:59] <Ravi> mru: can you provide linkage?
  • [01:15:11] <Ravi> mru: looking but having a hard time navigating
  • [01:15:15] <banderson> Ravi: not sure if dikikey sells it but look at bottom of this page for links to other peripherals. http://beagleboard.org/hardware
  • [01:15:15] <mru> sorry, I don't have a link
  • [01:15:52] <Ravi> banderson: got it
  • [01:15:56] <Ravi> mru: what is this used for?
  • [01:16:10] <Ravi> mru: oh, Isee
  • [01:21:26] * rsalveti_ (n=salveti@189.119.117.6) Quit (Read error: 110 (Connection timed out))
  • [01:22:28] <Ravi> hey guys, I have a 15" LCD from Hanss-G. It has a DVI input, would I need an adapter to hook up the beagle to it?
  • [01:23:09] <banderson> Ravi: yes. you need a hdmi-dvi adapter cable
  • [01:23:29] <mru> or an hdmi-dvi adapter if you already have a cable
  • [01:23:46] <Ravi> I have a dvi cable
  • [01:23:54] <Ravi> same thing as hdmi dvi cable I presume?
  • [01:24:36] <mru> some of the pins in the hdmi connector correspond with some pins in the dvi connector
  • [01:24:40] <Ravi> this is what I have: http://www.global-b2b-network.com/direct/dbimage/50270101/Dvi_Cable.jpg
  • [01:25:05] <mru> hdmi is single-link dvi plus sound
  • [01:25:16] <Ravi> right, mru but is what I have good enough?
  • [01:25:24] <Ravi> mru: or do I need an adapter?
  • [01:25:48] <mru> if one end can be plugged into the beagle, and the other end into the monitor, it will work
  • [01:26:16] <Ravi> mru: the image I just sent, that's what I have, that female version fits into the monitor - will it work?
  • [01:27:35] <banderson> Ravi: you will need something that has connectors with the two ends pictured in image on this link
  • [01:27:39] <mru> the male end fits the monitor
  • [01:27:54] <Ravi> mru: yes, sorry - so the female goes to beagle
  • [01:27:56] <banderson> http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=AE10260-ND
  • [01:28:06] <mru> Ravi: then you need an adaptor
  • [01:28:18] <Ravi> mru: I'll take a look actually - hold on :)
  • [01:28:33] <ds2> wtf did classpath drop jikes support $@%#$@#%A$!#!@#
  • [01:28:50] <banderson> Ravi: Here is a bigger image: http://media.digikey.com/photos/Assmann%20Photos/AK639-3,5-R.jpg
  • [01:30:02] <Ravi> ok, so just looked - both sides of the dvi cable I have are male - the monitor takes male too
  • [01:30:35] <Ravi> banderson: in this picture, which side goes to the beagle?
  • [01:31:41] <banderson> Ravi: the hdmi side (smaller connector)
  • [01:31:59] <Ravi> awesome - ok - I will just need this cable
  • [01:32:04] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit (Read error: 110 (Connection timed out))
  • [01:32:08] <mru> Ravi: with a male-male dvi cable you need an adaptor like this: http://www.tvcables.co.uk/cgi-bin/tvcables/DVI-HDMI-PRO.html
  • [01:35:20] <Ravi> http://www.amazon.com/gp/product/B0009ZDBF2
  • [01:35:38] <Ravi> that'll probably do fine?
  • [01:36:38] <banderson> looks good to me
  • [01:38:08] <Ravi> any nice cases for this?
  • [01:39:10] <banderson> Ravi: somebody mentioned some this week. Search through irc logs.
  • [01:39:16] <Ravi> banderson: ok
  • [01:41:13] <Ravi> found it: http://specialcomp.com/beagleboard/
  • [01:42:20] <Ravi> "The RS-232 serial console connector (DB9M-IDC10 cable) is now included with each case."
  • [01:42:34] <Ravi> does this mean I don't need that adapter you suggested mru ?
  • [01:48:56] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [01:50:24] * cian_ (n=cianh@cian.ws) Quit ()
  • [01:53:17] <Ravi> heheh... that would be one geeky wedding registry - amazon wishlist is asking me if I want to add all these items to a wedding registery
  • [01:53:21] <calculus> I saw one of the cases at LWE, it was nice
  • [01:53:39] <calculus> the one that looks like a jigsaw puzzle
  • [01:53:45] <Ravi> calculus: you're talking aboutthe sc board?
  • [01:53:51] <calculus> no, the case
  • [01:53:57] <Ravi> sorry, I meant case
  • [01:54:02] <Ravi> calculus: the one I linked to above?
  • [01:54:17] <calculus> yep, the first one on the page is the one I saw
  • [01:54:45] <Ravi> calculus: does it require a lot of work to get the board in it and secured?
  • [01:54:56] <Ravi> calculus: or is it snap and go kind of deal?
  • [01:55:14] <calculus> didn't look like a lot of work, but it already had a board inside
  • [01:55:48] <Ravi> calculus: cool - you see I have my brithday coming up and I'm gonna ask my folks to get me whatever they can out of my wish list
  • [01:56:29] <calculus> heh, I can see you get all the cables, peripherals, cases and no the board :)
  • [01:56:35] <calculus> not*
  • [01:56:53] <Ravi> I'm trying to organize all the things I might need using Amazon's new Universal Wish List feature (bookmarklet). very cool stuff.
  • [01:56:55] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) has joined #beagle
  • [01:56:57] <Ravi> yeah, I can see that happening :)
  • [01:57:13] <calculus> add SD cards to the list
  • [01:57:27] <Ravi> calculus: good point, almost forgot!
  • [01:59:16] <Ravi> is 8 GB good enough?
  • [01:59:21] <Ravi> or should I go for 16?
  • [01:59:41] * calculus passes the question to the rest of the channel
  • [01:59:45] <Ravi> :P
  • [02:00:08] <Ravi> and also, which brand are most people using with Beagle?
  • [02:01:18] <mru> I'm using a sandisk 8GB card
  • [02:03:29] <ds2> finally
  • [02:03:34] <ds2> some semblience of java
  • [02:03:54] <Ravi> ds2: ?
  • [02:04:04] <mru> ds2: about as close as your spelling?
  • [02:04:07] <mru> ;-)
  • [02:05:56] * bobbarker (n=bob@adsl-69-109-116-44.dsl.pltn13.pacbell.net) has joined #beagle
  • [02:05:59] <dcordes> mru: :( my cable doesn't seem to work too well
  • [02:06:26] <Ravi> how long do these SDHC cards last in most cases?
  • [02:06:37] <bobbarker> What's up all! I just got the usb cable for my beagle board!
  • [02:07:22] <asterick> ravy: conservative estimate: 1000 writes per sector. resonable 10~20k per sector.
  • [02:08:04] <asterick> most SDHC cards have a 1~3 year warrenty on them, so as long as you maintain backups of important data, they'll replace them.
  • [02:08:12] <dcordes> mru: I'm quite sure I connected all pins correctly. is there a way to just probe if something is on the cable? my pc detects ttyS1 and ttyS0, I tried both in minicom with the options given at http://code.google.com/p/beagleboard/wiki/outofthebox
  • [02:08:17] <Ravi> asterick: k, thanks for the info
  • [02:08:22] <dcordes> mru: and it just say modem offline
  • [02:08:45] <ds2> mru: yep, that was intentional for addd effect ;)
  • [02:09:02] <mru> dcordes: you could check the connections with an ohm meter
  • [02:09:59] <dcordes> I had one but it's broken
  • [02:11:09] <dcordes> at least the beagle didn't go up in flames
  • [02:11:51] * dozerl (n=dozerl@68-188-163-73.dhcp.bycy.mi.charter.com) Quit ("Leaving")
  • [02:11:53] <ds2> Hmmm a flaming beagle
  • [02:12:54] * Crofton|work (n=balister@66-207-66-26.black.dmt.ntelos.net) Quit ("Leaving")
  • [02:13:48] <dcordes> mru: that's so frustrating. I even sawed off the rest of the connector, it looks just like the one on the most photos
  • [02:16:51] <dcordes> mru: is there anything else I can try?
  • [02:17:25] <mru> a different cable ;-)
  • [02:17:32] <dcordes> noooo
  • [02:18:17] * Crofton|work (n=balister@66-207-66-26.black.dmt.ntelos.net) has joined #beagle
  • [02:21:04] <haggisX> goodnight
  • [02:21:10] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has left #beagle
  • [02:25:04] <dcordes> all my passion is in that cable
  • [02:30:14] <NishanthM> gnite folks..
  • [02:30:24] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit ("Aloha")
  • [02:38:25] <dcordes> bye
  • [02:38:28] * dcordes (n=dcordes@unaffiliated/dcordes) Quit ("leaving")
  • [02:38:31] <bobbarker> later
  • [02:50:08] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [02:53:32] <jkridner> good morning khasim
  • [03:05:31] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [03:05:46] * odesus (n=Widmung@201.165.138.120) has joined #beagle
  • [03:06:31] <odesus> hi..
  • [03:06:45] <bobbarker> hello
  • [03:06:54] <odesus> as a newbie I want to know if there is
  • [03:07:06] <odesus> a IDE for working with arm compiler.
  • [03:07:30] <odesus> I want to install the gnu toolchain arm from code sourcery..
  • [03:08:15] <odesus> as you know the last release is free but doesn't have IDE
  • [03:08:28] <odesus> any suggestion?
  • [03:08:31] <bobbarker> I'm a noob too, it looks like you're ahead of me... what cable do I need to connect to the board via computer?
  • [03:08:54] <odesus> well you can connect your board..
  • [03:09:05] <odesus> with 2 possibilities..
  • [03:09:13] <odesus> the first one is RS232 ..
  • [03:09:17] <asterick> USB through the OTG port, and the RS232
  • [03:09:33] <odesus> with a flat cable to RS232
  • [03:09:50] <odesus> and the second as asterick told us is thru USB otg port..
  • [03:10:43] <odesus> Does anybody knows how to work comfy with an appropiate IDE for arm toolchain?
  • [03:10:45] <odesus> I mean..
  • [03:10:47] <bobbarker> Okay so I've got USB A to usb mini-b, tera term doesn't have a usb selection.
  • [03:11:10] <odesus> I suggest you to use the flat cable..
  • [03:11:22] <odesus> flat cable to RS232..
  • [03:11:35] <odesus> or better flat cable db9 female to RS232..
  • [03:12:05] <odesus> when you're using the USB in your first try..
  • [03:12:29] <odesus> the USB behaves like other power source..
  • [03:12:38] <odesus> I suggest you use the flat cable to RS232..
  • [03:12:56] <bobbarker> k, thanks
  • [03:13:07] <odesus> no problem bobbarker :)
  • [03:13:11] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [03:13:29] <khasim> hi Jason, sorry missed your ping
  • [03:13:32] <odesus> Does anybody know about an appropiate IDE for working with arm compiler.
  • [03:13:50] <odesus> Do I have to use eclipse or something like that?
  • [03:14:05] <asterick> That or pay for code warrior or TI's official kit. ;)
  • [03:14:15] <asterick> there are worse things than eclipse (not many though)
  • [03:14:30] <odesus> what IDE are you using asterick?
  • [03:14:35] <asterick> Me? none
  • [03:14:54] <odesus> uh? then are you using on line compilation?
  • [03:15:06] <odesus> online command compilation from prompt?
  • [03:15:23] <asterick> not using anything at the moment
  • [03:15:37] <asterick> Don't have a beagleboard yet. ;)
  • [03:15:48] <odesus> oh.. I have it..
  • [03:15:58] <odesus> that's way I need to know what IDE to work..
  • [03:16:06] <odesus> I don't like console compilation..
  • [03:16:17] <asterick> What host platform are you using
  • [03:16:20] <odesus> It's kind of annoying.
  • [03:16:21] <asterick> (linux, win32, etc)
  • [03:16:26] <odesus> linux fedora 9
  • [03:16:27] <khasim> Today India celebrates its 61st Independence Day - Happy Independece day India....
  • [03:16:43] <asterick> yea, eclipse is your best bet, you can probably configure it to do ant\make based projects
  • [03:16:43] <odesus> Viva mahatma Gandhi!!
  • [03:17:11] <odesus> could you use Eclipse with code sourcery?
  • [03:17:18] <odesus> I mean..
  • [03:17:26] <odesus> Don't you need to purchase all from the same source?
  • [03:17:34] <asterick> I'm sure you can, eclipse is pretty flexable as far as it's target tool chain is concerned.
  • [03:17:58] <asterick> lemmie see what I can find.
  • [03:18:18] <odesus> but how do you integrate eclipse to codesourcery gnu arm compiler?
  • [03:19:14] <asterick> http://elinux.org/BeagleBoard#Development_environments
  • [03:19:51] <odesus> oh..
  • [03:19:56] <odesus> but that's my confusion..
  • [03:20:27] <asterick> I believe the topic links you to a forum post on configuring it.
  • [03:20:48] <odesus> mmm..
  • [03:20:51] <odesus> My confusion..
  • [03:20:57] <odesus> is about
  • [03:21:09] <odesus> development eviroments..
  • [03:21:11] <odesus> I mean..
  • [03:21:25] <odesus> let me explain the panorama..
  • [03:21:41] <odesus> there's many RTOS available for beagle..
  • [03:21:53] <odesus> but just a few were tested on beable..
  • [03:22:04] <odesus> Armstrong and maesmo..
  • [03:22:12] <odesus> maemo sorry :)
  • [03:22:28] <odesus> as we know Armstrong is GNU..
  • [03:22:42] <odesus> so you can use it in your own application without no restrictions..
