**** BEGIN LOGGING AT Mon Mar 05 02:59:58 2012 Mar 05 03:02:29 lol, and agan, killing browserd reboots the device... boah, whats that shit? i give up. this cant be! i feel like a dumb windows user using this phone Mar 05 03:02:35 except kismet is working nice Mar 05 03:02:41 thx and good night Mar 05 03:12:25 me again Mar 05 03:12:31 how funny.... problem solved. Mar 05 03:13:00 i had moved /usr/microb-engine to opt and symlinked to it, sill yicrob seems not to like that Mar 05 03:13:22 after wasting space for microb-engine in /usr/lib on rootfs again, it works Mar 05 03:13:50 just to let you know :) thx Mar 05 03:14:22 it eats up to 20mb on / Mar 05 03:15:01 interesting Mar 05 03:15:19 isnt that silly? :) Mar 05 03:15:38 could be a permissions thing...? Mar 05 03:15:44 i will try a loopbackmount like for python stuff Mar 05 03:15:59 Symlink should work fine, so long as permissions are good Mar 05 03:16:09 robbiethe1st, i moved as root, so couldnt be a permission problem, symlinks have no permissions imho Mar 05 03:16:37 hmmm, wait, ill check if the /opt/usr/lib is restricted, but i guess not, or should not Mar 05 03:16:39 Dunno, wasn't symlinked quite right maby? Linked to the wrong thing, or a bad path, or...? Mar 05 03:16:58 no, symlink was ok, pointed to the correct place Mar 05 03:17:44 hmm, should be not a permisson problem, just the symlink Mar 05 03:17:44 Odd. Try it again? Mar 05 03:18:00 robbiethe1st, oyu mean copy over to opt and symlink again? Mar 05 03:18:00 also, try moving it to somewhere else on the rootfs, see if you can get it to work from there? Mar 05 03:18:35 i could check, but first need a coffee :) its 4am here, hehe Mar 05 03:18:35 * ShadowJK wonderas if microb is started before /opt is mounted Mar 05 03:19:24 Though many things break with symlink thing anyway :P Mar 05 03:19:50 but 20mb is quite heavy in /, so its the first candidate to move :) Mar 05 03:20:23 ill get a coffee and then try again, move on / then to opt... lets see what happens Mar 05 03:21:45 hey Mar 05 03:22:38 i am trying to launch mediaplayer through terminal, but neither do i get an error, nor does the player start Mar 05 03:22:51 i am using /usr/bin/mediaplayer Mar 05 03:23:14 what am i doing wrong? thanks ;) Mar 05 03:35:47 ok, moved /usr/lib/microb-engine to microb-engine.x and symlinked Mar 05 03:36:02 works. symlinking seems not to be the problem. Mar 05 03:36:16 now trying to move it to opt.... Mar 05 03:43:13 moving to /opt/usr/lib and symlink breaks microb Mar 05 03:43:45 wtf ? :)) its mounted in fstab, so ... i dont understand this Mar 05 03:44:40 moving back to /usr/lib works, without rebooting Mar 05 03:45:08 a symlink to another partition breaks it, how can that be? Mar 05 03:51:05 my guess is that microb doesn't like the fact that it lives on /opt for some reason Mar 05 04:21:56 jonwil, sure, otherwise it wont fail. but whats if i only move the subdirs and libs to opt :)) hehe, ill see. 20mb is just too much on / and i cant accept that. at next i'll try a bind mount to /usr/lib, but not this night Mar 05 04:22:04 ill let you know. Mar 05 04:26:38 when I connect my n900 to PCSuite in XP, in both Ovi suite and Nokia PCsuite the filesystem in the phone shows but is really only useful for drag and drop 'cos there's no volume label I can use. Anyone come up with a solution to this? Mar 05 04:28:10 fwiw I use cygwin and rsync to keep good old-fashioned backups, but I have to unplug and replug in mass stoge mode Mar 05 04:28:46 vizzy; btw fstab is backwards on maemo, it's autogenerated *after* things have been mounted Mar 05 04:29:00 oh! Mar 05 04:29:25 to contain what the startup scripts tried to do, regardless if they succeeded or not. Mar 05 04:29:33 so where to set up mounts for