**** BEGIN LOGGING AT Thu Apr 21 02:59:58 2011 Apr 21 07:09:31 JaMa|Zzz: great, thanks... trying now Apr 21 07:21:54 mrmoku: yw, can you please try if /OE/tmpdir-shr/sysroots/x86_64-linux/usr/bin/valac --version , returns 0.12.0-dirty for you? Apr 21 07:22:06 mrmoku: here it doesn't but on buildhost it does Apr 21 07:22:22 mrmoku: and waf/vala.py is failing for midori because of that Apr 21 07:37:54 JaMa|W: 0022 mok@ladyt[pts/1]:~-> shrbuild/shr-chroot/OE/shr-unstable/tmp/sysroots/x86_64-linux/usr/bin/valac --version Apr 21 07:37:57 Vala 0.12.0-dirty Apr 21 07:42:38 ok, thanks Apr 21 07:44:33 anybody can explain me how elfe works? Apr 21 07:51:33 JaMa|W: hmm... ppp problem is still there though Apr 21 07:53:31 :/ Apr 21 07:56:03 mrmoku: btw while rebuilding from scratch libfreesmartphone-glib-2011.01.13.1 is now failing, because of this: Apr 21 07:56:06 #define FREESMARTPHONE_GLIB_ERROR g_quark_from_static_string("org.freesmartphone.Application") Apr 21 07:56:09 #define FREESMARTPHONE_GLIB_ERROR g_quark_from_static_string("org.freesmartphone.Audio") Apr 21 07:56:12 #define FREESMARTPHONE_GLIB_ERROR g_quark_from_static_string("org.freesmartphone") Apr 21 07:56:31 is it libfreesmartphone-glib fault or is something wrong in specs xml files (I don't see anything wrong there Apr 21 07:57:10 hmm Apr 21 07:57:30 except that org.freesmartphone.Application org.freesmartphone.Audio are last 2 additions.. Apr 21 07:57:40 and daniele was not seen for some time... Apr 21 07:58:47 and here is grep in xmldir http://paste.pocoo.org/show/375648/ Apr 21 07:59:46 I think we can live without it.. I was more concerned about possible specs issue Apr 21 08:00:54 no, I don't think it's spec issues Apr 21 08:00:54 hehe italian comments in tools/dbus-rapid-bindings.py are not really helpfull to me ;) Apr 21 08:01:00 ok, thanks Apr 21 08:01:34 and the italian comments are less disturbing for me than the fact it's python code ;) Apr 21 08:02:55 I think I've found it.. Apr 21 08:12:20 freesmartphone.org: 03Martin.Jansa 07libfreesmartphone-glib * re2ba3e01a1b1 10/ (3 files in 2 dirs): Apr 21 08:12:21 freesmartphone.org: specs bump, add cfg file for Application and Audio manager Apr 21 08:12:21 freesmartphone.org: * otherwise FREESMARTPHONE_GLIB_DBUS_ERROR is redefined 3 times, because we have newer fso-specs Apr 21 08:12:21 freesmartphone.org: Signed-off-by: Martin Jansa Apr 21 08:18:01 freesmartphone.org: 03mok 07cornucopia * r92605a31957f 10/fsogsmd/src/plugins/modem_nokia_isi/plugin.vala: Apr 21 08:18:01 freesmartphone.org: modem_nokia_isi: add missing call to gpio_prope Apr 21 08:18:01 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 08:22:41 heh shr buildhost load > 7 and git daemon refuses my git push :/ Apr 21 08:48:19 SHR: 03Martin.Jansa 07shr-chroot * r96287f0e23ae 10/etc/passwd: passwd: add uucp user to fix uucp build (calling chown uucp) Apr 21 08:58:26 mickeyl: now I could need some help :) Apr 21 08:58:45 mickeyl: where is the correct place to hook in the poweron/poweroff inside the modem class? Apr 21 08:59:26 mickeyl: constructor/destructor? or something else? Apr 21 09:04:31 pespin: hey, ping :) Apr 21 09:05:57 mrmoku, pong! Apr 21 09:06:22 pespin: the bumping did not fix PDP for me Apr 21 09:06:42 Morning all. Apr 21 09:07:08 mrmoku, it's not suppoused to fix the shr-settings, but I think there are changed which I need to implement it well in shr-settings :) Apr 21 09:07:16 I have only 1 error building an SHR-U image: http://pastebin.com/ar3StmqL What can I do to fix the compile error? Apr 21 09:07:21 s/changed/changes/ Apr 21 09:07:21 pespin meant: mrmoku, it's not suppoused to fix the shr-settings, but I think there are changes which I need to implement it well in shr-settings :) Apr 21 09:07:26 pespin: settings is not the problem Apr 21 09:07:37 problem is that one can't activate it Apr 21 09:07:48 mrmoku, using mdbus? Apr 21 09:07:54 using quick-settings Apr 21 09:08:23 gives an 