  • [03:23:00] <odesus> then If think about construct a proyect for ultrasound monitoring..
  • [03:23:07] <odesus> like in structures..
  • [03:23:20] <odesus> you can use beagle board as a development platform..
  • [03:23:33] <odesus> and use armstrong as you RTOS..
  • [03:23:47] <odesus> then all applications you will development run over armstrong..
  • [03:24:01] <sakoman> khasim: Happy Independence Day
  • [03:24:23] <sakoman> Do you get a day off from work?
  • [03:30:57] <khasim> sakoman: yes :)
  • [03:31:34] <sakoman> Excellent! Three day weekend :-)
  • [03:31:59] <sakoman> Bad for me though cause I can't ask you to use jtag to see where my x-load is crashing ;-)
  • [03:35:02] <khasim> sakoman: yeah, I was thinking of going to office some time in this weekend.
  • [03:35:48] <khasim> sakoman: I can take a diff and help you in finding out the issue
  • [03:36:05] <sakoman> Would you mind taking a look if you get a chance? I'll send you my most recent attempt
  • [03:36:15] <khasim> sakoman: will it be possilbe for you to upload a tar ball of you x-loader
  • [03:36:38] <khasim> sakoman: yes I will work on your x-loader
  • [03:37:19] <sakoman> Thanks! To get a tarball, go to:
  • [03:37:30] <sakoman> http://www.sakoman.net/cgi-bin/gitweb.cgi?p=x-load-omap3.git;a=summary
  • [03:37:32] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) has joined #beagle
  • [03:37:59] <sakoman> Then click on snapshot for the most recent checkin
  • [03:38:18] <sakoman> Hi NishanthM!
  • [03:38:43] <sakoman> khasim: I will push my most recent x-load change this evening
  • [03:38:45] <NishanthM> hi sakoman
  • [03:39:13] <sakoman> NishanthM: how goes u-boot-v2?
  • [03:39:31] <NishanthM> just came back from a long hack week
  • [03:39:35] <NishanthM> rather weeks
  • [03:39:42] <NishanthM> new board on CSST :)
  • [03:40:25] <sakoman> new boards are a *lot* of work!
  • [03:40:28] <NishanthM> http://www.i-glassesstore.com/i-3d.html i just got it working.. wondering if i can lay the claim to the first person with head mounted display with beagle :)
  • [03:40:37] <sakoman> especially when there is only one of you :-)
  • [03:41:02] <sakoman> cool! do you have plans for it?
  • [03:41:18] <NishanthM> yep.. next is to add camera + a usb hub and a gps
  • [03:41:32] <NishanthM> + get angstrom functioning with voice recognition + tts :D
  • [03:42:01] <NishanthM> i recollect koen saying he had sphinx package on angstrom..
  • [03:42:23] <khasim> NishanthM: your long dream came true :)
  • [03:42:34] <NishanthM> aye.. touch wood
  • [03:42:38] <NishanthM> :D
  • [03:42:58] <khasim> does it take s-video?
  • [03:43:07] <khasim> I sett only composite
  • [03:43:16] <sakoman> that ought to keep you busy for a few weeks!
  • [03:43:29] <NishanthM> heh heh :) my old laptop had s-video -> composite convertor
  • [03:43:44] <NishanthM> used it along with the cable i had used for my ps3 extender :D
  • [03:44:06] <NishanthM> i am waitin for my rebel's battery to get charged.. will put out a snap :D
  • [03:44:12] <khasim> on 3430 SDP the converter was resulting in black & white screen, I hope this is not the case with beagle
  • [03:44:20] <NishanthM> nope i can see color
  • [03:44:30] <khasim> great...
  • [03:44:39] <NishanthM> but i see noise.. which is kinda bad.. but i have my cable length to blame too
  • [03:45:19] <NishanthM> it feels great to get back on beagle :)
  • [03:45:48] <khasim> sakoman: I will try your x-loader and let you know
  • [03:46:10] <sakoman> khasim: thanks!
  • [03:46:19] <NishanthM> sakoman, on u-boot v2.. just send a mail to sascha to discuss on directory arch.. probably will have some devels soon.. :)
  • [03:47:01] <NishanthM> oh guys by the way.. remember the discussion on hdmi vs dvi for beagle?
  • [03:48:26] <sakoman> NishanthM: your uboot-serial was the first code to run on the Overo proto
  • [03:48:26] <NishanthM> sakoman: glad to hear that :)
  • [03:48:27] <sakoman> It has been the only code I was able to use with USB boot!
  • [03:48:27] <sakoman> I used it to bootstrap u-boot onto the hw
  • [03:48:27] <NishanthM> tounge-in-cheek -> always write the right code ;)
  • [03:48:38] <NishanthM> oh on hdmi -> we got this new board with hdmi -> and it does work :) albeit at 720p
  • [03:48:38] <sakoman> khasim and I have been trying to figure out how to build an x-load that will USB boot
  • [03:48:58] <NishanthM> hdmi+omap that is.. something to think for a future rev of beagle...
  • [03:50:33] <sakoman> khasim: I will be pushing a couple of changes to the x-load git during the next hour or so, so you might want to wait before grabbing the tarball
  • [03:52:26] * gulick (n=gulick@rrcs-208-125-50-2.nys.biz.rr.com) Quit (Read error: 110 (Connection timed out))
  • [03:52:58] <NishanthM> khasim, the A5 rev board -> audio out.. i suppose you have a stereo out? standard 4 pin?
  • [03:53:20] * NishanthM wondering why audio only on one channel.. :(
  • [03:56:01] <NishanthM> gnite guys...
  • [03:56:10] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit ("Aloha")
  • [04:11:50] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [04:14:49] * Ravi (n=chatzill@ip68-100-201-32.dc.dc.cox.net) has left #Beagle
  • [04:16:33] <odesus> good nite for everybody!!
  • [04:16:36] <odesus> bye.
  • [04:16:39] * odesus (n=Widmung@201.165.138.120) has left #beagle
  • [04:28:48] * Olipro (n=Olipro@unaffiliated/olipro) Quit (Connection timed out)
  • [04:31:51] <sakoman> khasim: OK, my x-load changes are now pushed. If you grab a tarball you will be using exactly the same code that I have been using to test USB boot of x-load
  • [04:32:08] <asterick> http://www.bustedtees.com/greece
  • [04:32:54] <sakoman> khasim: The only things I do differently when building for usb peripheral boot:
  • [04:33:09] <sakoman> 1. the config.mk change you recommended
  • [04:33:17] <sakoman> 2. use and unsigned image
  • [04:33:25] <sakoman> s/and/an/
  • [04:53:56] <t_s_o> hmm, i wonder how easy it would be to turn a beagleboard into a voip phone ;)
  • [04:55:43] <ds2> a case, a keypad, display, plus a wifi interface
  • [04:56:39] <t_s_o> hmm, wifi, or ethernet...
  • [04:56:48] <t_s_o> i see now that there is no built in network connection
  • [04:57:14] <t_s_o> would be kinda fun if a future version would come with a ethernet port that support PoE ;)
  • [04:59:33] <t_s_o> thing is that with video out and usb, this board can be hooked up to a tv, or similar, and a webcam to have video chats
  • [05:01:26] <rsalveti> t_s_o: you can add more devices at your usb port, like a wireless network adapter, and be happy :-)
  • [05:02:09] <t_s_o> maybe so, but i was thinking of it as a kind of replacement to the old phone. and being able to be powered from that one socket in the wall is kinda nice
  • [05:04:15] <t_s_o> anyways, say one stuff it to the brim with im support, astrix (or whatever its called) and similar?
  • [05:05:36] <t_s_o> i guess it could even do 3G with the right dongle in the usb ;)
  • [05:12:58] <t_s_o> hmm, guess one could hide a usb hub in the case...
  • [05:16:20] * bobbarker (n=bob@adsl-69-109-116-44.dsl.pltn13.pacbell.net) Quit ()
  • [05:16:22] <ds2> t_s_o: you can add a ethernet
  • [05:16:38] <ds2> t_s_o: how important is ethernet + POE for you?
  • [05:17:06] <ds2> hubs are small
  • [05:17:12] <t_s_o> none, im more brainstorming then anything else
  • [05:17:29] <t_s_o> its more of a "would have been fun to have" kind of idea
  • [05:17:29] <ds2> heh
  • [05:17:40] <ds2> it is easy enough to design an add on for one
  • [05:18:31] <t_s_o> thing is that it would have been fun to pack it all up so that at first glance it just looks like a orginary desk phone, but with some extra ports on the back
  • [05:19:19] <ds2> that's not too hard to do
  • [05:19:28] <t_s_o> indeed
  • [05:19:36] <ds2> hardest part I think is finding a decent shell
  • [05:20:01] <t_s_o> heh, most likely
  • [05:42:48] * Beagle0 (n=Beagle0@cpe-70-112-43-135.austin.res.rr.com) has joined #beagle
  • [05:43:24] <Beagle0> hi
  • [05:44:13] <Beagle0> need help with git
  • [05:44:14] * Beagle7 (n=Beagle7@117.192.128.89) has joined #beagle
  • [05:46:01] <Beagle0> "git commit -a" does not work for some reason
  • [05:47:41] * chakie (n=chakie@dsl-aur-fee3de00-172.dhcp.inet.fi) has joined #beagle
  • [05:50:46] * Beagle0 (n=Beagle0@cpe-70-112-43-135.austin.res.rr.com) has left #beagle
  • [05:51:55] * Beagle7 (n=Beagle7@117.192.128.89) Quit ()
  • [05:52:31] * Beagle1 (n=Beagle1@117.192.128.89) has joined #beagle
  • [05:57:14] * Thanatos_desktop (n=rwvens@p578FEFA2.dip.t-dialin.net) Quit (Read error: 110 (Connection timed out))
  • [06:03:22] * diego_ (n=diego@host-84-222-18-165.cust-adsl.tiscali.it) has joined #beagle
  • [06:07:52] * asterick (n=unicdk@24-117-80-129.cpe.cableone.net) Quit ()
  • [06:08:31] <grummel> Beagle1: try git-commit -a -m "Whatyouhavedone"
  • [06:16:46] * diego (n=diego@host-84-223-121-229.cust-adsl.tiscali.it) Quit (Read error: 101 (Network is unreachable))
  • [06:43:19] * Beagle1 (n=Beagle1@117.192.128.89) Quit ()
  • [06:45:04] * trickie (n=trickie@basesoft.xs4all.nl) has joined #beagle
  • [07:42:30] * NTNP[U] (n=u@193.10.100-84.rev.gaoland.net) has joined #beagle
  • [08:06:41] * Beagle5 (n=Beagle5@hunter.netmodule.com) has joined #beagle
  • [08:06:45] * Beagle5 is now known as sprotch
  • [08:17:38] * incandescant (n=incandes@foo.stupids.org) has left #beagle
  • [08:26:40] * NTNP[U] (n=u@193.10.100-84.rev.gaoland.net) Quit ("Quitte")
  • [08:42:14] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [08:44:53] * RogerMonk (n=a0740758@nat/ti/x-b6df9f60d0dee099) has joined #beagle
  • [09:12:20] * Olipro__ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [09:12:35] * valhalla (n=valhalla@81-174-21-16.dynamic.ngi.it) has joined #beagle
  • [09:28:53] * Olipro_ (n=Olipro@unaffiliated/olipro) Quit (Read error: 110 (Connection timed out))
  • [09:45:13] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [10:05:29] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [10:29:29] * gulick (n=gulick@rrcs-208-125-50-2.nys.biz.rr.com) has joined #beagle
  • [11:08:39] * sprotch (n=Beagle5@hunter.netmodule.com) Quit ()
  • [11:09:11] * ade_ (n=ade@80.177.64.36) has joined #beagle
  • [11:30:30] <ade_> hi. newbie question. does anyone know how to drive an LCD TV with the beagle board at 720p?
  • [11:34:53] * RogerMonk (n=a0740758@nat/ti/x-b6df9f60d0dee099) has left #beagle
  • [11:54:13] * docelic (n=docelic@78.134.199.31) has joined #beagle
  • [12:06:47] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [12:08:41] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) has joined #beagle
  • [12:15:05] * cbrake_away is now known as cbrake
  • [12:16:15] * dcramer (n=davec@dcdsl.ebox.com) Quit ()
  • [12:17:02] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
  • [12:17:30] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [12:18:49] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) has joined #beagle
  • [12:21:23] * BThompson (n=BThompso@nat/ti/x-4ac837277fc4fc3a) has joined #beagle
  • [12:26:14] * robclark (n=robclark@nat/ti/x-271a1384c1fa9311) has joined #beagle
  • [12:28:45] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit (Read error: 104 (Connection reset by peer))
  • [12:29:15] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [12:31:55] <jkridner> ade_: today, you have to recompile your kernel.
  • [12:32:16] <jkridner> further, it depends on which kernel sources you are using what you must do.
  • [12:33:11] <jkridner> I believe the linux-omap git tree has a 720p option that is selectable through the Kconfig (selectable when you run make menuconfig)
  • [12:33:14] <ade_> ok, thanks. I'm bit of a noob to kernel sources. I followed some instructions to build using OE
  • [12:33:49] <jkridner> the "diagnostic" kernel would require that you edit the sources for entirely new resolution and clock settings.
  • [12:33:50] <jkridner> ah...
  • [12:34:04] <jkridner> using OE, there is a 'defconfig' file that gets brought in.
  • [12:34:16] <jkridner> you'll want to change the settings in the 'defconfig'.
  • [12:34:54] <jkridner> also, I believe the kernel built by the head of OE does not support many 720p TVs because it uses a reduced vertical blanking interval.
