**** BEGIN LOGGING AT Tue Feb 15 02:59:56 2011 Feb 15 06:47:22 moin Feb 15 07:29:37 Been doing some rooting around in the N9i00 GPS stack and if we can find the header file with the right ISI GPS definitions in it (to do that we need to ask the N900 ISI location server what version it is) then there is nothing standing in the way of supporting GPS in FreeSmartPhone Feb 15 07:30:02 even the SUPL stuff can be sorted thanks to http://gitorious.org/maemo-5-certificate-manager/ Feb 15 07:30:12 which incudes all the certificates and keys Feb 15 07:30:28 including any that might be used as part of the supl server communications Feb 15 07:35:25 jonwil: cool :-) Feb 15 07:37:37 jonwil: it's said that qt sdk from nokia has some very useful isi-related header, have you heard about it? Feb 15 07:37:49 It was me that found it :) Feb 15 07:37:53 hehe :) Feb 15 07:38:30 jonwil: do you already have an idea how to integrate n900 gps stuff with gpsd? Feb 15 07:38:48 no idea Feb 15 07:41:20 mrmoku, PaulFertser: moin.. the u-boot issue with gcc-4.5 was partially fixed by http://gitorious.org/u-boot-shr/u-boot/commit/4ccb9a930caf716a5285d3f3138682a691a4b303 Feb 15 07:41:44 JaMa|Off: hi :) can you probably just provide me with a working u-boot and >= 2.6.35 kernel? Feb 15 07:41:48 mrmoku, PaulFertser: after that booting maemo worked fine even with u-boot built by gcc-4.5 (that is what I was testing) Feb 15 07:41:51 I can try it this evening when i get home. Feb 15 07:42:46 PaulFertser: I have power46 + our u-boot http://build.shr-project.org/shr-unstable/images/nokia900/u-boot-nokia900-2010.06+gitr0+bd2313078114c4b44c4a5ce149af43bcb7fc8854-r68.bin.dos-like in nand Feb 15 07:43:04 PaulFertser: and current OE kernel (2.6.35) in /boot on 1st partition Feb 15 07:43:55 PaulFertser: without any boot.scr currently as saved env works for me fine (except booting nitdroid.. but I didn't debug it enough - still cannot find rootfs with their kernel) Feb 15 07:44:05 * jonwil is looking for someone who has a working isi library and can talk to the cell modem Feb 15 07:44:20 JaMa|Off: thanks, will try that in the evening. Feb 15 07:44:28 jonwil: have you tried libgisi? Feb 15 07:44:34 jonwil: it seems to be working. Feb 15 07:44:42 libgisi wont work on a stock meego install Feb 15 07:44:58 so I cant use it Feb 15 07:45:14 jonwil: stock meego should be ok, why not? Feb 15 07:45:20 I mean stock maemo Feb 15 07:45:22 sorry Feb 15 07:45:41 stock maemo doesnt have the right version of something (not sure whatP) Feb 15 07:45:46 PaulFertser: only change in "dos-like" is that maemo is booted by default with kbd closed and shr with kbd open Feb 15 07:46:10 jonwil: stock maemo has old glib version, so yes, it's difficult to compile. But stock maemo has working phonet forwarding, so you can debug ISI from host as well. Feb 15 07:46:31 except that I dont have the right setup to properly compile gisi on a host PC Feb 15 07:46:49 I run Gentoo (which is notorious for being wierd) Feb 15 07:47:07 jonwil: that really is easy, it has no real dependencies. Can you just try to "emerge vala" and to use that to compile libgisi? Feb 15 07:47:20 in any case mickeyl said he would run the tests for me :) Feb 15 07:48:31 mrmoku, PaulFertser: I've also sent an e-mail to Al http://al.robotfuzz.com/~al/maemo/u-boot/ about my u-boot master attempts and asked if he has newer version, but no reply in 2 months or so :/ Feb 15 07:48:55 jonwil: well, if you simply open another terminal, emerge vala in it and then try to compile libgisi, chances are it will just work. Feb 15 07:49:52 JaMa|Off: strange diff, i can't make sense of it. Feb 15 07:53:19 PaulFertser: it worked ok with -O0, but later I removed -O0 again and used patch from khem Feb 15 08:04:27 JaMa|Off: ok Feb 15 08:57:27 PaulFertser: renamed tested u-boot version to u-boot-nokia900-2010.06+gitr0+bd2313078114c4b44c4a5ce149af43bcb7fc8854-r68.bin.ok as there is small patch in OE which will rebuild it now and now I noticed there are also untested r70 and r71 (built with n900 unrelated changes but still untested with toolchain which was current while building them) Feb 15 08:57:28 JaMa|wrk: it would probably be good if someone corrected SHR wiki to point to the working u-boot and shr kernel for n900, i'll tell you in the evening the results. Feb 15 08:59:01 updated link in wiki Feb 15 10:10:56 ibot: message to jonwil: i have checked in the list you wanted to see in cornucopia/docs/fsogsmd/modems/ Feb 15 10:11:08 arr, wrong bot Feb 15 10:11:15 apt: message to jonwil: i have checked in the list you wanted to see in cornucopia/docs/fsogsmd/modems/ Feb 15 10:11:45 mickey|office: btw, just fyi, since i lowered the at timeout value, my phone is stable about enough for me to not get frustrated. Feb 15 10:12:41 Now i have another more important problem: battery capacity dramatically decreased to ~300mAh :) Feb 15 10:12:41 But that's of course up to me to solve. Feb 15 10:12:41 PaulFertser: that's good news. which timeout are you operating at now? Feb 15 10:12:54 mickey|office: i need to check but most probably about 5 seconds. Feb 15 10:13:20 i have meant to look at the AT backend for a while, but ISI kept me a bit busy. Hopefully I can soon carry out that change Feb 15 10:14:17 mickey|office: yes, 5s. Feb 15 10:15:03 mickey|office: that progress on ISI is rather encouraging, it's cool you finally found enough time to enroll in it. Feb 15 10:15:20 +1 Feb 15 10:16:27 :) Feb 15 10:16:53 i hope that someone can investigate libcmtspeech though Feb 15 10:58:21 mickeyl: ping Feb 15 11:19:37 jonwil: pong Feb 15 11:19:54 I saw your message, where exactly did you say that info was? Feb 15 11:20:56 jonwil: let me get you the exact link Feb 15 11:21:10 thanks Feb 15 11:21:29 Its looking very much like it should be possible to support GPS on n900 without anything funky Feb 15 11:22:07 http://git.freesmartphone.org/?p=cornucopia.git;a=blob_plain;f=docs/fsogsmd/nokia_isi/n900-servers-and-versions;hb=f5cd9a71a98e3272abf40b1cd2a3e47e4aaf4085 Feb 15 11:22:11 yes, i agree Feb 15 11:22:14 it looks pretty good Feb 15 11:25:21 Someone said something before about "AGPS may be hard because the N900 is using secret keys to talk to the supl server. But I checked and all the crypto and key functions called by the location stuff are open Feb 15 11:25:34 so anything location-proxy does to talk to the SUPL server we can replicate Feb 15 11:25:47 http://gitorious.org/maemo-5-certificate-manager/ is of value here Feb 15 11:26:25 since it has not only the code but all the certificates and keys that come with maemo Feb 15 11:26:26 and its LGPL Feb 15 11:27:00 also do you anything about something related to netlink and isi/phonet? Feb 15 11:32:12 Also, I think your code is broken, I doubt PN_SIM has versions v-01.