system boot? Mar 05 04:32:47 i forget Mar 05 04:33:33 probably one of the big scripts at start, with calls to the *mmc* scripts Mar 05 04:34:19 I'd probably start at /etc/inittab and /etc/init.d/rcS though Mar 05 06:53:44 preinit Mar 05 06:55:21 mount_device() { in /sbin/preiit Mar 05 06:56:53 or ask pali, for further places where mounts happen Mar 05 06:57:06 vizzy: ^^^ Mar 05 06:58:42 btw microb depends on browserd which probably gets preloaded before /opt gets mounted, or sth Mar 05 06:58:53 there's a command to optify Mar 05 06:59:10 can't recall the exact name right now Mar 05 07:01:44 /usr/sbin/maemo-optify-make-mountbind.sh Mar 05 07:01:46 /usr/sbin/maemo-optify-firstboot.sh Mar 05 07:01:47 /usr/sbin/maemo-optify-auto-opt.sh Mar 05 07:01:49 /usr/sbin/maemo-optify-lib.sh Mar 05 07:14:41 Pali, ping Mar 05 07:15:24 freemangoron, sorry but now I'm going to school :-) later Mar 05 07:15:30 ok Mar 05 07:15:32 :) Mar 05 09:50:15 holy crap, I DID IT Mar 05 09:50:31 yes? Mar 05 09:50:49 I used the osso-backup-1.4.4 code found by Pali to write a tool that runs in scratchbox and decrypts encrypted osso-backup zip files(!) Mar 05 09:50:58 zip files from Fremantle Mar 05 09:51:59 so they are not that much encrypted? Mar 05 09:52:37 they are using AES Mar 05 09:52:41 so the encryption is good Mar 05 09:52:47 For my tool you still need the password Mar 05 09:52:58 This is intended for people who have an encrypted backup Mar 05 09:53:07 but dont have a working N900 Fremantle image Mar 05 09:53:15 and need to recover their backups Mar 05 09:53:43 \o/ Mar 05 09:54:24 Rather neat! Mar 05 09:54:29 ah, ok. Congrats, then :) Mar 05 10:35:43 ok, its posted Mar 05 10:36:01 now people have a way to decrypt backups without using osso-backup Mar 05 10:36:09 i.e. if you dont have a working N900 Mar 05 10:36:25 or if you want to decrypt the zip files without restoring the backup Mar 05 10:43:03 it also provides enough information to allow someone to figure out the actual osso-backup encryption Mar 05 11:18:03 what's the use of that encryption stuff? would you really want to encrypt your backups when you want to use them on PC? Mar 05 11:19:23 well, I suppose some do Mar 05 11:20:47 anyway, if it's already implemented then all thumbs up for this nice fancy Mar 05 11:22:57 The point is that the stock firmware provides a way to encrypt the backups Mar 05 11:23:12 and some people might have an encrypted backup and want to decrypt it without having to restore it via osso-backup Mar 05 11:25:19 true Mar 05 11:25:56 full ack, a backup is useless if you cant restore it.... e.g. device got a brick or broken Mar 05 11:33:41 jonwil: yoh Mar 05 11:34:31 nice part we can build for x86 too Mar 05 11:35:46 I just wondered how useful is encryption in the first place Mar 05 11:37:26 I am not Nokia and I dont know why they added the encryption Mar 05 11:38:27 now that I see the source, it doesn't look as complex as it looked before Mar 05 11:38:34 when I was staring at the ARM assembler Mar 05 11:38:46 it's ok that it's there, but I cn't find a reason why I'd use it Mar 05 11:39:22 arm asm is a bitch Mar 05 11:39:27 yeah it is Mar 05 11:40:15 ok, now for code quality part 2/6 Mar 05 11:40:30 nm Mar 05 11:40:37 o/ Mar 05 11:45:14 nice part is that we finally know another of the mysteries of the N900 :) Mar 05 11:49:11 ~seen pali Mar 05 11:49:15 pali <~quassel@unaffiliated/pali> was last seen on IRC in channel #maemo, 4h 33m 50s ago, saying: 'freemangoron, sorry but now I'm going to school :-) later'. Mar 05 12:38:43 ok, so now that I have done osso-backup stuff, what should I tackle next? Mar 05 12:43:38 profiled Mar 05 12:47:57 already looked into profiled Mar 05 12:48:12 and yes we can use the MeeGo profiled source on maemo