  • [12:35:25] <jkridner> I've provided mru with some clock settings that I think should get the right 74.25MHz pixel clock required for many 720p TVs.
  • [12:35:40] <jkridner> normal computer monitors, however, are much more forgiving.
  • [12:36:00] <jkridner> most people are using Beagle today with computer monitors, as far as I'm aware.
  • [12:36:31] <jkridner> there are some people that show up on the channel that are much more experienced with tweaking the code within OE.
  • [12:36:50] <jkridner> This project is the first time I've used OE.
  • [12:37:01] <ade_> me too
  • [12:37:28] <ade_> only got my beagle board on monday so it's bit of a learning curve
  • [12:38:03] <jkridner> we just reached 600 members on the mailing list!
  • [12:38:10] <ade_> The TI Tech guy came to see us yesterday and he said 720p shouldn't be a problem and even 1080p at reduced frame rate
  • [12:38:17] * banderson (n=irc@69.71.183.7) Quit (Read error: 110 (Connection timed out))
  • [12:38:21] <jkridner> we've reached a bit of an equilibrium point, as the traffic has gone up and some people have left.
  • [12:38:36] <jkridner> ade_: it is true...
  • [12:39:59] <jkridner> I know that it can do 720p with the full 60fps and required vertical blanking interval.
  • [12:40:25] <jkridner> The limitation on 1080p is with the pixel clock rate.
  • [12:40:57] <jkridner> The 74.25MHz setting required for full 720p60 standard timing is the max pixel clock.
  • [12:40:59] <ade_> I don't think that will be a problem for my application
  • [12:41:14] <BThompson> you could do 1080p at a reduced frame rate if you had a display that could accept a reduced frame rate (pixel clock)
  • [12:41:20] <jkridner> you can reduce the blanking interval and the frame rate to get resolutions up to 2048x2048.
  • [12:41:33] <BThompson> which is non standard so may not be something you could find easily...
  • [12:42:02] * jeh (n=jeh@vpn-pub2.hard.ware.fi) has joined #beagle
  • [12:43:47] <jkridner> I'm waiting for someone to tell me that http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=586-1016-ND works well.
  • [12:44:08] <ade_> ok, thanks. we currenly manufacture our own controller using an Altera with Nios II core but can only go to 1280x576i and with limited features
  • [12:51:06] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit (Read error: 110 (Connection timed out))
  • [12:51:13] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit ()
  • [12:51:37] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) has joined #beagle
  • [12:54:33] * NishanthM (n=Nishanth@cpe-24-27-74-89.tx.res.rr.com) Quit (Client Quit)
  • [12:57:21] <ade_> jkridner: any idea what I'm looking for in 'defconfig' to help with 720p?
  • [13:00:57] * haggisX (n=haggis@cpc1-clyd1-0-0-cust187.renf.cable.ntl.com) Quit ("Leaving")
  • [13:02:40] * Beagle6 (n=Beagle6@CPE00095bc22225-CM001a6682b616.cpe.net.cable.rogers.com) has joined #beagle
  • [13:03:40] * Beagle6 (n=Beagle6@CPE00095bc22225-CM001a6682b616.cpe.net.cable.rogers.com) Quit (Client Quit)
  • [13:04:46] * khasim (n=a0393720@192.163.20.231) has joined #beagle
  • [13:16:26] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [13:26:43] * JuanG (n=Juan@nat/ti/x-c6af773392ace3dd) has joined #beagle
  • [13:26:54] * JuanG (n=Juan@nat/ti/x-c6af773392ace3dd) has left #beagle
  • [13:33:57] * GeorgeB (n=chatzill@bas1-toronto10-1279555770.dsl.bell.ca) has joined #beagle
  • [13:35:13] * GeorgeB (n=chatzill@bas1-toronto10-1279555770.dsl.bell.ca) has left #beagle
  • [13:43:11] * t_s_o (n=tso@173.84-49-129.nextgentel.com) Quit ("Konversation terminated!")
  • [13:43:27] * rsalveti (n=salveti@189.70.119.44) Quit (Connection timed out)
  • [13:43:57] <khasim> sakoman: I see lot of difference in your version of x-load/board/omap3530beagle/omap3530beagle.c
  • [13:44:15] * dcramer (n=davec@missdsl.ebox.com) has joined #beagle
  • [13:45:12] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
  • [14:12:06] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [14:12:29] <sakoman> khasim: anything that you find to be a potential source of the hang?
  • [14:13:46] <sakoman> It's different because there is added functionality
  • [14:21:19] * dirk2 (n=dirk@F31aa.f.strato-dslnet.de) has joined #beagle
  • [14:22:30] <khasim> sakoman: I didnt get access to the debugger today, will be trying out tomorrow.
  • [14:22:53] <sakoman> khasim: thanks for looking at this
  • [14:23:25] <sakoman> that code seems to work well from both mmc and nand -- doesn't require 2 different versions
  • [14:24:03] <khasim> sakoman: yes, I see that, this is how it should be... :)
  • [14:24:09] <sakoman> would be nice if the unsigned version also worked for usb/peripheral boot ;-)
  • [14:24:27] <khasim> I suspect config_3430sdram_ddr in omap3530beagle.c
  • [14:24:27] * dclarke_atwork (i=3b6030f5@gateway/web/ajax/mibbit.com/x-3c466176e0b25935) has joined #beagle
  • [14:24:34] * dclarke_atwork is now known as dclarke_work
  • [14:24:36] <khasim> this is configuring memory
  • [14:25:24] <sakoman> is that necessary, or has the rom code already done that?
  • [14:26:37] <khasim> it is necessary as it is configuring DDR
  • [14:26:51] <khasim> Can you just replace this function from the one from code.google and try
  • [14:26:52] <sakoman> I assumed it was
  • [14:27:00] <sakoman> OK, I will do that
  • [14:27:04] <dcramer> Hi
  • [14:27:07] <dcramer> I just got my board
  • [14:27:13] <khasim> otherwise, you can just wait till tomorrow, I will give the exact location where it fails
  • [14:27:16] <dcramer> ls just sits there scanning jffs ?
  • [14:27:24] <dcramer> is it supposed to work ?
  • [14:27:30] <sakoman> khasim: I believe that routine is from the latest evm x-load
  • [14:27:48] * banderson (n=irc@69.71.183.7) has joined #beagle
  • [14:27:51] <sakoman> That was my starting point since it seemed to be the most recent
  • [14:28:40] <khasim> sakoman: as you have commented MMC, the only place I think it could hang is here, I took a diff of your source v/s code.google, didnt find any other critical area
  • [14:28:54] * Olipro__ (n=Olipro@unaffiliated/olipro) Quit (Read error: 110 (Connection timed out))
  • [14:29:10] <sakoman> OK, I will try replacing that routine
  • [14:29:39] <khasim> I still see TEXT_BASE = 0x40200800
  • [14:29:47] <khasim> In "board/omap3530beagle/config.mk"
  • [14:30:01] <khasim> make sure you enable TEXT_BASE = 0x80e80000
  • [14:30:25] <sakoman> khasim: people are building from that git, so I can't break it with the TEXT_BASE change :-)
  • [14:30:36] <sakoman> I change it locally when I build
  • [14:30:36] <khasim> :)
  • [14:30:43] <khasim> ok
  • [14:31:18] <sakoman> downloading latest from code.google
  • [14:31:36] * dclarke_work (i=3b6030f5@gateway/web/ajax/mibbit.com/x-3c466176e0b25935) Quit ("http://www.mibbit.com ajax IRC Client")
  • [14:31:38] <dirk2> sakoman: Any idea regarding the uboot nand breakage? I'm not through uboot list yet.
  • [14:32:33] <sakoman> dirk2: u-boot mtd implementation was "modernized" to linux 2.6.22 vintage so nand drivers need to change
  • [14:33:03] * dclarke_work (i=3b6030f5@gateway/web/ajax/mibbit.com/x-2721f9ecf691e4e5) has joined #beagle
  • [14:33:17] <sakoman> If it weren't for the ecc stuff we have in out driver it would probably be fairly straightforward
  • [14:33:40] <sakoman> Perhaps khasim's team will be able to have a look
  • [14:33:54] <khasim> team ??? :)
  • [14:33:58] <sakoman> s/out/our/
  • [14:34:10] <dclarke_work> i get errors such as "Error: reading boot sector" in x-loader
  • [14:34:11] <sakoman> khasim: ;-)
  • [14:34:19] <sakoman> it's like my team, eh?
  • [14:34:27] <sakoman> me, myself, and I
  • [14:34:27] <khasim> ha ha ha
  • [14:34:31] <dirk2> I see a mail at uboot list from Kumar reporting onenand breakage at apollon
  • [14:35:04] <sakoman> I'll try again: perhaps khasim or one of the other TI experts can have a look ;-0
  • [14:35:50] <khasim> dirk2: I will try to check with Mani or others to quickly comment on the same.
  • [14:36:30] <dirk2> currently building steves uboot git top, lets see
  • [14:37:33] <sakoman> dirk2: it will fail for beagle, but succeed for Overo and EVM since they use onenand
  • [14:38:06] <dirk2> khasim: here are the errors: http://rafb.net/p/ZbrW2C40.html
  • [14:38:58] <sakoman> khasim: Mani is having troubles getting his patches accepted, I try to do hand edits but oftentimes it has been too extensive and I worry I might get it wrong
  • [14:39:44] <dirk2> sakoman: Any idea of a "nand.c" from other board which is already converted correctly we can use as example?
  • [14:39:53] <khasim> sakoman: =-O
  • [14:39:54] <sakoman> dirk2: looks familiar ;-( I started editing the code, but the changes will be fairly extensive
  • [14:40:16] <khasim> sakoman: very sorry to know this, I will work with him to get it resolved.
  • [14:40:30] <sakoman> dirk2: the ecc stuff is the challenge. other machines don't have this
  • [14:40:53] <sakoman> the nand driver from linux will have some of the appropriate code
  • [14:41:14] <sakoman> (like the ALE/CLE stuff)
  • [14:41:26] <sakoman> but the ecc stuff is the major item of work
  • [14:41:57] <sakoman> structures changed pretty significantly and I have too much other stuff on my plate to sort through it
  • [14:43:05] <sakoman> khasim: if you look at the beagle mailing list you will see the current status of his submitted patches
  • [14:43:19] <sakoman> I really want to encourage him because he is doing great work
  • [14:43:20] <khasim> hw ecc was added for Micron NAND support, I dont think this is done in u-boot, I had seen some patches to OMAP GIT from a Micron team member
  • [14:43:25] <dcramer> hi
  • [14:43:34] <dcramer> what usb-ethernet device works properly
  • [14:43:55] <jkridner|work> I know the Linksys USB200M works. See http://beagleboard.org/hardware.
  • [14:44:04] <khasim> sakoman: yes Mani is doing really good work in getting all EVM stuff working on latest u-boot and also cleaning most of the stuff.
  • [14:44:31] <sakoman> I think we just need to get him to break patches into smaller chunks
  • [14:44:34] * Beagle0 (n=Beagle0@hunter.netmodule.com) has joined #beagle
  • [14:44:39] <khasim> sakoman: I will work with him and try to get things completed asap
  • [14:44:45] <dcramer> thanks
  • [14:45:10] <sakoman> He tries to pack lots of stuff in and that increases the odds that something upstream will change and cause a merge conflict
  • [14:45:34] <khasim> ok
  • [14:45:56] <sakoman> We should also encourage him to consider beagle and overo when he makes changes that might also be suitable for them
  • [14:46:20] <sakoman> I've been extending his patches to those machines as appropriate
  • [14:46:46] <sakoman> But it would be nice not to have to do the extra work to modify his patches ;-)
  • [14:46:46] <khasim> yeah, I have communicated this to him.
  • [14:47:12] <sakoman> For the greater glory of TI!
  • [14:47:16] <sakoman> ;-)
  • [14:47:42] <Crofton> hmmm, sakoman is preparing for the Russian invasion of California ....
  • [14:48:20] <sakoman> brushing up on my Russian and plating lots of fall beets
  • [14:48:46] <sakoman> planting, sheesh I should give up and get a career where I don't have to type!
  • [14:48:47] <dclarke_work> im having trouble booting out of mmc - x-load complains "Error: reading boot sector"
  • [14:49:02] <Crofton> I need to get "The Russians are Coming" from Netflix, if the fix their shipping problem
  • [14:49:24] <sakoman> dclarke_work: that means it can't read the mmc boot sector
  • [14:49:52] <dclarke_work> has anyone seen this problem before
  • [14:49:55] <dclarke_work> yeah i know. but i wonder why?
  • [14:50:01] <sakoman> what version of x-load are you using?
  • [14:50:33] <dclarke_work> i must add i have a evm
  • [14:50:56] <dclarke_work> x-load 1.41
  • [14:52:01] <khasim> dclarke_work: can you try new images uploaded on linux.omap.com, just see my recent mail on linux-omap-vger
  • [14:52:57] <khasim> dclarke_work: also make sure you have proper DIP switches selected for MMC Booting
  • [14:53:33] <dclarke_work> is x-loader same for one nand booting and mmc booting
  • [14:53:36] <khasim> dclarke_work: If you are already on recent images from extranet then no need to try the one on linux.omap.com, they are the same 1.0
  • [14:53:59] <khasim> dclarke_work: Does your board have ONENAND for sure
  • [14:54:26] <khasim> dclarke_work: all recent boards are going with Micron NAND
  • [14:54:30] <dclarke_work> yes
  • [14:59:31] * dcramer (n=davec@missdsl.ebox.com) Quit ()
  • [15:02:27] <dclarke_work> khasim:tried with latest x-loader, still get the same error
  • [15:03:30] <dclarke_work> i might have formatted it wrongly; wonder why x-loader has problem picking up u-boot
  • [15:03:50] * BThompson (n=BThompso@nat/ti/x-4ac837277fc4fc3a) Quit (Remote closed the connection)
  • [15:04:28] <dclarke_work> are there any restrictions regarding partition sizing?