'ppp immediately exited' error or something like that Apr 21 09:09:01 If it matters, I'm using the 32bit version of the chrooted env. Apr 21 09:09:40 ChristW: read wiki Building SHR Apr 21 09:10:23 ChristW: same as ffmpeg here http://wiki.shr-project.org/trac/wiki/Building%20SHR#ffmpeg Apr 21 09:10:29 freesmartphone.org: 03mok 07cornucopia * r3900855c0b17 10/fsogsmd/src/plugins/modem_nokia_isi/plugin.vala: Apr 21 09:10:29 freesmartphone.org: modem_nokia_isi: add a destructor to poweroff the modem Apr 21 09:10:29 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 09:10:32 freesmartphone.org: 03mok 07cornucopia * r3a9e59460cc3 10/fsogsmd/src/plugins/modem_nokia_isi/plugin.vala: Apr 21 09:10:32 freesmartphone.org: modem_nokia_isi: only reset cmt_rst_rq once per poweron on netlink changes Apr 21 09:10:32 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 09:10:41 mrmoku, hmm no idea. I can't look at it right now. Is new fso already in feeds? Apr 21 09:10:49 yup Apr 21 09:11:01 nvm then... thought you might have some insight :) Apr 21 09:15:35 * JaMa|W checkouting 32b chroot to upgrade it Apr 21 09:18:00 Can I edit de Wiki? I want to add a comment that ffmpeg _and_ libav have the same problem/solution. Apr 21 09:18:57 yes you can Apr 21 09:19:36 Ah, the 'edit page' at the bottom... Apr 21 09:19:49 I'm used to having an 'Edit' link per paragraph... Apr 21 09:22:09 Ok, updated. Apr 21 09:24:14 package libav-2_0.6.2+r0.2+gitr0+c6c2dfcf15c1d93b2189adff6f71c5c4b6b05338-r0.2: task do_compile: Succeeded Apr 21 09:29:06 ChristW: thanks Apr 21 09:33:41 Good morning gents! :) Looking for HW-aware hackers here. Apr 21 09:34:31 To anyone who has had any fun with i2c communication and the Freerunner or Neo.... Apr 21 09:35:13 Why is there a mandatory 20ms waiting window whenever the SW tries to do consecutive I2C read/writes?... Apr 21 09:37:19 DocScrutinizer, you may know this? ^ Apr 21 09:37:56 pespin: actually he pointed me here to find wpwrak... :) Apr 21 09:38:12 ah ok, sorry then :) Apr 21 09:38:44 actually, it was PaulFertser, sorry. :) Apr 21 09:46:59 NOTE: package shr-image-2.0-r18: task do_rm_work_all: Succeeded Apr 21 09:47:01 NOTE: Tasks Summary: Attempted 10990 tasks of which 10949 didn't need to be rerun and 0 failed. Apr 21 09:47:03 OE om-gta02@shr ~/shr-unstable $ Apr 21 09:47:27 So, now for the next step... editting and getting the .ipk on the FR! Apr 21 09:58:56 Ok, so now I edit conf/local-builds.inc. But, it's a bit unclear... The //path/to/shr-apps is always the same, innit? Apr 21 10:10:43 sorry, no instant answer Apr 21 10:10:49 maybe after a coffee Apr 21 10:11:20 wpwrak: hi there :) Here's a guy mokomoko who faced some unneeded delays with I2C communication with his homebrew devices, are you aware of some gotchas? Apr 21 10:11:24 DocScrutinizer: hey hey :) Apr 21 10:12:04 usually you got no delays on I2C Apr 21 10:12:42 mickeyl: interesting... root@nokia900 ~ # sendisi phonet0 0x15 0x01 0x01 0x00 Apr 21 10:12:50 is causing the link to go down and come back up Apr 21 10:13:00 repeatably Apr 21 10:13:01 mrmoku, hi lol Apr 21 10:13:08 hi GNUtoo Apr 21 10:13:31 so it's MTC_STATE_REQ ... Apr 21 10:13:43 and I still think it 'might' be normal Apr 21 10:14:18 mrmoku, singe you are now able to register you can get the log of registering fsogsmd and compare Apr 21 10:14:22 *since Apr 21 10:14:23 GNUtoo: btw. I eliminated the lowlevel plugin and integrated it into modem_nokia_isi Apr 21 10:14:32 yes I saw Apr 21 10:14:38 does it work now? Apr 21 10:14:55 no :/ Apr 21 10:15:12 ok Apr 21 10:15:31 did you try to lower the possibilities in ofono? Apr 21 10:15:35 like comment stuff Apr 21 10:15:40 to see what makes it register Apr 21 10:18:37 GNUtoo: not yet... as it's a pain to do Apr 21 10:19:03 btw. now it's more difficult to test Apr 21 10:19:13 as I can't just disable the lowlevel anymore ;) Apr 21 10:19:38 maybe I should add a parameter to modem_nokia_isi Apr 21 10:19:42 config