**** BEGIN LOGGING AT Tue Sep 11 02:59:56 2007
Sep 11 03:02:51 <Writchie> no libelf-locale
Sep 11 03:03:16 <Writchie> what does the word commit mean in openmoko speak?
Sep 11 03:05:48 <abraxa_> Writchie: I'd assume this is in OE's court, not OpenMoko's
Sep 11 03:06:08 <Writchie> you're probably right 
Sep 11 03:06:28 <abraxa_> Aside from that, I fully agree
Sep 11 03:59:21 <aloril> (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_North_Texas]]
Sep 11 04:33:31 <CM> rwhitby: I agree with your mail to the ML, set rm_work as default in the makefile
Sep 11 04:33:45 <CM> The ones who doesn't want it know where to remove it. ;)
Sep 11 04:35:16 <CIA-23> openmoko: 03rwhitby * r158 10mokomakefile/trunk/Makefile: Enabled rm_work by default.
Sep 11 04:41:53 <CM> :-)
Sep 11 04:43:24 <rwhitby> CM: that's enough of a consensus for me ;-)
Sep 11 04:44:08 <CM> Hehe
Sep 11 04:44:30 <CM> I think it's a logical thing to do, especially when so many "new" ppl are using the makefile
Sep 11 04:44:39 <roh> japp
Sep 11 04:45:11 <roh> *sigh* .. /me was playing with pulseaudio, jack and darkice tonight... unsuccessful
Sep 11 04:45:23 <roh> everything that touches jack seems to go b0rk
Sep 11 04:46:09 <roh> good that we use pulse on the moko
Sep 11 04:48:41 <Writchie|zzzZZZz> rwhitby: with rm_work the default does that mean no patched sources or am I missing something?
Sep 11 04:49:57 <CM> Writchie|zzzZZZz: It just means that the temporary build files are removed after the build has finished
Sep 11 04:50:01 <CM> Saving space
Sep 11 04:52:40 <roh> hm.. somehow i think we should get rid of patches in patches/
Sep 11 04:53:07 <roh> is patches/openmoko-HEAD/libmokogsmd.patch tested? and/or is there a bug for it?
Sep 11 04:53:37 <rwhitby> yes and yes
Sep 11 04:55:10 <roh> hm.. lets see if i could commit that
Sep 11 04:59:37 <Writchie> CM: doesn't removing everything in build/tmp/work/* remove all of the patched sources leaving only the binaries??
Sep 11 05:00:11 <roh> rwhitby hm.. when i see that right i could commit it.. do you have a comment/bugnumber for that?
Sep 11 05:00:27 <rwhitby> http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=758
Sep 11 05:00:39 <rwhitby> Writchie: yes
Sep 11 05:01:59 <rwhitby> roh: comment is "Provide null default behaviours for CB and DS events to prevent unhandled event type critical errors being reported (we have enough real critical errors already)."
Sep 11 05:02:32 <blindcoder> moin
Sep 11 05:03:13 <CIA-23> openmoko: 03roh * r2948 10/trunk/src/target/OM-2007.2/libraries/libmokogsmd2/libmokogsmd/moko-gsmd-connection.c: commit patch from rwhitby - workaround for bug #758 - still needs proper implementation
Sep 11 05:04:00 <CIA-23> openmoko: 03rwhitby * r159 10mokomakefile/trunk/patches/openmoko-HEAD/libmokogsmd.patch: Removed libmokogsmd.patch (applied upstream)
Sep 11 05:05:19 <roh> rwhitby sorry.. i was too fast... i added a comment in the buglog
Sep 11 05:05:39 <rwhitby> np
Sep 11 05:52:49 <bedboi> did anyone use deps (graph-includes)?
Sep 11 05:53:23 <roh> bedboi ?
Sep 11 05:53:32 <roh> you mean bitbake -g?
Sep 11 05:53:34 <bedboi> to show dependencies in a code
Sep 11 05:55:28 <roh> nope.. i know lxr, ctags and similar stuff for cross-referencing code
Sep 11 05:56:13 <bedboi> i just need something to show dependency as a graph
Sep 11 05:56:16 <roh> but for graphing stuff i only know what ida pro gives you ;) and things on a buildsys-level like bitbake -g (spits out files to use with graphviz)
Sep 11 05:57:13 <roh> the only sad thing is, that the .dot files which bb generates on openmoko-devel-image are so fat and complex that ive never succeeded in generating a graph from it. i read it with vim
Sep 11 05:58:12 <bedboi> roh: i need it just for a single project (approx. 30 .c files and 20 includes)
Sep 11 05:58:29 <roh> hm.. you know ctags?
Sep 11 05:59:05 <bedboi> yep
Sep 11 06:03:14 <bedboi> btw what does bitbake use for plotting dep graphs?
Sep 11 06:03:27 <roh> propably generating a dot file from a perl script is the fastest way
Sep 11 06:04:04 <roh> bedboi it does not plot. it just spits out all dependencys and task-dependencys as dot file which the can be read by graphviz
Sep 11 06:04:32 <roh> s/can/could ;)
Sep 11 06:05:23 <bedboi> yep
Sep 11 06:08:52 <aloril> (script) wiki RecentChanges: [[OpenMoko_Local_Groups:_San_Francisco]] [[GPS]] [[GPS_Navigation]]
Sep 11 06:13:32 <Hopscotch> good morning
Sep 11 06:26:28 <wehttamb> counter?
Sep 11 06:26:29 <aloril> http://londerings.novalis.org/wlog/index.php?title=Neo1973_counter Got it, get yours from openmoko.com (and read topic)  (1808;252)
Sep 11 06:27:26 <bedboi> wehttamb: you should not believe aloril
Sep 11 06:27:28 <CM> Heh, I almost miss the constant "counter" asking ;)
Sep 11 06:27:36 <wehttamb> lol
Sep 11 06:27:44 <aloril> hehe, well I haven't updated that page for a while
Sep 11 06:27:55 <wehttamb> when will the mass market release be released?
Sep 11 06:28:05 <aloril> nobody knows
Sep 11 06:28:20 <CM> I'm guessing it's quite a while away
Sep 11 06:28:32 <CM> It needs to be stable or it will fail miserably
Sep 11 06:29:53 <wehttamb> im looking forward to it being released so i can get one
Sep 11 06:30:13 <bedboi> i guess that by taking into account the delays of the device release (2 months?) and the shipment, i will forecast 6 months
Sep 11 06:30:40 <vallor> y'alls gotsta go tah sleep now, but first, don't forget to untangle the cord for the mouse:  http://funtarded.com/pics/show/1449
Sep 11 06:30:50 <bedboi> and i think that december is too optimistic
Sep 11 06:31:06 <wehttamb> k
Sep 11 06:31:08 <rwhitby> my entry in the pool is Nov for GTA02 P0, Jan for GAT02 P1 and who knows for GTA02 P2 mass market release.
Sep 11 06:31:30 <bedboi> eheh
Sep 11 06:32:04 <bedboi> the GTA02 P2 could be a gaussian variable with huge standard deviation
Sep 11 06:32:52 <bedboi> btw if the orders@openmoko.com guys  do not reply my mail, i'll cancel the order very soon
Sep 11 06:33:36 <roh> bedboi there are just ~980 untouched tickets *couch*
Sep 11 06:33:41 <bedboi> AHAHHA
Sep 11 06:33:53 <roh> hey.. thats already quite some progess
Sep 11 06:34:04 <bedboi> so i guess i should cancel.
Sep 11 06:34:25 <bedboi> roh: who gave you this info?
Sep 11 06:34:42 <roh> that does not speed up things... if you got a rt id# youre in the queue
Sep 11 06:34:51 <roh> bedboi i can see things in rt
Sep 11 06:35:01 <bedboi> roh: i'll soon leave USA
Sep 11 06:35:06 <bedboi> in 3 weeks
Sep 11 06:35:18 <bedboi> i guess 3 weeks is still too optimistic
Sep 11 06:35:21 <roh> i just don't do sales, and have no clue of their exact state and plan..
Sep 11 06:35:34 <roh> bedboi so you ordered a gta01?
Sep 11 06:35:39 <bedboi> yep
Sep 11 06:36:11 <bedboi> the thing that piss me off is that the website should write WE DON'T KNOW WHEN WE ARE SHIPPING
Sep 11 06:36:16 <roh> hm.. 3 weeks...  dunno.. depends mostly on the time the next shipment hits fica
Sep 11 06:36:19 <bedboi> the shipping date could be in 1 or 2 years
Sep 11 06:37:35 <bedboi> "Any orders placed now will not be shipping before September 10, 2007." this is what the openmoko.org online shop states
Sep 11 06:37:43 <bedboi> it is way too generic imho
Sep 11 06:37:51 <bedboi> not before
Sep 11 06:37:57 <bedboi> you should give a bound
Sep 11 06:38:11 <bedboi> between september 10, 2007 and september, 2010
Sep 11 06:38:37 <bedboi> so people can just change their mind
Sep 11 06:40:47 <roh> bedboi i would be happy to know myself, really
Sep 11 06:41:16 <bedboi> :)
Sep 11 06:44:38 <daMaestro> ah yes, i do also like the new front page of the wiki
Sep 11 06:49:53 <CM> Me too, thanks rejon :)
Sep 11 06:57:10 <notserpe> how do I get into GTA02 P0 for working on graphics stuff?
Sep 11 07:04:43 <bedboi> yea the frontpage wiki looks really good now
Sep 11 07:09:39 <doc|home> notserpe: they don't exist
Sep 11 07:09:41 <doc|home> afaik
Sep 11 07:10:00 <notserpe> I'm just wondering about the process of getting a 
Sep 11 07:10:13 <notserpe> GTA02 before retail
Sep 11 07:10:43 <CM> My guess is that it will be the same way as it was with GTA01
Sep 11 07:11:02 * CM pushes http://wiki.openmoko.org/wiki/Community_nominees_for_GTA02_P0 once again
Sep 11 07:11:10 <notserpe> a nightmare?
Sep 11 07:11:13 * notserpe ducks
Sep 11 07:11:18 <CM> notserpe: :P Hehe
Sep 11 07:14:00 <notserpe> thats an interesting rumor. I'll try to start making friends here ;-)
Sep 11 07:14:20 <CM> notserpe: More than a rumour actually, mickeyl said it here
Sep 11 07:25:09 <notserpe> so who's started implementing opengl es for openmoko?
Sep 11 07:42:20 <doc|home> so, any update on when mass-release will be?
Sep 11 07:57:05 <bedboi> Your order is still fine so far. But we are temporarily short of stock.
Sep 11 07:57:05 <bedboi> Our next batch have enough amount to fulfill everyone, the ETA is
Sep 11 07:57:05 <bedboi> September 20.
Sep 11 07:57:54 <bedboi> if the date is that it would be fine
Sep 11 07:58:59 <Fatal> bedboi: heh, first they said 10th, then they said 24th to me. glad to see it decrease again :)
Sep 11 08:09:31 <wibbit> bedboi: is it possible to be told the number of GTA01's that have been shipped?
Sep 11 08:09:32 <aloril> (script) wiki RecentChanges: [[GPS_Navigation]] [[User:Alessandro]] [[Community_nominees_for_GTA02_P0]] [[Talk:MacOS_X]] [[User:Wisp]] [[Wishlist:Set_Local_Time]]
Sep 11 08:10:10 <bedboi> wibbit: i don't know, i've just sent a mail to orders@om.com
Sep 11 08:10:23 <wibbit> bedboi: ahh fair enough
Sep 11 08:10:48 <bedboi> Fatal: ahaha, so maybe that 20th Sept is just a rough approximation of the actual shipping date
Sep 11 08:11:08 <Fatal> bedboi: no idea, FIC tires me :/
Sep 11 08:11:11 <bedboi> if this is the case i will be quite pissed-off
Sep 11 08:11:43 <bedboi> Fatal: when you placed your order?
Sep 11 08:11:47 <Fatal> first day
Sep 11 08:12:16 <bedboi> lol
Sep 11 08:12:22 <bedboi> i would be really pissed off
Sep 11 08:12:53 <bedboi> *would have been 
Sep 11 08:13:27 <bedboi> well time to sleep, it's 2 am here.
Sep 11 08:13:51 <Fatal> I'm extremly unimpressed, especially when they said "Don't worry, I'll reserve the phones for your order." when there was a problem with my CC, so sent them another CC the day after, they then decided to not reply for over a week, had to mail jouston at his private mail, then they say they sold the units they "reserved" during that week
Sep 11 08:14:38 <bedboi> lol
Sep 11 08:14:55 <bedboi> i guess i'll cancel my order
Sep 11 08:15:14 <bedboi> 20 is way too close to my departure date from here
Sep 11 08:15:20 <Fatal> I still want mine, so I'm sticking with it
Sep 11 08:15:43 <Fatal> but I had expected to get more than 2 months out of it before the new hw comes out
Sep 11 08:16:04 <bedboi> Fatal: well, it's too bad, but i can't do that. the package would arrive and i would not be here.
Sep 11 08:16:11 <Fatal> ouch
Sep 11 08:16:21 <Fatal> mail them and change the delivery address then?
Sep 11 08:16:25 <bedboi> no way
Sep 11 08:16:32 <bedboi> i'll be in paris, then in my city.
Sep 11 08:16:49 <bedboi> i'm quite nomadic in these months
Sep 11 08:18:51 <Fatal> ah
Sep 11 08:19:38 <bedboi> this is the longest period in a place until January
Sep 11 08:20:07 <bedboi> well i'll definitely go to sleep
Sep 11 08:20:17 <squalyl> hi all
Sep 11 08:21:14 * bedboi is away: I'm busy
Sep 11 08:21:21 * bedboi is away: i'm sleeping
Sep 11 08:21:51 <squalyl> er... I noticed that bitbake forgot to build uboot image here! how is it possible??
Sep 11 08:22:08 <abraxa_> squalyl: Using MokoMakefile?
Sep 11 08:22:17 <squalyl> non, direct bitbake openmoko-devel-image
Sep 11 08:22:21 <abraxa_> Ah
Sep 11 08:22:28 <squalyl> and running bitbake u-boot fails :(
Sep 11 08:22:31 <CM> I'm not sure, but I think it was removed as a dependency to the kernel or something like that
Sep 11 08:22:39 <abraxa_> Yeah CM
Sep 11 08:22:53 <squalyl> so, we need to build it separately?
Sep 11 08:23:07 <CM> squalyl: You have a debug board I presume?
Sep 11 08:23:18 <squalyl> I have qemu :'(
Sep 11 08:23:39 <CM> Ah, ok. Then why do you need uboot to build? Is it needed?
Sep 11 08:23:47 * CM hasn't build a qemu image in a while
Sep 11 08:23:47 <roh> squalyl bitbake uboot-openmoko
Sep 11 08:23:56 <squalyl> good tip :D
Sep 11 08:24:02 <squalyl> I'll try this
Sep 11 08:24:16 <roh> squalyl you can also do bitbake -b uboot-openmoko_svn.bb to have a nonfrozen version.
Sep 11 08:24:30 <squalyl> thanks
Sep 11 08:24:48 <squalyl> i'm updating OE now
Sep 11 08:26:44 <rwhitby> uboot-openmoko fails to build for me
Sep 11 08:29:02 <roh> rwhitby ive just built svn ones recently..
Sep 11 08:29:16 <El_Salvador> hello everybody
Sep 11 08:30:23 <squalyl> hi
Sep 11 08:37:01 <roh> rwhitby i'll do some tests..
Sep 11 08:42:51 <squalyl> uboot-openmoko_svn compiled fine here
Sep 11 08:45:31 <rwhitby> hmm
Sep 11 08:46:17 <roh> squalyl when? yesterday?
Sep 11 08:46:25 <squalyl> 5 minutes ago
Sep 11 08:46:31 <roh> weird
Sep 11 08:47:52 <squalyl> and I just finished using it to flash quemu
Sep 11 08:48:24 <squalyl> OK it works fine
Sep 11 08:48:48 <squalyl> wow my first OM2007.2 image :)
Sep 11 08:56:09 <Malebob> is there a public svn repo for u-boot?
Sep 11 09:01:51 <CIA-23> openmoko: 03abraxa * r2951 10/trunk/src/target/OM-2007.2/applications/openmoko-mediaplayer2/src/persistent.c: Fixed default file chooser location
Sep 11 09:03:15 <RalphEichelberge> well i am here again
Sep 11 09:03:51 * SpeedEvil stabs RalphEichelberge with a sryinge and verifies his DNA.
Sep 11 09:04:01 <RalphEichelberge> and i have that same problem again that i had bevore
Sep 11 09:04:18 <RalphEichelberge> sorry i dont get that joke
Sep 11 09:04:32 <RalphEichelberge> a movie?
Sep 11 09:04:41 <SpeedEvil> Just checking you are in fact back, and it's not an imposter.
Sep 11 09:04:53 <RalphEichelberge> ah
Sep 11 09:05:07 * SpeedEvil has just woken up.
Sep 11 09:05:09 <RalphEichelberge> well, i am here in flesh and blood
Sep 11 09:05:20 <SpeedEvil> Or ratehr I woke up some time ago, and am not awake.
Sep 11 09:05:29 <RalphEichelberge> i see
Sep 11 09:05:34 <RalphEichelberge> cofe?
Sep 11 09:05:46 <RalphEichelberge> i would meet if i could
Sep 11 09:05:54 <RalphEichelberge> but u know
Sep 11 09:06:32 <RalphEichelberge> hey, maybe you can give me a hint..
Sep 11 09:07:13 <RalphEichelberge> i just have started a new build after "make update"
Sep 11 09:07:13 <SpeedEvil> If I can.
Sep 11 09:07:18 <RalphEichelberge> but
Sep 11 09:07:24 <SpeedEvil> I've not done the build environment yet.
Sep 11 09:07:32 <RalphEichelberge> oh
Sep 11 09:07:45 <SpeedEvil> just say - maybe someone is lurking.
Sep 11 09:07:57 <RalphEichelberge> well than --> no coffee :)
Sep 11 09:08:20 <RalphEichelberge> Speed where are you from?
Sep 11 09:08:59 <SpeedEvil> Scotland.
Sep 11 09:09:36 <RalphEichelberge> OK: SOMEONE !  does ".../install/cups-locale"  ring a bell?
Sep 11 09:10:09 <jgm> RalphEichelberge: what's the actual error?  Could you pastebin it?
Sep 11 09:10:15 <wibbit> Has there been any update regarding the gps chips in the GTA02?
Sep 11 09:10:18 <RalphEichelberge> Speed where are you working on in Moko?
Sep 11 09:10:27 <SpeedEvil> Power managment, and GPS
Sep 11 09:10:37 <SpeedEvil> and no wibbit.
Sep 11 09:10:53 <wibbit> SpeedEvil: bummer
Sep 11 09:11:00 <RalphEichelberge> wibbit: NO such File or Dir
Sep 11 09:11:15 <Malebob> wibbit: I'm eagerly waiting too
Sep 11 09:11:26 <RalphEichelberge> sorry i ment jgm
Sep 11 09:11:35 <wibbit> I think we are all eagerly waiting on this particular piece of news
Sep 11 09:11:40 <thos> wibbit, harald sent an e-mail to community
Sep 11 09:11:48 <SpeedEvil> When?
Sep 11 09:11:58 <SpeedEvil> ->thos
Sep 11 09:12:05 <RalphEichelberge> JGM: it doent exists
Sep 11 09:12:20 <wibbit> thos: do you have a link to the entry in the mailing list?
Sep 11 09:12:58 <RalphEichelberge> if i go "ln -s cups-locale-de cups-locale" it works till the next ...locale#
Sep 11 09:13:20 <thos> SpeedEvil, 4th Sept
Sep 11 09:13:39 <jgm> RalphEichelberge: sounds some other problem, let me check with the latest updates
Sep 11 09:16:19 <wibbit> thos: to the announce or community list?
Sep 11 09:17:36 <RalphEichelberge> SpeedEvil: is GPS working?
Sep 11 09:17:54 <SpeedEvil> The hardware works.
Sep 11 09:18:03 <thos> wibbit, community
Sep 11 09:18:07 <SpeedEvil> The binary driver works in OM2007.1
Sep 11 09:18:26 <SpeedEvil> This has not yet been officially released though.
Sep 11 09:18:28 <RalphEichelberge> what did change in 2007.2?
Sep 11 09:18:37 <SpeedEvil> it does not work on 2007.2 due to ABI changes
Sep 11 09:18:41 <CM> RalphEichelberge: Have you tried to disable locales or changed that in some way?
Sep 11 09:18:51 <SpeedEvil> It may work with a 2007.1 chroot
Sep 11 09:19:05 <RalphEichelberge> CM: no - how can i do this?
Sep 11 09:19:14 <RalphEichelberge> i dont need locals
Sep 11 09:19:17 <CM> SpeedEvil: Have you tried pavelm's chroot instructions?
Sep 11 09:19:31 <SpeedEvil> I haven't tried - I assume it will.
Sep 11 09:19:36 <SpeedEvil> I've not installed 2007.2
Sep 11 09:19:46 <CM> RalphEichelberge: Add ENABLE_BINARY_LOCALE_GENERATION = "0" to your local.conf
Sep 11 09:20:05 <RalphEichelberge> at /etc/..?
Sep 11 09:20:12 <CM> No, in bitbake
Sep 11 09:20:17 <RalphEichelberge> i c
Sep 11 09:20:22 <RalphEichelberge> ok i try
Sep 11 09:20:48 <CM> RalphEichelberge: Mine is /home/moko/build/conf/local.conf
Sep 11 09:21:14 <CIA-23> openmoko: 03abraxa * r2952 10/trunk/src/target/OM-2007.2/applications/openmoko-mediaplayer2/src/ (5 files): 
Sep 11 09:21:14 <CIA-23> openmoko: Fix doubled unref of the alsa sink (the playbin already does that)
Sep 11 09:21:14 <CIA-23> openmoko: Prevent file chooser path from being empty
Sep 11 09:21:14 <CIA-23> openmoko: Fix playlist list icon
Sep 11 09:21:14 <CIA-23> openmoko: Fix playlist editor list icon
Sep 11 09:21:20 <CM> RalphEichelberge: Add INHERIT += "rm_work" too if you don't already have it
Sep 11 09:21:57 <RalphEichelberge> CM: sorry but i cant find local.conf 
Sep 11 09:22:10 <RalphEichelberge> i found bitbake.conf but without that tag
Sep 11 09:22:25 <CM> RalphEichelberge: That's odd. In what dir do you run the makefile?
Sep 11 09:22:42 <CM> Maybe you should try a fresh setup from scratch ;)
Sep 11 09:22:59 <RalphEichelberge> oh well its in build found it!
Sep 11 09:23:05 <CM> :)
Sep 11 09:23:30 <RalphEichelberge> MACHINE = "fic-gta01"
Sep 11 09:23:30 <RalphEichelberge> DISTRO = "openmoko"
Sep 11 09:23:30 <RalphEichelberge> BUILD_ARCH = "i686"
Sep 11 09:23:30 <RalphEichelberge> conf/local.conf (END) 
Sep 11 09:23:43 <RalphEichelberge> thats all
Sep 11 09:23:54 <wibbit> thos: Okay, found it, I'd read that one before :\
Sep 11 09:24:04 <RalphEichelberge> :)
Sep 11 09:24:04 <CM> Ok, add the lines I suggested too. 