as-is Mar 05 12:48:14 or so it looks Mar 05 12:49:29 :) Mar 05 12:50:46 I need some realtime help. I wanted to change bootdelay on Nokia N900's U-Boot so executed "setenv bootdelay 1; saveenv". and now it stands with "Saving Environment to OneNAND..." for three minutes or so. Mar 05 12:50:55 is it normal? Can I reset it? Mar 05 12:51:47 int_ua: does it finally boot? Mar 05 12:52:30 no, it just stands so. I've disabled watchdog to be able to do something in uboot Mar 05 12:52:51 (I used flasher to disable watchdog) Mar 05 12:53:22 I mean. I didn't try to switch it off yet. Mar 05 12:54:09 mmm... no idea, sorry. If it booted you could setenv bootdelay 0; saveenv and it might sort itself out... Mar 05 12:54:58 Thanks. Anyone else have ideas? Mar 05 12:56:07 ok, resetting it Mar 05 12:57:15 it looks fine, but it didn't save environment Mar 05 13:04:07 ok so what should I do next? Mar 05 13:08:23 I just don't know what licence to pick... Mar 05 13:08:47 I have this pretty cool software project going, will probably do my initial release this summer, and pondering under what licence I'd like to publish it Mar 05 13:09:15 Knowing that I'd like to sell some services around this package, and probably start a company around it some time in the future Mar 05 13:09:50 When I first started writing code, I had just met up with Stallman, so I obviously just plunked a GPL licence on it, and that was that... Mar 05 13:10:15 A couple weeks ago, I was more in favour of LGPL, and now, I'd like to just put a MIT or BSD licence on it. Mar 05 13:10:23 (l)gpl should be the first choice Mar 05 13:10:39 or any other cool free license :) Mar 05 13:11:05 One should note that this is server software, so even if a company uses it, they'd most probably wouldn't distribute their changes per se, hence, no need to open source their stuff Mar 05 13:11:43 if you want feedback and a great community around it, opensource is best, imho Mar 05 13:12:25 vizzy, if you say "open source" to RMS, he'll bite your head off. Mar 05 13:12:42 rms? Mar 05 13:12:49 Richard M. Stallman? Mar 05 13:13:02 i dont care Mar 05 13:13:11 GPL = free software, BSD/MIT = Open Source. Mar 05 13:13:22 ah, ok Mar 05 13:13:48 There's one thing I still can't get a straight answer on Mar 05 13:14:01 Does GPLv3 mean that anyone linking to my libraries has to GPL their code? Mar 05 13:14:02 i personally dont see a difference, i use gpl for my project, and call it open source and free :) Mar 05 13:14:28 crashanddie, no, but for lgpl i guess, (imho) Mar 05 13:14:52 haha Mar 05 13:14:58 you don't know anything about licences, do you? Mar 05 13:15:00 lgpl is meant for hosted apps where users change code for their needs and have to give back changes and contributions Mar 05 13:15:10 No, it's not. Mar 05 13:15:14 "hosted apps"?! Mar 05 13:15:15 That's Affero. Mar 05 13:15:41 also, the free software foundation doesn't claim that BSD-style licenses are not free software, they're just non-copyleft Mar 05 13:15:42 LGPL is: You can link libraries with proprietary software, and don't have to release your code. Mar 05 13:16:07 free software and open source licenses are the same thing, the split in terminology is ideological Mar 05 13:16:11 Robot101, aye, I was just pointing out that all open source licences aren't free software. Mar 05 13:16:19 according to rms. Mar 05 13:16:38 they are - any OSS license is also free, but non-copyleft licenses are just weaker at preserving those freedoms Mar 05 13:17:12 well, I would say that preventing people from doing whatever they want with your code is restrictive... Mar 05 13:17:37 Which is why I'm more inclined to go for BSD/MIT. Mar 05 13:18:01 I'm just scared that my project might get the cred it deserves because it's badly