  • [15:04:43] * trickie (n=trickie@basesoft.xs4all.nl) Quit ("Leaving")
  • [15:06:01] * dcramer (n=davec@missdsl.ebox.com) has joined #beagle
  • [15:06:37] * BThompson (n=BThompso@nat/ti/x-d3899b3f11a6194b) has joined #beagle
  • [15:07:23] <sakoman> khasim: your suspicion was correct the ram init stuff was causing the hang
  • [15:07:54] <sakoman> It would be very good to know which part of the init caused the hang
  • [15:08:37] <sakoman> I believe the dram init code in my tree came from the EVM 0.98 SDK
  • [15:09:42] <sakoman> We should try to put together an init that works for all boards for nand, onenand, mmc and peripheral boot
  • [15:09:57] <sakoman> do you think it is possible, or just a crazy dream?
  • [15:11:20] <khasim> it should be possible. The way we were coding before was - just thinking we were only one in this world :)
  • [15:11:51] <khasim> with mulitple versions of boards on OMAP today, these are big concerns and code should be that modular
  • [15:13:12] <sakoman> The EVM code looks reasonable -- we should try to find out what causes the hang when used w/ peripheral boot
  • [15:13:52] <sakoman> Then we will be well on the way to a universal x-load for all forms of boot
  • [15:15:10] <sakoman> If we could squeeze usb support into x-load then we would really have an interesting tool
  • [15:22:13] * dclarke_work (i=3b6030f5@gateway/web/ajax/mibbit.com/x-2721f9ecf691e4e5) has left #beagle
  • [15:22:43] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [15:25:23] <khasim> sakoman: As I had mentioned we have USB support in u-boot EVM 1.0 rls, we can copy the same to x-loader, but it should not increase the size.
  • [15:26:46] * docelic (n=docelic@78.134.199.31) Quit ("http://www.spinlocksolutions.com/")
  • [15:27:21] <sakoman> khasim: I have a feeling usb support will increase the size a little bit :-)
  • [15:27:30] <khasim> :(
  • [15:27:45] <sakoman> unless we can use the boot rom stack
  • [15:27:53] <sakoman> not likely I would expect
  • [15:28:28] * dcordes (n=dcordes@unaffiliated/dcordes) has joined #beagle
  • [15:29:10] <khasim> can we write a x-loader that initializes DDR and untars a u-boot/x-loader on DDR and execute the same.
  • [15:30:24] <khasim> we should have only minimal code to configure DDR and the uncompressing utility, rest should be moved to DDR
  • [15:30:32] <sakoman> should be possible if one had the time to devote to it
  • [15:30:44] <khasim> yeah... :(
  • [15:31:09] <sakoman> my "team" is booked with lots of work
  • [15:31:13] <khasim> by what percentage can we zip the image ?
  • [15:31:25] <khasim> even my team :)
  • [15:32:29] <sakoman> I doubt you would see even 50% compression
  • [15:32:47] <sakoman> but perhaps that would be enough
  • [15:32:52] <khasim> actually lot of functionality is present in ROM CODE that we can leverage but due to restrictions the map table is not open
  • [15:33:20] <khasim> we should ask ROM Code team to implement the code such that we can re-use all the existing APIs
  • [15:34:00] <khasim> complete MMC, UART, USB stack is already present in ROM code, we just have to call those addresses
  • [15:34:11] <sakoman> Would be nice!
  • [15:34:37] <sakoman> probably years away though :-(
  • [15:34:52] <sakoman> I imagine there would be lots or internal debate and review
  • [15:35:02] <sakoman> s/or/of/
  • [15:35:40] <khasim> yeah.. I had started one when I was doing the MMC booting stuff, but thought coding is better than debate :)
  • [15:36:31] <khasim> I will have to restart the same. Every one thinks u-boot and x-loader are simple :(
  • [15:37:56] <khasim> Actually with every ES version change the ROM table might change as well, but if coded properly then it will be just a simple task to accomplish.
  • [15:37:56] <dirk2> khasim: u-boot nand: If you need a starter: http://rafb.net/p/xBSVs240.html
  • [15:38:11] <dirk2> Did the easy stuff and disabled oob
  • [15:38:36] <dirk2> oob is disabled in http://www.sakoman.net/cgi-bin/gitweb.cgi?p=u-boot-omap3.git;a=blobdiff;f=cpu/arm926ejs/davinci/nand.c;h=43041b635c0fdcc39a032036ca7f09e5a2e4c8a0;hp=36468e6c3a00920d95a9c0f08c2e75e7bf160d87;hb=cfa460adfdefcc30d104e1a9ee44994ee349bb7b;hpb=cd82919e6c8a73b363a26f34b734923844e52d1c too
  • [15:39:28] * NTNP[U] (n=u@193.10.100-84.rev.gaoland.net) has joined #beagle
  • [15:39:31] <khasim> dirk2: will look into this
  • [15:39:46] <dirk2> The comment for this commit is: "... Many
  • [15:39:46] <dirk2> drivers were trivial to port, but some were not so trivial"
  • [15:39:46] <sakoman> dirk2: does that build, or is it just a jump start for khasim's virtual team?
  • [15:40:05] <khasim> dirk2: will you be posting this patch to steve's tree
  • [15:40:07] <dirk2> It builds. But don't know if it will run
  • [15:40:19] <NTNP[U]> hi all
  • [15:40:28] <khasim> dirk2: ok, will start from here
  • [15:40:29] <sakoman> dirk2: the comment also said something like "I haven't tested any of this tuff cause I don't have the hw" ;-)
  • [15:40:53] <NTNP[U]> It is impossible for me to bring up the beagle board for the 1st time with mmc. Anyone can help me please ?
  • [15:40:59] <dirk2> sakoman: Yes, and "The
  • [15:40:59] <dirk2> following drivers must be examined carefully and maybe rewritten to
  • [15:40:59] <dirk2> some degree:"
  • [15:41:07] <dirk2> cpu/arm926ejs/davinci/nand.c
  • [15:41:49] <sakoman> dirk2: do you want to submit a patch or should we wait for khasim?
  • [15:41:50] <NTNP[U]> I formatted the mmc card in FAT32, made one partition, copied MLO of Koen, u-boot and uImage to MMC. After I inserted it in the board and powered up
  • [15:42:04] <dirk2> khasim: posting: No, until we have an idea what to do with the oob stuff
  • [15:42:08] <NTNP[U]> but nothing shown in the screen.... no ...40T...
  • [15:42:24] <dirk2> sakoman: I really don't have an idea what this oob is. Will the driver work without?
  • [15:42:33] <Crofton> NTNP[U], is your serial cable correct?
  • [15:42:39] * dcramer (n=davec@missdsl.ebox.com) Quit ()
  • [15:42:46] <Crofton> 40T comes from the hw, without the SD card
  • [15:43:05] <Crofton> bbl, lunch time
  • [15:43:12] <dirk2> If the driver will work without what I disabled, then we can push it.
  • [15:43:18] <NTNP[U]> yes Crofton. I have an IDC 10 converter to DB9 and a null modem cable
  • [15:43:40] <sakoman> dirk2: It is the "out of bounds" storage in the nand, typically used for bad block management. I have no idea what side effects would come from disabling it
  • [15:43:51] <NTNP[U]> but I do not have an external power supply 5V so I powered the board up with a cable USB
  • [15:44:19] <sakoman> dirk2: are you brave enough to try it?
  • [15:44:20] <dirk2> sakoman: thanks :) Then we better have to check how to re-enable it in new driver style
  • [15:45:23] <NTNP[U]> Crofton: If 40T comes from hw, so I have to see it in minicom all the times when the board is powered up with or without MMC ?
  • [15:45:43] <NTNP[U]> I had to try to press user/reset button, but nothing....
  • [15:46:07] <dcordes> mru: ping?
  • [15:46:37] <dirk2> NTNP[U]: Are the LEDs enabled after power up and ~20s wait?
  • [15:46:49] <sakoman> dirk2: I hate having to simultaneously deal with upstream breakage on both u-boot and linux-omap :-(
  • [15:47:01] <khasim> sakoman: just to confirm you dont want me to try debbugger on your x-loader code right?
  • [15:47:11] <NTNP[U]> dirk2: yes all leds enabled, 4 exactly
  • [15:47:30] <sakoman> khasim: I think it would be good to see what is causing the hang so that we can fix it
  • [15:47:43] <dirk2> NTNP[U]: Then most probably x-laoder and uboot start from NAND, but your serial connection is broken
  • [15:48:02] <sakoman> I think the current setup code attempts to be more "universal" than the Beagle code, so it would be good to debug it
  • [15:48:09] <dirk2> NTNP[U]: If USR0 and USR1 LEDs are on, uboot is running
  • [15:48:27] * Olipro_ is now known as Olipro
  • [15:48:45] <sakoman> khasim: so if you are willing, I would love to know where it breaks
  • [15:48:52] <dirk2> sakoman: linux-omap breakage: Didn't try recent git yet after all the warnings here ;)
  • [15:48:55] <khasim> sakoman: you replaced the config_3430sdram_ddr(void) only right?
  • [15:49:02] <NTNP[U]> dirk2: yeah; usr1 and usr0 were on. When U pressed reset button, they were down and up again.
  • [15:49:19] <NTNP[U]> So u think i have a prob with serial cable ?
  • [15:49:25] <sakoman> khasim: correct, I just copied over the beagle code from code.google
  • [15:49:31] <dirk2> NTNP[U]: Then uboot is running. Check serial connection, and try with or without null modem cable
  • [15:49:32] <sakoman> just that one routine
  • [15:49:52] <dcordes> I have a problem setting up rs232. I connect like this: (mainboard com1 pins) <-> (idc10 to d-sub) <-> (null modem cable) <-> (idc10 to d-sub) <-> (beagle rs232 pins). I configured minicom as explained in the google wiki: "BAUD RATE - 115200, DATA - 8 bit, PARITY- none, STOP - 1bit, FLOW CONTROL - none"
  • [15:50:00] <dcordes> minicom only says modem offline
  • [15:50:50] <khasim> sakoman: so the problem should be only in this initialization, It will be difficult to use debugger for this
  • [15:51:05] <khasim> sakoman: I think we are missing some register configuration here
  • [15:51:06] <NTNP[U]> dirk2: ok. I will check it again. Is there any way to get ouput of boot phase without serial connection ?
  • [15:51:21] <khasim> let me review only this function
  • [15:51:23] <dcordes> is there
  • [15:51:31] <dcordes> sorry, what's a good alternative to minicom?
  • [15:52:02] <dirk2> NTNP[U]: If you use uboot from code.google.com (~700k) one, you should get something to speakers and logo at DVI attached monitor
  • [15:52:37] <sakoman> khasim: OK, I will do the same a bit later today. I am in the midst of getting the latest linux-omap kernel to build
  • [15:52:57] <dcordes> dirk2: it plays a sound when booted?
  • [15:53:01] <NTNP[U]> dirk2: my uboot comes from this link http://code.google.com/p/beagleboard/wiki/BeagleSourceCode (the second one for MMC)
  • [15:53:16] <khasim> sakoman: ok, I will look into it as well. thanks
  • [15:53:44] <NTNP[U]> dirk2: is there any thing on NAND flash by defaut (comes from factory) ?
  • [15:54:47] <NTNP[U]> because, when I powered up the board without MMC, i also have 4 leds enabled
  • [15:55:12] <khasim> NTNP[U]: have you tried reading this link http://code.google.com/p/beagleboard/wiki/outofthebox
  • [15:55:35] <dirk2> NTNP[U]: Yes, xloader and uboot are stored in NAND
  • [15:55:37] <dcordes> is there any place that details the minicom usage for beagleboard connection?
  • [15:55:47] <khasim> NTNP[U]: this should answer all your queries
  • [15:56:21] <khasim> dcordes: minicom should be configured for (BAUD RATE - 115200, DATA - 8 bit, PARITY- none, STOP - 1bit, FLOW CONTROL - none)
  • [15:56:23] <NTNP[U]> khasim: nope, I didnot got it. I will read now
  • [15:56:35] <dcordes> khasim: I did this, and it says modem offline.
  • [15:56:39] <dcordes> khasim: see above
  • [15:58:11] <dirk2> dcordes: offline is okay. Mine says offline, too, and works
  • [15:58:20] <dcordes> interesting
  • [15:59:21] <dcordes> I have 115200 8N1, no, no
  • [15:59:33] <khasim> dcordes: to test your minicom and serial cable interface, just don't connect the cable to board, on your serial connector (in cable) just short pin 3 & 2 and type some thing on host keyboard it should show the charaters that you typed
  • [15:59:34] <dcordes> and I hope /dev/ttyS0 is the right serial device
  • [15:59:58] <khasim> sometimes I see /dev/cua0 working as well
  • [16:00:25] <dcordes> khasim: I grep dmesg for tty and it only says ttyS0 so I think that's the right one
  • [16:00:31] <khasim> ok
  • [16:00:35] <NTNP[U]> I got color bars on TV and sound when reset the board
  • [16:00:36] <dcordes> also when I hexdump that device, it doesn't terminate
  • [16:00:49] <NTNP[U]> i think i have to change another serial cable :) thx all
  • [16:00:50] <dcordes> when I try the other ttyS devices, it termiantes directly
  • [16:00:57] <dcordes> khasim: does that inidcate anything?