Sep 11 09:24:07 <wibbit> For any one else that may be interested. http://marc.info/?l=openmoko-community&m=118891977802077&w=2
Sep 11 09:25:00 <CM> wibbit: I know mickeyl has an eabi binary for the GTA01 phones
Sep 11 09:25:15 <CM> SpeedEvil: You saw that too didn't you?
Sep 11 09:25:29 <CM> GPS binary that is
Sep 11 09:25:42 <RalphEichelberge> oh no! not again! 
Sep 11 09:25:53 <SpeedEvil> yes.
Sep 11 09:26:05 <SpeedEvil> It would ahve been a bit nice to mention this earlier.
Sep 11 09:26:27 <CM> Yes, but I think he got it quite recently
Sep 11 09:28:38 <SpeedEvil> I mean the 'we may be changing the GPS' in somewhere official.
Sep 11 09:29:37 <mjr> it would.
Sep 11 09:32:21 <Malebob> wibbit: thanks
Sep 11 09:32:50 <wibbit> Malebob: no problem
Sep 11 09:33:27 <Malebob> the 'may' is a little bit sad ;)
Sep 11 09:39:29 <XorA> woowoo, I can call my neo
Sep 11 09:53:47 <RalphEichelberge> well its compiling still :) i hope i can start developing soon
Sep 11 09:54:34 <RalphEichelberge> SpeedEvil: my mission is to get GPS working - so how about working 2gether on that beast?
Sep 11 09:54:51 <RalphEichelberge> I have tht neo GTA02
Sep 11 09:56:27 <SpeedEvil> What do you mean by GPS working?
Sep 11 09:57:12 <XorA> RalphEichelberge: GTA02?
Sep 11 09:58:46 <SpeedEvil> Well - as I don't have a GTA02 - I can't do much with that.
Sep 11 09:59:52 <aloril> RalphEichelberge: what GPS chip your GTA02 has? (or did you mean GTA01B_v04?)
Sep 11 10:02:30 <RalphEichelberge> wenn u are right its GTA01V4
Sep 11 10:02:50 <RalphEichelberge> i meant the 2nd not the 0 version
Sep 11 10:04:05 <RalphEichelberge> i PM u Speed when i am ready to start hacking
Sep 11 10:04:30 <SpeedEvil> What do you want to do?
Sep 11 10:09:52 <aloril> (script) wiki RecentChanges: [[Bootloader]] [[GPS_Navigation]]
Sep 11 10:10:18 <RalphEichelberge> we are using gps-mouses or cellphones with java-api and client-software to do various applications
Sep 11 10:11:10 <RalphEichelberge> me, i have to get neo1973 working in order to develop our clients
Sep 11 10:11:35 <SpeedEvil> For the moment, just use the gpsd fake thing
Sep 11 10:11:51 <RalphEichelberge> for development?
Sep 11 10:12:18 <SpeedEvil> yes. you don't really need real GPS positions at least initially.
Sep 11 10:12:26 <RalphEichelberge> is there a java-api on moko?
Sep 11 10:12:29 <SpeedEvil> No.
Sep 11 10:12:34 <SpeedEvil> Not yet anyway.
Sep 11 10:12:47 <RalphEichelberge> well, thats not a problem
Sep 11 10:13:16 <RalphEichelberge> ok, and that fake thing is a wrapper for what you are developing?
Sep 11 10:13:41 <SpeedEvil> yes. gpsd can either take input from 'real' hardware, or it can take input from a file.
Sep 11 10:13:48 <SpeedEvil> Or it can produce random positions.
Sep 11 10:13:55 <SpeedEvil> well - a random wander
Sep 11 10:14:03 <RalphEichelberge> is there a api-doc for gpsd?
Sep 11 10:15:09 <SpeedEvil> yes - look for the project homepage - it's got all of the documentation. 
Sep 11 10:15:34 <RalphEichelberge> ok - thx
Sep 11 10:15:54 <SpeedEvil> It's pretty simple, you basically connect to a named pipe, ans send simple commands.
Sep 11 10:19:17 <RalphEichelberge> i ll go get something to eat now - cu later
Sep 11 10:29:30 <tuukkah> can i find the uboot image my phone shipped with somewhere?
Sep 11 10:30:53 <xkr47> tuukkah, wassup ?-O
Sep 11 10:31:21 <tuukkah> the new uboot crashes when i try to boot from or access the sd card
Sep 11 11:01:59 <tuukkah> my server died right after i said the new uboot crashes when trying to boot from sd :-(
Sep 11 11:04:06 <CM> tuukkah: Take the 2007.08 uboot, it works quite ok
Sep 11 11:05:17 <tuukkah> CM, that's the new one i'm talking about. the bug is like this: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=799
Sep 11 11:08:04 <CM> Oh, ok
Sep 11 11:08:11 <rwhitby> hooray - uboot now builds for me
Sep 11 11:08:17 <rwhitby> time to add it back into mokomakefile
Sep 11 11:08:42 <CM> My uboot have been broken from start... In the beginning I could get the boot menu back, but I haven't been able to do that lately
Sep 11 11:09:27 <tuukkah> CM, what about the usb console, can you access that?
Sep 11 11:09:38 <CM> No, not without the boot menu
Sep 11 11:09:47 <tuukkah> you should be able to
Sep 11 11:10:11 <CM> What I did before was to start the neo, ssh in, scp mtdblock1 over an then dd it to /dev/mtdblock1
Sep 11 11:10:21 <CM> And then restart
Sep 11 11:10:50 <CM> Also, I've only ever managed to boot using Factory Reset
Sep 11 11:11:02 <CM> Plain boot always fail on "Wrong magic number"
Sep 11 11:11:06 <tuukkah> there should be a short delay when uboot starts before it boots the kernel. during that time, you can gain access to the bootloader console
Sep 11 11:11:20 <CM> Ok. I'll give that a try
Sep 11 11:11:29 <CIA-23> openmoko: 03rwhitby * r160 10mokomakefile/trunk/Makefile: Added uboot-openmoko back into the image build
Sep 11 11:12:03 <tuukkah> by default the time period is unfortunatly just a couple of seconds
Sep 11 11:12:21 * CM starts a new makefile build
Sep 11 11:12:32 <Maledictus> rwhitby: can I browse the repo with the mokomakefile?
Sep 11 11:12:42 <rwhitby> which repo?
Sep 11 11:13:05 <tuukkah> the mokomakefile repo i assume :-)
Sep 11 11:13:13 <Maledictus> yep
Sep 11 11:14:00 <rwhitby> http://www.openembedded.org/filebrowser/org.openembedded.dev or http://www.openembedded.org/filebrowser/org.openembedded.dev or http://svn.nslu2-linux.org/svnroot/mokomakefile/ ??
Sep 11 11:14:23 <rwhitby> oops, second paste should be http://svnweb.openmoko.org/
Sep 11 11:14:44 <rwhitby> mokomakefile repo is the last one
Sep 11 11:14:48 <Maledictus> the 2nd one is it I guess
Sep 11 11:15:01 <rwhitby> is projects.openmoko.org working properly yet does anyone know?
Sep 11 11:16:14 <rwhitby> (as the only reason I have it on the nslu2-linux site is cause projects.openmoko.org anon svn access was broken for so long)
Sep 11 11:16:44 <Maledictus> ah, ok
Sep 11 11:18:46 <raster> there will come a day
Sep 11 11:19:11 <raster> wehn  we break all bonds of brotherhoot and my openmoko build will compile
Sep 11 11:19:16 <raster> but today is not this day
Sep 11 11:19:20 <CM> Hehe
Sep 11 11:19:30 <raster> today... we make update and try again!
Sep 11 11:23:06 <CM> NOTE: build 200709111313: completed
Sep 11 11:23:13 <tuukkah> of course i can run strings on the backup to see the image i'm after is "U-Boot 1.2.0-moko9_r8_70124c2602ae2d4c5d3dba05b482d91548242de8_0_2040 (May 19 2007 - 06:59:48)"
Sep 11 11:29:23 <squalyl> (not) funny http://theinquirer.net/?article=42235
Sep 11 11:29:55 <raster> actually
Sep 11 11:29:58 <raster> silly user most likely
Sep 11 11:30:12 <Maledictus> rwhitby: I don't understand why you use mkdosfs on make qemu?
Sep 11 11:30:34 <Maledictus> ahh, mom, I see
Sep 11 11:30:45 <Maledictus> you insert it as a SD card
Sep 11 11:31:28 <rwhitby> yep, I can't take credit for that - it was someone else's contribution
Sep 11 11:31:38 <Maledictus> problem is: under freebsd there is no mkdosfs, so I just touch the file
Sep 11 11:32:00 <CM> rwhitby: uboot_openmoko built fine on amd64 :)
Sep 11 11:32:22 <rwhitby> CM: builds fine here too now - I think I just had a bad git checkout or something
Sep 11 11:32:30 <CM> *nod*
Sep 11 11:32:53 <CM> Have not been much updates to om in the last two days
Sep 11 11:32:57 * wibbit thinks he should have checked to see how spicy his food was, before adding a load of chillies
Sep 11 11:33:02 <Maledictus> without the mkdosfs (and a minor patch to qemu-neo1973) the MokoMakefile runs just fine under freebsd
Sep 11 11:34:19 <raster> woot
Sep 11 11:34:23 <rwhitby> Maledictus: if you submit a patch that detects FreeBSD and does the right thing without affecting the Linux operation, I'll be happy to incorporate it.
Sep 11 11:34:24 <raster> there it goes again
Sep 11 11:34:47 <raster> filling up my disk
Sep 11 11:34:49 <Maledictus> rwhitby: ok, that's fair. Have to look how gmake works...
Sep 11 11:35:19 <CM> raster: Haven't heard anyone else having that problem. Weird
Sep 11 11:35:20 <raster>  27G -rw-r--r-- 1 raster raster  27G Sep 11 20:34 log.do_compile.6792
Sep 11 11:35:34 <raster> in ~/moko/build/tmp/work/armv4t-angstrom-linux-gnueabi/pcmciautils-014-r1/temp
Sep 11 11:35:45 <raster> ooh
Sep 11 11:35:47 <raster> up to 28g now
Sep 11 11:35:51 <raster> and i already ctrl-c'd it
Sep 11 11:35:54 <CM> Maybe time to bug some people in #oe?
Sep 11 11:36:00 <raster> it just may be...
Sep 11 11:36:48 <raster> ccache arm-angstrom-linux-gnueabi-gcc -march=armv4t -mtune=arm920t -I/home/raster/moko/build/tmp/work/armv4t-angstrom-linux-gnueabi/pcmciautils-014-r1/pcmciautils-014/src -isystem/home/raster/moko/build/tmp/staging/arm-angstrom-linux-gnueabi/include -fexpensive-optimizations -fomit-frame-pointer -frename-registers -Os -c -o src/pccardctl.o src/pccardctl.c
Sep 11 11:36:56 <raster> billions of lines of that...
Sep 11 11:36:58 <raster> niiiiiice
Sep 11 11:37:29 <raster> 32G log.do_compile.6792
Sep 11 11:37:36 <raster> it seems to have gotten caught in a yes loop
Sep 11 11:39:27 <rwhitby> and the best thing is that pcmciautils is not even used in the image, nor is it a dependency of anything in the image ;-)
Sep 11 11:41:16 <raster> yeah
Sep 11 11:41:20 <raster> greeeeeaaaaaat!
Sep 11 11:41:24 <raster> grrr
Sep 11 11:41:29 <raster> my patience has run thin for today
Sep 11 11:44:37 <zedstar> sometimes my icons/menu at the top vanishes and i need the keyboard...anyone know how to get it back without rebooting?
Sep 11 11:45:24 <raster> mmmm crashes
Sep 11 11:46:42 <tuukkah> zedstar, can you restart X from a terminal?
Sep 11 11:48:01 <zedstar> yeh...i cud try that..i notice openmoko-keyboard launches the old keyboard...wonder if there was a quick way of launching that icon bar
Sep 11 11:48:38 <raster> no idea
Sep 11 11:48:42 <raster> i need to replace that stuff...
Sep 11 11:48:47 <raster> if my om every builds...
Sep 11 11:48:49 <raster> ever
Sep 11 11:49:18 <tuukkah> zedstar, has to be something in /etc/matchbox/session
Sep 11 11:49:33 <zedstar> thanks tuukkah will have a look
Sep 11 12:05:39 <zedstar> tuukah: great found it there... matchbox-panel-2 --start-applets keyboard
Sep 11 12:07:55 <xkr47> :)
Sep 11 12:09:34 <aloril> (script) planet: Thomas Wood: OpenMoko Dates - First shots! http://blogs.gnome.org/thos/2007/09/11/openmoko-dates-first-shots/
Sep 11 12:09:34 <aloril> (script) wiki RecentChanges: [[User:Maledictus]] [[User:Marsan]] [[Manually_re-flashing_the_bootloader]]
Sep 11 12:19:02 <CM> Looks nice
Sep 11 12:26:31 <borg_> indeed
Sep 11 12:32:28 <josch_> to have python build into my openmoko image by the makefile - do i just have to edit a .bb file in openembedded/packages/tasks?
Sep 11 12:32:37 <Basher> :C
Sep 11 12:32:46 <Basher> shit, wrong channel - again
Sep 11 12:32:59 <Basher> sorry =)
Sep 11 12:36:03 <CM> josch_: That, or install via ipkg
Sep 11 12:38:51 <josch_> there has to be a wiki article or a howto somewhere how to do this? if not, I will try it with both ways and then create an article for the openmoko wiki myself
Sep 11 12:40:09 <Maledictus> rwhitby: how about this? http://bsdpaste.bsdgroup.de/583 (patch got mangled)
Sep 11 12:40:35 <Maledictus> this way I can use gmake MKDOSFS="truncate -s 500m openmoko/openmoko-sd.image; true" qemu
Sep 11 12:42:57 <zilt> josch_: just extend openmoko-devel-image.bb and make your own bb with the packages you want.
Sep 11 12:43:11 <zilt> josch_:  on the wiki, see: http://wiki.openmoko.org/wiki/User:CesarB/Using_a_local_overlay
Sep 11 12:43:54 <zilt> josch_: the last paragraph on that page " Changing what's included on the build"
Sep 11 12:45:01 <Sup3rkiddo> hey all, can anyone take a look at this and tell me whatsup..http://pastebin.ca/691852
Sep 11 12:45:13 <CM> josch_: Or, alternatively: http://wiki.openmoko.org/wiki/Package_management
Sep 11 12:46:19 <CM> Sup3rkiddo: That's a new one
Sep 11 12:46:59 <Sup3rkiddo> CM, was getting this since 2 days even after updating my tree...
Sep 11 12:47:21 <CM> Strange. I've build it two days ago, and updated this morning and built again
Sep 11 12:47:42 <CM> Sometimes I think svn or git or something just eats files when used with bitbake
Sep 11 12:48:23 <Sup3rkiddo> CM, maybe i will do a clean of the package in question, didnt try that yet :)
Sep 11 12:53:28 <CIA-23> openmoko: 03rwhitby * r161 10mokomakefile/trunk/Makefile: Allow mkdosfs to be overridden (for building on FreeBSD)
Sep 11 12:54:59 <Sup3rkiddo> CM, kick me in the butt....cleaning up solved it :|
Sep 11 12:57:34 <CM> Sup3rkiddo: Well, good thing you got around it :)
Sep 11 12:58:00 * CM continues to spam the channel with http://wiki.openmoko.org/wiki/Community_nominees_for_GTA02_P0
Sep 11 12:58:12 <mmazur> Why?
Sep 11 12:58:52 <CM> mmazur: Hoping that it will be of some use
Sep 11 12:58:54 <buz> who was doing the GPS Reverse Engineering?
Sep 11 12:59:03 <buz> Ian something?
Sep 11 12:59:15 <CM> Yes, and Elrond I think
Sep 11 13:00:26 <XorA> Ian == SpeedEvil I think
Sep 11 13:00:46 <jeddy3> what about dmitri?
Sep 11 13:00:56 <buz> can i get a neo for trolling the chan?
Sep 11 13:01:05 <CM> Think so,just looked at [Hammerhead/Protocol]
Sep 11 13:01:38 <CM> jeddy3: Yes, that's [User:Shadowjack]
Sep 11 13:01:48 <jeddy3> ah, nod
Sep 11 13:07:14 <RalphEichelberge> hey guys!
Sep 11 13:07:39 <RalphEichelberge> i dont understand how you build any rootfs
Sep 11 13:07:54 <RalphEichelberge> i cant even get over that locales problem
Sep 11 13:08:06 <RalphEichelberge> no matter what i am trying
Sep 11 13:08:21 <RalphEichelberge> i allways get this:
Sep 11 13:08:33 <RalphEichelberge> OSError: [Errno 2] No such file or directory: '/home/openmoko/moko/build/tmp/work/armv4t-angstrom-linux-gnueabi/cups-1.2.10-r2/install/cups-locale'
Sep 11 13:08:33 <RalphEichelberge> ERROR: Task 1430 (/home/openmoko/moko/openembedded/packages/cups/cups_1.2.10.bb, do_package_write_ipk) failed
Sep 11 13:08:33 <RalphEichelberge> NOTE: Tasks Summary: Attempted 3160 tasks of which 3160 didn't need to be rerun and 1 failed.
Sep 11 13:08:33 <RalphEichelberge> ERROR: '/home/openmoko/moko/openembedded/packages/cups/cups_1.2.10.bb' failed
Sep 11 13:08:33 <RalphEichelberge> make: *** [openmoko-devel-image] Fehler 1
Sep 11 13:08:35 <RalphEichelberge> openmoko@openmoko-celsius:~/moko$ 
Sep 11 13:08:40 <CM> ~pastebin
Sep 11 13:08:41 <apt> from memory, pastebin is a place to paste your stuff without flooding the channel - try  http://pastebin.ca, or http://channels.debian.net/paste, or http://rafb.net/paste/, or http://pastebin.com is usually painfully too slow and unresponsive to use, use one of the other pastebin sites, or dpaste.com is a very nice pastebin as well
Sep 11 13:10:21 <CM> RalphEichelberge: I haven't heard anyone have that problem, and I just finished a build after lunch today. 
Sep 11 13:10:22 <RalphEichelberge> ok
Sep 11 13:10:45 <CM> So there's something strange with your computer
Sep 11 13:10:51 <RalphEichelberge> well i got this thing have configured from my boss
Sep 11 13:11:31 <RalphEichelberge> maybe i should start all over again delete that moko dir and install it from skratch
Sep 11 13:11:37 <RalphEichelberge> scratch*
Sep 11 13:11:44 <RalphEichelberge> i will
Sep 11 13:11:46 <CM> RalphEichelberge: That's what usually works for me
Sep 11 13:11:49 <CM> :)
Sep 11 13:11:53 <RalphEichelberge> :)
Sep 11 13:12:15 <CM> Remember to set that line in the local.conf so you don't build all locales
Sep 11 13:12:24 <CM> it saves time, more than an hour for me
Sep 11 13:12:41 * CM is off to a meeting
**** BEGIN LOGGING AT Tue Sep 11 13:31:46 2007
Sep 11 13:46:38 <nnpiggy> Are there any other phase 0 folks out there? Have you managed to run gllin with a newer kernel?
Sep 11 13:47:30 * XorA never even tried with old kernel
Sep 11 13:50:58 <nnpiggy> The file command identifies it as a "2.4.0" executable. 
Sep 11 13:52:18 <XorA> ah, are you confusing the fact gllin is totally unable to run on 2007.2 openmoko with kernel problems?
Sep 11 13:52:55 <nnpiggy> Perhaps. Is there some other reason for it's inability to run?
Sep 11 13:53:13 <nnpiggy> I assume were just missing a module for the particular executable format.
Sep 11 13:53:39 <nnpiggy> Or did the Linux ARM folks break the ABI going from 2.4 to 2.6?
Sep 11 13:54:00 <aloril> OABI vs EABI?
Sep 11 13:54:23 <nnpiggy> I don't know either of the acronyms.
Sep 11 13:54:31 <aloril> different floating point emulation
Sep 11 13:54:41 <nnpiggy> Ah.
Sep 11 13:55:01 <aloril> copy old system to media/card, run old gllin chrooted?
Sep 11 13:56:24 <nnpiggy> That would still leave you with a newer kernel. It's clear that the kernel does not recognize gllin as a valid executable.
Sep 11 13:56:28 <CM> nnpiggy: http://wiki.openmoko.org/wiki/Gllin
Sep 11 13:56:43 <CM> At least it seems to work for pavelm
Sep 11 13:56:59 <nnpiggy> Hmm... It's just the libraries? Interesting...
Sep 11 13:57:42 <mickeyl> EABI-aware gllin is in progress
Sep 11 13:59:03 <nnpiggy> Drat. We have no ldd...
Sep 11 13:59:14 * CM waits patiently
Sep 11 13:59:26 * nnpiggy chomps bit :o)
Sep 11 13:59:40 <nnpiggy> Thanks, mickeyl.
Sep 11 13:59:57 <mickeyl> np. hopefully within this week
Sep 11 14:00:19 <nnpiggy> My primary interest in the neo1973 is as a GPS device.
Sep 11 14:00:25 <nnpiggy> Great!
Sep 11 14:00:27 <CM> nnpiggy: Same here
Sep 11 14:00:30 <SpeedEvil> #00
Sep 11 14:00:32 <SpeedEvil> Err
Sep 11 14:00:34 <SpeedEvil> Woo
Sep 11 14:00:36 <SpeedEvil> :)
Sep 11 14:00:59 <nnpiggy> I'm interested in using it for specimen notes for field collections.
Sep 11 14:01:12 <CM> nnpiggy: What are you collecting?
Sep 11 14:05:35 <SpeedEvil> I'm interested in doing openstreetmap stuff.
Sep 11 14:05:43 <cb22> SpeedEvil, same here
Sep 11 14:06:21 <nnpiggy> CM: mostly fungi.
Sep 11 14:06:43 <nnpiggy> I'm an amateur field mycologist.
Sep 11 14:08:48 <SpeedEvil> natural:mushroom
Sep 11 14:09:10 <nnpiggy> Yep.
Sep 11 14:09:14 <aloril> (script) wiki RecentChanges: [[Wishlist:Camera]]
Sep 11 14:09:48 <nnpiggy> Yeah, that would be handy too. But location information is sufficient for a lot of cool applications.
Sep 11 14:10:16 <SpeedEvil> Especially with a bluetooth thingy to do voice tags
Sep 11 14:10:43 <nnpiggy> Oh, I like that.
Sep 11 14:11:03 <nnpiggy> For that matter, the mike in the phone is sufficient for voice tagging.
Sep 11 14:11:32 <nnpiggy> I could see the birders getting excited about that...
Sep 11 14:11:44 <baird> nnpiggy: I've done a bit of that.. No-one in the Australian Fungi group has yet to ID this huge purple mushroom I found :) .. http://www.flickr.com/photos/chris_j_b/519954794/ ..&.. http://www.flickr.com/photos/chris_j_b/497808809/
Sep 11 14:12:22 <XorA> baird: those are cool assed mushrooms
Sep 11 14:12:49 <nnpiggy> It's a Cortinarius, but getting more specific than that is really difficult.
Sep 11 14:13:07 <nnpiggy> That is a rusty-brown spore print on the upper stipe, right?
Sep 11 14:13:36 <baird> My ex-girlfriend, who I was geocaching with at the time, wants me to name them after her if they're a new species.. -_-
Sep 11 14:13:39 <nnpiggy> Was it viscid?
Sep 11 14:14:18 <SpeedEvil> baird: if you were gay, that would make the obvious joke a lot better.