licensed. Mar 05 13:18:19 I don't want to do dual licensing, because that's just bullshit. Mar 05 13:21:05 yes, the GPL is more restrictive in terms of you have to do extra stuff to comply with licenses, but the FSF argues that preserving those freedoms are more important Mar 05 13:21:24 generally as a company we use LGPL 2.1 (or later) Mar 05 13:22:15 LGPLv3 is kind of annoying as GPL2 programs can't use it Mar 05 13:23:04 I think I'll just go with LGPL. Mar 05 13:23:44 Now, if my app is LGPL, can I use GPL'd code? Mar 05 13:24:00 no Mar 05 13:24:08 if it's an app rather than a library, the LGPL doesn't make that much sense Mar 05 13:24:22 Well... Mar 05 13:24:37 60% of app is libraries :) Mar 05 13:25:21 In most cases, they won't technically be linking with it, but they'll still be including some of my headers, inheriting from my classes, implementing my interfaces... Mar 05 13:25:53 umm... that sounds exactly like linking Mar 05 13:26:24 Well, no. Mar 05 13:26:29 Linking = binaries. Mar 05 13:26:37 Most of the stuff I'll provide will be headers only. Mar 05 13:27:59 what are you actually implementing? Mar 05 13:28:17 The application, you mean? Mar 05 13:28:30 yes, and what are the headers for? Mar 05 13:30:24 It's an application server, which is capable of running self-contained apps that can be composed of C++ or Python, or a mix of both. Mar 05 13:30:25 per the GPL, including such headers into another program makes a derived work. the LGPL explicitly defines what in a header file can be used by an application and /not/ be a derived work of the library. Mar 05 13:30:50 ok, you definitely want the LGPL then - otherwise you cannot have any non-GPL'd plugins / apps / etc Mar 05 13:30:57 Aye Mar 05 13:31:35 But as I mentioned earlier, it's server-side software, so most of the users could potentially develop their own modules/plugins/whateveryawannacallthem and never have to think about licensing, considering they don't distribute the app. Mar 05 13:31:45 right? Mar 05 13:32:38 as I understand GPL, yes Mar 05 13:32:45 but ianal Mar 05 13:32:50 yeah, pretty much Mar 05 13:33:24 crashanddie: also it means they will not distribute the plugins Mar 05 13:33:34 their own plugins Mar 05 13:33:40 well, obviously. Mar 05 13:34:04 so, hmm, would agpl then 'force' the users to distribute the new code? Mar 05 13:34:32 as I understand it, AGPL forces you to distribute the code, even if the code is "hosted", and not distributed. Mar 05 13:34:54 ah, ok, thats what i need then :) Mar 05 13:37:56 i saw some cms systems and other apps just copied from a gpl version, released and sold under a new name. hell. they just searched and replaced headers and names. but this will happen anyway, some people just dont care about any license.... :( and sell it under their own name. Mar 05 13:39:06 vizzy: did you notify the original copyright holders? Mar 05 13:39:26 yes, indeed. i contacted them, in this case it was xoops. Mar 05 13:39:43 there is a lot of quite successful GPL enforcement Mar 05 13:40:09 agencies like the FSF, SFLC, SFC, etc getting source code released Mar 05 13:40:35 things like openWRT and samsung TV hacking projects are all because of GPL enforcement activities Mar 05 13:41:24 I suspect large chunks of Android hacks rely on the GPL kernel source released by vendors Mar 05 13:41:33 either because they are shipping a new kernel Mar 05 13:41:48 or because the kernel contains important hardware information used to hack open the device Mar 05 13:41:49 yep, cyanogen mod etc Mar 05 13:42:05 and some vendors also provide their bootloader, making the modding a lot easier Mar 05 13:43:20 its (not)fun when you get (optional) install mails from your own software (e.g. php webapplication) and it comes with a