  • [16:01:41] <ade_> dcordes: is your idc10 to d-sub 1 to 1 wiring? a lot of PCs come with these and they are not correct for beagle I believe
  • [16:02:04] <dcordes> ade_: you mean no twisted cables?
  • [16:02:23] * NishanthMenon (n=gnat@nat/ti/x-d6765b91229776fd) has joined #beagle
  • [16:02:51] <ade_> yeh, all I could find was a cable with a ribbon with no twists so I had to rewire
  • [16:03:11] <dcordes> both of my idc10 to d-sub have no twists.
  • [16:05:20] <ade_> sorry, actually, it should be pin1 to pin1, pin2 ro pin2 etc, but some cables are wired with IDC fittings on each end so the pin mappings are wrong
  • [16:05:40] <chakie> where can one get such an idc->db9 cable for the board?
  • [16:06:25] <ade_> I just hacked the cable (or rather our hardware guy did)
  • [16:07:02] <dcordes> I made one myself yesterday so I could easily rewire
  • [16:07:23] <dcordes> I bought two new ones and nullmodem cable because I thought I made it wrong
  • [16:08:21] <ade_> I have a null modem connected to my desktop, then the d-sub->idc10 connected to the beagle
  • [16:08:30] <chakie> ok, so one has to be made manually. i think that before the device is up and running serial access is very interesting
  • [16:08:33] <dcordes> ade_: I'm looking in the manual of my mainboard right now and it looks like the com1 pinout there matches exactly the one in the wiki
  • [16:09:09] <dcordes> ade_: ah so it's the beagle side that needs to be different?
  • [16:09:42] * Beagle0 (n=Beagle0@hunter.netmodule.com) Quit ()
  • [16:10:00] <dcordes> ade_: http://www.pccables.com/07120.htm picture two matches the com1 pinout description from my mainboard
  • [16:10:06] <ade_> yeh, well, I only have one idc-10 to d-sub, the null modem cable connects to the d-type on the back of the PC
  • [16:10:38] <dcordes> so what did you need to change on that one adapter?
  • [16:10:46] <dcordes> http://www.pccables.com/07120.htm mine is liek that
  • [16:10:53] * Ragha (n=Ragha@nat/ti/x-c394ac4b689e2883) has joined #beagle
  • [16:11:54] <ade_> sorry, yours looks correct
  • [16:12:00] <dcordes> ok
  • [16:12:04] <ade_> mine is not like that
  • [16:12:27] <dcordes> good so I can move on, I should try to figure if really ttyS0 is correct
  • [16:12:30] <ade_> sorry to waste your time in that case
  • [16:12:33] <dcordes> is that possible without cable fiddling?
  • [16:12:43] <dcordes> no thanks a lot. I'm happy about any hint
  • [16:13:01] <dcordes> it's good knowing that in case I have a similar problem again
  • [16:15:32] <dcordes> serial8250: ttyS0 at I/O 0x3f8 (irq = 4) is a 16550A <- that's the only device mentioned in the system logs and since my board has only that one, I think I can assume it's right
  • [16:17:56] * [-ip-] (n=[-ip-]@mnhm-590f6978.pool.einsundeins.de) has joined #beagle
  • [16:22:10] <dcordes> brb
  • [16:22:20] * dcordes (n=dcordes@unaffiliated/dcordes) Quit (Remote closed the connection)
  • [16:27:23] * dcordes (n=dcordes@unaffiliated/dcordes) has joined #beagle
  • [16:28:34] <ade_> if I am using OpenEmbedded to build the kernel, do I edit system/angstrom/work/beagleboard-angstrom-linux-gnueabi/linux-omap2-2.6.26-r60/git/arch/arm/configs/omap3_beagle_defconfig to change the settings?
  • [16:29:14] <dcordes> ade_: I would do "bitbake virtual/kernel -c devshell"
  • [16:29:28] <dcordes> the .config should be in place by then so you can edit it
  • [16:29:38] <dcordes> that's the common way
  • [16:30:03] <ade_> cool. cheers. I'm a real noob when it comes to OE and bitbake
  • [16:30:04] <dcordes> assuming virtual/kernel builds your beagle kernel, otherwise specify the according pname
  • [16:30:22] <dcordes> but that should be the case if the configured MACHINE is beagle
  • [16:30:24] <chakie> heh, i found a idc10->dsub cable in my "closet full of old cruft"
  • [16:31:04] <dcordes> ade_: I just looked in my bios setup what it has to offer about serial. I could just edit the base address for the port or disable it
  • [16:31:13] <ade_> yeh, my machine is beagle. I've gone through the tutorials and built a kernel and fs and they seem to work ok
  • [16:31:52] <dcordes> ok then do the devshell thing. it will open a terminal with workdir as beagle kernel source
  • [16:32:14] <dcordes> afterwards only compile using "bitbake virtual/kernel -c compile"
  • [16:33:04] <dcordes> personally, when I need to change something about a kernel source in OE, I always copy the patched and configured workdir and edit and compile manually, w/o bitbake
  • [16:33:10] <dcordes> and just use the OE toolchain
  • [16:33:26] <ade_> I'm running this through ssh
  • [16:34:09] <ade_> so it can't open the display
  • [16:34:33] <ade_> basically I'm doing this on my home PC from work because it's faster
  • [16:34:44] <ade_> maybe I'll wait until I get home
  • [16:35:05] <dcordes> so probably devshell task will not work because it can't open the new terminal. in that case you can do "bitbake virtual/kernel -c patch" so it will stop after patching, then you can cd to the source, edit it and "bitbake virtual/kernel -c compile"
  • [16:35:15] <ade_> cross-compiling outside of bitbake it my next learning curve
  • [16:35:46] <dcordes> only specify ARCH=arm and CROSSCOMPILE=where/is/OE/toolchain-
  • [16:36:04] <dcordes> that's all the magic about cross-compiling a kernel
  • [16:36:52] <dcordes> chakie: are you connecting the pins 1:1 or twist somethign?
  • [16:37:27] * dcramer (n=davec@missdsl.ebox.com) has joined #beagle
  • [16:37:37] <chakie> dcordes: i haven't connected anything yet, i'm just gatherin up supplies. i haven't even ordered a board yet...
  • [16:38:12] <dcordes> I see
  • [16:38:37] <dcordes> that's so screwed the whole serial conenction is a complete mystery to me
  • [16:38:53] * dcramer (n=davec@missdsl.ebox.com) Quit (Client Quit)
  • [16:39:13] <chakie> are you sure the cable is connected the right way?
  • [16:39:21] <chakie> i did that mistake at work once
  • [16:39:24] <dcordes> the pin connection is 1:1 I think, only good quality bought cables
  • [16:39:41] <chakie> i mean, rotated the right way
  • [16:40:00] <dcordes> on the desktop mainboard, it was only possible to connect it in one way
  • [16:40:20] <chakie> but on the beagle?
  • [16:40:25] <chakie> there you can do it wrong
  • [16:40:29] <dcordes> and my adapter looks just like the one suggested in the mainboard documentation
  • [16:40:44] <dcordes> I already tried it in both ways on the beagle
  • [16:40:58] <dcordes> I will just retry both positions
  • [16:41:04] <chakie> ok
  • [16:41:39] <dirk2> C70 remove fix for serial hang/32kHz clock issue is "official" now: "Work-around of using MPU timer has been confirmed. Board revision B5 removes capacitor C70, which improves the 32kHz clock quality and avoids this hang-up" http://code.google.com/p/beagleboard/issues/detail?id=22
  • [16:41:51] <dcordes> what should happen in minicom if it would work?
  • [16:43:08] <dirk2> dcordes: http://code.google.com/p/beagleboard/wiki/outofthebox
  • [16:43:21] <dirk2> -> Experience
  • [16:44:47] <dcordes> nade de nada
  • [16:44:56] <dcordes> it doesn't print anything
  • [16:45:08] <dcordes> btw I'
  • [16:45:21] <dcordes> btw I'm sure the board is not broken. it displays the logo to dvi
  • [16:45:31] <dcordes> maybe the rs232 port is?
  • [16:46:41] <dirk2> anybody with a flickr picture of C70 location?
  • [16:47:02] <dcordes> I can see 71
  • [16:49:10] <dcordes> software and hardware flow control must both be off, right?
  • [16:52:01] <mru> yes
  • [16:52:25] <dcordes> no joy.
  • [16:53:10] <chakie> tried with some other app than minicom?
  • [16:53:14] <dcordes> is it necessary to change other options than serial setup? I read something about the phone stuff setup in a similar context.
  • [16:53:23] <dcordes> chakie: no, I don't know any other apps
  • [16:53:55] <dcordes> can you suggest one?
  • [16:54:07] <ade_> I'm using gtkterm and it works ok if thats any help (on ubuntu)
  • [16:54:27] <chakie> dcordes: i had one that i used when i couldn't get minicom configured. let me see if i can find it. linux, i hope?
  • [16:54:37] <dcordes> ubuntu
  • [16:54:52] <chakie> http://cutecom.sourceforge.net/
  • [16:54:57] <mru> minicom should be fine
  • [16:55:01] <chakie> *very* bare bones, but also easy
  • [16:55:31] <chakie> mru: yes, it's the de facto standard. but it also does *a lot*
  • [16:56:26] <sakoman> dirk2: I might as well apply your nand patch -- beagle build is broken anyway and this will at least make it easier for khasim/others who might want to leverage your work
  • [16:56:32] <dcordes> gtkterm seems to behave just like minicom
  • [16:56:34] <sakoman> OK with you?
  • [16:58:00] <dirk2> sakoman: yes. What else is broken? U-Boot builds fine with this patch
  • [16:59:17] <sakoman> dirk2: I haven't checked beagle, but Overo built from the tree seems just fine. I've been using it since yesterday with no issue
  • [16:59:47] <sakoman> So the upstream onenand changes are good
  • [17:00:02] <sakoman> Sadly they broke Mani's patch though
  • [17:00:39] <sakoman> I extracted what I could from his patch and asked him to give it another try
  • [17:01:18] <dirk2> sakoman: yes, read it. I just rebuilt U-Boot beagle and it builds
  • [17:01:20] <dcordes> I'm going nuts on this
  • [17:01:41] <dirk2> dcordes: Do you have a null modem cable to test?
  • [17:01:59] <chakie> isn't a null modem cable required?
  • [17:02:01] <dcordes> dirk2: yes, I use a null modem cable to conenct. above I pointed out my connection
  • [17:02:14] <dcordes> don't have it in scrollback
  • [17:02:22] <sakoman> dcordes: FWIW I use kermit on Ubuntu 8.04
  • [17:02:39] <dirk2> chakie: Depends on the serial cable ;)
  • [17:02:52] <dcordes> dirk2: the box says null modem connect cable
  • [17:03:12] <chakie> dirk2: assuming a dc10->db9 cable i guess a null modem would be required to connect to a pc?
  • [17:03:14] <dirk2> dcordes: Or maybe try a USB serial converter
  • [17:03:16] <dcordes> it is female on both ends and the d-sub to idc10 adapters both have female d-sub
  • [17:03:37] <dirk2> dcordes: Try all combinations you have ;)
  • [17:03:41] <dcordes> dirk2: I don't believe the onboard comport is broken
  • [17:03:43] <dcordes> I did
  • [17:04:03] <dcordes> I even have a custom made cable I made yesterday aynd that doesn't work either
  • [17:04:32] <chakie> could you have some software running that "steals" the tty?
  • [17:05:01] <dcordes> sakoman: I install gkermit and ckermit, how do you run/configure it?
  • [17:05:13] <dcordes> chakie: how can I check? lsof?
  • [17:05:27] <chakie> dcordes: could try that at least
  • [17:05:40] <chakie> dcordes: just guessing though
  • [17:05:41] <banderson> I had that happen once with a usb-serial converter...some gpsd latched onto it and couldn't understand why serial wasn't working
  • [17:05:51] <sakoman> dcordes: I do: 'kermit -l /dev/ttyUSB0'
  • [17:06:20] <sakoman> then 'take kermit-setup' at the kermit prompt
  • [17:06:29] <sakoman> will post that file contents in a moment
  • [17:06:51] <banderson> I would also suggest putting a loopback connector on it and see if that works...
  • [17:07:27] <dcordes> banderson: is that==shorting pin 1 and 2, pin 3 and 4 etc?
  • [17:07:53] <banderson> dcordes: for this you only need pins 2,3 ..true loopback has more but not need
  • [17:07:59] <sakoman> dcordes: http://rafb.net/p/jVXdOT98.html for kermit-setup file
  • [17:08:00] <banderson> ..for this test
  • [17:08:38] <dcordes> thanks, ok I'm testing kermit with sakoman's settings, then short pin 2 and 3 on the connector on my pc
  • [17:09:08] <banderson> dcordes: I use a loopback because then I can use it to figure out where the problem is...you can start with it right at the usb-serial converter
  • [17:10:15] <dcordes> I have no usb-serial, but onboard serial port
  • [17:10:20] <dcordes> on my desktop's mainboard
  • [17:10:25] <banderson> dcordes: ok
  • [17:10:48] <chakie> do you need to modprobe something?
  • [17:11:41] <dcordes> the serial module should be there or it wouldn't say [ 22.742003] serial8250: ttyS2 at I/O 0x3e8 (irq = 4) is a 16550A
  • [17:11:59] <dcordes> (ttyS2 was changed from ttyS0 because I edited my bios settings accordingly)
  • [17:12:15] <banderson> chakie: I do think it would be worth doing a lsof | grep "ttyS" just to be sure...