-01 Feb 15 11:32:37 it has Feb 15 11:32:44 it's actually documented somewhere Feb 15 11:32:50 PN_SIM has a broken version field Feb 15 11:34:00 oh ok Feb 15 11:34:13 I am trying to match the versions to the various headers I have Feb 15 11:34:40 ok, cool Feb 15 11:36:41 if it helps, i can write a 'testmessage' test later this week that allows you to send arbitrary messages and see the response. i have something like this already in fsogsmd, but it has a bunch of more dependencies Feb 15 11:41:14 Would this be a phone-side tool or a PC-side tool? Feb 15 11:41:18 I am assuming PC-side Feb 15 11:41:40 anywhere Feb 15 11:41:43 it doesn't matter Feb 15 11:41:45 ok Feb 15 11:41:49 you just name the interface Feb 15 11:42:02 phonet0 if running on the device, usbpn0 if running on the PC Feb 15 11:42:56 Do you know anything about something to do with netlink and isi/phonet? Feb 15 11:45:17 i'm afraid not I, paulfertser should, he looked into the source code and "talked" to some of the guys in #ofono about it Feb 15 11:50:35 yo GNUtoo|htcdream Feb 15 11:50:55 hi Feb 15 11:51:07 any progress with the modem? Feb 15 11:52:13 * GNUtoo|htcdream is in an airport with free wifi(15min per mac address, doesn't seem to block irc once the 15min are over) Feb 15 11:52:39 maybe i should go in the queue for my ariplane... Feb 15 11:53:13 no, waiting for mrmoku's log or paul's success with forwarding Feb 15 11:53:21 where are you flying to? Feb 15 11:53:32 milan Feb 15 11:54:01 anyway....I should *really* get dvfs Feb 15 11:54:05 for n900 Feb 15 11:54:33 but i always ask for help at the wrong hour... Feb 15 11:55:11 is the press release out? Feb 15 11:55:37 btw I've news about the iphone port Feb 15 11:56:09 they think our gui is counter intuitive and isn't pretty Feb 15 11:56:22 so they want to make a new gui in gtk Feb 15 11:56:42 there weren't a lot of activity theses time Feb 15 11:56:52 becuse they needed some guidance Feb 15 11:56:58 whick I provided Feb 15 11:57:18 such as meego won't work without 3d Feb 15 11:57:42 and such as ditch desktop distros such as ubuntu,debian,fedora Feb 15 11:57:56 because they are not customizable enough for phones Feb 15 11:57:58 Good luck getting 3G on the iPhone Feb 15 11:58:15 or PMU Feb 15 11:58:16 I mean 3D Feb 15 11:58:20 ah ok Feb 15 11:58:26 indeed.... Feb 15 11:58:28 although cell modem access will be a nightmare too Feb 15 11:58:41 The iPhone is the LAST phone I would try to port a new OS to Feb 15 11:58:47 its the least documented phone on the planet Feb 15 11:58:55 well one of the least Feb 15 11:59:08 at least with an Android phone you get a kernel you can usually use Feb 15 11:59:10 yes Feb 15 11:59:35 but the #iX people could give us a prettier GUI Feb 15 11:59:45 hmmm Feb 15 12:00:08 depend on your definition of use Feb 15 12:00:18 if you mean base upon it....yes Feb 15 12:00:35 i'm pragmatic.... Feb 15 12:00:46 if there is 1 dev on htcdream Feb 15 12:00:51 or less than 1 Feb 15 12:01:09 and tons od devs for idroid and iX.... Feb 15 12:01:39 as long as it's free software I'm pragmatic Feb 15 12:02:01 I mean in real life examples.... Feb 15 12:02:14 the starting point of the palm pre.... Feb 15 12:04:05 GNUtoo|htcdream: aaaw, please not gtk Feb 15 12:04:16 GNUtoo|htcdream: please recommend QML Feb 15 12:04:31 or EFL Feb 15 12:08:43 http://pastebin.com/6tWsiUv3 Feb 15 12:08:57 Thats the notes for the isi/phonet servers Feb 15 12:10:03 The ones with a -1 version number may match headers I have, I have no way to know though Feb 15 12:11:01 sounds like we need to get Nokia to release correct headers for some servers e.g. PN_CALL, PN_GPDS, PN_DSP_AUDIO Feb 15 12:11:29 the main thing is that we DO have correct definition for PN_LOCATION Feb 15 12:12:55 right Feb 15 12:16:46 so we need to find someone with an "in" Feb 15 12:17:52 "in" at Nokia who can push for the headers we are missing and/or confirm that the header files we have match what we have (e.g. correct PN_CALL, correct PN_GPDS, correct PN_LOCATION etc) Feb 15 12:18:04 I mean correct PN_DSP_AUDIO not PN_LOCATION Feb 15 12:25:24 kc Feb 15 12:25:28 oops Feb 15 12:25:37 mickey|office: I have an image... but no working