new name and copyright Mar 05 13:43:52 theyre not just copying it, theyre aso just stupid in some case, hahaha Mar 05 13:47:12 me, I never did get the kernel and open source bits for my linux based NetComm NB5Plus4W router (not that I mind so much, a router is a router and does what I need it to :) Mar 05 13:48:14 jonwil: do you know what's on it? you could notify the SFLC if it has Busybox on it. Mar 05 13:48:51 I think the code was out there somewhere at one point, I just never took a great interest in it Mar 05 13:49:25 I think its based on the Texas Instruments AR7 router chipset Mar 05 13:49:39 but I dont much care since I dont need or want to hack my router :) Mar 05 13:50:01 I got a N900 and a Gentoo box for my linux hacking needs :) Mar 05 13:57:35 There's a busybox (and the Linux kernel, and GNU Zebra, and some GLP'd FTP server, and ...) on the "ZTE ZXDSL 931WIIa" router my ISP shipped me; someone was asking for the sources, but ZTE's Finnish folks just deferred to the ISP (they don't distribute any firmware updates or anything), and the ISP said nothing. I believe the person doing the asking did mail the people doing GPL enforcement of BusyBox, no idea if that's going to lead to anything. Mar 05 14:17:37 what pisses me off the most is vendors *cough*HTC*cough* that continue to violate the GPL time after time by delaying the release of GPL source code for months after the binaries are published or worse, not including the exact source code (including kernel drivers etc) that was used to build the binaries Mar 05 16:18:05 hello Mar 05 16:19:20 new to IRC, but my name is not right ;o Mar 05 16:19:40 arf Mar 05 16:20:00 actually wanted to write /nick newnick to explain how you do it :D Mar 05 16:20:28 ah thank you =) Mar 05 16:21:13 merlin1991: // will let you write /things as messages, in most clients Mar 05 16:21:15 /like this Mar 05 16:21:35 / // in irssi Mar 05 16:21:40 er Mar 05 16:21:47 i'm confused myself now Mar 05 16:21:56 / / even Mar 05 16:21:57 I just installed irggu Mar 05 16:22:19 as I wanted to see, what happens on maemo IRC :P Mar 05 16:23:51 but are there some more things that might be useful while using IRC? Mar 05 16:25:50 i use xchat, and thinking of porting a speechplugin to maemo for xchat... hihi Mar 05 16:26:48 I tried xchat too but I find it's gui not adapted for hildon/n900 Mar 05 16:27:33 the coolest app i found is kismet for n900 :D which is running all the time and i captured almost the whole cities wifi networks while walking around , hehe, "warwalking" Mar 05 16:28:14 Yeah I used it too, awesome app =) Mar 05 16:28:47 MrPingu, really? i found its adaption is great and works like a charm, xchat Mar 05 16:30:02 Yeah, I didn't like it Mar 05 16:30:21 you really should look at irggu http://www.elisanet.fi/irggu/ Mar 05 16:30:50 it's in extras-devel though Mar 05 16:36:56 irggu, i guess i already have it :) will try Mar 05 16:38:21 I think it looks much cleaner, as its just designed for N900 nothing less en nothing more Mar 05 16:40:26 Btw does dbus-scripts have bad impact on batterylife/cpu usage? Mar 05 16:40:54 I mean it does monitor dbus Mar 05 16:42:25 monitoring does cost resources, but isn't it already monitored by default and reacting on that? Mar 05 16:43:07 talking about dbus, is there a way to get the currently playing title in mediaplayer using dbus? Mar 05 16:55:28 You tried running dbus-monitor and then opening a song? Mar 05 16:55:52 yup, and i remember i did find that right value Mar 05 16:55:59 You might get the dbus-string for the mediaplayer for opening ;) Mar 05 16:56:08 but couldn't find a way to get it as an output Mar 05 16:56:55 hanning: see the mafw docs Mar 05 16:57:02 somewhere on the maemo wiki Mar 05 16:57:15 thanks Mar 05 