  • [17:12:34] <dcordes> sakoman: I ran kermit -l /dev/ttyS0 and did a "take <your-conf-file>"
  • [17:12:36] <chakie> banderson: of course
  • [17:12:47] <dcordes> banderson: did that, nothing does access it.
  • [17:12:57] <banderson> dcordes: dang
  • [17:13:01] <dcordes> lsof | grep ttyS is empty
  • [17:18:18] <dcordes> banderson: ok, I shorted the 2 and 3 on the d-sub connector of the null modem cable. the other side is connected to the back of my pc
  • [17:19:20] <dcordes> banderson: now how can I try?
  • [17:19:48] <sakoman> dcordes: if you are sure you are using /dev/ttyS2, than that is what you should use on the command line
  • [17:19:55] <banderson> dcordes: if you type into com program you should see the letters you type
  • [17:21:02] <chakie> could some cable be broken? can you check it?
  • [17:22:10] <banderson> dcordes: the idea of the loopback is that the data you send is routed back. So if you type the leter 'a' you should get a letter 'a'
  • [17:23:26] <dcordes> banderson: minicom behaves the same as if the beagle was connected with 2 and 3 bridged
  • [17:24:06] <banderson> dcordes: I assume that means you don't see anything on screen?
  • [17:25:16] <dcordes> yea it says welcome, serial device /dev/ttyS0
  • [17:25:21] <dcordes> and a cursor is blinking
  • [17:25:35] <dcordes> while I type things, that cursor does stop blinking, but nothing happens.
  • [17:25:36] <banderson> dcordes: try ttyS2
  • [17:26:15] <dcordes> sorry, I'm confusing the two. I'm using /dev/ttyS2 all the time. and that's the one shown in dmesg
  • [17:27:05] <dcordes> OPTIONS: I18n
  • [17:27:05] <dcordes> Compiled on Dec 10 2007, 10:36:19.
  • [17:27:05] <dcordes> Port /dev/ttyS2
  • [17:27:05] <dcordes> Press CTRL-A Z for help on special keys
  • [17:27:26] <dcordes> below that, the blinking curosr, and on the bottom a status line:
  • [17:27:30] <dcordes> CTRL-A Z for help |115200 8N1 | NOR | Minicom 2.3-rc | VT102 | Offline
  • [17:27:44] * valhalla (n=valhalla@81-174-21-16.dynamic.ngi.it) Quit ("Leaving")
  • [17:28:55] * NTNP[U] (n=u@193.10.100-84.rev.gaoland.net) Quit (Read error: 110 (Connection timed out))
  • [17:28:57] <dcordes> also I'm quite sure on the /dev/ttyS2 now because after I changed the base address in the computer's bios, it changed from 0 to 2
  • [17:29:40] <chakie> try them all, 0-3
  • [17:29:58] <dcordes> ok
  • [17:31:29] <dcordes> with ttyS0 minicom terminates saying Minicom: Device disappeared, aborting!
  • [17:31:37] <dcordes> as soon as I type one letter
  • [17:32:12] <dcordes> same with 1
  • [17:32:42] <dcordes> and 3
  • [17:34:04] <chakie> so 2 seems to be the one that does something :)
  • [17:34:26] <dcordes> yea
  • [17:35:01] <chakie> time for chips, beer, popcorn and some tv
  • [17:37:41] * khasi1 (n=a0393720@192.163.20.231) has joined #beagle
  • [17:38:56] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [17:39:59] <dcordes> that's odd: when I have the minicom session runnin on /dev/ttyS2 , lsof | grep ttyS still won't show anything
  • [17:40:21] <banderson> dcordes: Well I am running out of ideas quickly...maybe try making loopback that connects directly to pc...do you have solder equipment? and a female dsub?
  • [17:40:26] <banderson> dcordes: that is odd
  • [17:40:32] <banderson> dcordes: are you root?
  • [17:43:12] <dcordes> banderson: yes I have both. on the board itself there are exactly the same pins as on the beagle. I use a idc10-d-sub male
  • [17:43:45] <dcordes> if the pins would be neighbours I could short them with a jumper on the board
  • [17:44:03] <dcordes> and yes, I am root
  • [17:44:40] <dcordes> ah, but the lsof was ran as user. lsof with root reveals "minicom 7561 root 3u CHR 4,66 1729 /dev/ttyS2
  • [17:44:51] <banderson> dcordes: sorry not making myself clear. What I am talking about is just a female 9-pin dsub connector ...no wires attached to it at all....
  • [17:45:01] <dcordes> I have that, yea
  • [17:45:24] <dcordes> actually with wires, but they can be removed quickly
  • [17:46:17] <banderson> if you solder pins 2-3 together then just plug that and only that into pc then test if you see any characters
  • [17:48:33] * khasim (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [17:48:51] <dcordes> ok got you. so I can sort out it's not the null modem cable?
  • [17:49:31] <banderson> yes
  • [17:50:25] <banderson> dcordes: just like to troubleshoot by narrowing down where the problem is
  • [17:52:01] <banderson> dcordes: make sure that you disable any flow control
  • [17:52:05] <dcordes> ok I will try it tomorrow. if that doesn't work either I will try a serial device with usb connection
  • [17:52:19] <dcordes> I have flow control turned off all the time: both soft and hardware
  • [17:58:25] <dcordes> bye I'm off. thank you very much everybody and have a good day
  • [18:04:05] <ade_> dcordes: thanks for your help. my kernel is now compiled :)
  • [18:05:18] * dcramer (n=davec@dcdsl.ebox.com) has joined #beagle
  • [18:23:29] * BThompson (n=BThompso@nat/ti/x-d3899b3f11a6194b) Quit ("Trillian (http://www.ceruleanstudios.com")
  • [18:27:26] * khasi1 (n=a0393720@192.163.20.231) Quit (Remote closed the connection)
  • [19:14:19] * NTNP[U] (n=u@193.10.100-84.rev.gaoland.net) has joined #beagle
  • [19:23:37] * asterick (i=lemoncar@svn.rkiq.net) has joined #beagle
  • [19:23:40] * ade_ (n=ade@80.177.64.36) Quit ()
  • [19:27:44] * dirk2 (n=dirk@F31aa.f.strato-dslnet.de) has left #beagle
  • [19:58:34] <dcramer> where does one find a standard-a to mini-a connector
  • [19:58:40] <dcramer> or is there another way to connect this
  • [20:00:05] <banderson> dcramer: beagle site lists this one: http://www.vernier.com/accessories/access.html?usb-mini&template=basic.html
  • [20:00:16] <dcramer> and that's the only one there is ???
  • [20:00:27] <banderson> dcramer: not sure...
  • [20:00:38] <banderson> dcramer: try google :)
  • [20:00:48] <dcramer> I did
  • [20:00:53] <dcramer> and that one comes up first
  • [20:01:15] <banderson> dcramer: other option is to use one of these http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=AE1474-ND
  • [20:01:36] <dcramer> yeah, that wold work
  • [20:01:58] <banderson> dcramer: the last link is what I use
  • [20:02:04] <dcramer> thanks
  • [20:02:10] <banderson> np
  • [20:02:17] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
  • [20:02:37] <dcramer> they certainly managed to find some hard to get components on this board
  • [20:07:48] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [20:31:52] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) has joined #beagle
  • [20:32:58] * NTNP[U] (n=u@193.10.100-84.rev.gaoland.net) Quit ("Quitte")
  • [20:39:40] * guillaum1 (n=Guillaum@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) has joined #beagle
  • [20:40:46] * _Ade_ (n=Ade@78-86-195-216.zone2.bethere.co.uk) has joined #beagle
  • [20:43:24] * prpplague (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [20:44:00] * joshin (n=joshin@unaffiliated/joshin) has joined #beagle
  • [20:44:06] * GeneralAntilles (n=GAN8001@pdpc/supporter/active/generalantilles) has joined #beagle
  • [20:44:27] * GeneralAntilles just backordered his Beagle.
  • [20:45:24] <mru> these dogs are really popular
  • [20:45:55] <joshin> Where can I find a good description of the expansion bus?
  • [20:45:58] <GeneralAntilles> It's a damn impressive (and affordable!) piece of hardware.
  • [20:46:20] <joshin> I can finally replace my NSLU2's
  • [20:46:29] <GeneralAntilles> Plus, I'd like to get a head start on OMAP3 before Nokia ships an OMAP3 tablet.
  • [20:46:49] <mru> GeneralAntilles: do you some interesting project in mind?
  • [20:47:16] <banderson> joshin: http://www.beagleboard.org/uploads/Beagle_HRM_B4.pdf
  • [20:47:22] <joshin> thanks banderson
  • [20:47:38] <mru> I reckon it'll be at least a year before nokia ships an omap3 tablet
  • [20:48:12] <GeneralAntilles> mru, unlikely.
  • [20:48:20] <GeneralAntilles> No, nothing in particular in mind
  • [20:48:31] <GeneralAntilles> I might play with a STB approach just for kicks.
  • [20:48:31] <mru> what's unlikely?
  • [20:49:12] <GeneralAntilles> A year until an OMAP3 tablet.
  • [20:49:17] * jpao79 (n=a0193294@nat/ti/x-2550c41ae7f63fb8) has joined #beagle
  • [20:49:26] <Crofton> joshin, HRM
  • [20:49:52] <mru> GeneralAntilles: do you suppose it'll be more or less than a year?
  • [20:49:56] <GeneralAntilles> My money's on December-March.
  • [20:49:58] * Crofton just hunted down a digikey catalog so he can interface the beagle to an FPGA dev board :)
  • [20:50:01] <GeneralAntilles> 2008-2009
  • [20:50:18] <GeneralAntilles> We're WAY overdue for a generational upgrade to the tablets
  • [20:50:58] <mru> nokia are probably developing something
  • [20:51:13] <GeneralAntilles> The N800 shipped in January 2007.
  • [20:51:22] <mru> but I doubt they'll ship it with the current revision of omap3 silicon
  • [20:51:49] <GeneralAntilles> The N810 wasn't really an upgrade
  • [20:51:59] <GeneralAntilles> It was a little over a year between the 770 and the N800
  • [20:52:43] * asterick (i=lemoncar@svn.rkiq.net) Quit ()
  • [20:52:58] <mru> n810 has the same omap chip as n800, true
  • [20:53:20] <mru> the main difference there is the additional integrated peripherals
  • [20:53:43] <GeneralAntilles> Transflective screen, GPS, keyboard, and MiniSD
  • [20:54:08] <mru> the minisd proved unpopular though
  • [20:54:49] <GeneralAntilles> Very
  • [20:54:57] <GeneralAntilles> Dual-SD is a very nice thing to have
  • [20:56:12] <GeneralAntilles> Word on the street is that Fremantle is likely to have some very non-OMAP2-compatible stuff.
  • [20:56:23] <GeneralAntilles> 3D, etc.
  • [20:56:33] <mru> fremantle?
  • [20:56:38] <GeneralAntilles> Maemo 5
  • [20:56:47] <mru> I don't like maemo at all
  • [20:56:56] <jkridner|work> I only have an N800. I really want an N810 for the GPS and slide-out keyboard....
  • [20:56:59] <GeneralAntilles> http://en.wikipedia.org/wiki/Maemo_(operating_system)#Release_history
  • [20:57:08] <GeneralAntilles> What's not to like about Maemo?
  • [20:57:17] <jkridner|work> but, I've been holding out for hopefully better controls and an OMAP3.
  • [20:57:19] <mru> you have to everything just so
  • [20:57:26] <mru> *do
  • [20:57:30] <chakie> GeneralAntilles: the horrible toolkit?
  • [20:58:03] * jkridner|work wouldn't have likely been involved in Beagle if there was already an OMAP3-based Nokia Internet Tablet.
  • [20:58:27] <mru> beagle is much more fun
  • [20:58:29] <mru> no rules
  • [20:58:34] <GeneralAntilles> Just so?
  • [20:59:14] <mru> on maemo you have to stay within their frameworks
  • [20:59:28] * GeneralAntilles shrugs
  • [20:59:32] <GeneralAntilles> Just install Debian, then.
  • [20:59:43] <mru> it's probably fine for those who only want to knock up a gui quickly
  • [20:59:49] <mru> I don't like debian either
  • [20:59:57] <chakie> gtk is not for any sane developer :)
  • [20:59:58] <mru> don't like, and don't trust
  • [21:00:02] * Crofton|work wonders what mru likes?
  • [21:00:14] <mru> the bare metal
  • [21:00:18] <Crofton|work> :)
  • [21:08:28] * prpplague^2 (n=dave@mail.americanmicrosystems.com) has joined #beagle
  • [21:08:55] * prpplague (n=dave@mail.americanmicrosystems.com) Quit (Nick collision from services.)
  • [21:08:57] * prpplague^2 is now known as prpplague
  • [21:09:32] <GeneralAntilles> Bare metal is great and all, but I frequently like to have something I don't have to poke at too much to make work. ;)
  • [21:09:44] <mru> the fun is all in the poking
  • [21:10:16] <sakoman> dealing with header files that have moved is not fun
  • [21:10:33] <sakoman> finally back to building from linux-omap head
  • [21:10:37] <sakoman> :-)
  • [21:11:13] <mru> GeneralAntilles: my idea of fun: http://git.mansr.com/?p=ffmpeg.mru;a=blob;f=libavcodec/armv4l/h264dsp_neon.S;hb=arm
  • [21:11:40] <GeneralAntilles> Hehe, yeah, I can only handle stuff like that in small doses. ;)
  • [21:12:00] <sakoman> mru: you are one sick puppy ;-)
  • [21:12:01] <Crofton|work> the dsp/bridge/link/gw discusion on linux-omap is amusing
  • [21:12:24] <Crofton|work> It is highlighting alot of the issues we will have doing "open" work that uses the DSP effectively
  • [21:13:07] <mru> I have yet to start playing with the dsp
  • [21:13:18] <mru> never enough time...