kernel yet :-/ Feb 15 12:28:37 what's missing in 2.6.37, or should we switch OE recipe to it now? Feb 15 12:40:21 JaMa|wrk: good question Feb 15 12:40:37 DVFS is missing... but that's missing in 2.6.35 too (which is built now) Feb 15 12:40:51 and a good defconfig Feb 15 13:11:14 ok, how to continue to get the PR out? Whom to contact? Feb 15 13:11:18 radekp: ping Feb 15 13:11:37 mrmoku: pong Feb 15 13:11:38 ping replicant + AoF Feb 15 13:11:49 radekp: I would like to get the PR thing done Feb 15 13:11:57 oki Feb 15 13:12:18 want to take a final look? Feb 15 13:12:19 http://wiki.freesmartphone.org/index.php/Feb11PRDraft Feb 15 13:12:29 mickey|office: ping Feb 15 13:14:20 i'd like to mention the GTA04 project Feb 15 13:14:38 can we add that right behind the "HTC phones" ? Feb 15 13:14:40 ok, good idea Feb 15 13:14:49 let me do it right now Feb 15 13:14:53 great Feb 15 13:16:01 for me it looks good Feb 15 13:16:56 whom to contact for replicant... probably gnutoo who is in an airplane right now Feb 15 13:17:11 and whom for AoF? Feb 15 13:18:16 about d-bus Feb 15 13:18:20 done Feb 15 13:18:23 maybe also note about efl? Feb 15 13:18:38 will a 'server' notice 'clients' to listen to a signal? Feb 15 13:19:17 (my guess is "no", as e.g dbus-monitor is listening to merely everything) Feb 15 13:19:22 yup Feb 15 13:19:33 you = "no"? Feb 15 13:19:39 yup Feb 15 13:20:00 from what I know, yes, the server does *not* know that Feb 15 13:20:04 or "yup it will notice"? Feb 15 13:20:11 aah, k Feb 15 13:20:14 tjnx Feb 15 13:20:29 yw (and hopefully I'm right :P) Feb 15 13:21:07 I'm trying to reproduce jonwil's findings Feb 15 13:21:39 findings for what? Feb 15 13:21:41 about SMS-CB Feb 15 13:22:13 ahh... finally the damn thing boots again :-) Feb 15 13:22:41 JaMa|wrk: where would you like to add efl? Feb 15 13:25:33 Any luck getting dbus-monitor to work? Feb 15 13:25:42 DocScrutinizer: correct. IMO a major oversight in the dbus architecture. and one of the reasons why more and more projects avoid signals and adopt an agent-based approach, where they use application-specific protocols (subscribe, unsusbscribe) to replicate something that should have been in dbus from day 1. Feb 15 13:25:50 http://paste.debian.net/107708/ uint32 "221" is the CB-channel Feb 15 13:26:21 mrmoku: maybe in "ABOUT SHR" something like "providing fast user interface based on enlightenment libraries and illume which are also used in LiMo4" Feb 15 13:26:48 the rest is bogus shit, i.e. no f'ing clue how there's going to be a string like "347434234754" in that Feb 15 13:27:13 mickey|office: agreed Feb 15 13:27:53 That fits with what I found Feb 15 13:28:01 i.e. that there is a bug in the SMS code Feb 15 13:29:16 yep Feb 15 13:29:36 we need to write some new SMS code then Feb 15 13:30:02 which means we need libisi-dev, libisi-glib-dev and csd-plugin-dev (or reverse engineered pieces matching the same) Feb 15 13:30:12 mickey|office: dunno, maybe this as well is interesting for you: http://paste.debian.net/107709/ (recall you mentioned sth about operator name) Feb 15 13:30:40 jonwil: that's the clean approach Feb 15 13:31:04 the dirty and ugly approach is to patch libsms Feb 15 13:31:07 which I would rather avoid Feb 15 13:31:13 for all sorts of reasons Feb 15 13:31:15 a dirty approach would be a complete bypass, using "tcpdump" plus some decoding shit Feb 15 13:31:56 no, we wouldn't need to go that far, we could use the ofono isi code and go from there if we dont want to (or cant Feb 15 13:31:58 plug into CSD Feb 15 13:32:00 DocScrutinizer: yes, thanks. i have since then found out that while the modem always reports the virtual operator