option that is Apr 21 10:19:55 something like 'power_on_modem' Apr 21 10:20:17 yes Apr 21 10:20:20 could be an idea Apr 21 10:20:28 hmm... have to get local builds working again then Apr 21 10:20:47 mkaes no sense to blindly commit just to be able to build Apr 21 10:23:30 PaulFertser: hmm, you mean in the code or in actual behaviour ? i didn't notice any major delays in the behaviour. wouldn't have noticed small unnecessary delays, though Apr 21 10:26:47 there might be I2C slaves that need a delay to initialize Apr 21 10:27:34 I'm not aware of any offhand, and also this shouldn't result in a general delay in lowlevel I2C Apr 21 10:28:46 hi wpwrak Apr 21 10:29:36 ~seen morphis Apr 21 10:29:41 morphis <~morphis@ip-109-44-179-82.web.vodafone.de> was last seen on IRC in channel #openmoko-cdevel, 1d 55m 52s ago, saying: 'antrik: and there will be a aurora-components which takes some things from meego-ux-components, nothing more :)'. Apr 21 10:32:43 mickeyl: hi, is there some AbstractModem method getting called on request/release that I could override? Apr 21 10:33:06 let me take a look Apr 21 10:34:05 JaMa|W: my memory sucks... was there some trick to make local builds work again? Apr 21 10:34:20 mrmoku: you can override Modem::poweron(), but please call the base implementation Apr 21 10:34:36 ok, and poweroff() too then? Apr 21 10:34:38 yes Apr 21 10:34:44 good, thanks Apr 21 10:36:40 actually, no need to call the base implementation Apr 21 10:36:44 it's empty Apr 21 10:36:59 but it won't hurt, in case i will change the base Apr 21 10:37:18 mokomoko: ^^^ Apr 21 10:37:19 hmm... can't find them though Apr 21 10:37:53 powerOn and powerOff, actually Apr 21 10:38:01 they're listed in modem.vala as overridable Apr 21 10:38:23 mrmoku: it worked for me just as before when I needed srctree+gitver.. Apr 21 10:38:33 mickey: yeah, my search was case sensitive :P Apr 21 10:38:38 :) Apr 21 10:39:19 mickey: AbstractModem.powerOn(); Apr 21 10:39:26 is the way to call it from the base class? Apr 21 10:39:35 no, that's pythonic :) Apr 21 10:40:05 try base.powerOn(); Apr 21 10:40:20 hehe, ok Apr 21 10:40:36 you should evaluate the return though Apr 21 10:40:42 if base.powerOn() already fails, no need to continue Apr 21 10:43:46 yup Apr 21 10:47:40 freesmartphone.org: 03mok 07cornucopia * r9365f1c63ccd 10/fsogsmd/conf/nokia_n900/fsogsmd.conf: Apr 21 10:47:40 freesmartphone.org: fsogsmd: disable the vanished lowlevel plugin for nokia900 Apr 21 10:47:40 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 10:47:41 freesmartphone.org: 03mok 07cornucopia * re76833a48c85 10/fsogsmd/src/plugins/modem_nokia_isi/plugin.vala: Apr 21 10:47:41 freesmartphone.org: fsogsmd: modem_nokia_isi: override powerOn / powerOff Apr 21 10:47:41 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 10:48:37 I've made changes (locally) to shr-apps/libframeworkd-phonegui-efl/src/view/dialer-view.c. What do I need to re-build to try the .ipk on my FR>? Apr 21 10:49:11 PaulFertser: hi, I'm back! :) Apr 21 10:49:52 mokomoko: please repeat your question with relevant details and hilight Werner :) Apr 21 10:50:02 wpwrak: my device is really simple, it doesn't (theoretically) need any delays, it's ready to answer. Apr 21 10:51:05 wpwrak: there's a 20ms delay forced between consecutive read()s and/or write()s. no idea where it comes from, since the slave does not block the bus. Apr 21 10:51:49 wpwrak: (i remember you told me how the OpenMoko project is a nice opportunity for me to get into kernel programming) Ironically, by now i've recevied 3 potential job offers as a kernel dev, but all usa-based. Apr 21 10:52:00 wpwrak: I'm using /dev/i2c-0, then ioctl() to set the slave address and it works, minus the delay. Apr 21 10:53:56 PaulFertser: (3 offers) congratulation ! now you should educate them on the benefits of teleworking (or consider the joys of seeing the world :) Apr 21 10:54:16 wpwrak: the delay is verifiable both at the