Sep 11 14:14:26 <SpeedEvil> 'was he a fungi'?
Sep 11 14:14:31 <baird> nnpiggy: no, that's very close to the original colour in the underside photo-- more orange than rust.
Sep 11 14:14:48 <SpeedEvil> Wacky shroom.
Sep 11 14:15:11 <nnpiggy> That's still within the range of colors for Corts. Did you take a spore print?
Sep 11 14:17:33 <baird> No, unfortunately-- didn't have the paper on me, plus I was blushwalking with whining people. Also, I was in a National Park, where removing flora and fuana is a big no-no.
Sep 11 14:18:10 <nnpiggy> Are you allowed to collect fruit in the National Parks?
Sep 11 14:18:28 <baird> Nope, not in NSW Australia, anyway.
Sep 11 14:18:33 <vallor> http://ap.google.com/article/ALeqM5iT1KAi6UEPN8LqZlvLnfsxP7ToKw  <--- I hope this isn't another "cold fusion hoax"
Sep 11 14:18:34 <nnpiggy> Bummer.
Sep 11 14:18:43 <vallor> ...Burning Seawater
Sep 11 14:18:45 <SpeedEvil> I wonder when portable rapid inexpensive DNA analysis tools will hit.
Sep 11 14:19:05 <vallor> http://news.google.com/news?hl=en&ned=us&q=burning+sea+water&btnG=Search+News
Sep 11 14:19:07 <XorA> SpeedEvil: depends how you define analysis
Sep 11 14:19:10 <SpeedEvil> poke it at the plant/animal, and it gives you a species name and care info.
Sep 11 14:19:13 <SpeedEvil> for $20
Sep 11 14:19:24 <SpeedEvil> In a pocket-sized device.
Sep 11 14:19:44 <baird> Cortinarius sounds very likely, now. There was an incredible colour in the mushrooms, which might be explained by the fluorescent orellanine toxin..
Sep 11 14:20:03 <nnpiggy> Tom Volk says there are about a dozen purple corts--it's not one of the four he mentions specificly...
Sep 11 14:20:38 <SpeedEvil> Be some time.
Sep 11 14:20:41 <baird> It's said that in Australia, only about 15% of fungi species have been identified and classified to date..
Sep 11 14:20:59 <nnpiggy> SpeedEvil--aside from having to wait a bit, there are services that will do the analysis for you for a little as $5 per sample. We're getting there...
Sep 11 14:21:09 <jeddy3> vallor, if true, it's pretty damn wicked
Sep 11 14:21:18 <madwoota> we work on identifying the ones that can kill you first
Sep 11 14:21:20 <madwoota> spiders, snakes, sharks :P
Sep 11 14:21:40 <nnpiggy> Yeah, even in North America there are a significant number of unnamed species. If you want to study Amanitas for Russulas, you can name species left and right...
Sep 11 14:21:58 <vallor> indeed jeddy3 -- I'm already thinking about how something like this could power our datacenter, hehe
Sep 11 14:22:01 <nnpiggy> I'm working on a new subspecies of Xylaria polymorpha now...
Sep 11 14:22:02 <SpeedEvil> nnpiggy: I know - I was more meaning an actual portable device - which is some way off. A full sequencer would be nice too. But not needed for most stuff.
Sep 11 14:22:22 * SpeedEvil adds to hardware:wishlist
Sep 11 14:22:43 <nnpiggy> If you just want a signature, a portable device could be built today. It would take a few hours for the device to do it's job, but you could build it.
Sep 11 14:23:06 <nnpiggy> DNA sequencer? Yeah, I want a full tricorder...
Sep 11 14:23:24 <SpeedEvil> I want holograms to work like they do in the movies.
Sep 11 14:24:09 <nnpiggy> baird, have you been studying fungi for long?
Sep 11 14:24:23 <nnpiggy> I developed my interest while living in Tassie.
Sep 11 14:25:50 <nnpiggy> http://www.fungiphoto.com/CTLG/IMAGES/IM/0408-06.jpg
Sep 11 14:26:04 <nnpiggy> I LOVE Taylor Lockwood's photos...
Sep 11 14:26:48 <baird> No, and it's hardly serious-- it came about while geocaching and bushwalking with a friend who drags his SLR everywhere, and we werw taking advantage of the March-April season.
Sep 11 14:27:25 <nnpiggy> Cortinarius is another gigantic genus--odds are very good your species in unnamed.
Sep 11 14:28:02 <nnpiggy> You should get yourself a permit and look for it again next Fall.
Sep 11 14:28:52 <nnpiggy> Oh well. Off to interview another candidate...
Sep 11 14:28:53 <vallor> wow, that fungus looks like Lovecraft's "The color out of space"
Sep 11 14:28:56 <vallor> or something like that
Sep 11 14:31:54 <baird> Could do... I would say it's very likely that the plant has a permanent home in the area, going on the 6-8 mushrooms of that species I found within 15 metres of each other.
Sep 11 14:33:29 <baird> Actually.. I have a friend who is a Biology Postgrad, and I believe she already has a State-wide collection permit.
Sep 11 14:34:40 <nnpiggy> Excellent. You should arrange to collect a few and dry them. I can't think of a cort person off the top of my head, but I know there are several in the US....
Sep 11 14:35:14 * CM listens to Fungus Amongus ;)
Sep 11 14:38:04 <nnpiggy> baird, do you have names for all of your "Glendale mushrooms"? I recognize a few...
Sep 11 14:41:13 <baird> nnpiggy: nothing beyond what I've mentioned with the posts. 
Sep 11 14:41:53 <nnpiggy> That ball looks like a Scleroderma. It's not S. citrinum--too smooth...
Sep 11 14:42:14 <nnpiggy> I agree with the assessment of Gymnopilus.
Sep 11 14:42:36 <nnpiggy> http://www.flickr.com/photos/chris_j_b/499213424/in/set-72157600217056554/ looks like another cort.
Sep 11 14:43:24 <nnpiggy> http://www.flickr.com/photos/chris_j_b/499213416/in/set-72157600217056554/ I know this genus, but I'm drawing a blank at the moment...
Sep 11 14:43:47 <baird> nnpiggy: I was just thinking that. "webcaps"
Sep 11 14:44:12 <nnpiggy> Nipple-topped Mushroom is a Mycena. Species may come to me in the next few minutes...
Sep 11 14:44:43 <nnpiggy> "Small Yellow Fungi" is Dacromyces palmatus (sp?)
Sep 11 14:44:43 * CM enjoys #OpenMushroom
Sep 11 14:44:47 <CM> :D
Sep 11 14:45:06 <nnpiggy> Has CM renamed this channel? :-)
Sep 11 14:45:37 <CM> I know chantrelle (sp?) and toadstool, but there my skillz ends
Sep 11 14:45:51 <nnpiggy> "Tall Brown Mushroom" is a Xerula, probably X. radicata (=Clitocybe radicata, Oudomansiella radicata)
Sep 11 14:46:13 <nnpiggy> p5152479 is another cort.
Sep 11 14:46:51 <nnpiggy> Purple-grey mushroom looks like Lepista nuda (= Clitocybe nuda, the blewit).
Sep 11 14:47:34 <torpor> w00t
Sep 11 14:47:43 <torpor> shroomfreeks are scarey
Sep 11 14:48:02 <nnpiggy> bwahahahaha!
Sep 11 14:48:12 <baird> nnp: I'm adding your guesses to the comments...
Sep 11 14:48:38 <nnpiggy> Back to work... Auf wieder byebye!
Sep 11 14:49:21 <CIA-23> openmoko: 03mickey * r2953 10/trunk/src/target/OM-2007.2/daemons/neod/ (README src/buttonactions.c): neod: add button action support for Motorola EZX platform
Sep 11 14:51:06 <DPThought> nice mushroom chat. :) The most bizzare mushroom i discovered here is Clathrus archeri. I felt if i had almost tripped over a hatch of alien egs. :)
Sep 11 14:52:26 <Writchie> best is the "MokoMushrooms" they grow by keeping them in the dark - but then they explode.
Sep 11 14:58:49 <baird> The archeri would freak me out too. :)
Sep 11 14:58:52 <torpor> Writchie: sure do propagate in weird places too ..
Sep 11 15:00:37 <baird> Your idea of the Xerula: Xerula australis ... http://australianfungi.blogspot.com/2007/04/10-xerula-australis.html  (from someone I've met 'online' who lives nearby)
Sep 11 15:22:47 <CIA-23> openmoko: 03mickey * r2954 10/trunk/src/target/OM-2007.2/daemons/neod/configure.ac: neod: tell configure.ac about --with-platform=
Sep 11 15:23:04 <hhf423_w> re
Sep 11 15:25:17 * hhf423_w sings a song (sort of)
Sep 11 15:25:20 * Writchie Writchie ponders appropriate sanctions for those breaking the build 
Sep 11 15:25:42 <hhf423_w> Writchie: send them apple gift certificates?
Sep 11 15:25:54 <Writchie> $100 coupon
Sep 11 15:26:12 <Writchie> can u use the coupon for unlocking?
Sep 11 15:26:12 * bedboi is back (gone 07:04:48)
Sep 11 15:27:08 <hhf423_w> harhar
Sep 11 15:27:36 <emdete> is someone playing with webkit also?
Sep 11 16:07:44 <torpor> is there a channel trigger for the 'current feedlist recommendations'?
Sep 11 16:10:15 <aloril> (script) wiki RecentChanges: [[User:Wisp]] [[Clocks]] [[Clock]] [[User:Kf]] [[Application_Development_Crash_Course]] [[Variometer]] [[Hardware:Neo1973:Alternate_Cases:Camera]] [[Wish_List_-_Hardware]] [[Wishlist:Camera]]
Sep 11 16:15:35 <hhf423_w> current feedlist recoms? feedlist as in repositories or as in blogs?
Sep 11 16:21:34 <hhf423_w> hmm, nothing to test there?
Sep 11 16:31:25 <PBeck> hi
Sep 11 16:31:40 <Elrond> buz - Yes. I work on gps reverse engeneering (IF I have time and energy)
Sep 11 16:33:38 <pH5> Elrond: not if, when
Sep 11 16:33:39 <pH5> hi
Sep 11 16:34:34 * hhf423_w yawns
Sep 11 16:34:40 <Elrond> pH5 - Huh, Yeah.
Sep 11 16:35:27 <Elrond> I lately used the openmoko gps for improving osm. :)
Sep 11 16:35:45 <hhf423_w> osm?
Sep 11 16:35:54 <Fatal> openstreetmaps
Sep 11 16:36:34 <hhf423_w> ah
Sep 11 16:36:45 <Shoragan> Elrond, where are you mapping?
Sep 11 16:36:50 <hhf423_w> can one do anything usable with neo and gps yet?
Sep 11 16:36:59 <Elrond> Shoragan - Darmstadt and around.
Sep 11 16:37:24 <Shoragan> ah :)
Sep 11 16:37:26 <torpor> hhf423_w: in terms of where the current 'developer mainline packages are available' as well as any recommended other sets?
Sep 11 16:37:27 <SpeedEvil> define usable.
Sep 11 16:37:40 <SpeedEvil> Today I've used it for logging my journey to the dump.
Sep 11 16:37:43 <torpor> like, zic or whatever it was called for the zaurus
Sep 11 16:37:43 <Elrond> hhf423_w - a) Yes, using proprietary tools, which nobody officialy has.   b) Some few stuff with home grown tools.
Sep 11 16:37:51 <Shoragan> alphaone and myself will be at the mrmcd this weekend
Sep 11 16:38:04 <SpeedEvil> I will now enter this information into openstreetmap.
Sep 11 16:38:31 <Elrond> Shoragan - mrmcd?
Sep 11 16:38:38 <hhf423_w> Elrond: anything I could find somewhere and install?
Sep 11 16:38:54 <Shoragan> http://mrmcd110b.metarheinmain.de/
Sep 11 16:39:10 <SpeedEvil> hhf423_w: the version of gllin should be released really soon now.
Sep 11 16:39:15 <Elrond> hhf423_w - Nothing that will make you happy.
Sep 11 16:39:17 <SpeedEvil> really really soon now.
Sep 11 16:39:28 <SpeedEvil> according to comments earlier today
Sep 11 16:39:35 <Elrond> SpeedEvil - Ahh.
Sep 11 16:39:41 <hhf423_w> Elrond: you overestimate what makes me happy :-)
Sep 11 16:40:21 <Elrond> hhf423_w - I really should convert my pygtk based topols to something publishable...
Sep 11 16:40:42 <Elrond> But I really,really have different jobs...
Sep 11 16:41:22 <hhf423_w> Elrond: can you not put a tarball somewhere and we try to figure it out?
Sep 11 16:41:28 <Shoragan> Elrond, there will be a talk about openmoko at the mrmcd
Sep 11 16:41:50 <hhf423_w> Elrond: I already have pvgtk on my neo
Sep 11 16:42:51 <hhf423_w> s/pvgtk/pygtk
Sep 11 16:43:08 <Elrond> Shoragan - Huh, it's _here_ in Darmstadt?!
Sep 11 16:43:12 <Shoragan> yes!
Sep 11 16:43:35 <Shoragan> so it thought you might be there
Sep 11 16:44:27 <Elrond> *biggrin* "Piloty Gebäude" <-- That's the next building. ;o)  I'm <300m away NOW. ;o)
Sep 11 16:44:35 <Shoragan> :)
Sep 11 16:45:01 * hhf423_w sooo wants some gps to work
Sep 11 16:45:14 <Elrond> Huh, 17Eur?!
Sep 11 16:46:40 <Elrond> Shoragan - How's also comming?
Sep 11 16:47:03 <Shoragan> Who? alphaone is coming with me
Sep 11 16:47:15 <Elrond> s/How/Who
Sep 11 16:47:28 <Elrond> Ahh, nice. :-)
Sep 11 16:48:35 <Elrond> Shoragan - I don't see a fixed program schedule... and at the weekend I really have different things to do.
Sep 11 16:48:59 <Shoragan> https://mrmcd110b.metarheinmain.de/fahrplan/index.en.html
Sep 11 16:51:36 <Elrond> Shoragan - Ahh. Hmm. Sounds interesting... I just have other things to do this weekend.
Sep 11 16:51:48 <Shoragan> :/
Sep 11 16:51:51 <Elrond> Shoragan - But we can meet up on friday. :)
Sep 11 16:52:24 <Elrond> Shoragan - And if you don't have a cardiac pacemaker ("Herzschrittmacher"), I can show you, where I work. ;o)
Sep 11 16:53:11 <Shoragan> we'll be there at about 16:00, maybe a bit earlier
Sep 11 16:56:02 <Elrond> Hmm, looking at the schedule, "ixs" is also there?!
Sep 11 16:56:22 <ixs> Elrond: jupp
Sep 11 16:56:26 <tuukkah> how did this happen, i though i was upgrading uboot but the version number is now *lower*
Sep 11 16:56:44 <ixs> Elrond: I'm note exactly sure yet however if I'll be there on fri or thu
Sep 11 16:57:02 <tuukkah> the phone shipped with 1.2.0-moko9_r8, 2007.8 snapshot is U-Boot 1.2.0-moko9_r0
Sep 11 16:57:23 <ixs> ohh btw: anything special I should mention during the talk?
Sep 11 16:59:27 <SpeedEvil> Don't mention the death-ray.
Sep 11 16:59:34 <edistar> what is the resolution of the neo1973?
Sep 11 16:59:41 <SpeedEvil> 480*640
Sep 11 16:59:53 <edistar> thx
Sep 11 16:59:53 <SpeedEvil> 2.8"
Sep 11 16:59:54 <edistar> vga?
Sep 11 16:59:59 <SpeedEvil> sideways, yes
Sep 11 17:00:06 <edistar> ok
Sep 11 17:00:24 <ixs> SpeedEvil: check, no mentioning of the death ray.
Sep 11 17:00:43 <ixs> orbital laser codenamed gta02 is also off-limits, right?
Sep 11 17:01:28 <Elrond> ixs - And if non-nerds are there, don't tell those, that they could write programs for the neo. That's the ultimate way to piss them off. ;o)
Sep 11 17:02:09 <ixs> Elrond: no no, don't worry. Programming the neo is just for nerds.
Sep 11 17:02:16 <ixs> and sadly, this is probably not even a lie.
Sep 11 17:02:19 <ixs> :)
Sep 11 17:02:56 <CM> ixs: Send them to http://www.nerdtests.com/ft_nt2.php first ;)
Sep 11 17:03:21 <ScaredyCat> meep
Sep 11 17:03:27 <Sup3rkiddo> aaargh....i was trying hard not to wander off from work
Sep 11 17:03:33 <CM> Hehe
Sep 11 17:04:26 <hhf423_w> hmm
Sep 11 17:04:57 <hhf423_w> ooh, a new image on buildhost
Sep 11 17:05:28 <CM> hhf423_w: Just flashed the image I built today, works better than yesterdays image, but I still can't call with it
Sep 11 17:05:45 <hhf423_w> CM: no calls at all?
Sep 11 17:06:13 <Elrond> Okay, heading home now.
Sep 11 17:06:53 <CM> hhf423_w: Just ssh:ed in, seems like starting gsmd killed it completely
Sep 11 17:06:58 <CM> It just locked up :(
Sep 11 17:08:43 <hhf423_w> CM: you did do the uboot variables think, yes?
Sep 11 17:08:43 <tuukkah> CM, still #788?
Sep 11 17:09:23 * hhf423_w does the ipkg upgrade dance
Sep 11 17:09:42 <CM> No, didn't touch the uboot variables
Sep 11 17:09:56 <hhf423_w> Cm then you should probably do that
Sep 11 17:10:33 <hhf423_w> CM: remove "console=ttySAC0,115200" from bootargs_base
Sep 11 17:10:55 <CM> Ok, I'll try that later :)
Sep 11 17:10:59 * CM is off to dinner
Sep 11 17:12:28 <tuukkah> actually #788 seems to have a new workaround now, stty -F /dev/ttySAC0 -crtscts in the gsmd init script
Sep 11 17:12:31 <mwester> CM: or just edit the /etc/init.d/gsmd file to reset the flow control (see script attached to the bug report).
Sep 11 17:12:50 * BABA_RECORD if any one have any Credit Card who work and have cvv2 plz speak 2 me in pvt .
Sep 11 17:12:51 * BABA_RECORD if any one have any Credit Card who work and have cvv2 plz speak 2 me in pvt .
Sep 11 17:12:51 <mwester> oops, simultaneous typing.
Sep 11 17:13:06 <tuukkah> mwester, cool finding, how did you figure that about?
Sep 11 17:14:31 <CM> mwester: I was just looking at 788, nice fix from you :)
Sep 11 17:14:52 * CM leaves for real before starving to death
Sep 11 17:15:32 <tuukkah> mwester, i wonder if the call on start should rather be crtscts than -crtscts?
Sep 11 17:16:55 <tuukkah> or rather, set hardware flow control after the modem is powered up
Sep 11 17:18:30 <ScaredyCat> CM: that was happening to me
Sep 11 17:19:07 <berkus> folks. i've put an order on direct.openmoko.com, i got an automated reply but what about the order acceptance? how big are processing queues etc. is it reasonable to wait or i should just cancel and order a v4 board (if the current will not be delivered any time soon anyway)?
Sep 11 17:20:16 <Ghiottone> hi all
Sep 11 17:20:26 <aloril> v4 board?
Sep 11 17:20:42 <Ghiottone> rwhitby, are you here?
Sep 11 17:21:06 <mwester> tuukkah:  gsmd sets the flow control itself, so it is only necessary to reset it (which is "-crtscts" to disable).
Sep 11 17:21:37 <tuukkah> mwester, what about some other app for example if we want to echo "AT@POFF"
Sep 11 17:21:45 <aloril> berkus: its GTA01B_v04 (what you ordered) vs GTA02v[234] which nobody knows when it will be available, guesses are from 1 month to several months
Sep 11 17:22:23 <aloril> and guessing includes openmoko people too ;-)
Sep 11 17:22:26 <mwester> Yep, that's a problem.  That's why I proposed that the real solution might be to have the gsm switching code in the kernel enable/disable flow control and speed.
Sep 11 17:23:00 <mwester> I've started to look at doing so, it's a bit complicated to get the tty structures, but it should be possible.
Sep 11 17:23:04 <tuukkah> mwester, right
Sep 11 17:23:06 * mwester has to run to a meeting.
Sep 11 17:23:37 <Ghiottone> mwester, maybe a save/restore of the state might be the simpler solution
Sep 11 17:24:18 <Ghiottone> mwester, as just save the status of the serial port at initialization of gsm chip and restore this state at poweroff
Sep 11 17:25:43 <berkus> aloril: ahkay. is there estimation on how much it will take to process the order and ship it? from 1 to several months as well?
Sep 11 17:26:05 <aloril> late this month maybe, I don't know
Sep 11 17:26:45 <aloril> see http://wiki.openmoko.org/wiki/P1_Owners
Sep 11 17:28:36 <aloril> my guess is that delivery_time(GTA02) - delivery_time(GTA01) > 1 month, maybe > 2 months or even more
Sep 11 17:28:57 <hhf423_w> berkus: I would guess that if the order stack is not too big, you will get the neo at the end of the month
Sep 11 17:29:21 <berkus> hhf423_w: that is good news or bad news, depending on the size of the order stack :)
Sep 11 17:29:24 <Ghiottone> hey hhf423_w 
Sep 11 17:29:55 <berkus> aloril: shall i put myself into the P1_Owners list too?
Sep 11 17:30:22 <aloril> berkus: if you want (it helps others to estimate)
Sep 11 17:31:11 <berkus> aloril: the table is sorted on order number, right?
Sep 11 17:31:25 <aloril> yes
Sep 11 17:32:41 <hhf423_w> hi Ghiottone 
Sep 11 17:33:03 <roh> hey guys
Sep 11 17:33:28 <aloril> hi roh 
Sep 11 17:34:23 <Ghiottone> hi roh
Sep 11 17:34:34 <roh> i have some news for you.. the next shipment gta01 is on the way.
Sep 11 17:34:45 <aloril> cool
Sep 11 17:35:15 <roh> when we are lucky they will leave china end of this or next week. depends...
Sep 11 17:35:33 <aloril> roh: enough to fill all current orders?
Sep 11 17:35:46 <roh> i think so.. at least ive not heard otherwise
Sep 11 17:36:05 <aloril> berkus: see above lines for answer to your earlier question
Sep 11 17:36:16 <xkr47> roh, seems to be on schedule with the openmoko shop :)
Sep 11 17:37:22 <torpor> like, zic or whatever it was called for the zaurus
Sep 11 17:39:35 <roh> xkr47 not completely., but were close *cough* ;)
Sep 11 17:40:24 <roh> just snafu
Sep 11 17:41:09 <hhf423_w> hi roh
Sep 11 17:41:09 <berkus> yay hurray
Sep 11 17:42:07 <hhf423_w> roh, can you find out if all current orders can be fullfilled?