in URC, it actually responds with both the virtual and the real network operator when you specifically ask for it. Feb 15 13:32:22 unfortunately not for ListProviders though Feb 15 13:32:28 it always overrides the real ones in that case :/ Feb 15 13:32:36 umm ok Feb 15 13:32:47 strange modem :) Feb 15 13:33:57 mickey|office: their first approach to hammer BB5 firmware into some shape where it's really useful for this exotic (from their perspective) architecture Feb 15 13:34:53 jonwil: I'm not familiar with all those libs/subsystems/APIs so can't comment on it Feb 15 13:41:23 I still wonder if there maybe is a method call to path=/com/nokia/phone/SMS; interface=Phone.SMS; member=IncomingCBS(?), to pull the actual content of that SMS-CB Feb 15 13:42:57 nope Feb 15 13:42:57 btw the first uint32 seems to be some index, or flag. I've seen 0, 1, 0 for 3 consecutive SMS, with the 3rd one probably indentical content to first one Feb 15 13:43:09 I am sure exactly what the bug is and why its failing Feb 15 13:43:17 I can see it in the disassembly of libsms Feb 15 13:43:34 that wont ever work Feb 15 13:43:41 IncomingCBS that is Feb 15 13:45:33 umm, you should document that to the byte and address, in https://bugs.maemo.org/show_bug.cgi?id=8347 ; and demand Nokia either is fixing it and releasing a fixed version, or alternatively open up sources to enable community to fix this bug Feb 15 13:48:04 without exact documentation WHAT is wrong, and optimally where it's wrong and maybe even HOW to fix it (not possible here yet, as we can't offer a patch), chances is it gets ignored by Nokia no matter how valid it is. Just too much effort for them to scrutinize their closed code to spot the bug themselves, so they will assume the request is invalid or at least NEEDINFO Feb 15 13:50:00 I cant see Nokia fixing the bug and pushing a fix OR opening the sources of libsms.so (and all the things that opening up libsms.so depends on) Feb 15 13:50:20 Although I will document the bug Feb 15 13:53:11 mickey|office: what do you want me to do? Feb 15 13:53:16 * mrmoku has a booted n900 now :-) Feb 15 13:57:50 2011-02-14T10:36:03.322022Z [DEBUG] fsogsmd : GLib : isitransport.vala:53: ISI OPEN ASYNC ENTER PHASE 1 Feb 15 13:57:53 2011-02-14T10:36:03.323273Z [DEBUG] fsogsmd : GLib : gisicomm.vala:107: NETLINK STATE = PN_LINK_UP Feb 15 13:57:56 2011-02-14T10:36:03.323578Z [DEBUG] fsogsmd : GLib : isitransport.vala:63: ISI OPEN ASYNC ENTER PHASE 2 Feb 15 13:58:34 hmmm, its hard to document this bug without knowing the names of internal data structures etc :( Feb 15 14:00:48 | configure: error: GLIB 2.27.5 or later is required to build libsoup Feb 15 14:00:49 NOTE: package libsoup-2.4-2.33.6-r0: task do_configure: Failed Feb 15 14:00:50 grr Feb 15 14:01:57 hmm 2.28.0 is out :) Feb 15 14:10:45 mrmoku: that's all? :/ Feb 15 14:11:08 JaMa|wrk: oh, good Feb 15 14:11:34 mickey|office: after that it segfaults :P Feb 15 14:13:53 mickey|office: hmm... libgisi is not in fso-autorev? Feb 15 14:14:21 hmm... it is Feb 15 14:14:42 ok, only the last commit is missing Feb 15 14:15:48 https://bugs.maemo.org/show_bug.cgi?id=8347 Feb 15 14:15:54 I posted something to the bug Feb 15 14:15:57 describing a bit more Feb 15 14:16:41 Now we just need to find someone at Nokia who can talk to whoever owns libsms and csd-sms and find out what, if anything, can be done to solve this problem. Feb 15 14:17:25 mickey|office: http://shr.pastebin.com/MgV486dG (the complete log) Feb 15 14:18:23 mickey|office: it does *not* segfault Feb 15 14:18:27 it just seems to hang Feb 15 14:18:29 MDBUS2> org.freesmartphone.ousaged /org/freesmartphone/Usage