  • [21:13:58] <mru> damn dayjob always has to come first ;-)
  • [21:14:28] <mru> sakoman: at least it's not intel sse code
  • [21:14:39] <sakoman> so true!
  • [21:15:16] <sakoman> hopefully that's not what you do for your day job!
  • [21:15:25] <mru> no
  • [21:15:32] <sakoman> good!
  • [21:15:40] <mru> I work on settop boxes
  • [21:15:52] <mru> mostly mips, some arm
  • [21:16:16] <mru> I avoid the various ST based ones when I can
  • [21:16:36] <sakoman> ah, so this is almost like therapy!
  • [21:17:10] <mru> yesterday I spent poking display controller registers over jtag
  • [21:17:49] <mru> the worst problems are the compiler bugs
  • [21:20:40] <uberfry> wow, arm asm is lots worse than ppc :O
  • [21:21:03] <mru> uberfry: what are you looking at?
  • [21:21:05] <mru> my code?
  • [21:21:08] <uberfry> http://git.mansr.com/?p=ffmpeg.mru;a=blob;f=libavcodec/armv4l/h264dsp_neon.S;hb=arm
  • [21:21:22] <mru> that's neon code, the new arm simd unit
  • [21:21:34] <uberfry> pretty complicated instruction set
  • [21:21:56] <mru> it's not very unlike the ppc altivec instruction set
  • [21:22:07] <uberfry> ahhhh I hate altivec :(
  • [21:22:17] <uberfry> the only instructions I know are tlbiel and tlbie
  • [21:23:02] <mru> I don't know altivec
  • [21:23:23] <mru> I've only flipped through the manual to get an idea of what it can do
  • [21:23:51] <mru> all modern simd units seem to have pretty much the same basic functions
  • [21:24:03] <mru> the more exotic instructions vary
  • [21:24:30] <uberfry> I had an idea
  • [21:24:35] <uberfry> but not sure if the arm is fast enough...
  • [21:24:43] <uberfry> how fast is the gpio?
  • [21:24:44] <mru> let's hear it anyway
  • [21:24:50] <mru> don't know
  • [21:24:58] <uberfry> I don't know if you know how much I hate dvi-d
  • [21:25:12] <mru> I don't
  • [21:25:13] <uberfry> I want to connect beagle to a psp lcd
  • [21:25:20] <mru> ah
  • [21:25:25] <uberfry> so instead of using a dvi-to-vga
  • [21:25:29] <uberfry> and adcs
  • [21:25:48] <mru> there are dvi to lvds converters
  • [21:25:57] <uberfry> well...I need parallel...
  • [21:25:59] <uberfry> anyways :)
  • [21:26:19] <uberfry> maybe you've heard yesterday that I want to have meters to see how much torque my car can do
  • [21:26:20] <uberfry> and so on
  • [21:26:24] <uberfry> high speed, etc.
  • [21:26:38] <uberfry> so I was thinking to use gpio + fpga
  • [21:26:40] <uberfry> + ram
  • [21:26:41] <mru> no, I missed that bit
  • [21:26:57] <uberfry> well anyway, I'll code an app that takes in the values using gpio
  • [21:27:14] <uberfry> and then writes differences of the pixels displayed into memory
  • [21:27:38] <uberfry> I'll split the screen into 4x4 tiles
  • [21:27:59] <uberfry> or even 16x16
  • [21:28:19] <uberfry> the software will update the screen every 5 seconds
  • [21:28:25] <uberfry> but differences are immediately updated
  • [21:28:54] <uberfry> it'll take the first image and the following image, substract one from another, write the differences into memory
  • [21:28:59] <uberfry> using spi or similar
  • [21:29:22] <GeneralAntilles> Somewhat related: http://openbossa.indt.org/carman/
  • [21:29:42] <uberfry> OOOOOOOOHHHHHHHHH nice!
  • [21:29:44] <uberfry> thanks!!!
  • [21:29:55] <GeneralAntilles> That's just the old one, too
  • [21:29:56] <GeneralAntilles> One sec
  • [21:30:29] <GeneralAntilles> http://www.marceloeduardo.com/blog/design/graphical-design/carman-is-also-alive-news-upcoming-update
  • [21:30:36] <GeneralAntilles> The update may be out this month
  • [21:30:45] <GeneralAntilles> Opensource, so you can gank whatever you need.
  • [21:30:54] <uberfry> looks really really nice
  • [21:31:01] <uberfry> thanks for the links
  • [21:31:05] <GeneralAntilles> INdT are the guys who did Canola
  • [21:34:50] <uberfry> hmmm for some reason I want to start from scratch tho
  • [21:35:27] <uberfry> but anyway
  • [21:35:38] <uberfry> is there any webbrowser for beagle that works well?
  • [21:35:53] <mru> is there any web browser that works well?
  • [21:36:14] <uberfry> I just need one for hotmail
  • [21:36:16] <uberfry> that's all
  • [21:36:27] <mru> you can run firefox
  • [21:36:37] <mru> whether it works well is a matter of opinion
  • [21:36:56] <dcramer> will the usb otg power a device, or does it need external power
  • [21:37:00] <jkridner|work> epiphany-webkit seems to be preferred for performance reasons, but I'm struggling with both a bit right now.
  • [21:37:02] <uberfry> hmmm I heard it works really really bad
  • [21:37:07] <jkridner|work> firefox dependencies didn't all install for me.
  • [21:37:19] <jkridner|work> I'm still stuck with my libfb issue today.
  • [21:37:35] <jkridner|work> uberfry: from whom?
  • [21:37:37] <mru> I haven't run any browser on my beagle
  • [21:37:49] <mru> you should talk to koen about these things
  • [21:37:59] <uberfry> jkridner: it was a link from beagle
  • [21:38:04] <GeneralAntilles> uberfry, you could grab GTKMozEmbed and find a reasonable front-end
  • [21:38:04] <jkridner|work> minimo and epiphany worked fine for me for a while, but I've screwed up my OE install and Epiphany is now crashing on me.
  • [21:38:17] <GeneralAntilles> Fennec'll probably be somewhere by winter for touchscreen stuff.
  • [21:38:22] <jkridner|work> koen is picky(TM)
  • [21:38:24] <uberfry> thanks for the tip GeneralAntilles
  • [21:38:25] <uberfry> btw
  • [21:38:34] <uberfry> is it possible to run windows mobile 2003 on beagle?
  • [21:38:41] <GeneralAntilles> lol
  • [21:39:04] <jkridner|work> taking a Beagle to the moon is possible.
  • [21:39:16] <jkridner|work> I'd think that running Windows Mobile would likely be a bit easier.
  • [21:39:29] <rsalveti> midori was running quite fine when I tested at n770 with mamona
  • [21:39:39] <rsalveti> guess that will also work well for beagle
  • [21:39:41] <prpplague> jkridner|work: hmm, i didn't think the beagle used the radhard version of omap
  • [21:39:46] <GeneralAntilles> s/n770/770/g
  • [21:39:51] <jkridner|work> radhard?
  • [21:39:59] <prpplague> radiation hardened
  • [21:40:10] <prpplague> <jkridner|work> taking a Beagle to the moon is possible.
  • [21:40:30] <jkridner|work> ah. well, you can take it with you, but you wouldn't necessarily hook it into the navigation system. :)
  • [21:40:58] * NishanthMenon dreams of beagle barking in a lead box ;)
  • [21:41:56] <Crofton> koen is off at a party all weekend
  • [21:42:03] <Crofton> and I am going "diving"
  • [21:42:19] * jkridner|work pounds head on desk one more time thinking it will free up some brain cells to figure out the next step to fix the pixman_fill issue without risking another 'opkg upgrade'.
  • [21:42:32] * Crofton is skeptical of a lake in West Virginia claiming to be as good as the Bahamas
  • [21:42:35] <uberfry> good night everybody
  • [21:42:38] <Crofton> gn
  • [21:42:44] <jkridner|work> good night uberfry.
  • [21:42:49] <prpplague> jkridner|work: i use that method frequently
  • [21:44:37] <jkridner|work> anyone have versions of libfb.so and libpixman-1.so
  • [21:44:44] <jkridner|work> that they believe work?
  • [21:45:12] * robclark (n=robclark@nat/ti/x-271a1384c1fa9311) Quit ()
  • [21:45:12] <mru> your problem might not be with either of those, actually
  • [21:45:47] <jkridner|work> I'm also getting a failure to load "glx", but I don't see where that is in my xorg.conf.
  • [21:46:10] <jkridner|work> it is commented out of Section "Module"
  • [21:46:22] <mru> glx errors are generally harmless unless you intend to use it
  • [21:46:33] <jkridner|work> just don't know why it is even there.
  • [21:47:13] <mru> libfb.so has no reference to libpixman
  • [21:47:18] <jkridner|work> jamvm and GNU classpath seem to be working rather well for me.
  • [21:47:22] <jkridner|work> that is my good news for the day.
  • [21:47:26] * BThompson (n=BThompso@cpe-76-185-93-11.tx.res.rr.com) has joined #beagle
  • [21:47:32] <mru> so it must be relying on libpixman being loaded already
  • [21:48:07] <mru> X module loading contains some rather dark magic
  • [21:48:10] * JoeBorn__ (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) has joined #beagle
  • [22:09:51] * prpplague (n=dave@mail.americanmicrosystems.com) Quit ("Leaving")
  • [22:14:55] <jkridner|work> anyone know why cc1 doesn't get installed?
  • [22:18:12] * JoeBorn__ is now known as JoeBorn
  • [22:20:51] * cbrake is now known as cbrake_away
  • [22:36:47] <jkridner|work> joy. Jetty works on Beagle!
  • [22:37:07] <jkridner|work> and Rhino.
  • [22:39:55] <jkridner|work> Helma is next to try. Quickly running out of space on my 1GB SD. :)
  • [22:39:55] * likewise (n=likewise@82-171-51-231.ip.telfort.nl) Quit ()
  • [22:40:57] <_Ade_> are there any browsers and/or flash9 that work well on beagle?
  • [22:41:57] <mru> no flash9
  • [22:42:26] <mru> jkridner|work seems to have gone mad with java
  • [22:51:03] * [-ip-] (n=[-ip-]@mnhm-590f6978.pool.einsundeins.de) Quit ("and thanks for all the fish")
  • [22:52:46] <jkridner|work> :)
  • [22:54:06] <jkridner|work> The java stuff is for a presentation on rapid UI development.
  • [22:54:26] <vlad_> is it possible to use Xomap on the beagleboard instead of the framebuffer server?
  • [22:54:30] <vlad_> (are there any advantages in doing so?)
  • [22:54:33] <jkridner|work> not the best route for those who know what they are doing, necessarily.
  • [22:54:54] <ds2> java *choke* *gag*
  • [22:55:18] <mru> those who don't know what they are doing shouldn't be doing
  • [22:57:18] <jkridner|work> I agree java sucks, but if it sucks so bad, tell me where else I can get an HTTP server API simpler than the servlet API?
  • [22:57:31] <ds2> shell script
  • [22:57:32] <ds2> =)
  • [22:57:36] <jkridner|work> I imagine there are some nice C implementations--and I'd rather code in C.
  • [22:57:41] <mru> you call that simple?
  • [22:57:43] <jkridner|work> now you are just being silly.
  • [22:57:43] <ds2> libcurl
  • [22:57:53] <ds2> <-- used inetd to write http servers
  • [22:57:53] <jkridner|work> libcurl? ack!
  • [22:58:07] <ds2> netcat makes a fine client
  • [22:58:10] <jkridner|work> have you written servlets?
  • [22:58:17] <jkridner|work> server, not client.
  • [22:58:28] <ds2> servers, yep. wrote a little man page server a while ago
  • [22:58:41] <ds2> also custom ones as a control interface
  • [22:58:55] <ds2> even did a entire HTTP proxy in perl
  • [22:58:57] <jkridner|work> inetd is a possibility.
  • [22:58:57] <ds2> =)
  • [22:59:05] <jkridner|work> but not a pretty one.
  • [22:59:16] <jkridner|work> perl is easy, but I don't see it doing so much better than Java.
  • [22:59:23] <ds2> my code works... asthetics is a whole different thing ;)
  • [22:59:33] <jkridner|work> actually, I'm writing my servlets in JavaScript!
  • [22:59:34] * NishanthMenon (n=gnat@nat/ti/x-d6765b91229776fd) Quit ("Friends help you move. Real friends help you move bodies.")
  • [22:59:44] * jkridner|work 's reputation completely obliterates.
  • [22:59:47] <ds2> not requiring the whole class path baggage is a big plus
  • [23:00:48] <jkridner|work> class path issues are horrible, but I'm promoting a canned environment (like Helma or others) where you simply drop jar files into a folder, then program in JavaScript.
  • [23:01:04] <jkridner|work> What I don't like about Helma is some of the boilerplate junk in the framework.
  • [23:01:16] <jkridner|work> argh. Now tar is giving me headaches.
  • [23:01:19] <jkridner|work> tar: invalid number ' 11442'
  • [23:01:28] <ds2> busybox tar or full gtar?
  • [23:01:34] <mru> if java apps are so portable, why do all them come with their own jvm?