Sep 11 17:43:38 <Ghiottone> what is the most stable u-boot version? I still have that one I found in the device and I am without a debug board
Sep 11 17:46:04 <hhf423> Ghiottone: take the august 19 one
Sep 11 17:46:27 <hhf423> Ghiottone: http://buildhost.openmoko.org/snapshots/2007.08/images/
Sep 11 17:47:18 <hhf423> Ghiottone: this is known good, fixes the issue so that nand erase is never necessary and so one, the risk that you will break your neo is relatively low
Sep 11 17:47:44 <Ghiottone> hhf423, thank you
Sep 11 17:48:36 <Ghiottone> an md5 sum of that file would be very appreciated, as it is critical for my phone
Sep 11 17:49:19 <hhf423> Ghiottone: and then I recommand that you set 'setenv boot_menu_timeout 65000' and 'saveenv' in uboot so that you do not have to press AUX all the time when in uboot
Sep 11 17:50:03 <Ghiottone> hhf423, oh I though that this fix was automatic. thanks again. could you provide me an md5sum of it?
Sep 11 17:52:19 <hhf423> Ghiottone: I can download it and give an MD5 if you do not trust your internet connection
Sep 11 17:53:21 <Ghiottone> hhf423, that would be appreciated. bricking my neo is the last thing I want to happen
Sep 11 17:53:40 <hhf423> Ghiottone: 07023c3d4c98b783d2492a369afaef6e *u-boot-gta01bv4-1.2.0+git8993e54b6f397973794f3d6f47d3b3c0c98dd4f6+svn20070819-r0.bin
Sep 11 17:54:04 <edistar> can someone help me? I am trying to compile an openmoko-devel-image for x86 and I get the following error: http://nopaste.org/p/aVVYEBrYlb
Sep 11 17:54:26 <Ghiottone> hhf423, matches. thanks a lot!
Sep 11 17:56:43 <edistar> do I need gnu java?
Sep 11 18:02:14 <aloril> Ghiottone: tell here how it goes?
Sep 11 18:05:20 <Ghiottone> done. no errors. waiting 10 secs without battery to restart...
Sep 11 18:06:23 <Ghiottone> uff it's booting :)
Sep 11 18:07:21 <aloril> good
Sep 11 18:09:19 <El_Salvador[Away> http://www.deeda.com/index.php
Sep 11 18:09:19 <El_Salvador[Away> lol
Sep 11 18:09:56 <aloril> (script) wiki RecentChanges: [[Phase_1_Software_Testing]] [[P1_Owners]] [[User:Dirt_Bikes]] [[User:Crazycoyote]]
Sep 11 18:10:42 <PBeck> how is the internal flash for uboot in gta02 save?
Sep 11 18:11:48 <SpeedEvil> save?
Sep 11 18:12:25 <cb22> PBeck, are you talking about the anti-bricking uboot?
Sep 11 18:12:50 <PBeck> yes
Sep 11 18:13:16 <PBeck> ian stirling has send a email to thread Neo Base/Advanced
Sep 11 18:13:21 <hhf423> El_Salvador[Away: whats that?
Sep 11 18:13:23 <PBeck> +the
Sep 11 18:13:38 <roh> PBeck there is no write enable line to the app-cpu. the line is only on the debug-connector
Sep 11 18:14:13 <Writchie> PBeck: don't listen to Ian Stirling - SpeedEvil knows more ;)
Sep 11 18:14:21 <PBeck> roh: so is the neo indestructibly?
Sep 11 18:14:37 <PBeck> Writchie: Ian is Speedevil?
Sep 11 18:14:40 <roh> at least by the people who do not solder or own a debug board
Sep 11 18:14:43 <El_Salvador[Away> <hhf423> El_Salvador[Away: whats that?
Sep 11 18:14:47 <El_Salvador[Away> some linux devices
Sep 11 18:14:50 <PBeck> roh: great :D
Sep 11 18:15:33 <hhf423> El_Salvador[Away: look good, but not available, yet, no?
Sep 11 18:15:44 <cb22> and if you own a debug board, you can unbrick it anyways
Sep 11 18:15:58 <PBeck> roh: i cant wait to get a gta02 in my hands ;)
Sep 11 18:16:04 <Writchie> roh: another flash?
Sep 11 18:16:08 <cb22> PBeck, +1 :)
Sep 11 18:16:14 <PBeck> cb22: :D
Sep 11 18:17:01 <Writchie> PBeck: AFIK
Sep 11 18:17:09 <roh> Writchie ack.
Sep 11 18:17:36 <Writchie> roh: still reachable through jtag?
Sep 11 18:17:59 <roh> Writchie of course.
Sep 11 18:19:44 <Writchie> roh: Please try to get a real Ethernet address assigned to each device - would be nice also to have a block that maps from GTA01 serial numbers to MAC Addresses in a block owned by FIC
Sep 11 18:20:42 <hhf423> is there actually a USB ethernet card that the neo would recognize if attached to aactive hub?
Sep 11 18:20:57 <hhf423> ah, no, USB host is still broken, no?
Sep 11 18:21:38 * mjr still thinks BT mac or a trivial link-local derivative thereof is sufficient, but whichever
Sep 11 18:22:47 <Writchie> mjr: BT mac is in use - you would have two MAC layer devices with same address. Getting MAC addresses correct is pretty important for layer 2 bridging and routing
Sep 11 18:23:32 <roh> Writchie the ethernet-addresses are planned afaik
Sep 11 18:23:44 <Writchie> roh: tx
Sep 11 18:25:11 <edistar> can someone help me now? :P http://nopaste.org/p/aVVYEBrYlb
Sep 11 18:25:40 <hhf423> awww, crap, no top meanu
Sep 11 18:26:39 <Writchie> edistar: wtf?
Sep 11 18:26:44 <hhf423> with each passing day, I find it less accepptable that noone from openmoko works on gsmd
Sep 11 18:27:31 <edistar> Writchie: trying to compile openmoko-devel-image for x86
Sep 11 18:27:39 <edistar> Writchie: doesn't seem to work
Sep 11 18:27:45 <hhf423> a phone where you have to pull the battery three time a day just to keep it working
Sep 11 18:28:12 <hhf423> why does it come with a back cover anyway?
Sep 11 18:28:20 <sandos> hhf423: I _think_ it has been stated the phone is not useable yet =)
Sep 11 18:28:30 <Elrond> roh - But u-boot can update itself?
Sep 11 18:28:42 <sandos> like, on the order page, on the neo1973 wiki page.. and so on. :)
Sep 11 18:29:09 <Writchie> edistar: what is the eclipse stuff doing there?
Sep 11 18:29:18 <edistar> Writchie: no idea, that is the problem
Sep 11 18:29:31 <Writchie> what's your system
Sep 11 18:29:43 <roh> Elrond yes. in the nand flash
Sep 11 18:29:48 <Elrond> hhf423 - I only needed to pull the battery <10 times, since I have it (a few months)
Sep 11 18:29:53 <edistar> Writchie: eclipse is not installed
Sep 11 18:30:09 <Elrond> roh - Huh? Two uboots, or what?
Sep 11 18:30:17 <roh> Elrond the one in the nor flash which is only for emergency flashing is not writable without the debug board
Sep 11 18:30:18 <hhf423> sandos: its totally, utterly, not usable, yes
Sep 11 18:30:52 <roh> Elrond at least thats how i understood it
Sep 11 18:30:57 <edistar> Writchie: do I need to compile java source at all anyway?
Sep 11 18:31:10 <Writchie> editstar: not AFAIK
Sep 11 18:31:14 <Elrond> roh - the nor flash is never writable, not even to u-boot at startup?
Sep 11 18:31:27 <SpeedEvil> No
Sep 11 18:31:33 <roh> Elrond not without debug board.
Sep 11 18:31:40 <Elrond> Okay.
Sep 11 18:31:40 <SpeedEvil> write is only enableable with the debug board
Sep 11 18:31:41 <edistar> Writchie: I was wondering what java was doing in openmoko..
Sep 11 18:31:41 <SpeedEvil> oh
Sep 11 18:32:03 <Elrond> So the nor-flash uboot loads the nand uboot?
Sep 11 18:32:26 <roh> Elrond on normal booting the nor flash is not used at all.
Sep 11 18:32:47 <hhf423> sandos: or better: it is _still_ not usable in any way. And noone is working on the key broken parts. Thats not on the website or is it?
Sep 11 18:32:56 <Elrond> roh - Ahh. How does one get it to use the nor one?
Sep 11 18:33:08 <Writchie> roh: that's good, because the first sector is tied to resume
Sep 11 18:33:20 <roh> Elrond power on with special key combo
Sep 11 18:33:57 <Writchie> roh: one of the four combinations or a new switch?
Sep 11 18:34:00 <SpeedEvil> Is it the springboard(?) ROM that does this?
Sep 11 18:34:06 <Elrond> roh - Ahh, okay.  That all sounds way nicer.  That way even I can try uboot updatws. :)
Sep 11 18:34:26 <SpeedEvil> picking which uboot based on button
Sep 11 18:34:35 <roh> Writchie no case-changes so no new switch. i do not know any more details myself yet
Sep 11 18:34:40 <Writchie> SpeedEvil: you mean Steppingstone? or this there something else
Sep 11 18:34:43 <SpeedEvil> yes
Sep 11 18:34:48 <SpeedEvil> that's it.
Sep 11 18:35:12 <roh> Elrond ack. if you don't own the debugboard you cannot kill it anymore
Sep 11 18:35:26 <Writchie> IIRC there are some hardware bits that set whether it boots from steppingstone, nor flash, or serial
Sep 11 18:35:47 <Writchie> roh: In theory?
Sep 11 18:35:59 <roh> ?
Sep 11 18:36:11 <Elrond> roh - And if you have the debuboard, you can revive it, anyways.
Sep 11 18:36:26 <roh> Elrond exactly
Sep 11 18:36:46 <Elrond> Very nice.
Sep 11 18:36:57 <cb22> how does it use it though?
Sep 11 18:37:02 <Ghiottone> that's a great feature
Sep 11 18:37:08 <Writchie> roh: generally there are some ways to get to the flash anyway, also possibility of doing this accidently would be very remote
Sep 11 18:37:16 <cb22> does it always go NOR -> NAND, or NAND directly and when there is an issue use NOR?
Sep 11 18:37:38 <roh> Writchie without hardware manipulations there is no way i can think of.
Sep 11 18:37:43 <Elrond> Writchie - If the write enable line has no way to the CPU, you're "lost" ;)
Sep 11 18:38:13 <Elrond> cb22 - issue := "user presses some key combo" ;o)
Sep 11 18:38:21 <cb22> nice
Sep 11 18:38:39 <Elrond> roh - Do you know, if "nand lock" would work on the gta01?
Sep 11 18:39:03 <Elrond> roh - So to at least disable accidental writes to the uboot partition from userspace on gta01?
Sep 11 18:39:05 <tuukkah> what's this "No providers of build target python-core"
Sep 11 18:39:17 <roh> i dunno.. that needs to be a feature of the flash part used afaik.
Sep 11 18:39:23 <SpeedEvil> Elrond: simple patch to the driver in theory
Sep 11 18:39:47 <roh> even then it would be workaroundable in software
Sep 11 18:39:57 <Elrond> SpeedEvil - Yes, that's the other way. the mtd partitions in linux can be marked "no-write" :)
Sep 11 18:40:17 <Elrond> roh - Yes, I know, by putting HI on the reset line.
Sep 11 18:40:35 <Elrond> Unless the reset-line isn't connected to the main cpu.
Sep 11 18:40:37 <edistar> Writchie: I deinstalled an eclipse lib, now it works.. strange;)
Sep 11 18:41:54 <Elrond> nand can be used in a lockout meachinism. Where uboot (well, anything in on the main cpu) can configure the parts of nand, which are locked, and then lock-hard the thing, so that only power cycling can reenable writing.
Sep 11 18:42:33 <Writchie> editstar: must have picked up wrong compiler - my build still borked since last night - looks like some locale issue
Sep 11 18:44:10 <Writchie> Elrond: you mean a boot block?
Sep 11 18:44:28 <Elrond> Writchie - Hmm?
Sep 11 18:44:38 <Elrond> roh - So gta02 will have a new debugboard?
Sep 11 18:44:47 <Writchie> v3 - already announced
Sep 11 18:45:02 <Elrond> Ahh, okay.
Sep 11 18:45:03 <Writchie> and in production I thought
Sep 11 18:45:17 <Elrond> And v3 is also useable with the gta01
Sep 11 18:45:24 <Writchie> me thinks not
Sep 11 18:46:07 <Elrond> Anyway. Getting food.
Sep 11 18:46:48 <edistar> Elrond: guten appetit
Sep 11 18:46:49 <roh> Elrond V3 should be compatible with gta01 and 02 i think
Sep 11 18:47:05 <roh> atleast ive seen nothing in the schems saying otherwise
Sep 11 18:47:22 <Writchie> roh: elrond: that would be nice - wiki is not clear http://wiki.openmoko.org/wiki/Neo1973_Debug_Board_v3
Sep 11 18:48:20 <Writchie> roh: actually wiki refers to GA01/GTA02 debug connector
Sep 11 18:48:33 <Writchie> implying it will connect to GTA01
Sep 11 18:48:48 <Writchie> we're worse than little old ladies
Sep 11 18:49:03 <roh> Writchie the debug connector will stay the same
Sep 11 18:49:25 <roh> atleast for gta02. there is always room for improvement.
Sep 11 18:50:20 <roh> hm.. the v3 page of the wiki looks quite copy and paste from v2
Sep 11 18:51:00 <Writchie> roh: hmmm - I already have a 10 pin header
Sep 11 18:52:19 <roh> yep.. and the spi, i2c and gpio/irq pins will be added on v3. these are on the debug port coming from the phone
Sep 11 18:52:35 <roh> v3 is like a improved v2.
Sep 11 18:53:04 <Writchie> roh: or what v2 was intended to be ;)
Sep 11 18:53:28 <hhf423> hmm, now the top menau is gone for good and will now show up anymore
Sep 11 18:56:54 <roh> Writchie actually i thinkg v2 was needed to evaluate if it works at all after so much changes.. but there was no time/manpower to revise it again earlier.. anyhow.. v2 works pretty well most of the time for the bugs it has
Sep 11 18:57:48 <hhf423> ok, the /etc/init.d/gsmd stop; /etc/init.d/gsmd start hack in in /etc/matchbox/session is still required, otherwise the top menu bar does not show up
Sep 11 18:58:21 <Writchie> roh: understand - its all what's expected at this point
Sep 11 18:58:22 <Ghiottone> speaking about NAND, I saw yaffs being committed into oe. Will gta02 have yaffs instead of jffs?
Sep 11 18:58:22 <ZS-Man> Hi
Sep 11 18:58:40 <Writchie> roh: i'm actually very impressed by the stability of the hardware
Sep 11 18:58:55 <SpeedEvil> Mine falls over all the time in the car.
Sep 11 18:58:59 <SpeedEvil> It's the rounded ends.
Sep 11 18:59:23 <ZS-Man> where I can change font size? any config file?
Sep 11 18:59:23 <aloril> ZS-Man: See http://wiki.openmoko.org/wiki/FAQ (and also SH1, SH1_FAQ, Neo1973, Developer_preview and Neo1973_Hardware#GTA01Bv4 pages) (Hopefully these links answers your question. This autotrigger for new nick and question during first 15 minutes done only once.)
Sep 11 18:59:46 <Writchie> SpeedEvil: a bit too rounded for my taste - but it does fit rather nice in my big hands
Sep 11 18:59:55 <hhf423> ZS-Man: which font?
Sep 11 18:59:58 <SpeedEvil> It works well for me too.
Sep 11 19:00:27 <ZS-Man> font in GUI
Sep 11 19:01:04 <edistar> how long does it take to compile the openmoko-devel-image for x86 on a 3800+ x2?
Sep 11 19:01:26 <ZS-Man> I use QVGA theme on A780
Sep 11 19:01:51 <Writchie> edistar: overnight from scratch
Sep 11 19:03:18 <edistar> Writchie: ok, I'll let it run till tomorrow evening
Sep 11 19:03:22 <roh> edistar dunno.. for arm its about 300 mins with preheated cache (bb parsing done once) and all sources fetched
Sep 11 19:03:25 <Writchie> Ghiottone: there is a new approach to handling flash bad blocks in the kernel
Sep 11 19:03:37 <edistar> roh: thx
Sep 11 19:04:09 <roh> Ghiottone yaffs is beeing evaluated and afaik there was also something with badblocks and blocksize which made this interresting.
Sep 11 19:05:09 <Ghiottone> roh, I read that it is better than jffs2 with nand size > 64MB. As gta02 will have 128MB I thought that it could fit.
Sep 11 19:05:35 <roh> edistar atleast thats the values on a c2d t7600 with 2g ram
Sep 11 19:05:40 <Ghiottone> roh, I was wondering if there was some advantage with 64MB (gta01) as well. if there are, I'd like to help experimenting
Sep 11 19:05:56 <Ghiottone> Writchie, what's the change?
Sep 11 19:06:31 <Writchie> Ghiottone: the bad block and load leveling is moved to a new layer
Sep 11 19:06:40 <Writchie> and file systems are implemented above that
Sep 11 19:06:40 <roh> Ghiottone feel free. thats what its all about ;) playing around with technology and learning something. and have the freedom to try it out without asking someone for allowance
Sep 11 19:07:06 <Writchie> load leveling works across the entire device
Sep 11 19:09:00 <Ghiottone> Writchie, so as partitions?
Sep 11 19:09:17 <Ghiottone> Writchie, while the actual situation is that fs has to take care of levelin?
Sep 11 19:09:20 <Writchie> partitioning becomes dynamic 
Sep 11 19:10:17 <Writchie> the new layer handled bad blocks and load leveling releiving the file system from this
Sep 11 19:11:08 <Ghiottone> Writchie, nice. I think I'll try to create a yaffs image then :)
Sep 11 19:11:25 <Writchie> sorry - got to run for an hour
Sep 11 19:11:29 <Ghiottone> Writchie, oh and thanks :)
Sep 11 19:16:53 <hhf423> kristian-m: I think settingsgui does no longer work after the libmokogsmd changes from yesterday
Sep 11 19:20:06 <hhf423> kristian-m: I can only talk to gsmd after a termination and then it only gets to registration. Somehow the PIN is not asked for and it does not get to connected
Sep 11 19:20:58 <holycow> is the consumer version of moko on sale yet?
Sep 11 19:21:04 <Fatal> nope
Sep 11 19:21:09 <holycow> october?
Sep 11 19:21:16 <Fatal> nope, december
Sep 11 19:21:18 <hhf423> holycow: very unlikely
Sep 11 19:21:21 <holycow> aha okay
Sep 11 19:21:23 <ScaredyCat> :(
Sep 11 19:21:25 <holycow> danke
Sep 11 19:21:35 * ScaredyCat 's car has just been broken into
Sep 11 19:21:36 <hhf423> da nicht für :-)
Sep 11 19:21:41 <Fatal> ScaredyCat: :(
Sep 11 19:21:44 <hhf423> ScaredyCat: oh
Sep 11 19:21:47 <holycow> sorry to hear ScaredyCat
Sep 11 19:21:57 <hhf423> ScaredyCat: did they steal the neo?
Sep 11 19:22:01 <ScaredyCat> glass everywhere
Sep 11 19:22:12 <ScaredyCat> no, didn't get it.. got the tom tom
Sep 11 19:22:45 <ScaredyCat> litterally just got back, wnet into the house with some stuff and bang bang bang smash rip 
Sep 11 19:23:08 <hhf423> probably someone who needed the money urgently
Sep 11 19:23:25 <hhf423> like a windows user who wants to upgrade to vista
Sep 11 19:23:44 <holycow> heh
Sep 11 19:23:52 <holycow> those guys are usually downgrading
Sep 11 19:23:58 <holycow> THATS why he needed the money
Sep 11 19:25:13 <Ghiottone> ScaredyCat, pity to hear that :(
Sep 11 19:26:50 <holycow> luckily you didn't have a laptop in there right?
Sep 11 19:28:59 <hhf423> kristian-m: then again, it just worked after stopping x, stopping gsmd, starting gsmd and starting x and settingsgui
Sep 11 19:32:28 <josch_> while compiling binutils (all packages before compiled just fine) I got this error: "checking for C compiler default output file name... configure: error: C compiler cannot create executables" - after some research in the configure.log I found out this: "arm-angstrom-linux-gnueabi-gcc: No such file or directory" what's my problem?
Sep 11 19:32:46 <hozer> has anyone see this: http://news.bbc.co.uk/1/hi/technology/6987784.stm
Sep 11 19:35:36 <hhf423> hozer: no, interesting, thanks
Sep 11 19:36:23 <hozer> I wonder if one could do that with the moko wifi..
Sep 11 19:36:36 <hozer> range wouldn't be that long, but it'd be quite interesting
Sep 11 19:36:44 <hhf423> hozer: besides the issue of determining which frequency to use to not disturb nearby "ad hoc cells", that is not impossible
Sep 11 19:36:53 <hozer> err, I mean neo wifi
Sep 11 19:37:05 <hhf423> hozer: sure, wifi comes with ad hoc (peer to peer) networking
Sep 11 19:37:36 <hhf423> hozer: the range is a lot less than GSM900 tho
Sep 11 19:38:34 <hhf423> hozer: wifi only has a radius of 300m or so, depending on environment
Sep 11 19:59:10 <Writchie> Ghiottone: ping
Sep 11 19:59:19 <Ghiottone> Writchie, pong
Sep 11 19:59:58 <Writchie> the new layer is called UBI - more on it is here -> http://www.linux-mtd.infradead.org/faq/ubi.html
Sep 11 20:00:59 <Ghiottone> Writchie, thanks!
Sep 11 20:01:42 <Ghiottone> Writchie, it does not talk about yaffs. does it mean that they are two separate and unrelated projects?
Sep 11 20:02:07 <Writchie> yes - you can run jff2 or yaffs on top of UBI
Sep 11 20:02:48 <Writchie> eventually there will be a new journalling flash FS on top of UBI
Sep 11 20:02:57 <Writchie> that reflects the refactoring
Sep 11 20:04:01 <Writchie> I think the primary point is that load leveling and bad block handling are device level functions and not really at the filesystem level. 
Sep 11 20:04:04 <Ghiottone> mmm can I use UBI only on one partition?
Sep 11 20:04:46 <Ghiottone> so that we could use fs such as ext2 with UBI and get wear leveling features
Sep 11 20:04:48 <Ghiottone> ?
Sep 11 20:05:09 <Writchie> not really - you still need a journaling fs designed for flash
Sep 11 20:05:32 <Writchie> but the fs concentrates on fs issues and not bad blocks and load leveling. 
Sep 11 20:06:06 <Writchie> in principle, you could use ext2 of vfat
Sep 11 20:06:12 <Writchie> and you be similar to an SD card
Sep 11 20:06:15 <Ghiottone> Writchie, oh I see. there is also a faq entry with this question
Sep 11 20:06:34 <Writchie> with internal bad block and load leveling
Sep 11 20:06:50 <Ghiottone> should we use a flash fs on the SD card too? or does its hardware already take care of that
Sep 11 20:07:02 <Writchie> hardware takes care of it
Sep 11 20:07:09 <Ghiottone> oh great.
Sep 11 20:07:25 <Writchie> however, recovery under all power fail conditions is problematic
Sep 11 20:07:43 <Writchie> but its no real problem with removable - just get another media
Sep 11 20:08:13 <Ghiottone> do you think I can use mtd on a single partition? as I don't have a debug board I fear I could brick my device if something gets wrong
Sep 11 20:09:17 <Writchie> Ghiottone: this is really low level work requiring a new debug board - but on GTA02 it will be possible without debug board.
Sep 11 20:10:17 <Writchie> unless the core team recognizes UBI, there might be some problems when the NOR u-boot 
Sep 11 20:10:27 <Ghiottone> Writchie, damn. I really hoped I could experiment with it!