SW level and at the bus lines with the scope. Apr 21 10:55:20 mokomoko: hmm, sounds a bit odd. 20 ms could even be a short sleep that gets increased to timer granularity Apr 21 10:55:45 wpwrak: i wouldn't like to work for a usa-based company anyway, yet alone moving there even temporarily is out of the question :) Apr 21 10:55:56 that's my feeling too, the granularity is towards 2 jiffies. but I've no sleeps in my code.. Apr 21 10:56:36 mokomoko: but nothing comes to mind that would cause this. i've one only very little playing with i2c (mainly for the gta03 camera), but i don't recall having encountered gratuitous large delays on simple reads/writes Apr 21 10:57:01 wpwrak: could Python be to blame? Apr 21 10:57:18 mokomoko: have you tried strace ? Apr 21 10:57:26 no, good idea. Apr 21 10:57:46 wpwrak: what are you doing lately, btw? Haven't heard from you for ages. Apr 21 10:58:00 PaulFertser: why not work for them ? they're not that barbarian, no matter what cold war propaganda has told you ;-) Apr 21 10:58:05 wpwrak: but it's going to completely alter the timings.... Apr 21 10:58:26 mokomoko: but you'll still see any attempted sleeps :) Apr 21 10:58:34 right! :) Apr 21 10:58:41 mokomoko: (that is, if they come from system calls) Apr 21 10:59:01 hmm... I'm afraid there may be mischief in the i2c driver.... Apr 21 10:59:16 wpwrak: G.W. Bush for 8 years straight was more impressive than any loosy propaganda. Apr 21 10:59:35 PaulFertser: i'm now busy at qi-hw. mainly having fun with wireless these days. critters like these: http://downloads.qi-hardware.com/people/werner/wpan/tmp/2boards-20110305.jpg Apr 21 11:00:22 wpwrak: nice tiny boards, do you diy those? Apr 21 11:00:42 PaulFertser: oh, and I got the first PCB that's all designed by myself "mass-produced": http://www.tuxbrain.com/en/oscommerce/products/1880 Apr 21 11:00:58 PaulFertser: a humble beginning for sure, but hey' it's a start ;-) Apr 21 11:01:52 SHR: 03Martin.Jansa 07shr-chroot * r19dba1d9186b 10/OE/.bashrc: synchronize configs with 64bit chroot Apr 21 11:01:52 SHR: 03Martin.Jansa 07shr-chroot * r2c42f2e78091 10/OE/.bashrc: synchronize configs with 64bit chroot Apr 21 11:01:56 SHR: 03Martin.Jansa 07shr-chroot * r3feaad3abce8 10/ (309 files in 42 dirs): synchronize configs with 64bit chroot Apr 21 11:02:29 PaulFertser: yup, i designed and made them. they're ieee 802.15.4 devices. kinda like wifi, but shorter range, less demanding, and with less IPR issues Apr 21 11:02:39 SHR: 03Martin.Jansa 07shr-chroot * r4c1de2255087 10/ (5312 files in 219 dirs): system: drop .pyo and .pyc files and add them to .gitignore Apr 21 11:03:02 ChristW: I've just pushed bigger 32b chroot upgrade, please test and report any issues if you can Apr 21 11:03:14 SHR: 03Martin.Jansa 07shr-chroot * r7d31fb684e50 10/ (9896 files in 603 dirs): system upgrade Apr 21 11:03:32 JaMa|W: Just 'make update' to update the chroot? Apr 21 11:03:41 SHR: 03Martin.Jansa 07shr-chroot * r9e25859815df 10/ (51 files in 9 dirs): bitbake upgrade Apr 21 11:04:08 wpwrak: cool interestinf stuff indeed Apr 21 11:04:57 heyho Apr 21 11:06:02 wpwrak: how on earth do you solder that QFN chips? Apr 21 11:06:25 PaulFertser: with a soldering iron, flux, ... and patience :) Apr 21 11:06:40 PaulFertser: to make things a little more interesting, i used lead-free solder Apr 21 11:06:44 wpwrak: i thought you were using some thermo-controlled roster or something. Apr 21 11:07:09 s/roster/oven/ Apr 21 11:07:10 PaulFertser meant: wpwrak: i thought you were using some thermo-controlled oven or something. Apr 21 11:07:20 ChristW: http://wiki.shr-project.org/trac/wiki/Building%20SHR#Updatingshr-chrootenvironment Apr 21 11:08:04 PaulFertser: i also have my toaster oven (although only with manual temperature control), but i tend to get better results with manual soldering Apr 21 11:10:53 wpwrak: but qfn components seem to be almost as hard as bga :) Apr 21 11:11:13 JaMa|W: How do I save my