org.freesmartphone.Usage.RequestResource GSM Feb 15 14:18:32 org.freedesktop.DBus.Error.NoReply: Message did not receive a reply (timeout by message bus) Feb 15 14:24:54 bbiab Feb 15 15:09:29 re Feb 15 15:11:12 :) Packaged contents of webkit-efl into /OE/tmpdir-shr/deploy/ipk/armv7a/libewebkit0_1.3.11+svnr78547-r0.6_armv7a.ipk Feb 15 15:18:52 JaMa|wrk: btw, have you seen the gta01 advances? larsc remotely debugged everything that we could think of. Feb 15 15:20:23 PaulFertser: nope :/, but good news Feb 15 15:20:38 PaulFertser: as far as he debugged that everthing works as expected :) Feb 15 15:21:19 now I see new om-for-upstream branch on git.om :) Feb 15 15:25:53 JaMa|wrk: cool, with new u-boot linked from the wiki it works nicely. Feb 15 15:27:57 PaulFertser: after reading backlog I have to say wow :) Feb 15 15:28:29 PaulFertser: I'll try newer u-boot master I've prepared today Feb 15 15:29:42 PaulFertser: strangly it's much bigger now (200K for -Os and 350 for -O0/-O2) with current gcc-4.5 from OE Feb 15 15:30:44 and 350K is too big for current code as it keeps only 0x40000 gap for u-boot in combined image Feb 15 15:32:12 and matan's u-boot is using only 0x38000 :/ Feb 15 15:49:35 JaMa|wrk: the kernel from SHR (2.6.35) lacks ubifs support. Feb 15 15:50:51 PaulFertser: reading recipes/linux/linux-nokia900/defconfig it looks like it should support Feb 15 15:51:10 PaulFertser: can you check /proc/config.gz? Feb 15 15:56:31 JaMa|wrk: no, it's not enabled. Feb 15 15:57:36 weird, I'll check later Feb 15 15:58:34 rebuilding kernel on buildhost to see log Feb 15 15:59:49 JaMa|wrk: OE does not build nokia-900 kernel... but standard linux one Feb 15 16:11:44 PaulFertser: I have the following in dmesg Feb 15 16:11:45 [ 3724.514129] ADDRCONF(NETDEV_CHANGE): upnlink0: link becomes ready Feb 15 16:11:45 [ 3724.516235] No Phonet route to 00 Feb 15 16:13:35 mrmoku: yes, reasonable... Feb 15 16:13:42 JaMa|wrk: alsa i see no phonet0 interface on device. Feb 15 16:13:55 PaulFertser: with the 2.6.35 kernel? Feb 15 16:14:52 PaulFertser: the 2.6.35 as being built by the buildhost is a plain mainline linux one Feb 15 16:15:38 mrmoku: i booted with the one in /boot/ in the official shr rootfs tarball. Feb 15 16:16:22 sad but wrong :( Feb 15 16:17:18 PaulFertser: http://build.shr-project.org/tests/mrmoku/n900/uImage-2.6.37+gitr0+52b8e7248df1eafb122612936cdb4130cfc146fb-r58-nokia900.bin Feb 15 16:17:26 PaulFertser: and http://build.shr-project.org/tests/mrmoku/n900/modules-2.6.37+gitr0+52b8e7248df1eafb122612936cdb4130cfc146fb-r58-nokia900.tgz Feb 15 16:17:34 mrmoku: yo, great, thanks :) Feb 15 16:17:37 boots for me :-) Feb 15 16:24:35 mrmoku: it sees usd as mmcblk0p1 but root= is mmcblk1p1 Feb 15 16:26:01 PaulFertser: still CONFIG_UBIFS_FS=m,CONFIG_MTD_UBI=y,# CONFIG_MTD_UBI_GLUEBI not set in .config (in this rebuild) Feb 15 16:26:45 mrmoku: are you sure? Feb 15 16:28:25 mrmoku: ah right, sorry Feb 15 16:30:29 mrmoku: ah .35 is linux and .28 is linux-nokia900, too long for my short memory :/ I though vice-versa Feb 15 16:31:13 JaMa|wrk: yeah, we should change that as soon as we have a good defconfig for the newer nokia kernel Feb 15 16:31:36 PaulFertser: root@nokia900 ~ # cat /proc/cmdline Feb 15 16:31:37 snd-soc-rx51.hp_lim=42 snd-soc-tlv320aic3x.hp_dac_lim=6 console=tty1 root=/dev/mmcblk0p1 rootwait panic=20 debug Feb 15 16:32:36 bbl Feb 15 16:33:03 mrmoku: yes, but my cmdline comes from u-boot i use... Feb 15 16:33:19 mrmoku: or boot.scr but i do not know how to build it yet. Feb 15 16:33:47 PaulFertser: hmm... think I have an alias or script for that... moment Feb 15 16:34:57 hmm... probably just in bash history