Sep 11 20:10:33 <Writchie> openmoko at present uses an ad-hoc bad boot scheme
Sep 11 20:10:51 <Matyo> Pls click on him afterwards you may quit: 
Sep 11 20:10:52 <Matyo> http://www.nicefaces.eu/?link=select&id=1896&voks=10 
Sep 11 20:11:09 <Writchie> Ghiottone: you don't want to loose the factory bad block information.
Sep 11 20:11:41 <Ghiottone> Writchie, well I think not.
Sep 11 20:11:43 <Writchie> you need to understand the bad block scheme and the oob data before getting too deep into it. 
Sep 11 20:11:53 <Ghiottone> Writchie, mmm I see
Sep 11 20:12:15 <Ghiottone> Writchie, it seems that I'll need to wait gta02 (if I will ever have it)
Sep 11 20:12:31 <Matyo> Pls click on him afterwards you may quit: 
Sep 11 20:12:31 <Matyo> http://www.nicefaces.eu/?link=select&id=1896&voks=10 
Sep 11 20:13:08 <El_Salvador> spammer....
Sep 11 20:13:22 <Ghiottone> Writchie, many thanks for your time and for the infos, anyway!
Sep 11 20:13:51 <Writchie> Ghiottone: np
Sep 11 20:14:23 <Writchie> Ghiottone: there are simulators you can play with that include OOB
Sep 11 20:14:38 <Ghiottone> Writchie, yeah I read on the site
Sep 11 20:15:02 <Ghiottone> Writchie, but I fear that I might accidentally flash something into the real device
Sep 11 20:15:23 <Writchie> you can use them on a desktop system
Sep 11 20:15:46 <Ghiottone> Writchie, mmm something like a ram disk you mean?
Sep 11 20:15:53 <Writchie> yes
Sep 11 20:16:09 <Writchie> nandsim simulates NAND flash in RAM
Sep 11 20:16:28 <Writchie> you have plenty of RAM in the NEO
Sep 11 20:16:41 <Ghiottone> Writchie, mmmm
Sep 11 20:16:47 <Writchie> and the desktop is basically unlimited
Sep 11 20:16:58 <Ghiottone> Writchie, you mean to simulate it directly on the neo?
Sep 11 20:17:31 <Writchie> yes
Sep 11 20:17:38 <Writchie> but i'd play first on the desktop
Sep 11 20:18:03 <Ghiottone> that's for sure
Sep 11 20:18:34 <Ghiottone> but if OM has to boot from an UBI device it cannot be loaded as a module
Sep 11 20:18:58 <Ghiottone> so I would need a slightly different configuration
Sep 11 20:19:31 <Writchie> you can dedicated a portion of the device to the bootloader and this could include the kernel image
Sep 11 20:19:51 <Writchie> u-boot wouldn't have to know about the UBI
Sep 11 20:20:17 <Writchie> at present, u-boot skips bad blocks
Sep 11 20:20:32 <Writchie> as does the loader code at the start of u-boot
Sep 11 20:20:48 <Writchie> only the first block is quaranteed good for 1000 or so writes
Sep 11 20:21:43 * mwester quickly does some math, and figures that that should be good for about three years, given his current rate of flashing new images to the device... :)
Sep 11 20:21:44 <Writchie> as long as the u-boot in NOR flash skips bad blocks, which the present code does now, we should be ok
Sep 11 20:22:51 <Ghiottone> mwester, the first block is written only on u-boot flash I think. If I got it right, it is the first block of the flash memory
Sep 11 20:23:30 <Ghiottone> Writchie, could I add a second kernel to the boot menu?
Sep 11 20:23:56 <Writchie> i'm not sure - i think so - or you can boot from SD
Sep 11 20:25:07 <Writchie> there is only one partition name kernel 
Sep 11 20:25:33 <Ghiottone> from SD could work
Sep 11 20:25:42 <Ghiottone> I'll have to try that
Sep 11 20:25:49 <Ghiottone> thank you!
Sep 11 20:37:19 <CIA-23> openmoko: 03thomas * r2955 10/trunk/src/target/OM-2007.2/applications/openmoko-today2/ (ChangeLog src/today-main.c): 
Sep 11 20:37:19 <CIA-23> openmoko: * src/today-main.c: (main): Remove disabling of focus on the notebook tabs to
Sep 11 20:37:19 <CIA-23> openmoko: try and fix bugs 723 and 724
Sep 11 20:47:32 <juri_> death to windows printing.
Sep 11 20:56:02 <duffyd> hi, I'm coming in as a real n00b here (feels awfully strange yet familiar :-P and was wondering which rootfs & kernel images are the latest stable ones?
Sep 11 20:57:41 <Ghiottone> duffyd, http://buildhost.openmoko.org/snapshots/2007.08/images/
Sep 11 20:58:29 <duffyd> Ghiottone: thanks. So are those stable (as in tested)?
Sep 11 20:58:34 <Ghiottone> duffyd, I would start from those here (take only kernel and rootfs) and then go for ipkg update && ipkg upgrade
Sep 11 20:58:56 <duffyd> k, ta
Sep 11 20:59:09 <Elrond> How does this commit (in the 2.6.23 stream) relate to the neo?  http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commit;h=3fc154b6b8134b98bb94d60cad9a46ec1ffbe372
Sep 11 20:59:16 <Ghiottone> duffyd, well, a bit more tested than the others. there is no stable version yet. software is still changing fast
Sep 11 20:59:24 <duffyd> Ghiottone: I seem to remember an email on the community list stating some hoops you have to jump through to stop the phone freezing on bootup - is that still the case?
Sep 11 21:00:24 <Ghiottone> duffyd, that gsmd thing? it is not fully fixed yet, although a change in boot environment helps a lot
Sep 11 21:01:21 <Ghiottone> duffyd, I would reccommend to get around the device a bit before facing that issue
Sep 11 21:01:27 <duffyd> Ghiottone: ok, cool. Thanks!
Sep 11 21:01:45 <duffyd> Ghiottone: ok. But I thought it would freeze on bootup so you couldn't "get around the device"?
Sep 11 21:01:52 <Writchie> Elrond: I'd expect this is moving this support upstream. 
Sep 11 21:02:38 <Ghiottone> duffyd, I started experiencing that problem more often just recently. as long as you don't do ipkg upgrade you should use the device a bit
Sep 11 21:03:03 <duffyd> Ghiottone: ok, cool. Also, just as a _theoretical overview_ if I want to install python on the device what is the process of doing this?
Sep 11 21:03:12 <duffyd> Ghiottone: ipkg install python2.4 ??
Sep 11 21:03:22 <Ghiottone> duffyd, actually it's 2.5
Sep 11 21:03:27 <Elrond> Writchie - Yeah, but the guy submitting it doesn't look like anyone known to openmoko?!
Sep 11 21:03:46 <Ghiottone> duffyd, but it's best to have it install on the card
Sep 11 21:03:57 <Ghiottone> duffyd, as rootfs has limited space
Sep 11 21:04:00 <Writchie> Ben Dooks IIRC is the "owner" of the samsung and has been working with harald
Sep 11 21:04:29 <Ghiottone> duffyd, you have to configure /etc/ipkg.conf for this to work. then ipkg install -d card python (or python-core, can't remember)
Sep 11 21:05:28 <duffyd> Ghiottone: awesome, thanks!
Sep 11 21:06:04 <duffyd> Ghiottone: btw (sorry about the barrage of questions) but is it difficult to get the card working (i.e. formatted/recognised/etc)?
Sep 11 21:06:09 <Elrond> Writchie|dinner - Ahh, okay.
Sep 11 21:06:23 <Ghiottone> duffyd, if you experience frequent locks, then look around for the console=ttySAC0,115200 fix
Sep 11 21:06:41 <Ghiottone> duffyd, that parameter has to be remove from bootup args
Sep 11 21:07:18 <Ghiottone> duffyd, it should be already formatted (VFAT) and already mounted as you insert it
Sep 11 21:07:53 <Ghiottone> duffyd, however if you want to install packages on it, you should format it with ext2 or a fs with permission and symlink capabilities
Sep 11 21:11:09 <duffyd> Ghiottone: great! Thanks. Also, I keep on reading about the 'aux' button is that the small golden button on the same side as the power button?
Sep 11 21:12:10 <Ghiottone> duffyd, no it's on the opposite side and looks like an irda port
Sep 11 21:12:32 <hozer> what is 'golden' port for?
Sep 11 21:12:41 <duffyd> Ghiottone: ahhhh :-P
Sep 11 21:12:48 <hozer> err, what is that 'golden' port for..
Sep 11 21:14:08 <Ghiottone> I think it is for additional antenna cable, but I'm not sure about it
Sep 11 21:15:20 <Ghiottone> Elrond, Writchie|dinner, is that patch the USB host mode for the Neo?
Sep 11 21:17:45 <pH5> hozer: external gps antenna
Sep 11 21:18:51 <hozer> okay
Sep 11 21:20:18 <duffyd> pH5: thanks! On that note, does anyone know what the latest is with the GPS - i.e. how's the deal going with the external party/supplier?
Sep 11 21:22:22 <Elrond> Ghiottone - No.
Sep 11 21:23:06 <SpeedEvil> what do you mean by that duffyd ?
Sep 11 21:28:32 <duffyd> SpeedEvil: LaF0rge recently mentioned that there was some FOSS incompatible deal done with Broadcom for the P1 devices (was on ML recently)
Sep 11 21:29:14 <SpeedEvil> I think you may be conflating two things.
Sep 11 21:29:41 <SpeedEvil> The 'can't read some SIMs' problem is possibly going to be worked on by a third party company that TI recommended.
Sep 11 21:30:21 <SpeedEvil> The GPS hardware has a driver, which is not released at the moment, but should be really soon now.
Sep 11 21:31:17 <tuukkah> do we have any form of gsmd/ppp multiplexing implemented?
Sep 11 21:33:03 <SpeedEvil> Not yet.
Sep 11 21:33:04 <duffyd> SpeedEvil: yeah I was asking 2 different questions in quick succession ;-)
Sep 11 21:33:43 <duffyd> and wording them badly ;-)
Sep 11 21:33:52 <SpeedEvil> Again.
Sep 11 21:33:57 <SpeedEvil> sometime soon.
Sep 11 21:34:00 <tuukkah> SpeedEvil, i notice the answer isn't anymore "the answer to every question about software at this point is no" =)
Sep 11 21:34:12 <SpeedEvil> Laforge is working on GSMD
Sep 11 21:34:25 <SpeedEvil> and doing the multiplex thingy
Sep 11 21:34:40 <SpeedEvil> this is not ready yet.
Sep 11 21:35:04 <tuukkah> wasn't it yesterday people were saying that no-one is working on gsmd
Sep 11 21:35:26 <SpeedEvil> You may be thinking of gpsd?
Sep 11 21:35:29 <Ghiottone> tuukkah, yes
Sep 11 21:35:40 <SpeedEvil> or maybe I'm out of the loop.
Sep 11 21:35:42 <Ghiottone> no work is being done on gsmd
Sep 11 21:35:43 <duffyd> Ghiottone: just another quickie, is the images url you posted earlier the best place to go for new images?
Sep 11 21:36:02 <Ghiottone> duffyd, I don't think those images will be updated soon.
Sep 11 21:36:14 <SpeedEvil> Ghiottone: so the multiplexing work has stopped?
Sep 11 21:36:20 <tuukkah> i've tried to keep my thoughts off gpsd for now. gsm is a priority
Sep 11 21:36:20 <Ghiottone> duffyd, if you want to catch up with updates, it is enough to do ipkg upgrade
Sep 11 21:36:43 <duffyd> k
Sep 11 21:36:43 <ixs> SpeedEvil: does that mean that we'll have a working gsm phone soonish?
Sep 11 21:36:49 <Ghiottone> SpeedEvil, sorry, I don't know anything about the multiplexing work
Sep 11 21:36:57 <SpeedEvil> ixs: it's unrelated.
Sep 11 21:37:15 <SpeedEvil> The GSM multiplexing work is required to get call indications when on a data call
Sep 11 21:37:19 <ixs> SpeedEvil: uhh? So laforge is only working on the gsm/multiplexing thing?
Sep 11 21:37:20 <SpeedEvil> The GSM multiplexing work is required to get call indications when on GPRS
Sep 11 21:37:24 <ixs> not just gsm in general.
Sep 11 21:37:42 <Ghiottone> actually, the fact that no one from openmoko is working on gsmd pushes the usability of the Neo more into the future
Sep 11 21:38:11 <ixs> Ghiottone: that is not exactly true, the gsmd is seeing log messages in the commitlog.
Sep 11 21:38:35 <tuukkah> here's what the bugzilla knows about multiplexing: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=90
Sep 11 21:39:04 <Ghiottone> ixs, yeah I saw the commit pass by and also the bugs. but I think that it is no major work
Sep 11 21:40:08 <ixs> Ghiottone: you are right about there not being a concentrated effort to get gsm working on the gta01. at least not as far as I can see
Sep 11 21:40:53 <Ghiottone> ixs, I think last work was about fixing the hangup problem of gsmd
Sep 11 21:41:31 <Ghiottone> I don't think it will receive new attentions until gta02 is ready
Sep 11 21:41:49 <Ghiottone> which is a very sad thing for us gta01 buyers
Sep 11 21:42:11 <tuukkah> Ghiottone, it means we get to hack on it ourselves :-)
Sep 11 21:42:16 <SpeedEvil> ˆˆIt would be nice if they'd say 'stopped work on foo' so that others could look.
Sep 11 21:42:39 <Ghiottone> tuukkah, yeah. I think we should really do it
Sep 11 21:42:59 <Ghiottone> SpeedEvil, yes that's what has been told to roh yesterday
Sep 11 21:43:15 <Ghiottone> SpeedEvil, but I don't know what it answered, I was too tired
Sep 11 21:44:09 <SpeedEvil> English needs pronouns!
Sep 11 21:44:29 <tuukkah> 'it' should be enough for everyone :-)
Sep 11 21:45:17 <tuukkah> (says a finn)
Sep 11 21:45:42 <Ghiottone> s/it/he/
Sep 11 21:45:43 <apt> Ghiottone meant: SpeedEvil, but I don't know what he answered, I was too tired
Sep 11 21:46:05 <Ghiottone> sorry
Sep 11 21:46:12 <SpeedEvil> I don't actually know roh's gender, which is why I made the comment.
Sep 11 21:46:41 <SpeedEvil> As in needs a pronoun for a person of unknown gender.
Sep 11 21:46:49 <SpeedEvil> (and I know some have attempted it)
Sep 11 21:47:08 <mjr> mostly it needs not to have something as stupid as gendered pronouns
Sep 11 21:47:16 <mjr> (says another finn)
Sep 11 21:47:22 <tuukkah> not necessarily unknown but irrelevant
Sep 11 21:47:25 <Ghiottone> I see. actually Italian gives preference to male when there is uncertainity
Sep 11 21:47:47 <tuukkah> Ghiottone, there's a view that english does too
Sep 11 21:49:07 <tuukkah> "let's learn a language while webkit-gtk compiles" ;-)
Sep 11 21:49:16 <Ghiottone> :)
Sep 11 21:49:33 <Ghiottone> it seems that channel logs have some problems
Sep 11 21:49:47 <Ghiottone> at least, today's and yesterday's are almost not existant
Sep 11 21:51:37 <duffyd> Ghiottone: Thanks a lot - I know have a working UI/Neo! How do I know, though, if the installed kernel matches the version of the modules contained in the root image *before I install them*?
Sep 11 21:52:23 <Ghiottone> duffyd, the installed kernel is that one you flashed on the neo with dfu-util
Sep 11 21:52:41 <Ghiottone> duffyd, see the version of the modules package and flash the kernel accordingly
Sep 11 21:53:17 <Ghiottone> duffyd, at least I did this way when my neo did not load any module at all
Sep 11 21:53:28 <Ghiottone> duffyd, after an upgrade
Sep 11 21:55:30 <duffyd> Ghiottone: but theoretically doesn't the root image includes the /lib/modules dirs?
Sep 11 21:55:51 <tuukkah> indeed
Sep 11 21:55:53 <frma71> Any news on gllin availability ?
Sep 11 21:56:18 <tuukkah> and there can be modules tar and rootfs tar packages
Sep 11 21:57:13 <Ghiottone> yes. the problem only happens if you upgrade to a newer kernel version
Sep 11 21:59:08 <duffyd> Ghiottone: ahhh, ta
Sep 11 22:03:04 <Ghiottone> duffyd, you will probably have to reflash your kernel. I have just checked and the kernel you flashed is 2.6.21 while the most recent
Sep 11 22:03:46 <Ghiottone> duffyd, is 2.6.22.5 So when you'll do ipkg upgrade, you will install the modules for 2.6.22.5 automatically (I think)
Sep 11 22:05:16 <Ghiottone> duffyd, a newer kernel is automatically built every day or so here http://buildhost.openmoko.org/OM2007.2/tmp/deploy/glibc/images/fic-gta01/
Sep 11 22:05:32 <Ghiottone> duffyd, but that's most unstable stuff
Sep 11 22:05:58 <Ghiottone> duffyd, (in the sense that it is not tested)
Sep 11 22:16:11 <LittleIdiot> hmm, strange gsmd throws "Netreg registration denied", although the pin is correct...
Sep 11 22:16:32 <LittleIdiot> it worked with another sim card, but not with this one...
Sep 11 22:27:04 <duffyd> Ghiottone: thanks again :-)
Sep 11 22:27:08 <duffyd> cu all
Sep 11 22:28:43 <kristian-m> hi
Sep 11 22:29:35 <Ghiottone> hey kristian-m 
Sep 11 22:29:45 <frma71> Cool, the mediaplayer is actually working... reason enough to carry the neo around :)
Sep 11 22:32:40 <borg_> frma71: can it make a playlist now too?
Sep 11 22:32:52 <borg_> or do i have to make in on my pc before?
Sep 11 22:33:57 <borg_> s/in/it/
Sep 11 22:33:58 <apt> borg_ meant: or do i have to make it on my pc before?
Sep 11 22:39:14 <frma71> borg_: Yes, with the updates commited today !
Sep 11 22:39:19 <borg_> cool
Sep 11 22:39:22 <borg_> *upgrade* :)
Sep 11 22:41:09 <borg_> arghh
Sep 11 22:41:16 <borg_> ipkg upgrade restartet gsmd
Sep 11 22:41:18 <borg_> damn
Sep 11 22:41:38 <frma71> borg_ Mine didn't work together with pulseaudio (Resource busy) at first, I needed to create a /etc/asound.conf, but I might have lost it somewhere on the way !
Sep 11 22:43:31 <frma71> Anyone tried prelink on the neo ?
Sep 11 22:44:51 <frma71> Is there a reson the play-button is turned backwards in openmoko-mediaplayer2 ?
Sep 11 22:45:10 <rtm_> Does anyone know how I may add an application to the "utilities" menu under 2007.2?
Sep 11 22:47:49 <frma71> rtm_ Just put a .desktop file in /usr/share/applications
Sep 11 22:48:08 <borg_> mh
Sep 11 22:48:19 <borg_> what can i do to prevent ipkg from restarting gsmd?
Sep 11 22:48:21 <frma71> rtm_ Use mbinputmgr.desktop as a template !
Sep 11 22:48:31 <borg_> thats an odd situation ;)
Sep 11 22:48:42 <rtm_> frma71: Cool!   Thanks a bunch.
Sep 11 22:49:38 <borg_> i cant update because restarting gsmd kills my neo :)
Sep 11 22:50:17 <Ghiottone> borg_, fix the kernel cmdline
Sep 11 22:50:28 <borg_> how, where? ;)
Sep 11 22:50:29 <frma71> borg_ I would guess it calls /etc/init.d/gsmd start/stop... just put an exit 0 early in that file ..
Sep 11 22:50:30 <tuukkah> or the init script: http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=788
Sep 11 22:50:49 <borg_> frma71: ah ok, that will work definitly ;)
Sep 11 22:51:10 <frma71> borg_ remove console=/dev/tty in bootcmd_base environment variable of uboot.
Sep 11 22:51:18 <frma71> borg_ if I remember correctly.
Sep 11 22:51:40 <frma71> borg_ /dev/tty = /dev/ttySAC0
Sep 11 22:51:51 <tuukkah> frma71, there's a new workaround that doesn't need that change
Sep 11 22:52:00 <adminjs> hey kloeri
Sep 11 22:52:25 <Ghiottone> tuukkah, the new gsmd script?
Sep 11 22:52:29 <frma71> tuukkah: Whats that ?
Sep 11 22:52:47 <tuukkah> bottom of the bug report, some stty lines to the init script
Sep 11 22:52:49 <Ghiottone> tuukkah, I would better trust kernel cmdline, as gsmd could be killed or segfault
Sep 11 22:53:11 <borg_> where do i change the kernel cmdline?
Sep 11 22:53:14 <tuukkah> Ghiottone, it's not about gsmd, it's about powering off the gsm chip
Sep 11 22:53:41 <frma71> tuukkah: Is there a better workaround then disabling the console on serial ?
Sep 11 22:54:03 <frma71> borg_ in uboot.
Sep 11 22:54:19 <frma71> borg_: with printenv/setenv
Sep 11 22:54:21 <tuukkah> frma71, yes: disable hardware flow control before powerinf down the gsm chip
Sep 11 22:54:22 <borg_> ah, ok ;)
Sep 11 22:54:33 <borg_> i was searching for something like grub.conf :P
Sep 11 22:54:41 <frma71> tuukkah: Ahh, sweet.
Sep 11 22:54:41 <tuukkah> at least that's what the bug report now says
Sep 11 22:54:44 <mwester> The better workaround is to add the "stty" lines as outlined in the bug report for bug 788 -- the ideal solution is yet to be implemented.
Sep 11 22:54:51 <Ghiottone> tuukkah, you mean that it works if gsmd dies too?
Sep 11 22:55:11 <tuukkah> Ghiottone, it doesn't matter if gsmd dies
Sep 11 22:55:14 <mwester> Ghiottone:  the problem is not triggered by gsmd dying, it's triggered by the power-off of the GSM modem.
Sep 11 22:55:51 <mwester> At present, the only places the GSM modem is powered-off is in the /etc/init.d/gsmd script.  That could change, hence we need a better solution in the long term.
Sep 11 22:55:56 <Ghiottone> which is done in the script
Sep 11 22:55:58 <Ghiottone> I see
Sep 11 22:56:04 <Ghiottone> so that will work
Sep 11 22:56:11 <Ghiottone> thanks
Sep 11 22:56:16 <mwester> :)
Sep 11 22:57:25 <frma71> Well, I guess most of us don't need the console on serial anyway...
Sep 11 23:02:12 <mwester> frma71:  That is probably a true statement, but I intensely dislike the idea of giving up a console on the device.  So I'll be working on a better solution, anyway.
Sep 11 23:03:54 <borg_> hehe, ipkg is replacing /etc/init.d/gsmd before restarting it :P
Sep 11 23:04:29 <mwester> Oh, ouch!!
Sep 11 23:07:01 <borg_> bootargs_base=rootfstype=jffs2 root=/dev/mtdblock4 console=ttySAC0,115200 console=tty0 loglevel=8
Sep 11 23:07:22 <borg_> so i have to do setenv bootargs_base=rootfstype=jffs2 root=/dev/mtdblock4 console=tty0 loglevel=8
Sep 11 23:07:24 <borg_> ?