local changes before update-shr-chroot? Apr 21 11:11:22 Copy them to the Internet? :-) Apr 21 11:11:26 SHR: 03Martin.Jansa 07shr-chroot * r99b5e868cef5 10/OE/.bashrc: bashrc: sync with 32bit chroot Apr 21 11:11:51 SHR: 03Martin.Jansa 07shr-chroot * r12eba56eac3a 10/ (5328 files in 220 dirs): system: drop .pyo and .pyc files and add them to .gitignore Apr 21 11:12:37 ChristW: you can commit them and then export with git format-patch Apr 21 11:12:42 and then reapply Apr 21 11:12:59 or just commit and git pull --rebase and then check diff Apr 21 11:13:09 ie passwd/group should be different Apr 21 11:13:10 I'm not quite ready to commit them to SHR... I was just building and testing a change! Apr 21 11:13:29 if they are in /OE directory then they should be ignored Apr 21 11:13:39 except stuff like .bashrc etc Apr 21 11:13:43 Since it's only one file now, I'll just copy it somewhere else and copy it back... Apr 21 11:14:09 PaulFertser: naw. you can reach the QFN contacts from the side. what i do is this: 1) pre-tin the board. 2) apply flux, place the qfn component, hold it down with tweezers. 3) dab some solder at contacts all around. this usually be very bad (only soldering a small fraction of the pins and leaving shorts), but it will hold the component in place. 4) flux again and solder such that the pins have a bit of solder floating around them ( Apr 21 11:14:09 you don't need much). also, remove excess solder by wiping it onto the adjacent traces. Apr 21 11:14:43 PaulFertser: 5) repeat until all the pins look good. 6) measure for shorts. Apr 21 11:14:47 1970-01-01T00:05:12.732332Z [DEBUG] fsogsmd : GLib : gisicomm.vala:859: regstatus = NET_REG_STATUS_HOME Apr 21 11:14:47 1970-01-01T00:05:12.732942Z [DEBUG] fsogsmd : GLib : gisicomm.vala:860: regmode = NET_SELECT_MODE_AUTOMATIC Apr 21 11:14:52 can I assume that a register Apr 21 11:14:55 in my tests Apr 21 11:15:29 wpwrak: cool, will keep in mind this method works, thanks! :) Apr 21 11:16:13 PaulFertser: particularly with lead-free solder, you can still get some unsoldered pins, but it's usually not too bad. what sucks a little are large ground area, like in these RF boards, because they draw away the heat. so the iron has to be quite hot. (alternatively, you could heat up the whole area including the ground place, but that probably does more damage than good) Apr 21 11:17:51 wpwrak: yeah, i'm aware of that ground area on rf chips, probably a big enough (for the tip of the iron) hole in the board can help? Apr 21 11:19:49 ChristW: commited changes are still local... nobody said you have to push them :-) Apr 21 11:20:04 PaulFertser: hole ? no, you just need a bit of clearance from the ground. as long as you don't touch it directly, you're fine. alas, RF usually wants ground to be close to the signal. Apr 21 11:20:42 ah it is Apr 21 11:20:43 ok Apr 21 11:20:46 bbl Apr 21 11:20:57 PaulFertser: here you can see more details: http://downloads.qi-hardware.com/people/werner/wpan/tmp/2boards-20110305-full.jpg Apr 21 11:22:08 freesmartphone.org: 03mok 07cornucopia * r0112158843e5 10/fsogsmd/src/plugins/modem_nokia_isi/plugin.vala: Apr 21 11:22:08 freesmartphone.org: fsogsmd: modem_nokia_isi: don't release the power key twice Apr 21 11:22:08 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 11:22:09 PaulFertser: ah, and i did the bloody crystal with hot air. with an iron, i'd just short to the case (there's very little clearance between the case and the pcb) Apr 21 11:22:09 freesmartphone.org: 03mok 07cornucopia * r06c3855c111d 10/fsogsmd/src/plugins/modem_nokia_isi/plugin.vala: Apr 21 11:22:09 freesmartphone.org: fsogsmd: modem_nokia_isi: add config to make power handling of the modem optional Apr 21 11:22:09 freesmartphone.org: Signed-off-by: Klaus Kurzmann Apr 21 11:22:20 JaMa|W: Fixed the Wiki-page about updating the 32-bit environment. But: http://pastebin.com/s8VxsPUw Apr 21 11:24:57 I re-ran with 'sudo' (to see if that would fix