16:57:28 http://wiki.maemo.org/Documentation/Maemo_5_Developer_Guide/Using_Multimedia_Components/Media_Application_Framework_%28MAFW%29 Mar 05 20:16:40 Pali, ping Mar 05 20:18:59 freemangoron, pong Mar 05 20:19:20 what are current bugs in last kernel-power? Mar 05 20:19:29 none AFAIK Mar 05 20:20:03 it was you to report some missing symbols in isp, but i cannot reproduce it Mar 05 20:20:22 well the injection driver isn't compiled for the new kernel Mar 05 20:20:32 do you have your scratchox builds? Mar 05 20:20:37 but that's not really part of the kernel tho Mar 05 20:20:45 MrPingu, correct, but that is not a bug in KP Mar 05 20:20:54 Pali, yes Mar 05 20:21:04 MrPingu, there are new builds of packet injection driver compatible with v50 Mar 05 20:21:04 BTW, I have them uploaded Mar 05 20:21:13 Pali, where? Mar 05 20:21:33 lxp1 wrote link here on IRC or sent me mail Mar 05 20:21:33 Indeed where? Did we something? Mar 05 20:21:48 good Mar 05 20:21:51 try to find in IRC log, I will search in mailbox Mar 05 20:22:29 Pali, incompatibe injection drivers are not a bug in KP Mar 05 20:22:40 yes Mar 05 20:23:00 but new builds of packet injection driver for last kp will be usefull Mar 05 20:23:18 Yeah, indeed =) Mar 05 20:23:29 I do not see them in my mailbox Mar 05 20:23:30 sure Mar 05 20:23:41 so search in IRC log of #maemo Mar 05 20:24:16 Do you know when lxp wrote that? Mar 05 20:24:34 in feb I think Mar 05 20:24:49 that's long time ago :P Mar 05 20:25:17 Is it really that hard to compile? If not, I could try it in SB Mar 05 20:25:54 http://mg.pov.lt/maemo-irclog/%23maemo.2012-02-20.log.html#t2012-02-20T23:02:44 Mar 05 20:26:00 http://mg.pov.lt/maemo-irclog/%23maemo.2012-02-21.log.html#t2012-02-21T23:42:53 Mar 05 20:26:20 source code was updated Mar 05 20:26:53 well, but no binaries? Mar 05 20:27:26 Pali, did lxp promise to make .debs? Mar 05 20:27:49 also binaries, see first link Mar 05 20:28:01 he created som scripts for compiling Mar 05 20:28:04 We never had debs for that driver, did we? Mar 05 20:28:36 my idea is to build packet injection driver in kernel-power tree and ship it in other package (not in kernel-powe-modules) Mar 05 20:28:52 ok, so you can push KP in extras-devel and ask lxp to make .debs or binaries or whatever, as AIUI the script pulls KP sources from the repo. Mar 05 20:28:54 so have it in one dsc, but other deb Mar 05 20:29:29 So first thing to happen is KP50 entering repos Mar 05 20:30:52 Yeah, would love that. I can live some weeks without injection ;) Mar 05 20:31:37 :nod: Mar 05 20:31:54 afk for a couple of minutes Mar 05 20:33:25 my 2 problems with last pre50 version of kernel-power: Mar 05 20:34:01 1) missing symbos error message (but everything worked, only error messages) in dmesg Mar 05 20:34:29 2) too high battery drain and frequent reboot after fmtxd crashed Mar 05 20:35:03 Pali, how did you do .debs, with -nc otion? Mar 05 20:35:10 *option Mar 05 20:35:17 problem 2) I solved: I enabled FMTX, disabled FMTX, enabled and again disabled. After that reboot Mar 05 20:35:29 and no other problem with battery and fmtxd Mar 05 20:35:49 I'm creating debs in madde Mar 05 20:36:04 + with patch debian/madde.patch Mar 05 20:36:22 does not matter, what options did you pass to dpkg-buildpackage? Mar 05 20:36:35 mad dpkg-buildpackage -b Mar 05 20:36:48 but madde using host ubuntu dpkg-buildpackge Mar 05 20:36:57 not same version as in scratchbox! Mar 05 20:37:11 let me check where is my build Mar 05 20:37:35 btw can you do lsmod | grep isp Mar 05 20:38:32 ok, anyway we can ignore that problem 1) if your scratchbox has no problems, then also autobuilder will do not have problems Mar 05 20:38:53 sure Mar 05 20:39:00 I will see after my n900 power on :-) I connected it to charger