Sep 11 23:07:50 <borg_> i am a little bit scared of changing something in uboot ;)
Sep 11 23:08:14 <Marcel_M> no "=" after bootargs_base
Sep 11 23:08:20 <borg_> ok
Sep 11 23:08:59 <borg_> quit
Sep 11 23:09:03 <borg_> ups ;)
Sep 11 23:10:45 <Ghiottone> borg_, try with boot before saving the environment
Sep 11 23:11:24 <frma71> borg_ coward :)
Sep 11 23:13:41 <borg_> *reboot*
Sep 11 23:14:02 <frma71> Odd, I installed dillo (web browser) on my neo and that starts three times faster then any of the openmoko applications...
Sep 11 23:16:38 <borg_> worked ;)
Sep 11 23:20:22 <jseghers_> Hi all...
Sep 11 23:21:11 <jseghers_> I'm having problems with building the current version of 2007.2... openmoko-terminal2 has a dependency on "vte" which apparently doesn't exist in the build tree?
Sep 11 23:22:12 <frma71> jseghers_: Odd, in mine it vte does exist, but I havn't updated oe today !
Sep 11 23:22:45 <tuukkah> jseghers_, i was able to get bitbake vte separatel
Sep 11 23:22:48 <tuukkah> y
Sep 11 23:23:51 <jseghers_> tuukkah: hmmm yeah, I was able to build it manually, too... I wonder why it wouldn't build automatically?
Sep 11 23:24:15 <kristian-m> jseghers_ maybe you also want to use mrxvt as openmoko-terminal2 is still experimental
Sep 11 23:24:19 <tuukkah> jseghers_, isn't the dependency missing from the bb file?
Sep 11 23:25:23 <jseghers_> Actually... I think I know what the problem may have been... I was having a problem with terminal not building.  And since I have parallel build options set, it's not always clear what the root failing cause is.
Sep 11 23:25:39 <jseghers_> So I was building one package at a time... 
Sep 11 23:26:01 <jseghers_> Which doesn't necessarily build the dependencies...
Sep 11 23:26:46 <jseghers_> kristian-m: I'll try that if I can't get this to work...
Sep 11 23:27:11 <tuukkah> build a package does build dependencies too if you use bitbake <package>
Sep 11 23:27:34 <jseghers_> One of the really odd things is that when I go to build openmoko-devel-image, I'm getting errors on web-kit not building... but if I build webkit directly, it builds fine.
Sep 11 23:27:59 <tuukkah> same here
Sep 11 23:28:25 <tuukkah> although, i had to try many times until it passed
Sep 11 23:28:26 <jseghers_> tuukkah: you're seeing it too with webkit?
Sep 11 23:28:40 <tuukkah> yeah. i don't know what made it
Sep 11 23:29:13 <frma71> I've disabled webkit (actually the feed reader), it's not very intresting anyway !
Sep 11 23:29:13 <jseghers_> I built it fine... then tried restarting the overall build and it insisted it needed to rebuild it--and failed.
Sep 11 23:29:16 <tuukkah> could be another issue with parallel build
Sep 11 23:29:54 <tuukkah> frma71, is dillo more interesting?
Sep 11 23:31:03 <ckuethe> no
Sep 11 23:31:13 <ckuethe> dillo is kinda handy
Sep 11 23:31:16 <ckuethe> but pretty limited
Sep 11 23:31:35 <tuukkah> doesn't webkit come with some kind of a simple browser?
Sep 11 23:34:35 <frma71> tuukkah: Don
Sep 11 23:35:11 <frma71> tuukkah: Don't know really, but isn't webKit in the standard build just compiled because of the dependency of openmoko-feedreader.
Sep 11 23:35:36 <tuukkah> i suppose in time we'll have a browser based on the same engine
Sep 11 23:37:23 <frma71> Is webKit the engine used in safari ? 
Sep 11 23:37:47 <tuukkah> yes
Sep 11 23:38:34 <frma71> Hmm, webKit actually exports /usr/bin/GdkLauncher. 
Sep 11 23:38:50 <frma71> Maybe I need to compile it anyway then...
Sep 11 23:47:25 <orospakr> hm, I don't suppose there will be some sort of credit against a GTA02 for folks who purchased the GTA01 dev unit?
Sep 11 23:47:40 <orospakr> I dimly remember a post by Sean or Harald to that effect...
Sep 11 23:47:49 <tuukkah> yeah
Sep 11 23:48:17 <spikebike> I have a vague memory saying that any discount offered is likely to be less than what you could sell it for on ebay
Sep 11 23:48:17 <frma71> orospakr: I think they lowered the price of the GTA01 with $50 instead...
Sep 11 23:48:36 <cesarb> orospakr: the "credit" is already built into GTA01's price
Sep 11 23:48:42 <cesarb> orospakr: so they don't have to track it
Sep 11 23:48:53 <orospakr> hm.
Sep 11 23:48:59 <orospakr> thanks, guys. :)
Sep 11 23:49:24 <cesarb> (I, for one, am grateful for that, since it lowered the absurdly high taxes a little bit for me)
Sep 11 23:49:24 <orospakr> well, since my GTA01 hasn't shipped *yet*, I might just cancel the order and wait it out.
Sep 11 23:49:47 <anrp> hopefully not too long a wait
Sep 11 23:50:24 <frma71> No one heard anything about gllin, I saw something like "probably/hopefully next week" here last week !
Sep 11 23:50:29 <tuukkah> orospakr, more gta01 is going to ship soon though
Sep 11 23:50:47 <Maledictus> gllin?
Sep 11 23:51:19 <frma71> orospakr: Well, I'd wait... The software is (at least close to) miserable and the software required to use the GPS hasn't been released yet !
Sep 11 23:51:52 <anrp> not on purpose, at least
Sep 11 23:51:55 * anrp coughs
Sep 11 23:51:55 <frma71> Maledictus: The propritary software communicating with the GPS.
Sep 11 23:52:06 <Maledictus> ah, ok
Sep 11 23:52:17 <frma71> ... as I understand it !
Sep 11 23:52:24 <Maledictus> afaik there is no plan to ever release it
Sep 11 23:52:36 <anrp> that may be the case
Sep 11 23:52:43 <anrp> since v02 will have a different gps altogether
Sep 11 23:52:54 <anrp> but it would really suck if they never legally allowed v01's gps to work
Sep 11 23:52:55 <Maledictus> hooray for that :)
Sep 11 23:53:11 <anrp> it's not exactly cheap
Sep 11 23:53:13 <mjr> Maledictus, shows what you know
Sep 11 23:53:45 <Maledictus> ah sorry, with release you meant releasing the blob and not releasing it as opensource?
Sep 11 23:54:03 <mjr> I assume as much
Sep 11 23:54:12 * anrp pokes the batteries on his desk with a 40a discharge
Sep 11 23:54:25 <Maledictus> mjr, I was an a talk from michael lauer and he said the hope the GTA02 won't need a binary driver for the gps
Sep 11 23:54:59 <mjr> it's taken... too much time, but at least last week still there was a Broadcom tech on the community list and mickey said he's working on an EABI version of gllin. No source, of course.
Sep 11 23:55:32 <mjr> yeah, the last word was that they were evaluating another gps chip that does computation in firmware
Sep 11 23:56:10 <mjr> the final gps chip decision hadn't been properly made at least last week
Sep 11 23:56:54 <frma71> I don't care that much about source regarding gllin as long as I get the nessecary data out of it I'd just consider it firmware running on the same CPU.
Sep 11 23:57:18 <mjr> I care, but am willing to put up with it
Sep 11 23:57:44 <mjr> and my gta­01 will end up with my wife for some time, and she'd very much like the gps to work...
Sep 11 23:57:45 * mwester is an avowed skeptic, and points out that there is no positive reason for anyone to announce officially that the broadcom GPS is NOT part of GTA02 until broadcom at least delivers on the promised new driver for the GTA01.
Sep 11 23:58:11 <mjr> mwester, that crossed my mind too, but I decided not to start with that :]
Sep 11 23:58:22 <Maledictus> what is EABI?
Sep 11 23:58:49 <mjr> the binary interface for openmoko 2007.2
Sep 11 23:58:59 <Maledictus> I care very much about that special thing, as I plan to run FreeBSD on the neo.
Sep 11 23:59:01 <Maledictus> ok
Sep 11 23:59:15 <mwester> EABI is the mechanism for calling functions and getting return values; it is better than the predecessor, but unfortunately code compiled in the old format is not compatible.
Sep 12 00:00:24 <mwester> Although should broadcom fail to live up to their end of the agreement, I have little doubt that somebody in this community is smart enough to write a wrapper for the old library.
Sep 12 00:00:51 <mwester> Er, for the benefit of the few who have a copy of that old library, of course.
Sep 12 00:00:53 <cesarb> Maledictus: ouch, you will have a lot of work reimplementing all the drivers...
Sep 12 00:01:09 <tuukkah> i think there already is a wrapper
Sep 12 00:01:11 <mjr> mwester, or we'll end up with a gllin chroot...
Sep 12 00:01:29 <cesarb> mwester: no need for wrapper, the kernel can run both EABI and OABI binaries, you just need a OABI glibc
Sep 12 00:01:32 <tuukkah> ok, well, there's a chroot
Sep 12 00:01:35 <Maledictus> cesarb: at first I can take the GPL'ed ones I think. But documentation for the hw is necessary anyway
Sep 12 00:02:44 <cesarb> Maledictus: for some of the hardware, the sole documentation is the code, some random emails scattered over half a dozen mailing lists, and several wiki pages (which might or might not be correct or relevant to your hardware revision)
Sep 12 00:02:55 <cesarb> Maledictus: for instance, the GTA01 board itself
Sep 12 00:03:34 <Maledictus> thats a little bit sad. But I did not expect the neo to be perfect on first try ;)
Sep 12 00:04:03 <mjr> it's a small miracle that for the gta-02 gpu there apparently are at least plans to make docs
Sep 12 00:04:26 <Maledictus> mjr: yeah. I was very happy to read that!
Sep 12 00:05:12 <Maledictus> does anyone know if there is also a user manual for the s3c2442? couldn't find a link so far
Sep 12 00:05:14 <cesarb> Maledictus: try grokking the documentation on the power buses (there are about a dozen different ones, the best source is the wiki page) or the audio routing (two audio chips, one of them with a very high amount of toggles and switches, plus it goes to half a dozen chips/plugs/speakers/mics)
Sep 12 00:05:58 <cesarb> Maledictus: I think I read a mailing list post two days ago saying the s3c2442 docs disappeared in the horrible website shuffle the manafacturer made some time ago, and weren't found again yet
Sep 12 00:06:10 <Maledictus> cesarb: I set up this page http://wiki.openmoko.org/wiki/GTA02_Openness , I hope it will be more complete over time...
Sep 12 00:06:37 <Maledictus> cesarb: grml, I searched the whole day ;)
Sep 12 00:06:50 <cesarb> Maledictus: the same email said he would either ask them to put the docs back online or ask for permission for redistribution
Sep 12 00:07:49 <Maledictus> redistribution of a doc which has been made publicly available by the copyright holder shouldn't be a problem?
Sep 12 00:08:21 <cesarb> Maledictus: the whole point of copyright is to annoyingly forbid redistribution of something made publicly available
Sep 12 00:08:48 <Maledictus> hehe, you're right
Sep 12 00:09:07 * cesarb looking at your new wiki page right now
Sep 12 00:09:19 <mjr> indeed...
Sep 12 00:09:35 <cesarb> Maledictus: the bluetooth module will most probably be some standard HCI variant (GTA01 uses hci-usb IIRC)
Sep 12 00:10:09 <roh> bluetooth will be hci-usb, jap
Sep 12 00:10:14 <roh> standard csr module
Sep 12 00:10:15 <cesarb> Maledictus: the vibrator is nothing more than a small motor with an unbalanced weigth, driven by the CPU's PWM... no need for docs on that
Sep 12 00:10:16 <Maledictus> cesarb: sounds pretty standard (sorry I'm pretty new to this embedded world and hw specs, etc.)
Sep 12 00:10:31 <Maledictus> cesarb: yay, sounds good
Sep 12 00:10:35 <cesarb> roh: hci-usb? they won't change the link to uart or something even more random this time?
Sep 12 00:10:37 <Maledictus> I'll correct it
Sep 12 00:10:53 <Maledictus> HCI is an open standard with docs available?
Sep 12 00:11:22 <cesarb> Maledictus: yep
Sep 12 00:11:32 <cesarb> Maledictus: it's in the middle of a very thick PDF IIRC
Sep 12 00:11:46 <cesarb> Maledictus: in the official bluetooth specs
Sep 12 00:11:56 <Maledictus> thats ok.
Sep 12 00:12:07 <rwhitby> it's register-ware
Sep 12 00:12:14 <cesarb> Maledictus: and hci-usb is a hci variant over usb, which is also publicly documented
Sep 12 00:12:36 <Maledictus> wonderful. so at least one wireless thing is open enough ;)
Sep 12 00:13:05 <Maledictus> rwhitby: hi, have you read my proposol for the mokomakefile change?
Sep 12 00:13:18 <mjr> so you can do walkie-talkie over bluetooth ;)
Sep 12 00:13:26 <rwhitby> HCI is at http://www.bluetooth.com/Bluetooth/Learn/Technology/Specifications/
Sep 12 00:13:29 <cesarb> Maledictus: the touch screen... it's connected directly to either the CPU or the PMU, using the CPU/PMU's touch screen block (probably the CPU in the neo), so you just need the CPU/PMU docs
Sep 12 00:13:35 <Maledictus> rwhitby: thanks!
Sep 12 00:13:35 <rwhitby> Maledictus: you didn't see the CIA message here with the commit?
Sep 12 00:13:56 <Maledictus> rwhitby: sorry, I'm ill and have been in bed most of the time
Sep 12 00:14:07 <cesarb> Maledictus: the touch screen is in fact nothing more than a giant resistor, all the smarts is on the chip it's connected to
Sep 12 00:14:38 <Maledictus> PMU = power management unit??
Sep 12 00:15:25 <cesarb> Maledictus: JTAG is JTAG (a very common embedded standard; you have to pay for this one, or just check on the web - it's so common there are dozens of places explaining it), but you also need a (IIRC) BSDL file (which describes the chip as viewed by the JTAG); that file should come with the chip's documentation
Sep 12 00:15:39 <cesarb> Maledictus: yeah, the PFC* chip
Sep 12 00:16:05 <rwhitby> and the BSDL for openmoko will be found in the openocd sources
Sep 12 00:16:41 <Maledictus> what is openocd?
Sep 12 00:17:26 <cesarb> Maledictus: the LCD is proprietary, but the most important part driver-side would be the chip controlling it (the interface is not that nonstandard). On the GTA01 it's the CPU itself, on the GTA02 it's the SMedia chip
Sep 12 00:17:28 <Maledictus> does one need a BSDL file for every chip or just for the cpu?
Sep 12 00:18:28 <cesarb> Maledictus: AFAIK you would need at least a partial one for every chip _connected to the same JTAG chain_... There are probably several scan chains on the neo, the only one which would matter would be the one with the CPU on it (which probably has nothing but the CPU on it...)
Sep 12 00:18:50 <Maledictus> ok.
Sep 12 00:19:12 <Maledictus> your help and explanations are very much appreciated btw :D
Sep 12 00:19:22 * cesarb doesn't know how much would be needed... perhaps just some length, perhaps just "there's 2 chips here, skip them"... my JTAG isn't that good
Sep 12 00:20:35 <cesarb> For the SD card you would be able to use some specs the SD guys recently released for the public, or the Sandisk ones the linux kernel guys used earlier (check the comment at the top of the kernel's SD sources)
Sep 12 00:20:50 <cesarb> It's the same as a normal SD card, in fact, perhaps *BSD already have this one
Sep 12 00:20:53 <anrp> or run it in mmc mode, if it's wired up to allow spi
Sep 12 00:21:23 <cesarb> Ah, you would also need a driver for the SD controller... Which would be in the CPU (so, as you can see, the CPU docs are really the most important of them all)
Sep 12 00:21:52 <Maledictus> yes, i think SD could work ootb
Sep 12 00:22:07 <cesarb> And the flash chip should be easy to find, and they are almost all alike (the fact that the CPU can talk to it directly helps)
Sep 12 00:22:41 <cesarb> So, as you can see, it's a lot of work... Each driver individually would be easy (sorta), but there are tens of them!
Sep 12 00:22:45 <Maledictus> yes, that why I spent so many hours searching for it. I found the s3c2410 docs and thought "hopefully the s3c2442 docs are also that good"
Sep 12 00:23:16 <cesarb> Maledictus: they probably will be pretty similar. Most embedded SoCs have docs in that style. Just watch out for the errata.
Sep 12 00:23:27 <Maledictus> cesarb: I hope I will not be alone with the little project ;)
Sep 12 00:23:46 <anrp> yeah... errata...
Sep 12 00:24:20 <anrp> "yeah, you know that great feature that sold this chip? yeah, about that.... it doesn't actually work. but it will in silicon rev 2, don't forget to buy it!"
Sep 12 00:24:24 <cesarb> Maledictus: well, you probably will be alone unless you get some other guys from the BSD camp... Most people here will be too busy trying to make that damn gsmd work ;-)
Sep 12 00:24:31 <Maledictus> I'm currently active in advertising the neo for some other BSD guys. And this weekend will be the biggest BSD conference in europe where I will try to promote it too
Sep 12 00:24:34 <cesarb> anrp: lol
Sep 12 00:24:54 <Maledictus> lool
Sep 12 00:25:13 <anrp> at least they don't flat out say it
Sep 12 00:25:13 <cesarb> Maledictus: and the bootloader? Are you thinking of keeping the linux-derived uboot?
Sep 12 00:25:23 <anrp> uboot is not linux derived, really
Sep 12 00:25:30 <Maledictus> cesarb: Thats ok with me. I happy if no one is angry with me for trying the port and not using linux :)
Sep 12 00:25:42 <anrp> uboot can boot anything that can run on the hw
Sep 12 00:25:53 <cesarb> Maledictus: nah, we won't be angry, we'll just point and laugh ;-)
Sep 12 00:25:53 <anrp> up to and including another os :o
Sep 12 00:26:17 <anrp> like ... windows ce!
Sep 12 00:26:21 * anrp runs for the hills
Sep 12 00:26:25 <cesarb> anrp: urgh
Sep 12 00:26:55 <Maledictus> boot loader is interesting. ther is this redboot thing which already can boot a freebsd kernel on arm. but I think it's more work to port redboot than to get u-boot boot freebsd
Sep 12 00:27:03 <Maledictus> hehe
Sep 12 00:27:08 * anrp nods
Sep 12 00:27:24 <Maledictus> there is already much work in u-boot for the neo
Sep 12 00:27:51 <cesarb> Maledictus: ah, and you ask in the wiki "is that enough" for both audio chips... the docs are more than enough for the chips themselves, but it won't do any good if you don't know how they were wired to the other chips (the audio part has the second most complex routing I've seen in the neo, the first being the power buses...)
Sep 12 00:27:51 <Maledictus> I talked to a guy who used u-boot to boot a freebsd/ppc kernel, so it shouldn't be that hard
Sep 12 00:27:52 <mjr> (the namer of u-boot should be whacked on the head with a newspaper, then bought a beer)
Sep 12 00:28:15 <cesarb> mjr: ouch... just now I noticed the pun :P
Sep 12 00:28:19 <LittleIdiot> may it be that the version of gsmd coming with the august snapshot is... uhm quite disfunctional? ;)
Sep 12 00:28:20 <Maledictus> cesarb: the routing is described on other wiki pages, at least for GTA01
Sep 12 00:28:28 * cesarb had never noticed the awful pun before
Sep 12 00:28:50 <roh> hm.. somewhere i saw fbsd people in the wiki...
Sep 12 00:28:54 <cesarb> LittleIdiot: nah, it just gets confused so easily
Sep 12 00:29:53 * cesarb has an anti-confusing patch for gsmd on bug 766, which doesn't fix the fact that sometimes the modem doesn't answer it for some reason
Sep 12 00:30:03 <Maledictus> I would be interested if it's easily possible to dual boot linux and freebsd with u-boot. because I think it will be a long time before it is usable enough
Sep 12 00:30:19 <cesarb> Maledictus: well, it's possible to dual-boot linux and linux already
Sep 12 00:30:21 <mwester> LittleIdiot:  I personally don't think it's all the fault of gsmd; there seems to be some indication that there is something about the power-up and initialization of the GSM module itself that we don't know or understand just yet.
Sep 12 00:30:25 <cesarb> Maledictus: with the "boot from SD" menu item
Sep 12 00:30:36 <Maledictus> aaaaahh!
Sep 12 00:30:47 <Maledictus> hum
Sep 12 00:30:59 <cesarb> mwester: agreed, the "unconfusing" patch doesn't make much difference
Sep 12 00:31:09 <Maledictus> I don't have any hardware yet. Just the qemu. It doesn't show me that menu item :(
Sep 12 00:31:39 * SpeedEvil stabs samsungs website.
Sep 12 00:31:42 <roh> Maledictus if you for example boot from sdcard for fbsd and from nand for linux.. yeah.. should work
Sep 12 00:31:52 <Maledictus> sounds great!
Sep 12 00:32:09 <Maledictus> boot from sd just reads the first sector and executes it?
Sep 12 00:32:25 <roh> i just think somebody needs to put in quite some work to add all the drivers to fbsd
Sep 12 00:32:26 <cesarb> Maledictus: no, it reads a file from the SD
Sep 12 00:32:31 <cesarb> Maledictus: which must be formatted as ext2
Sep 12 00:32:42 <roh> Maledictus nope. uboot knows ext and vfat i think.
Sep 12 00:32:42 <Maledictus> uh
Sep 12 00:32:43 <Maledictus> ok
Sep 12 00:32:55 <Maledictus> ext2 should work. as opposed to jffs2!
Sep 12 00:32:55 <cesarb> Maledictus: look at openmoko/trunk/src/host/devirginator/environment.in, which is where the default uboot environment for the real hardware comes from, and thus where that menu item comes from
Sep 12 00:33:01 <roh> uboot is quite a monster, featurewise
Sep 12 00:33:11 <cesarb> roh: not enough
Sep 12 00:33:22 <Maledictus> yeah! at first I thought the "u" stood for micro ;)
Sep 12 00:33:42 * cesarb wants touchscreen input for uboot, so you don't have to plug the USB to play with it!
Sep 12 00:34:14 <Maledictus> jffs2 will be a real problem. no docs, just GPL'ed code. no idea if it's possible to integrate into fbsd kernel
Sep 12 00:34:21 <mwester> Gak!  it'll be an operating system on its own soon!
Sep 12 00:34:24 * bedboi wants a neo
Sep 12 00:34:42 <cesarb> Maledictus: but you _need_ either jffs2 or a wear-leveling layer
Sep 12 00:34:49 * mwester wonders when somebody will write a GUI for uboot...
Sep 12 00:34:54 <cesarb> Maledictus: the flash doesn't have builtin wear leveling and error checking/correction
Sep 12 00:35:05 <Maledictus> netbsd has an 20 year old log-structered FS ;)
Sep 12 00:35:08 <anrp> yay for dum flass
Sep 12 00:35:10 <frma71> Hmmm... just started openmoko-dates2, consumes 50% cpu by just showing the calendar :I
Sep 12 00:35:17 <cesarb> Maledictus: if you don't add one (either on the filesystem, like jffs2, or in a separate layer), you WILL wear the flash
Sep 12 00:35:34 <Maledictus> cesarb: I know. The SD too?