the permissions for dev/), but it still fails for proc/ and sys/ Apr 21 11:25:39 ChristW: did you umount all dirst first, right? Apr 21 11:26:17 looks like it according to path, but if you had 2nd terminal still in chroot... Apr 21 11:27:13 Had another terminal open... good call! Apr 21 11:27:15 ChristW: and all files should be owned by your "bitbake" user, if you did first setup-shr-chroot as root then fix it by chown -R your_uid:your_git shr-chroot-32bit Apr 21 11:28:00 So, that fixed it. I originally did the setup-shr-chroot as norman user, so ther shouldn't be a problem. Apr 21 11:28:35 yup it was because you were trying to writte to binded system directories.. Apr 21 11:28:57 * JaMa|W should make first sentence in that paragraph red :) Apr 21 11:28:58 Ok, I'm now back into the chrooted environment. What should I test? Apr 21 11:29:14 just build image as usuall Apr 21 11:29:31 Ok, 'make update' is running... Apr 21 11:30:31 bb -k shr-image too. So far, so good... I do get a 'Cache version mismatch', but I guess that's expected from a big update... Apr 21 11:30:48 yes Apr 21 11:30:56 that's from newer bitbake Apr 21 11:31:28 wpwrak: impressive-looking projects indeed, i'm glad to hear you're having fun :) Apr 21 11:31:34 thanks for test, if you find any issues, ping me Apr 21 11:32:38 Mkay, I will. Apr 21 11:33:43 PaulFertser: alas, it means a lot of experimenting. here's a collection of earlier attempts: http://downloads.qi-hardware.com/people/werner/wpan/tmp/boards-so-far.jpg Apr 21 11:34:09 PaulFertser: and here some more, from antenna testing: http://downloads.qi-hardware.com/people/werner/tmp/antfarm.jpg (those little boards are PCB antennas, with different sizes) Apr 21 11:37:59 wpwrak: you do have a lot of patience indeed Apr 21 11:38:29 PaulFertser: all the beginning is hard. especially if you don't know what you're doing ;-) Apr 21 11:38:46 (this is my first RF design) Apr 21 11:42:16 wpwrak: is there any specific project you're doing it for? Apr 21 11:43:10 PaulFertser: it's for adding wireless connectivity to the ben nanonote Apr 21 11:43:49 wpwrak: i understand but why 802.15.4, what's the usecase? Apr 21 11:46:32 PaulFertser: the usecase is just connectivity. sort of a poor man's WiFi. we picked 802.15.4 because it's reasonably well-supported by industry (i.e., you have a choice of properly documented chips), it's an open standard, and it operates in a global ISM band. Apr 21 11:48:04 PaulFertser: (usecase) some people are also thinking of using it with sensors and such, i.e., the stuff it is originally designed for. Apr 21 11:48:05 wpwrak: yes, but wifi has a clear usecase: i could (if i had nanonote) use my home wlan and then go to friend's and use his/her etc. Also, usb wifi dongles are available. Apr 21 11:51:35 mrmoku, pre_sim makes it registerhttp://wiki.freesmartphone.org/index.php/Phonet_Implementation#Testing_table Apr 21 11:54:09 PaulFertser: well, we don't have USB host. and even if we did, that would be quite a wasteful way of using it. wlan would be nice, but once you get to the chip level, the situation is fairly hopeless when it comes to openness. remember the ar6000 ? ;-) there's a chip from marvell, which is better supported, but it also isn't open. and it's just a question of time until that one disappears, too. (it's quite old - and the predecessor Apr 21 11:54:09 is broken and apparently was never fixed) Apr 21 11:55:02 PaulFertser: also, with 802.15.4, we have a better chance of being able to go SDR some day. that would be very very sweet. Apr 21 11:57:43 wpwrak: ath6kl with ar6003? Apr 21 11:58:41 mickey|office: ping Apr 21 11:58:52 wpwrak: (nanonote) no usb host? Damn, that's strange, i wasn't aware of that. Even my lousy wr741nd wifi router has that (usb1.1 only), as does FR and n900 :) Apr 21 12:00:38 PaulFertser: (ar6003) don't know that one. the ar6001 experience kinda spoiled my appetite for atheros :) Apr 21 12:01:07 PaulFertser: (usb