  • [23:01:39] <jkridner|work> busybox.
  • [23:01:56] <ds2> use a full tar, the busybox tar doesn't understand all the different subtle details of tar
  • [23:01:58] <jkridner|work> replacing busybox tar now.
  • [23:02:10] <ds2> posix tar options I think is the term, IIRC
  • [23:02:16] <mru> busybox, keeps you busy
  • [23:02:22] <jkridner|work> I never said java apps were portable.
  • [23:02:35] <mru> no, but the java advocates do
  • [23:02:36] <jkridner|work> those are just bad promises from a deluded set of developers.
  • [23:02:47] <mru> write once, debug everywhere
  • [23:02:55] <jkridner|work> they pretty much crushed all possibilities of "write once, run anywhere"
  • [23:03:08] <ds2> it doesn't help java's case when there is J2SE, J2ME, follow by Java2 vs original java vs all the different subset features like CLDC, etc
  • [23:03:29] <jkridner|work> no matter what they do, the java will never be compatible between platforms.
  • [23:03:36] <ds2> besides, writing javascript is basically writing active script ;)
  • [23:03:56] <mru> writing javascript is for masochists
  • [23:04:15] <ds2> but then I know people who made a lot money just doing active script stuff ;)
  • [23:04:23] <jkridner|work> I think the language is rather nice, if you forget about the DOM or any other existing library set.
  • [23:04:36] <jkridner|work> there are some nastinesses, but it a lot cleaner than C++.
  • [23:04:52] <jkridner|work> Java seems more masochistic to me.
  • [23:05:05] <ds2> heh, but that argument only flies if your audience actually thought something of C++ ;)
  • [23:05:46] <mru> c++ is all the evils in one box
  • [23:06:01] <mru> you can take away a few and still have plenty left
  • [23:06:10] <jkridner|work> I'd rather write C. For me, JavaScript is little more than interpreted C with dynamic types and objects.
  • [23:06:32] <ds2> until I can write processor startup code in javascript... ;)
  • [23:06:42] * mru is trying to find a better register allocation in an assembler function
  • [23:06:54] <mru> damn np-hard problems
  • [23:07:00] <jkridner|work> JavaScript doesn't force you into a bunch of crappy semantic sugar.
  • [23:07:11] <mru> neither does C
  • [23:07:19] <jkridner|work> I love C.
  • [23:07:22] <mru> and C actually works
  • [23:07:24] <jkridner|work> don't get me wrong.
  • [23:07:30] <mru> few other languages can make that claim
  • [23:07:34] <jkridner|work> Where C works, that is the language for me.
  • [23:07:59] <jkridner|work> C also has a ton of worthless libraries, but the quirks are quite minimal.
  • [23:08:12] <jkridner|work> I'm just waiting for a good C interpreter.
  • [23:08:18] <jkridner|work> there are a few, but none are very good.
  • [23:08:22] * cian (n=cianh@cian.ws) has joined #beagle
  • [23:08:23] <jkridner|work> at least that I've seen.
  • [23:08:34] <jkridner|work> I need a dynamic language for what I'm trying to show.
  • [23:09:08] <jkridner|work> btw, Helma runs on Beagle!
  • [23:09:18] <Pavlov> js is fine
  • [23:09:19] <jkridner|work> Helma is the server used to run BeagleBoard.org.
  • [23:09:31] <jkridner|work> now I just need to get git compiled.
  • [23:09:32] <mru> there are many useless libraries written in c
  • [23:09:45] <jkridner|work> mru: fortunately we can ignore them. :)
  • [23:09:48] <mru> the difference is that you're not required to use them all
  • [23:10:02] <jkridner|work> you can ignore them in js too.
  • [23:10:20] <mru> in java they come with the jvm...
  • [23:10:52] <jkridner|work> btw, one debate in the office today was if some flash utility routines needed to be rewritten since they were originally written in C#.
  • [23:10:53] <mru> and why do java programmer believe that make isn't good enough for them?
  • [23:11:16] <jkridner|work> we had some fun poking the guy for "hating freedom" for using 'mono' on Linux.
  • [23:11:35] <jkridner|work> I'd rather use 'bitbake' than make!
  • [23:11:44] <jkridner|work> make is horrible!
  • [23:11:49] <mru> bitbake isn't an alternative to make
  • [23:11:54] <jkridner|work> ant could be worse.
  • [23:11:54] * Olipro_ (n=Olipro@unaffiliated/olipro) has joined #beagle
  • [23:12:09] <mru> make is very good if you take the time to learn it
  • [23:12:10] <jkridner|work> well, it provides many aspects that make is missing.
  • [23:12:24] <mru> take a look at the ffmpeg makefiles
  • [23:12:37] <mru> there's some serious wizardry going on there
  • [23:12:39] <jkridner|work> well, taking a lot of time to learn it seems like wasting people's time to me.
  • [23:12:48] <jkridner|work> it is very difficult to make a very generic makefile.
  • [23:13:03] <mru> ever tried to write *any* kind of ant file?
  • [23:13:20] <jkridner|work> yup.
  • [23:13:21] <mru> they're so specific they only work on one machine
  • [23:13:41] <jkridner|work> they seem to completely mix-up procedural and declarative programming.
  • [23:13:54] <jkridner|work> at least it is clear that 'make' is procedural.
  • [23:14:11] <mru> that's a common misconception
  • [23:14:11] * jkridner|work assumes mru is much more fond of procedural programming.
  • [23:14:17] <mru> make is all but procedural
  • [23:14:26] <jkridner|work> oh?
  • [23:14:55] <jkridner|work> well, how about it is very clear where make shifts from declarative to procedural?
  • [23:15:01] <mru> a makefile describes the dependencies between targets, and how to rebuild targets if their dependencies have changed
  • [23:15:03] <jkridner|work> with ant, the shift is very fuzzy.
  • [23:15:27] <mru> make works out which rules to invoke, and in what order
  • [23:16:07] <jkridner|work> if they would just fix up the syntax for the procedural sections of a makefile, then I would like make.
  • [23:16:16] <mru> it's when people try to use makefiles as some sort of scripting language things go horribly wrong
  • [23:16:21] <jkridner|work> the declarative syntax portions are fairly clean (on certain implementations).
  • [23:16:25] * rsalveti (n=salveti@200.184.118.132) Quit (Read error: 104 (Connection reset by peer))
  • [23:16:34] * rsalveti (n=salveti@200.184.118.132) has joined #beagle
  • [23:16:37] <mru> you mean the rule syntax?
  • [23:16:47] <jkridner|work> yes.
  • [23:16:47] <mru> that's just shell commands prefixed with a tab
  • [23:16:51] <mru> couldn't be simpler
  • [23:17:14] <jkridner|work> except that what shell commands are available are completely dependent on the system installation.
  • [23:17:26] <jkridner|work> you never know what you will be hitting with those commands.
  • [23:17:43] <mru> there are standards for that
  • [23:17:44] <jkridner|work> and we don't need to discuss the 'tab'.
  • [23:17:55] <jkridner|work> standards that are broken all over the place.
  • [23:19:41] * JoeBorn (n=jborn@dsl017-022-247.chi1.dsl.speakeasy.net) Quit (Read error: 110 (Connection timed out))
  • [23:20:00] <jkridner|work> Helma is running pretty snappy on Beagle.
  • [23:20:51] <jkridner|work> whereas compiling 'git' is still a task left to do.
  • [23:21:29] * guillaum1 (n=Guillaum@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) Quit ("Leaving.")
  • [23:21:55] * guillaum1 (n=gl@AMontsouris-153-1-67-145.w90-2.abo.wanadoo.fr) has joined #beagle
  • [23:24:35] <_Ade_> on the subject of make, does anyone know if you have to do anything to a standard autotools project to allow it to cross compile?
  • [23:24:54] <mru> all too often
  • [23:25:41] <_Ade_> I've only just started to use autotools so I'm still learning
  • [23:26:43] <mru> take my advice and stop
  • [23:27:04] <mru> they create more problems than they solve
  • [23:28:36] * Olipro (n=Olipro@unaffiliated/olipro) Quit (No route to host)
  • [23:28:45] <_Ade_> it has been bit of a learning curve
  • [23:29:58] <davr> i have no experience with beagle or creating autotools projects, but i've cross compiled for other ARM-based projects & sometimes it's as simple as ./configure --host=arm-elf-linux && make
  • [23:30:23] <mru> recommended reading: http://miller.emu.id.au/pmiller/books/rmch/
  • [23:30:25] <mru> about make
  • [23:30:48] <mru> davr: yes, sometimes
  • [23:31:04] <mru> and when it's not, fixing the problem can be very tricky indeed
  • [23:31:24] <jkridner|work> Crofton, koen: any way for me to force OE to replace packages? gcc is giving me headaches (cannot exec cc1). This worked for me on an older OE install.
  • [23:31:27] <_Ade_> mru: thanks
  • [23:32:34] <dcordes> jkridner|work: in build time or with opkg?
  • [23:32:53] <dcordes> _Ade_: you can find lots of examples of cross compile patches in the OE metadata
  • [23:33:19] <_Ade_> dcordes: thanks, I'll have a look
  • [23:33:46] <jkridner|work> with opkg
  • [23:34:04] <dcordes> --force-downgrade?
  • [23:34:17] <jkridner|work> I'm trying to do some prep for presentations I'm giving on native development with Beagle (128MB RAM).
  • [23:35:11] <jpao79> no more cross compiling? cool!
  • [23:35:33] <mru> I think I'll stick to building on my quad-core
  • [23:35:53] <mru> but soon we'll have quad-core cortex-a9 ;-)
  • [23:36:35] <dcordes> jkridner|work: opkg --force-reinstall foobar.ipk ?
  • [23:37:45] <dcordes> maybe you are missing a package?
  • [23:37:46] <jkridner|work> root@beagleboard:~/git-1.5.6.5# opkg --force-reinstall http://www.angstrom-distribution.org/feeds/2008/ipk/glibc/armv7a/base/task-native-sdk_1.0-r2_armv7a.ipk
  • [23:37:56] <jkridner|work> above line doesn't work.
  • [23:38:02] <dcordes> wget the pack first
  • [23:38:11] <dcordes> and specify ipk filename
  • [23:38:13] <jkridner|work> single dash.
  • [23:38:18] <jkridner|work> nope, that wasn't it.
  • [23:38:41] <jkridner|work> forgot the word 'install'.
  • [23:38:44] <dcordes> what does it say?
  • [23:38:50] <dcordes> ah right.. sorry
  • [23:39:26] <jkridner|work> Multiple packages (task-native-sdk and task-native-sdk) providing same name marked HOLD or PREFER.
  • [23:39:32] <jkridner|work> so, I added -force-reinstall.
  • [23:39:42] <jkridner|work> root@beagleboard:~# gcc hello.c -o hello
  • [23:39:42] <jkridner|work> gcc: error trying to exec 'cc1': execvp: No such file or directory
  • [23:39:57] <jkridner|work> probably the task is missing some dependency.
  • [23:40:17] <dcordes> what image are you using?
  • [23:40:33] <dcordes> gpe-image ?
  • [23:41:19] <jkridner|work> what is 'gpe-image'?
  • [23:42:11] <dcordes> uhm it's a meta package in OE which gives you a complete rootfs. maybe you are using koen's demo image?
  • [23:42:33] <jkridner|work> koen's demo image.
  • [23:43:24] <dcordes> I'll look at it maybe I can make a new image with the missing packages included
  • [23:43:31] <dcordes> depending on how much he modified the image
  • [23:44:09] <jkridner|work> well, to keep rebuilding my image is a pain. it would be nice if I could do the right opkg steps to fix the one I've got.
  • [23:45:06] <dcordes> you can check the dependy tree of the package easily with the web interface of the angstrom feed
  • [23:45:37] <jkridner|work> root@beagleboard:~# g++ hello.c -o hello
  • [23:45:37] <jkridner|work> root@beagleboard:~# ./hello
  • [23:45:37] <jkridner|work> Hello, World!
  • [23:45:37] <jkridner|work> root@beagleboard:~# gcc hello.c -o hello
  • [23:45:37] <jkridner|work> gcc: error trying to exec 'cc1': execvp: No such file or directory
  • [23:45:56] * rsalveti (n=salveti@200.184.118.132) Quit (Read error: 113 (No route to host))
  • [23:46:01] <jkridner|work> so, g++ is working, but not gcc.
  • [23:46:21] <dcordes> http://www.angstrom-distribution.org/repo/?action=details&pnm=task-native-sdk see Depends:
  • [23:46:36] <mru> g++ invokes cc1plus
  • [23:46:45] <jkridner|work> right.
  • [23:47:07] <mru> use gcc -v and see exactly what it's trying to do
  • [23:47:17] <mru> compare with g++ -v, if that's working
  • [23:47:27] <jkridner|work> root@beagleboard:/usr/libexec/gcc/arm-angstrom-linux-gnueabi/4.3.1# ls
  • [23:47:27] <jkridner|work> cc1plus collect2
  • [23:49:12] <dcordes> is cc1 the same cc?
  • [23:49:14] * dcordes ducks
  • [23:49:22] <jkridner|work> bye!
  • [23:49:43] <dcordes> maybe it's just a missing symlink
  • [23:49:47] * rsalveti (n=salveti@189.70.119.44) has joined #beagle
  • [23:49:51] <dcordes> bye
  • [23:50:38] <jkridner|work> I'm thinking it could be.
  • [23:54:11] <jkridner|work> good night all.
  • [23:54:14] * jkridner|work (n=a0321898@nat/ti/x-ae3b30d234107489) has left #beagle