Sep 12 00:35:41 <anrp> sd does it internally
Sep 12 00:35:44 <anrp> so no worries there
Sep 12 00:35:46 <roh> Maledictus one thing i found out in the last few days, as a feature for the bsdies is: search for a copyright in the register definitions of our drivers.. ;)
Sep 12 00:35:50 <cesarb> Maledictus: a log-structured filesystem is not enough. You also need error correction.
Sep 12 00:36:10 <anrp> sd looks more or less like a normal block device
Sep 12 00:36:12 <roh> Maledictus basically the people writing the fbsd drivers can just copy them.
Sep 12 00:36:29 <bedboi> freelock: are you kidding?
Sep 12 00:36:32 <bedboi> 50%?
Sep 12 00:36:33 <anrp> ..?
Sep 12 00:36:36 <Maledictus> roh? sure? sounds to great to be true..
Sep 12 00:36:40 <bedboi> s/freelock/frma71
Sep 12 00:36:44 <bedboi> (damn autocomplation)
Sep 12 00:37:14 * rwhitby always wonders why people are so hung up on which OS they use, rather than just using Windows/Linux/FreeBSD/Angstrom/SlugOS etc for whatever applications they need to run.
Sep 12 00:37:18 <Maledictus> I should just start with using SD for porting. cheap enough to forget about wearing...
Sep 12 00:37:22 <freelock> Was wondering who was paging me here ;-)
Sep 12 00:37:26 <roh> Maledictus thats the answer i got. seems to be difficult to apply copyright on something like a tabelaric list of addresses and offsets... 
Sep 12 00:37:42 <frma71> bedboi: Yep, seems to be polling some socket... Well it's only the first release
Sep 12 00:37:42 <bedboi> rwhitby: it's kinda FAN thing
Sep 12 00:37:44 <hhf423> re
Sep 12 00:38:00 <cesarb> rwhitby: because operating systems are Serious Business
Sep 12 00:38:00 <roh> what the heck. i like gpl, but i think that bsdies should have a fair chance to redo the wheel with their licence
Sep 12 00:38:01 <bedboi> rwhitby: and that's the point of neo i guess.
Sep 12 00:38:08 <bedboi> stealing all linux nerds
Sep 12 00:38:11 <cesarb> (see http://tvtropes.org/pmwiki/pmwiki.php/Main/SeriousBusiness for a definition of the term)
Sep 12 00:38:23 <Writchie|dinner> cesarb: IIRC from mail with Harold the SD on the GTA02 is hung of the WiFi chip and the WiFi is connected to the SoC SD Port
Sep 12 00:38:32 <bedboi> roh: no man. it's not like that.
Sep 12 00:38:45 <anrp> the... what the hell?
Sep 12 00:38:46 <bedboi> roh: go and tell gnu guys. they will slaughter you
Sep 12 00:38:56 <anrp> sd is off the wifi??
Sep 12 00:39:14 <cesarb> Writchie|dinner: wow, they managed to make the routing for things EVEN MORE COMPLEX?
Sep 12 00:39:26 <anrp> that is unpossible!
Sep 12 00:39:27 * rwhitby equates OS fanaticism with religeous fanaticism
Sep 12 00:39:33 <anrp> unless they meant to say they're on the same SPI bus
Sep 12 00:39:36 <anrp> which could work
Sep 12 00:39:37 <cesarb> Writchie|dinner: so, like, you have to turn on the _wireless_ to read the card?
Sep 12 00:39:38 <roh> bedboi i am strongly pro gpl (v2) but i oppose the practise of just copying whole drivers and making them fit somehow. thats wrong and leads to bad code anyways.
Sep 12 00:39:39 <SpeedEvil> I thought the SD was hung off the display controller
Sep 12 00:39:41 <Maledictus> rwhitby: I'm just very interested in doing stuff like this. I am fine using linux though
Sep 12 00:40:07 <Maledictus> I have enough linux machines, it's not like I only use freebsd
Sep 12 00:40:10 <rwhitby> Maledictus: yeah, you have shown no fanatical outburts as far as I have seen :-)
Sep 12 00:40:25 <Writchie> SpeedEvil: Maybe you're right and i'm not recalling it correctly - let me google my mail
Sep 12 00:40:26 <bedboi> roh: yep, that does not have nothing to do with the license you are applying
Sep 12 00:40:27 <roh> bedboi on the other side its totally ok to read a gpl driver and copy only the register-mappings and implement it again, with proper architecture, fitting to the framework the os gives.
Sep 12 00:40:42 <cesarb> SpeedEvil: perhaps the SD is hung off the wifi which is hung off the display chip which is hung off...
Sep 12 00:40:58 * mwester is a fanatic for hire -- a job that generally makes one pretty much a pragmatist when it comes to things like operating systems...
Sep 12 00:40:59 <anrp> hung off of the touchscreen of course
Sep 12 00:41:01 <bedboi> roh: yep because you are rewriting it
Sep 12 00:41:07 <anrp> which is hung off the lanyard, which is hung off your neck
Sep 12 00:41:13 <anrp> not sure what that has to do with anything though
Sep 12 00:41:34 <Maledictus> roh: I'm happy you see it this way. I have heard harsh words when I told some people about my idea
Sep 12 00:41:46 <cesarb> rwhitby: it's like sports (ok, sports fanaticism also don't make much sense)
Sep 12 00:41:46 <roh> bedboi which also works into the other direction. reading a bsd driver and writing a gpl one. as long as you don't copy code its fine.
Sep 12 00:41:49 <Writchie> SpeedEvil: ar6k module will be connected via sdio with 4 data pins
Sep 12 00:41:55 <anrp> register mapping is in the same class of work as a telephone book
Sep 12 00:42:09 <anrp> no creativity -> almost impossible to assert copyright
Sep 12 00:42:10 <bedboi> roh:  that's the cool part of all the open source
Sep 12 00:42:15 <bedboi> you can learn a lot
Sep 12 00:42:16 <anrp> which gives it equiv usability as public domain
Sep 12 00:42:19 <cesarb> anrp: "dial 555-1234 for register 1..."
Sep 12 00:42:22 <Writchie> microSD slot will attach to the glamo3362 SD/MMC controller
Sep 12 00:42:28 <SpeedEvil> Writchie: I know - I thought I heard that the SD bus comes from the accel chip though.
Sep 12 00:42:36 <SpeedEvil> Oh - right
Sep 12 00:42:38 <SpeedEvil> I see.
Sep 12 00:42:47 <Writchie> i guess the 3362 must be dma
Sep 12 00:42:56 <anrp> no docs for the chip ;/
Sep 12 00:43:12 <roh> Maledictus its a fine line with the gpl/bsd stuff and porting between it. and there were some harsh words in the recent history due to some idiot stepping over the line and stealing basically all important code from some reverse-engineering based brmc driver i think. but to be fair, i like pluralism.
Sep 12 00:43:16 * hhf423 yawns
Sep 12 00:43:56 * mwester returns from the kitchen, having located his telephone directory and found (to his amusement) that it is indeed copyrighted.  :O
Sep 12 00:44:04 <roh> same with xine and mplayer/ffmpeg. its positive to have some concurrent developments which merge features and bugfixes either way. its just important to keep to some (not too complicated) rules.
Sep 12 00:44:31 <Lynet> mwester: Feist vs Rural. The telephone directory itself is not covered by copyright.
Sep 12 00:44:54 <Writchie> SpeedEvil: so SD will need new u-boot drivers as well as kernel drivers
Sep 12 00:45:04 <SpeedEvil> Indeed.
Sep 12 00:45:07 <frma71> Can anyone recomend a filemanager that is usable on the neo ? 
Sep 12 00:45:08 <Maledictus> roh: yep. I think Open Source is mainly about choice. And I like to have good choices :)
Sep 12 00:45:15 <Writchie> our boy harald gonna be busy 
Sep 12 00:45:26 <hhf423> frma71: gpe-filemanager
Sep 12 00:45:36 <mwester> Hmmm.  So apparently the (c) symbol is a ruse by the publisher, or it must cover some other part of the publication, then, I gues...
Sep 12 00:45:40 <frma71> hhf423: tnx!
Sep 12 00:45:55 <cesarb> Writchie: ...which means we must fix gsmd to give him some slack...
Sep 12 00:46:22 <Writchie> i've decided to just go ahead with with d-bus layer direct to the chip
Sep 12 00:46:25 <hhf423> frma71: and it should be already in one of the repositories, e.g. in the ScaredyCat repro, so you can just ipkg install it
Sep 12 00:46:53 <Writchie> cesarb: i'll have some docs and be in a position where others can help if they want in about two weeks. 
Sep 12 00:46:54 <hhf423> cesarb: lots of beer await the one who fixes gsmd
Sep 12 00:46:56 <Lynet> mwester: To be exact, the data in the directory (names, numbers, etc..) are pure facts and not covered by (C). The typesetting etc might be covered by (C). As long as you only copy the facts, you should be safe in the US.
Sep 12 00:47:09 <mwester> Writchie:  I've been toying with the idea of creating an in-kernel virtual tty device for it, to conceal the line speed, flow control, and the console mux functions.
Sep 12 00:47:41 <Writchie> there are differring opinions kernel vs. user space - i am in the user-space camp
Sep 12 00:47:48 <cesarb> Lynet: the US is not the whole of the world.
Sep 12 00:47:54 <Writchie> although this is not a fundamental detail. 
Sep 12 00:48:06 <hhf423> Lynet: mwester we had a case in .de a few years back. It was not ok to scan the phonebook and make a cd out of it, but is was ok to have chinese workers read it and type it into a computer again
Sep 12 00:48:08 <frma71> Is the sms support implemented in gsmd yet ? 
Sep 12 00:48:16 * cesarb is in the "whatever" camp
Sep 12 00:48:26 <Writchie> their is some crude support
Sep 12 00:48:32 <mwester> Writchie:  it would be good to get it working in user-space before we pursue any in-kernel effort!
Sep 12 00:48:40 <Writchie> mwester: yes
Sep 12 00:48:45 <cesarb> mwester: as a set of perl scripts!
Sep 12 00:48:46 * cesarb hides
Sep 12 00:48:51 <hhf423> frma71: for the moment the most elegant way to do sms is using smstool, see wiki
Sep 12 00:48:57 <SpeedEvil> what working /me missed it
Sep 12 00:48:57 <Lynet> cesarb: True, in EU databases are covered by sui generis. It is not copyright per se, but a 'related right' is given to those that assemble a database of facts.
Sep 12 00:48:58 <mwester> Perl - good idea.
Sep 12 00:49:21 <Writchie> hhf423: not in chinese
Sep 12 00:49:27 <mwester> Performance and overhead is secondary to just getting it working, and Perl has reasonable serial support and excellent pattern matching.
Sep 12 00:49:28 <Writchie> ;)
Sep 12 00:50:02 <cesarb> mwester: as long as we don't blow the flash... how large is perl nowadays?
Sep 12 00:50:06 <frma71> hhf423: Ok, looks like it is supported by gsmd and libgsmd then.
Sep 12 00:50:15 <frma71> What about GPRS support in gsmd ?
Sep 12 00:50:24 <cesarb> frma71: none
Sep 12 00:50:26 <hhf423> Writchie: in chinese it would be even see of an issue because a translation here has a copyright of it own
Sep 12 00:50:28 <mwester> Perl is huge, but if you remove all but the modules you need, you can get it down to a MByte or two...
Sep 12 00:50:43 * cesarb abstains from saying "Lua!"
Sep 12 00:50:47 <hhf423> frma71: yes, you could sms via libgsmd-tool
Sep 12 00:50:53 <Writchie> hhf423: i mean Chinese SMS
Sep 12 00:51:09 <hhf423> frma71: gprs see "manually using gprs" in the wiki
Sep 12 00:51:30 <hhf423> frma71: there are scripts there. for a gui based gprs see "settingsgui"
Sep 12 00:51:55 <mwester> Regarding the GSM itself, though, if roh could get one of us the docs for the device, that would greatly help figure out the initial power-up sequence.  I'd be willing to sign a non-disclosure (provided it was reasonable, of course).
Sep 12 00:52:03 <frma71> hhf423: Tried that, but I'd like to be able to receive calls/sms while keeping my gprs connection, so I guess that support need to be in gsmd !
Sep 12 00:52:48 <hhf423> frma71: probably. but that would be a nice feature of course
Sep 12 00:52:53 <Writchie> you pretty much have to use the mux and several channels to do things reliably and sanely
Sep 12 00:53:10 <SpeedEvil> Basically - as I understand it and in practice in my experience, doing echo 0 then echo 1 to the power_on file on the GSM module, with 1s seperating them works fine
Sep 12 00:53:16 <Maledictus> roh: could you give me an example of a filename where I can find those register definitions?
Sep 12 00:53:28 <cesarb> frma71: the mux is supposed to be in the kernel, according to harald
Sep 12 00:53:57 <Writchie> cesarb: premature optimization is the root of all evil
Sep 12 00:54:08 <Writchie> quote from harald
Sep 12 00:54:17 <Writchie> stolen from knuth
Sep 12 00:54:24 <frma71> cesarb: Ah, thats how its supposed to work.
Sep 12 00:54:25 <bedboi> Writchie: i definitely agree
Sep 12 00:54:35 <Maledictus> ha, theres another bsd guy :)
Sep 12 00:54:49 <Maledictus> daniel_bergamini: will you be in copenhagen this weekend?
Sep 12 00:54:54 <cesarb> Writchie: well, for all i care, you can do the mux in userspace, in javascript, as long as it works reliably and doesn't drain too much power
Sep 12 00:55:15 <roh> Maledictus nearly all .h files which openmoko adds for hardware drivers
Sep 12 00:55:24 <roh> Maledictus linux-2.6.22.5-moko/include/asm-arm/arch-s3c2410/gta01.h for example
Sep 12 00:55:45 <Maledictus> which should be somewhere under openmoko/trunk/src/target/kernel/patches ?
Sep 12 00:56:04 <Maledictus> gta01-core.patch got it :)
Sep 12 00:56:24 <roh> Maledictus ack. also gta02-core.patch has some headers
Sep 12 00:56:24 <Writchie> the mux has its own power up power down. perhaps it's not implemented correctly - we'll find out soon.
Sep 12 00:56:36 <Ghiottone> rwhitby, ping
Sep 12 00:58:09 <mwester> Writchie:  what mux speak you of?  Audio?
Sep 12 00:58:12 <rwhitby> Ghiottone: SYN|ACK
Sep 12 00:58:23 <Writchie> mwester: gsm 7.10
Sep 12 00:58:34 <Ghiottone> rwhitby, ACK :)
Sep 12 00:58:53 <Ghiottone> rwhitby, I'd like to ask you a question about MokoMakefile
Sep 12 00:59:06 <Ghiottone> rwhitby, is it you who did it, right?
Sep 12 00:59:45 <Ghiottone> rwhitby, well, more than a question it is a problem
Sep 12 00:59:57 <frma71> cesarb: Agree, as long as i can use pppd on one of the cannels.
Sep 12 01:00:08 <cesarb> Writchie: doesn't the whole thing have its own PM chip?
Sep 12 01:00:22 <rwhitby> Ghiottone: yes
Sep 12 01:00:44 <Writchie> cesarb: AFAIK the calypso handled its own PM. 
Sep 12 01:00:54 <Writchie> but you tell it to enter low power state. 
Sep 12 01:00:57 <Ghiottone> rwhitby, the openmoko tree is being checked out from the root of svn projects
Sep 12 01:01:07 <Writchie> it still needs to monitor network
Sep 12 01:01:17 <Ghiottone> rwhitby, and that checks out the whole trunk plus any other branch created
Sep 12 01:01:34 <Writchie> power consumption isn't too bad in this mode, 30ma i think
Sep 12 01:01:45 <frma71> Anyone tried to run some java (jme) on the neo ?
Sep 12 01:02:02 <Ghiottone> rwhitby, if it is kept this way it will grow exponentially as people start creating branches
Sep 12 01:02:12 <SpeedEvil> Writchie: more like 15mA - with the amp off
Sep 12 01:02:24 <Writchie> SpeedEvil: even better
Sep 12 01:02:39 <Writchie> about 40mw
Sep 12 01:02:44 <Ghiottone> rwhitby, was this intended?
Sep 12 01:03:41 <Writchie> cesarb: when you are going idle you put the mux into low power mode - if activity comes in it sends flags until you restablish link - all channelization remains intact
Sep 12 01:04:06 <hhf423> _
Sep 12 01:04:55 <Writchie> cesarb: when it wakes up IIRC it will singnal an RTS which is used to wakeup the SoC from deep sleep.
Sep 12 01:04:57 <rwhitby> Ghiottone: that's how the build process was specified by the OM core team at the start.  MokoMakefile simply automated that without changing any structure or semantics.
Sep 12 01:05:06 <Writchie> cesarb: which AFAIK is not implemented yet
Sep 12 01:05:10 <rwhitby> There is an intention to reduce it to just trunk/src in the future.
Sep 12 01:06:10 <mwester> Is the RTS line at all attached in such a fashion that it can generate the interrupt necessary to wake the CPU?
Sep 12 01:06:31 <Writchie> mwester: it is at least in GTA02 and I think also GTA01
Sep 12 01:06:35 <Ghiottone> rwhitby, ok. So the things outside trunk are needed by OM packages?
Sep 12 01:06:47 <rwhitby> no
Sep 12 01:07:01 <Writchie> unfortunately, the same is not true for Bluetooth
Sep 12 01:07:09 <Writchie> which moves backwards in GTA02
Sep 12 01:07:52 <Ghiottone> I changed it to checkout only trunk/ and built some packages. I had no problem.
Sep 12 01:08:11 <rwhitby> right
Sep 12 01:08:18 <Ghiottone> I think I'll leave it this way for a while.
Sep 12 01:08:36 <rwhitby> the trick is how to do it so that it happens automatically for people who already have a checkout.
Sep 12 01:08:45 <Ghiottone> I save around 50MB of disk space :)
Sep 12 01:08:52 <Ghiottone> thats easy
Sep 12 01:08:58 <Ghiottone> svn info openmoko
Sep 12 01:09:23 <Ghiottone> will print http://svn.openmoko.org/ for people who did the whole checkout
Sep 12 01:09:37 <Ghiottone> and http://svn.openmoko.org/trunk for the others.
Sep 12 01:10:16 <Ghiottone> if it is the first case, doing a cd openmoko ; svn switch http://svn.openmoko.org/trunk will do the change
Sep 12 01:10:37 <Ghiottone> (I am the subversion man at work ;)
Sep 12 01:11:35 <Ghiottone> I can create a patch for the makefile if you want
Sep 12 01:11:48 <rwhitby> please do - put it in the setup target
Sep 12 01:12:16 <tuukkah> wow feedreader works
Sep 12 01:12:19 <cesarb> Ghiottone: be careful, there's a symlink at build/conf IIRC, and the qemu stuff also symlinks to the openmoko tree
Sep 12 01:12:27 <Ghiottone> setup-openmoko ?
Sep 12 01:12:31 <cesarb> Ghiottone: (for flash.sh IIRC)
Sep 12 01:13:32 <Ghiottone> cesarb, oh right! I'll look there carefully
Sep 12 01:13:42 <Ghiottone> cesarb, thanks
Sep 12 01:14:57 <cesarb> Ghiottone: in fact, the whole qemu build probably depend on the tree (since an update to the openmoko tree, without any clobber-qemu, will recompile just what's changed)
Sep 12 01:15:06 <cesarb> s/depend/depends/
Sep 12 01:15:06 <apt> cesarb meant: Ghiottone: in fact, the whole qemu build probably depends on the tree (since an update to the openmoko tree, without any clobber-qemu, will recompile just what's changed)
Sep 12 01:16:41 <Ghiottone> mmm I see. maybe it's just a matter of changing the symlinks
Sep 12 01:16:53 <Ghiottone> or better, of finding them all
Sep 12 01:17:02 <hhf423> oh, gsmd is, like, less than 11 pages?
Sep 12 01:18:03 <Ghiottone> hhf423, documentation?
Sep 12 01:18:23 <Ghiottone> rwhitby, I'll give it a try tomorrow and let you know if I have success
Sep 12 01:18:43 <hhf423> Ghiottone: code of gsmd.c
Sep 12 01:19:01 <Ghiottone> rwhitby, about the packages you asked me to include in OE (it was you, right?)
Sep 12 01:19:18 <Ghiottone> hhf423, oh. not that much
Sep 12 01:19:35 <rwhitby> Ghiottone: yep
Sep 12 01:19:42 <hhf423> Ghiottone: no, I somehow got the idea that it was a monster
Sep 12 01:19:45 * Maledictus is hoping that there will be some blueprints of the gta02 board in the future (when the companies who want to copy the design already done it ;) )
Sep 12 01:20:16 <cesarb> Ghiottone: yeah, it's probably just changing symlinks (you'd just have to find them all)
Sep 12 01:20:23 * tuukkah is reading slashdot on the neo with GdkLauncher of webkit-gdklauncher =)
Sep 12 01:20:32 <cesarb> Ghiottone: if "make all" succeeds from an empty tree, you probably got all the symlinks which matter
Sep 12 01:20:37 <hhf423> tuukkah: why not use midori?
Sep 12 01:21:01 <Ghiottone> rwhitby, I filed a bug for ipkgorphan, it's #2954
Sep 12 01:21:59 <Ghiottone> cesarb, I have not got a computer powerful enough to try it. A from scratch build requires around a week on my pc
Sep 12 01:22:12 <tuukkah> hhf423, what's that?
Sep 12 01:22:53 <Ghiottone> rwhitby, and I'm working on fixing last issues with recipes for SettingsGUI and SMSTool (new entry)
Sep 12 01:23:26 <bedboi> bye
Sep 12 01:23:37 <hhf423> cesarb: Writchie: I mean, have a look at gsmd.c, thats chickenshit :-) reimplementation in perl to fix it should not be a big issue
Sep 12 01:23:38 <rwhitby> Ghiottone: one thing would be to make sure that the package names fit in with the rest of OE/OM naming conventions
Sep 12 01:23:54 <hhf423> tuukkah: a browser, based on webkit
Sep 12 01:24:13 <hhf423> tuukkah: you can ipkg install it from teh ScaredyCat repository
Sep 12 01:24:18 <Writchie> hhf423: present libgsmd/gsmd is only a fraction of what is needed
Sep 12 01:24:31 <Ghiottone> rwhitby, how?
Sep 12 01:24:34 <hhf423> Writchie: why is that?
Sep 12 01:24:53 <tuukkah> hhf423, it compiled already :-)
Sep 12 01:25:00 <cesarb> Ghiottone: you can get someone else to try your patch ;-)
Sep 12 01:25:07 <hhf423> Writchie: for what is needed in total or for what is needed to just make calls reliably?
Sep 12 01:25:14 * cesarb takes only a day or two for a full build without the feeds
Sep 12 01:25:15 <Writchie> hhf423: in total
Sep 12 01:25:23 <Writchie> making a call is trivial
Sep 12 01:25:53 <cesarb> Writchie: so trivial it works on first try when turning on the phone?
Sep 12 01:25:55 <hhf423> Writchie: well, you know: any big working system has once started as a small working system.
Sep 12 01:25:58 <rwhitby> Ghiottone: well, the question is whether settingsgui is specific to openmoko (and therefore one could make a case that it should have an openmoko- prefix), or it is a generic settings gui which will handle all sorts of embedded devices ...