host) hmm, i'm not sure you realize how small the ben is :) Apr 21 12:01:11 PaulFertser: yes, it's a pity. even more since the SoC has ohci support Apr 21 12:01:51 wpwrak: the avt2 has host support and fits nicely into the nanonote case Apr 21 12:01:55 PaulFertser: but i hope that we will be able to somehow cram a USB host port into a successor. just need maybe 1 mm more Apr 21 12:02:32 larsc: okay, but you normally need a cable with it Apr 21 12:02:38 wpwrak: atheros seems to be the most sane company in the industry, i hope it will last regardless of the qualcom deal. Apr 21 12:03:14 larsc: i would hope to have a full-sized connector. no loose external parts Apr 21 12:03:26 wpwrak: hm Apr 21 12:03:45 PaulFertser: as always with big global company: atheros != atheros ;) Apr 21 12:04:13 PaulFertser: yeah, there's easily worse than atheros. most of their chips have actually quite decent support. just the ar6k sucks. Apr 21 12:05:35 PaulFertser: anyway, with ieee 802.15.4, the chips do relatively little. so the companies are not nearly as close-fisted with their oh so precious intellectual property than in the wifi arena. Apr 21 12:06:21 afaik their plan for ar6k is to go a similar route marvel did for the libertas. write a small dummy firmware and a mac80211 driver Apr 21 12:06:37 instead of a fullmac firmware Apr 21 12:08:17 larsc: that would be the most sensible thing to do Apr 21 12:10:24 though every time i asked if there was progress in that area i was told that it is low priority... Apr 21 12:11:29 larsc: does wake on lan work with such dummy firmware? Apr 21 12:13:22 lindi-: wake on lan required a special wake packet, right? Apr 21 12:13:25 larsc: so not much has changed in the last 3-4 years ... Apr 21 12:15:21 larsc: or maybe they've just learned the chinese "no" :) Apr 21 12:15:23 lindi-: if you're interested in wowlan, the concept is in active development right now, there're related mails floating on their ml and the wiki is getting info. Apr 21 12:16:14 wpwrak: haha Apr 21 12:18:03 larsc: regular icmp worked here Apr 21 12:18:20 PaulFertser: for gta04? Apr 21 12:21:02 lindi-: PaulFertser seems to know more then me Apr 21 12:23:55 lindi-: no, wowlan as in "a kernel-level infrastructure to provide common framework for the drivers supporting it". Apr 21 12:32:18 mickeyl, mickey|office: ping Apr 21 12:33:06 does anyone know where to go to find out about hackable GPS's Apr 21 12:33:20 morphis: pong, morning Apr 21 12:33:35 like, is there custom linux firmware that you can do to get more features and better maps? Apr 21 12:33:53 mickey|office: we already have launch time :) Apr 21 12:34:05 indeed... tempus fugit Apr 21 12:34:10 hehe Apr 21 12:34:44 but anyway we have to talk a bit about the plan for aurora Apr 21 12:34:48 correct Apr 21 12:34:59 what are the steps, what are the features ... Apr 21 12:35:12 i have added a bit to the wiki page and created another one Apr 21 12:35:30 i also have one interested guy from Apr 21 12:35:32 I saw it Apr 21 12:35:37 #openezx who has knowledge w/ qml and want to help Apr 21 12:35:42 ah great Apr 21 12:35:47 i told him to subscribe smartphones-userland where an official announcement will come soon Apr 21 12:35:51 so add that to the agenda to talk about Apr 21 12:35:58 ok Apr 21 12:36:16 we should create the complete roadmap for 0.1 first before we do any annoucement Apr 21 12:36:27 and create tasks on which we can work Apr 21 12:36:39 so we still know what the other people are doing Apr 21 12:36:44 and what needs to be done Apr 21 12:36:46 yes. i would recommend first adding a bit up on the user stories Apr 21 12:36:52 ok Apr 21 12:36:55 then chosing the user stories we want to implement for 0.1 Apr 21 12:37:04 then seeing which consequences this has -> individual technology tasks Apr 21 12:37:19 then we can shuffle work items around Apr 21 12:37:24 jepp, it's a little bit like scrum Apr 21 12:37:31 and check who wants to take some