before reinstalling :P Feb 15 16:35:27 btw, any hint how to run it without the back cover? Feb 15 16:36:06 PaulFertser: http://www.denx.de/wiki/DULG/UBootScripts might help for boot.scr Feb 15 16:36:23 u-boot works fine wihtout the back cover for me Feb 15 16:36:34 I mean mounting uSD. Feb 15 16:36:44 maybe it's just the maemo kernel refusing ? Feb 15 16:37:01 mrmoku: probably, ok, i see, i've got mkimage :) Feb 15 16:41:28 mrmoku: no, without the back cover the kernel can't access usd it seems. Feb 15 16:42:54 ah, ok Feb 15 16:43:07 * mrmoku back after taxi service for the elder one Feb 15 16:57:54 mrmoku: wiki is wrong, proper command is mount -t ubifs ubi0:rootfs /mnt/ubi/ Feb 15 16:59:14 PaulFertser: /mnt/ubi breaks FHS? ;) Feb 15 16:59:18 * lindi- hides Feb 15 17:01:36 lindi-: yes, but it's only a temorary measure to fetch the bloody firmware. Feb 15 17:04:04 Finally connected over wifi. Such a pain! Feb 15 17:04:12 Mostly because "ip" is disfunctional. Feb 15 17:14:40 Damn, wifi with that 2.6.37 is unstable Feb 15 17:14:58 where? Feb 15 17:15:09 mickeyl: n900 Feb 15 17:16:52 at least you can boot into it Feb 15 17:16:59 Haha, yes Feb 15 17:17:27 Now to fight the default "full" shr idiocy: no functional "ip", no bash, no gcc... Feb 15 17:18:17 nothing a couple of opkg install shouldn't fix, right? Feb 15 17:18:28 Yes, no bash is really awful, you log in over ssh and feel like you're doing everything wrong, history doesn't work, some keybindings do not work etc etc. Feb 15 17:18:35 mickeyl: sure. Feb 15 17:18:57 mickeyl: but i'd consider working dhclient, ip, bash to be essential for any sane system. Feb 15 17:19:17 obviously the definition of sane varies ;) Feb 15 17:19:29 PaulFertser: well bash is only for command line users :) Feb 15 17:19:34 i for one, am still stuck in the mindset of ifconfig and friends Feb 15 17:19:36 never used ip Feb 15 17:19:41 i also quite like udhcpcd Feb 15 17:19:41 mickeyl: ip is great... Feb 15 17:19:46 ip is promising Feb 15 17:19:54 Ah, udhcpcd is ok, i didn't think of it. Feb 15 17:19:54 whole new CLI i must learn... Feb 15 17:20:00 unfortunately almost no distro integrates ip Feb 15 17:21:37 mickeyl: and bash is essential, especially for using it on device itself, having familiar keybindings saves keypresses on that tiny keyboard. Feb 15 17:25:39 Come on, vagalume is there but bash isn't? Feb 15 17:32:31 PaulFertser: well, i'm for adding bash by default Feb 15 17:33:24 and i think it shouldn't be hard to convince others about that, so adding it to SHR by default is propably as simple, as commiting small patch to OE ;) Feb 15 17:34:09 (while i think in shr-lite some people would say that it shouldn't be included, it is must have for shr-full) Feb 15 17:37:21 which would be problematic... for different reasons Feb 15 17:37:36 first shr-full is shr-lite + additional packages Feb 15 17:38:11 and more importantly it's not a nice thing to have some SHRs out there with bash and some with busycrap Feb 15 17:38:35 hmm Feb 15 17:38:36 i'm blind Feb 15 17:38:41 how can i send an SMS with maemo? Feb 15 17:38:50 though I'm suffering a lot from the crappy busybox shell too... so for me bash is fine :) Feb 15 17:38:53 can't see any icons for it Feb 15 17:38:58 hehe Feb 15 17:39:17 * mrmoku was searching that multiple times already too Feb 15 17:39:26 you can do it if you have an addressbook entry Feb 15 17:39:33 urghs Feb 15 17:39:36 and somewhere else too Feb 15 17:39:44 don't remember right now... and have shr booted :) Feb 15 17:41:35 heyho Feb 15 17:41:58 wuups ... double connection ... Feb 15 17:42:25 * mrmoku taxi again Feb 15 17:46:40 mickeyl: Messages app