Sep 12 01:26:07 <Ghiottone> cesarb, yeah:)
Sep 12 01:26:29 <hhf423> Writchie: so lets have the ability to make and receive calls reliably and then add to it
Sep 12 01:26:31 <Writchie> hhf423: absolutely untrue
Sep 12 01:26:46 <rwhitby> I've already posted a recipe for reliable calls.
Sep 12 01:27:02 <Writchie> you need a foundation and architecture that lets you build the whole building. 
Sep 12 01:27:08 <Writchie> then you build the first floor
Sep 12 01:27:15 <Writchie> other floors can take much longer.
Sep 12 01:27:21 <hhf423> Writchie: well, thats a law of systemantics
Sep 12 01:27:27 <anrp> then, if you're a software engineer, you skip right to the rooftop pool!
Sep 12 01:27:34 <anrp> then you grudgingl fill in the middle...
Sep 12 01:27:36 <Ghiottone> rwhitby, got the point. I think I saw it running on other OM devices... but I may be wrong
Sep 12 01:27:43 <Writchie> anrp: and then start over again
Sep 12 01:27:59 <anrp> true
Sep 12 01:28:14 <rwhitby> Ghiottone: but since you are the author, you have full rights to choose any name you want.  Just giving you my opinion from an OpenEmbedded packager's point of view.
Sep 12 01:28:30 <Writchie> unfortunately, writing specs and architecture is not as much fun as programming
Sep 12 01:30:04 <hhf423> Writchie: where would I find existing architecture documents?
Sep 12 01:30:21 <Writchie> hhf423: there are none
Sep 12 01:30:26 <Writchie> you have to read the code
Sep 12 01:30:27 <Ghiottone> rwhitby, I am not the author of SettingsGUI. I'm already writing him an email asking if he wants a name change
Sep 12 01:30:38 <anrp> hehe
Sep 12 01:30:43 <hhf423> Writchie: ah, right. of course.
Sep 12 01:31:09 <hhf423> my bad
Sep 12 01:31:16 <Writchie> as you can see there is not much really there and its add hoc. 
Sep 12 01:31:56 <rwhitby> Ghiottone: make sure you ask in the same way as I have done - it is a humble suggestion, not a demand.
Sep 12 01:32:47 <Writchie> hhf423: i can't make gsmd work because i think its approach is flawed - i'd rather spend time defining what I think is a better way and implemented the first part of it. 
Sep 12 01:32:56 <Ghiottone> rwhitby, could you tell me if there is a problem with bug 2954 or if it is ok?
Sep 12 01:33:16 <Ghiottone> rwhitby, sure. I'm attaching the logs :) he also is in this channel
Sep 12 01:33:52 <anrp> how is it a flawed approach?
Sep 12 01:33:54 <anrp> (curious)
Sep 12 01:33:58 <hhf423> Writchie: hmm, can you elaborate on that? whats the approach and why is it flawed? I have problems to derive an approach from the source
Sep 12 01:34:45 <roh> hm.. i fried my bike
Sep 12 01:35:03 <SpeedEvil> Oh no!
Sep 12 01:35:04 <roh> will need to buy some IRFB4710 tomorrow
Sep 12 01:35:45 <Maledictus> whats that?
Sep 12 01:35:50 <anrp> mosfet ftw
Sep 12 01:36:53 <roh> my bike has a brushless hub motor which needs 3-phase power. today is had water in the somehow not watertight case and that was it.. 'the magic smoke has left the device'
Sep 12 01:37:01 <anrp> oh no
Sep 12 01:37:05 <anrp> not the magic smoke D:
Sep 12 01:37:19 <anrp> i let that out of a wire at my desk by mistake
Sep 12 01:37:30 <anrp> 14awg does not handle 300 amps for very long ~_~
Sep 12 01:37:33 <roh> but hey.. i can go out and shop the parts for a few euros.. no real problem for a soldering-iron-terrorist
Sep 12 01:38:16 <hhf423> roh: I bet you learned that soldering in a terrorist camp!
Sep 12 01:38:48 <SpeedEvil> Only environmental terrorists use solder containing lead.
Sep 12 01:38:49 <Writchie> hhf423: present approach uses a library that talks over unix sockets to gsmd on one side, which will need a d-bus layer above duplicating the same thing for two round trips and then worries on the other side about high cost of a user-space mux versus kernel mux for 115Kbps
Sep 12 01:38:54 <hhf423> then again, visiting terrorist camps is not yet illegal in .de
Sep 12 01:39:25 <Ghiottone> ok really late here. nite all
Sep 12 01:39:37 <Writchie> hhf423: present approach ignores existing fax, ppp, and other stuff as well as bluetooth FAX, DUN, etc. 
Sep 12 01:39:39 <Maledictus> gn8 :9
Sep 12 01:39:40 <hhf423> Writchie: ah, so there comes in your comment about doing dbus right to it 
Sep 12 01:39:42 <hhf423> ?
Sep 12 01:40:05 <tuukkah> hhf423, midori's cool but how do i get to enter a uri?
Sep 12 01:40:06 <hhf423> Ghiottone: good night
Sep 12 01:40:08 * anrp strongly agree with second point
Sep 12 01:40:50 <hhf423> tuukkah: no idea, so far I always went via google, which it loads on start
Sep 12 01:43:08 <tuukkah> ok
Sep 12 01:43:18 <Writchie> hhf423: i think there are two camps - one says neo is a portable desktop that talks ip over wifi or the mythical mobile broadband - the other camp believes that neo could be the best implementation of a handset
Sep 12 01:44:00 <Writchie> hhf423: there is plenty of room for both camps
Sep 12 01:45:30 <hhf423> Writchie: doing dbus to gsm would a) eliminate some possible breakpoints and b) allow other apps to use the data announced by the gsm radio, so I think thats a good idea
Sep 12 01:45:42 <Writchie> hhf423: i would like to see a stable high level i/f that MMI apps can code to for at least the simple functions so that we don't create un-needed legacies. 
Sep 12 01:46:06 <Writchie> having every app talking to the modem is chaos
Sep 12 01:46:31 <Writchie> as is "watcher" apps to restart it after it crashes and other such nonsense
Sep 12 01:47:08 <hhf423> Writchie: err, I more though about apps listening to what gsm says, like location, cell-id, operator etc. not talking to
Sep 12 01:47:26 <Writchie> hhf423: i wasn't referring to your comment
Sep 12 01:47:38 <hhf423> ah, ok
Sep 12 01:47:43 <Writchie> yes, many listeners can note the events
Sep 12 01:47:48 <Writchie> and changes of state
Sep 12 01:47:55 <hhf423> exactly
Sep 12 01:48:11 <Writchie> and with d-bus they can do it with any bindings, glib, python, whatever
Sep 12 01:48:32 <Writchie> even shell scipts
Sep 12 01:48:41 <Writchie> and the d-bus part is already there
Sep 12 01:49:22 <Writchie> unfortunately, glib free d-bus is a bit of a bear and i'm trying to do it without imposing a requirement for glib 
Sep 12 01:49:49 <roh> hhf423 i think nobody will hinder you porting gsmd and libgsmd to dbus, just for trying it ou
Sep 12 01:49:49 <roh> t
Sep 12 01:50:13 <roh> basically thats what i would want to see... the best way of having a point is having a working prototype
Sep 12 01:50:42 <Writchie> roh: working prototype plus some architectural and api docs
Sep 12 01:50:51 <hhf423> roh: what hinders me is not nobody but my coding abilities
Sep 12 01:50:56 <frma71> True, "Talk is cheap, show me the code" (Linus T)
Sep 12 01:51:10 <frma71> :)
Sep 12 01:51:23 <anrp> linus
Sep 12 01:51:24 <anrp> who's that ?
Sep 12 01:51:27 * anrp >.>
Sep 12 01:51:51 <Maledictus> some guy who once got bitten by a penguin
Sep 12 01:51:57 <Writchie> linux had the benefit of existing API docs
Sep 12 01:52:02 <hhf423> roh: and I hope that this is not the OpenMoko position: "Reliable calls? -- thats a user option: if the user wants it, the user writes it."
Sep 12 01:52:02 <anrp> must've been painful
Sep 12 01:52:14 <roh> hhf423 no.. of course not.
Sep 12 01:52:40 <mwester> Roh: can you tap someone on the shoulder to review the latest workaround for bug 788, and commit the "stty -crtscts" workaround to the /etc/init.d/gsmd changes if they agree that is a reasonable workaround for the short-term?  [http://bugzilla.openmoko.org/cgi-bin/bugzilla/show_bug.cgi?id=788]
Sep 12 01:52:50 <roh> hhf423 its just that at the moment this is not a phone. its a embedded development kit in progress. we will complete it. just takes time
Sep 12 01:53:20 <hhf423> roh: I can see no progress on gsmd
Sep 12 01:53:25 <Writchie> i've had reliable calls and reliable operations for days on om2007.1 from command line. 
Sep 12 01:53:48 <hhf423> roh: I can see no plan to complete gsmd
Sep 12 01:54:26 <anrp> no written plan. you mean?
Sep 12 01:54:30 <anrp> or no intention?
Sep 12 01:54:33 <mwester> Then we should storm the citadel, and write it ourselves.  We just need a few docs, I think...
Sep 12 01:54:40 <Writchie> hhf423: people will keep hacking gsmd and eventually it will work. 
Sep 12 01:54:40 <roh> hhf423 you can see no progress because the company is still much smaller than you obviously think. people need to serialize tasks
Sep 12 01:54:42 <frma71> What worries me is the frequency of the commits and the fact that there are no commits during weekends :I
Sep 12 01:54:52 <anrp> worries?
Sep 12 01:54:55 <anrp> ...why?
Sep 12 01:55:05 <anrp> maybe they want time off
Sep 12 01:55:06 <Writchie> hhf423: or it will be replaced by a differnt approach 
Sep 12 01:55:13 <mwester> frma71:  That's the time I've had the most luck getting a build done!
Sep 12 01:55:14 <anrp> i'm certainly not going to hold them to working weekends
Sep 12 01:55:23 <hhf423> roh: less than 30 people I hear, no?
Sep 12 01:55:27 <roh> hhf423 and at the moment it is more important to complete the still in progress shipment of gta01 and complete work on the gta02 hardware to get it into mp
Sep 12 01:55:34 <roh> hhf423 ack.
Sep 12 01:55:43 <rwhitby> yeah, sunday nights are the time when I get calls working again ;-)
Sep 12 01:55:44 <Writchie> roh: I agree
Sep 12 01:55:59 <frma71> anrp: Thats the problems, they seem to have a life :I
Sep 12 01:56:04 <Writchie> i'd like to see less commits 
Sep 12 01:56:09 * anrp fails to see the problem
Sep 12 01:57:08 <Writchie> at this point i'd rather see harald insuring that the hardware, kernel drivers, u-boot, etc. are right. 
Sep 12 01:57:21 <Writchie> which is what he is doing. 
Sep 12 01:58:52 <Writchie> the only reason the phone part is working today is that the P1 hardware shipped months late.
Sep 12 01:58:58 <Writchie> s/is/is not/
Sep 12 01:58:59 <apt> Writchie meant: the only reason the phone part is not working today is that the P1 hardware shipped months late.
Sep 12 01:59:28 <hhf423> I think we need someone who stand up and makes comitments on milestones. like "reliable calls end of october". Even relative milestones would be ok, like "reliable calls four weeks after shipping batch 3 of gta01"
Sep 12 02:00:18 <hhf423> because if you do not create milestones, you lose sight. you do not make any sacrifices at all and then go nowhere
Sep 12 02:00:55 <Writchie> hhf423: well you will surely be able to make reliable calls and sms by end of October through the d-bus 
Sep 12 02:01:01 <hhf423> at the moment, we do not even know if "reliable phone calls" is in the top ten list of priorities.
Sep 12 02:01:18 <Writchie> hhf423: but it may not use the present libgsm/gsmd
Sep 12 02:02:20 <Writchie> hhf423: there will be another few months required of data collection from early adopters to refine things for all networks.
Sep 12 02:02:59 <Writchie> hhf423: openmoko is not doing too bad as mobile development goes. 
Sep 12 02:03:11 <hhf423> Writchie: months? I think if we tell people what data to collect, we'll have that in days :-)
Sep 12 02:03:52 <Writchie> it typically takes a couple of months to get it and digest it. 
Sep 12 02:03:55 <mwester> There's the catch: we won't know what data to tell people to collect until we begin collecting it, and discovering anomolies.
Sep 12 02:04:29 <Writchie> mwester: much more needs to working. 
Sep 12 02:04:49 <Writchie> SMS works across many different languages for example.
Sep 12 02:05:15 <Writchie> chinese SMS, 13 gsm languages, arabic, etc. 
Sep 12 02:05:39 <mwester> Yep.  So even a feature-limited initial version of something will be good -- it will let us gather meaningful datapoints on what works and doesn't sooner.
Sep 12 02:06:10 <Writchie> mwester: absolutely, as long as the architecture is sound. 
Sep 12 02:06:38 <mwester> So from that point-of-view, the current gsmd is a good thing -- we've learned lots of things that *don't* work with it.  Perhaps it *is* time to re-implement it.
Sep 12 02:06:38 <Writchie> actually, its either to get gsmd working right with the mux implemented because it simplifies many interactions. 
Sep 12 02:07:06 <Writchie> mwester: i think of it as a concept demo. 
Sep 12 02:07:16 <Writchie> i learned much more from talking to the modem directly. 
Sep 12 02:08:46 <Writchie> for any normally person, digesting gsm takes many months of study. 
Sep 12 02:08:50 * hhf423 thinks that it has been a long time since he talked to his U.S.Robotics Courier HST
Sep 12 02:09:54 <Writchie> i used to program a hayes 300 even before the AT command set was invented
Sep 12 02:10:02 <Writchie> from an S100 bus
Sep 12 02:10:07 <mwester> It looks like the Hayes command set.  Can't be that hard!  ;)
Sep 12 02:10:10 <Writchie> with 64K RAM in the system
Sep 12 02:10:38 <hhf423> Hayes did invent the AT command set, did they not?
Sep 12 02:10:54 <Writchie> not dennis but an engineer that worked for him
Sep 12 02:11:07 * hhf423 proclaims: "AT&T is a modem test command!"
Sep 12 02:11:26 <Writchie> it means eat my money
Sep 12 02:11:28 <hhf423> Writchie: right, I meant hayes as in company, not person
Sep 12 02:11:35 <mwester> It wasn't even a very good way to control a modem, but once again marketing beat technology.
Sep 12 02:12:00 <hhf423> +++
Sep 12 02:12:01 <hhf423> :-)
Sep 12 02:12:08 <Writchie> mwester: it was better than the autodial at the time which used an oddball interface
Sep 12 02:12:34 <baird> Writchie: did Hayes modems to run off a FORTH firmware, or was that just an Australian thing for the compatible modems over here..?
Sep 12 02:13:04 <orpheus> random question, a bit more grounded:
Sep 12 02:13:14 <Writchie> FORTH was used on some S100 systems at the time. 
Sep 12 02:13:18 <Writchie> talk about code
Sep 12 02:13:24 <hhf423> Writchie: what was good was the (more or less) standardization on AT command set
Sep 12 02:13:24 <orpheus> has anyone else had a problem where, if they ssh into the neo while it's booting, it locks up?
Sep 12 02:13:41 <orpheus> i've done it thrice now
Sep 12 02:13:56 <orpheus> no sshing, no lockup; sshing, lockup
Sep 12 02:14:16 <Writchie> orpheus: what exactly do you mean by sshing. 
Sep 12 02:14:17 <orpheus> and it sounds pitiful when it locks up while playing that intro sound
Sep 12 02:14:38 <Writchie> orpheus: leaving the usb attached?
Sep 12 02:14:41 <orpheus> well, twice i was ssh'ed in, once i was scping a file over
Sep 12 02:14:48 <orpheus> yeah, over usb
Sep 12 02:15:06 <orpheus> am i not supposed to let it sit on usb while booting?
Sep 12 02:15:34 <orpheus> i suspect it's just that i was trafficing over the usbnet stuffs, and that did it
Sep 12 02:15:58 <Writchie> orpheus: I think the kernel re-inits the USB
Sep 12 02:16:27 <Writchie> u-boot looses control for good when it turns over to the kernel
Sep 12 02:16:28 * cesarb wonders if he's the only one who doesn't get a startup sound anymore, or if it's just that only people using ipkg upgrade who get it
Sep 12 02:16:37 <orpheus> yeah
Sep 12 02:16:48 <orpheus> i'm actually getting ssh'ed in before X starts us
Sep 12 02:16:50 <orpheus> *up
Sep 12 02:16:56 <technil> orpheus: i usually get no response to a ssh request until the unit is fully booted
Sep 12 02:17:16 <Writchie> orpheus: it should work but there may be some race condition
Sep 12 02:17:25 <orpheus> and then, as X comes up, it'll lock up (while playing that happy little sound clip), and make some awful racket
Sep 12 02:17:56 <technil> cesarb: i also lost startup sound and i am having alsamixer trouble after ipkg update && ipkg upgrade
Sep 12 02:17:57 <jonnymac> wow
Sep 12 02:18:02 <jonnymac> I'm a linux baby
Sep 12 02:18:16 <orpheus> it just struck me as odd that it happened only while it was booting, but that it happened long after the boot got handed over to to userspace
Sep 12 02:18:48 <orpheus> technil: i've got sound, it's just that the kernel locks up mid-sound
Sep 12 02:19:16 <orpheus> alright, just making sure it wasn't some common thing
Sep 12 02:19:24 <Writchie> the gadget driver is in the kernel and prolly starts long before init has finished starting xinetd or whatever responds to ssh
Sep 12 02:19:35 <orpheus> yeah
Sep 12 02:20:13 <orpheus>  /usr/sbin/dropbear
Sep 12 02:20:21 <orpheus> but yeah
Sep 12 02:20:44 <orpheus> it's too weird of a problem to worry about too much right now
Sep 12 02:21:17 <hhf423_> ok, back to bed, tty all later, good night
Sep 12 02:21:22 <orpheus> but i was half expecting "oh, that... yeah, don't traffic on the usb too much while booting"
Sep 12 02:21:38 <spikebike> speaking of which
Sep 12 02:21:43 <spikebike> oops
Sep 12 02:22:08 <orpheus> that was an odd time to say oops
Sep 12 02:26:51 <orpheus> i have two dfu-capable devices on my system. I'm always tempted to flash the other one with an linux kernel, just to see what will happen
Sep 12 02:27:58 <Writchie> orpheus: what is the other device
Sep 12 02:28:08 <orpheus> i don't know
Sep 12 02:28:25 <Writchie> some BT adapters use DFU
Sep 12 02:28:57 <orpheus> or whatever responds to ssh
Sep 12 02:28:57 <orpheus> <orpheus> yeah
Sep 12 02:29:01 <orpheus> errrrr
Sep 12 02:29:11 <orpheus> Found Runtime: [0x044e:0x300c] devnum=5, cfg=0, intf=2, alt=0, name="UNDEFINED"
Sep 12 02:29:37 <orpheus> what the....?
Sep 12 02:29:41 <orpheus> it's my touchpad
Sep 12 02:29:43 <orpheus> :/
Sep 12 02:30:04 <Writchie> gateway?
Sep 12 02:30:16 <orpheus> what can a dfu runtime device do, anyways?
Sep 12 02:30:50 <Writchie> accept downloading through the alternate interface
Sep 12 02:31:26 <Writchie> some RAM based USB controllers use it for loading their initial program IIRC
Sep 12 02:31:38 <orpheus> odd
Sep 12 02:32:11 <orpheus> it's probably how you set all of the wierd options for it (there are dozens of paramaters you can poke at on some touchpads)
Sep 12 02:32:51 <orpheus> any chance I could read from it?
Sep 12 02:33:06 <orpheus> ooooh
Sep 12 02:33:07 <orpheus> coo.
Sep 12 02:33:07 <Writchie> vendor code 0x0443 is Alps Electric which is an OEM touchpad
Sep 12 02:33:13 <orpheus> yep
Sep 12 02:33:18 <orpheus> i'm familiar with it
Sep 12 02:33:27 <orpheus> in fact, you can even chain it with a track stick
Sep 12 02:33:44 <orpheus> which is odd, since i don't physically HAVE a touchpad...
Sep 12 02:33:58 <Writchie> is it a laptop
Sep 12 02:34:18 <orpheus> yeah, it becomes my desktop thanks to a dock
Sep 12 02:34:37 <orpheus> it's just funny to me that linux has to talk to a touchpad device to talk to the trackstick
Sep 12 02:35:09 <Writchie> legacy issues prolly
Sep 12 02:35:32 <Writchie> controler is prolly on the motherboard
Sep 12 02:35:57 <orpheus> i think it's actually that sony is just ... slow
Sep 12 02:36:38 <orpheus> having seen how the hardware is actually wired up, i'm not impressed
Sep 12 02:36:46 <orpheus> "Lost device after RESET?"
Sep 12 02:37:39 <orpheus> ack... i tried to read from my touchpad, and it reset and changed it's usb device id....
Sep 12 02:38:48 <cesarb> orpheus: some devices have several device ids for several different modes
Sep 12 02:39:01 <cesarb> orpheus: the neo itself has different ids for u-boot and the kernel
Sep 12 02:39:28 <orpheus> this is true
Sep 12 02:39:31 <cesarb> orpheus: perhaps it entered its firmware update mode now... with its own special id
Sep 12 02:39:38 <orpheus> it's in DFU mode now, instead of runtime mode
Sep 12 02:39:43 <orpheus> yep
Sep 12 02:40:29 <orpheus> i know that openmoko's dfu-util is somewhat modified, but should it still work for normal devices?
Sep 12 02:40:56 <Writchie> orpheus: you need the bit for bit correct firmware image. 
Sep 12 02:41:21 <Writchie> some low-end high volume usb controllers are ram based and may be loaded from the bios 
Sep 12 02:41:40 <orpheus> i'm trying to upload (device->file) the firmware/configuration/whatever it is
Sep 12 02:42:27 <orpheus> i'm hoping it's something cool like the configuration, since i've been wishing i could turn off the tap-to-click feature since i got it
Sep 12 02:44:10 <orpheus> this is to say  that i'm waiting for my neo to flash
Sep 12 02:45:32 <cesarb> orpheus: openmoko's dfu-util follow the standard too closely, and thus will not work with most devices
Sep 12 02:45:48 <cesarb> orpheus: most devices have their own take on the standards
Sep 12 02:47:15 <orpheus> ah
Sep 12 02:47:17 <orpheus> oh well
Sep 12 02:50:53 <roh> cesarb i think there were already some patches coming in to support different devices
Sep 12 02:51:57 <cesarb> roh: yeah, I've seen at least one of these
Sep 12 02:52:11 <cesarb> (patches, not devices... I didn't test dfu-util on my bluetooth dongle)
Sep 12 02:52:17 <orpheus> ah
Sep 12 02:52:31 <orpheus> anyways, i'm giving up for the night
Sep 12 02:52:49 <orpheus> thanks for the interesting tidbits about dfu
Sep 12 02:54:41 <roh> cesarb svn revisions 2765 2766 2767
Sep 12 02:56:45 <Moriarty_> ooh, new snapshot out.
Sep 12 02:56:49 * Moriarty_ snags it for the emulator.
Sep 12 02:59:14 <Moriarty_> does the 11 Sep image have the new openmoko-dates in it?
**** ENDING LOGGING AT Wed Sep 12 02:59:56 2007