load of our asses :D Apr 21 12:37:35 err... shoulders Apr 21 12:37:38 hehe Apr 21 12:37:41 s/of/off/ Apr 21 12:37:59 I am already started to think about the underlaying structure for having something like a "application" Apr 21 12:38:09 e.g. dynamic loading of qml stuff Apr 21 12:38:27 as we will have a memory problem when loading every stuff we need in and don't take care about it Apr 21 12:38:32 ok, cool. it's quite amazing that we can both load QML on demand or compile them in Apr 21 12:38:40 yep Apr 21 12:38:49 jepp, I think we should load it on demand Apr 21 12:38:54 in the launcher Apr 21 12:39:29 ok. what i don't understand yet is how the functional separation into different areas will work Apr 21 12:39:47 like, top bar, app area, bottom area Apr 21 12:40:00 can we compose that seperately or does every app need to know about these components? Apr 21 12:41:41 each app will completly independent of the statusbar/notification area Apr 21 12:42:09 you looked at the example I commited? Apr 21 12:42:15 no time yet Apr 21 12:42:23 but i will over the easter days Apr 21 12:42:42 ok, what you will do later for a application is the following Apr 21 12:42:46 you have a main.qml Apr 21 12:42:57 and several NameItScene.qml in your app directory Apr 21 12:43:39 a application will have several pages Apr 21 12:43:48 oh sorry, it should be NameItPage.qml Apr 21 12:44:01 in the main.qml you then: Apr 21 12:44:05 Application { Apr 21 12:44:13 name: "org.aurora.phone" Apr 21 12:44:31 onCompleted: loadPage(NameItPage.qml) Apr 21 12:44:32 } Apr 21 12:44:52 so when the app starts it loads first the NameItPage on demand and push it on stack Apr 21 12:45:18 then the NameItPage can go forward do any other page it wants within the application Apr 21 12:45:42 s/do/to/ Apr 21 12:45:42 morphis meant: then the NameItPage can go forward to any other page it wants within the application Apr 21 12:45:54 ok Apr 21 12:46:03 in the aurora application (the thing around all application, the master) there is a ApplicationStack Apr 21 12:46:21 where each running application will be put into when it launches and removed when it dies Apr 21 12:46:47 the applicationstack is some model we can then iterate with the application chooser in the launcher Apr 21 12:47:10 I will do some graphics about this and describe it in the wiki Apr 21 12:47:41 ok Apr 21 12:47:55 i like this Apr 21 12:48:36 i'm not sure how much of the actual components we can share for the ROKR E2 though, not having a touchscreen is quite a big impact on the UI design Apr 21 12:49:10 wuuups Apr 21 12:49:13 also, most of the QML examples i've seen take little care about key navigation :) Apr 21 12:49:26 so that's going to be quite a challenge for me Apr 21 12:49:31 as a touchscreen is up-to-date :) Apr 21 12:49:45 maybe we can do two different launchers Apr 21 12:49:50 one is very basic Apr 21 12:49:57 without expecting a touchscreen Apr 21 12:50:05 yep. that's my plan for tefosa Apr 21 12:50:07 and one is the fully bloated with all nice features Apr 21 12:50:29 it needs additional thinking to facilitate the two soft keys, where menu entries have to be used Apr 21 12:50:32 ok, I will put most components in aurora-components Apr 21 12:50:35 but anyways Apr 21 12:50:36 which you can then use two Apr 21 12:50:40 right Apr 21 12:50:42 s/two/too/ Apr 21 12:50:42 morphis meant: which you can then use too Apr 21 12:51:04 aurora will be the main application which does the application lifecycle and so on Apr 21 12:51:19 ya Apr 21 12:52:10 btw. there is now a android port for the Palm Pre ... Apr 21 12:52:15 ugh Apr 21 12:52:21 how do they access the modem? Apr 21 12:52:22 it's using our stuff like tslib, wifi, .... Apr 21 12:52:33 currently they don't Apr 21 12:52:36 heh Apr 21 12:52:43 they even don't released any source code Apr 21 12:52:50 mmh Apr 21 12:52:56 not nice Apr 21 12:53:11 http://www.precentral.net/android-dual-booted-pre-hardware Apr 21 12:53:53