Feb 15 17:47:40 ya, this one appears once you have one contact added Feb 15 17:47:46 PaulFertser, so it boots now? Feb 15 17:47:56 GNUtoo|laptop: ya Feb 15 17:48:39 * GNUtoo|laptop switched away from FreeBSD because of the no bash issue Feb 15 17:48:53 some stuff expect bash Feb 15 17:48:58 mickeyl: ah, probably yes - i had it connected to jabber, so it had contacts from my roster ;) Feb 15 17:48:59 some stuff expected the other shell Feb 15 17:49:27 indeed SHR is hard to work with it....but there is worse: Android Feb 15 17:49:35 there isn't even tab complettion in android Feb 15 17:49:47 unless you run ash Feb 15 17:49:52 mickeyl: but you don't need to have contact with some number to send SMS to it Feb 15 17:50:09 freesmartphone.org: 03morphis 07cornucopia * r060b0a7d96b4 10/fsodeviced/src/plugins/palmpre_powersupply/plugin.vala: fsodeviced: remove old palmpre_powersupply plugin Feb 15 17:50:15 freesmartphone.org: 03morphis 07cornucopia * ra6a870ab8222 10/fsogsmd/conf/palm_pre/fsogsmd.conf: fsogsmd: palm_pre conf: load lowlevel_palmpre plugin on startup Feb 15 17:50:16 freesmartphone.org: 03morphis 07cornucopia * r21252cd39e68 10/fsodeviced/conf/palm_pre/fsodeviced.conf: Feb 15 17:50:16 freesmartphone.org: fsodeviced: conf palm_pre: uncomment currently not used parts Feb 15 17:50:16 freesmartphone.org: Will be activated again later, but for now we don't want a big bang integration. We do it Feb 15 17:50:16 freesmartphone.org: step by step. Feb 15 17:50:16 freesmartphone.org: 03morphis 07cornucopia * r4e6fda9562fd 10/fsodeviced/src/plugins/palmpre_quirks/powersupply.vala: fsogsmd: palmpre_quirks: implement FreeSmartphone.Info for battery powersupply plugin Feb 15 17:50:17 freesmartphone.org: 03morphis 07cornucopia * rfaff930bdc30 10/fsodeviced/src/plugins/palmpre_quirks/router.vala: fsodeviced: palmpre_quirks: be more verbose in router plugin Feb 15 17:50:27 freesmartphone.org: 03morphis 07cornucopia * r14c2fcd44ef8 10/fsogsmd/src/lib/at/atchannel.vala: fsogsmd: AtChannel: if we're not the mainchannel detect modem status correctly Feb 15 17:50:35 freesmartphone.org: 03morphis 07cornucopia * r58c6b2c949c6 10/fsodeviced/src/plugins/palmpre_quirks/touchscreenmanager.vala: fsodeviced: palmpre_quirks: request display resource async Feb 15 17:51:49 mickeyl, do you need help with testing modem? Feb 15 17:52:18 GNUtoo|laptop: yes, but i have no idea how to continue Feb 15 17:52:32 where are you stuck? Feb 15 17:52:33 it hangs somewhere on non-maemo Feb 15 17:52:42 and i don't know why Feb 15 17:52:47 and i can't test, as you know... Feb 15 17:53:11 I can test Feb 15 17:53:12 we have a better test infrastructure now Feb 15 17:53:12 yeah, but i don't know what to change :) Feb 15 17:53:18 first let's find were it blocks Feb 15 17:53:18 ah? Feb 15 17:55:36 Got my n900 totally depleted running SHR... Feb 15 17:55:40 While connected to usb. Feb 15 17:55:54 depleted? Feb 15 17:55:58 what does that mean? Feb 15 17:56:05 GNUtoo|laptop: no juice :) Feb 15 17:56:10 ah Feb 15 17:56:10 ouch Feb 15 17:56:15 add that: Feb 15 17:56:27 to fsodeviced.conf Feb 15 17:59:22 +[fsodevice.powersupply_n900] Feb 15 17:59:26 remove the + Feb 15 17:59:47 copy /etc/freesmartphone/conf/nokia900/fsodeviced.conf in /etc/fsodeviced.conf Feb 15 17:59:50 and add that line Feb 15 17:59:52 between Feb 15 18:00:00 [fsodevice.kernel26_powersupply] Feb 15 18:00:01 and Feb 15 18:00:08 [fsodevice.powercontrol_ifconfig] Feb 15 18:00:12 charging should be slow Feb 15 18:00:16 but it would work Feb 15 18:01:40 PaulFertser, ^^^ Feb 15 18:02:04 basically do that and use the real charger Feb 15 18:02:07 not the usb cable Feb 15 18:02:11