Mar 05 20:39:13 but just in case, will you try my build? Mar 05 20:39:23 Where did you upload your version freemangordon? Mar 05 20:39:23 I can Mar 05 20:39:36 but what about problem 2)? Mar 05 20:39:45 I have no idea why I solved that problem... Mar 05 20:39:50 just a minute, trying to find it Mar 05 20:40:04 ...but I'm sure that fmtxd can crash kernel! Mar 05 20:40:36 (battery draining was because fmtx was spawned every minute and killed after minute - I think) Mar 05 20:40:39 Are you using default fmtxd binary or that modified one? Mar 05 20:40:48 modified Mar 05 20:41:04 :) mine is stock and i don't have that problem Mar 05 20:41:14 but application should not segfault kernel! Mar 05 20:41:40 but you said that fmtxd crashes, not the kernel Mar 05 20:41:41 and now after that my steps I cannot reprodute that problem... Mar 05 20:42:18 fmtxd was stopped and started every minute and once kernel crashed too Mar 05 20:42:26 see mtd2 log in TMO thread Mar 05 20:42:32 I will find my post Mar 05 20:42:37 hmm, do you have oops backtrace? Mar 05 20:42:42 aah, ok Mar 05 20:43:37 http://talk.maemo.org/showpost.php?p=1168630&postcount=2848 Mar 05 20:44:30 Pali, here http://dox.bg/files/dw?a=dcb2ef99a6 Mar 05 20:44:55 that is my build, could you try it to see if those missing symbols are still there Mar 05 20:45:52 Missing symbols reported in dmesg or mtd2? Mar 05 20:46:15 dmesg Mar 05 20:46:36 okie =) brb installing kp ;) Mar 05 20:46:48 ok I will try your build when my battery will be full :-) Mar 05 20:46:57 why full? Mar 05 20:47:08 that bug is fixed in KP50 :p Mar 05 20:47:41 hmm, can't download Mar 05 20:50:14 lsmod | grep isp show isp_mod module and 3 other which depends on isp_mod Mar 05 20:50:49 Pali, someone is trying to load isp_mod from a different kernel, NFC who Mar 05 20:51:06 MrPingu, what is the problem? Mar 05 20:51:19 it is in 16.91 sec Mar 05 20:51:32 that error in dmesg Mar 05 20:51:43 afk for a while Mar 05 20:51:53 freemengordon, can you write dmesg | grep isp ? Mar 05 20:51:57 ok Mar 05 20:54:27 I don't see any files, I know it should display files + checkboxes and you can download in zip... but I don't see that, and I can't read the language... Mar 05 20:56:34 MrPingu, you should enter the code under the text "Въведи кода от картинката:" Mar 05 20:57:07 the text means "enter the code on the picture" Mar 05 20:58:15 Nah it doesn't display that box Mar 05 20:59:23 MrPingu, you should enter the captcha Mar 05 21:00:40 maybe you do not have fonts with all characters Mar 05 21:01:01 Pali: Mar 05 21:01:04 [ 8.277526] omap-iommu omap-iommu.1: isp registered Mar 05 21:01:05 [ 8.737487] omap-iommu omap-iommu.1: isp: version 1.1 Mar 05 21:01:26 that's all Mar 05 21:01:28 both lines I have too Mar 05 21:01:44 in sec 19 (but because I have bootmenu) Mar 05 21:01:59 and special modified Maemo on eMMC :D Mar 05 21:02:15 hmm, could it be that someone is trying to load wrong kernel modues? Mar 05 21:02:17 I have to versions of Maemo: one in mtd and one on eMMC Mar 05 21:03:06 ah I know whats wrong, I clicked the link but it stopped after the = Mar 05 21:03:07 maybe... but I think that this problem should be on ALL kernels - not only by that builded by me Mar 05 21:03:40 and other people on TMO have same symbol problems with my build Mar 05 21:04:04 Pali, NFC what is wrong, just try my build. Mar 05 21:04:22 I will try after I charge battery :-) Mar 05 21:04:39 what the... :) Mar 05 21:04:56 that bug is fixed, trust me on that (tm) Mar 05 21:05:22 yes I trust you... I only think that problem can be in madde Mar 05 21:05:50 I was talking about "unable to reboot" bug Mar 05 21:06:32 bug was fixed Mar 05 21:18:10 Now I closed some fixed bugs on garage bugtracker Mar 05 21:19:07 freemangordon, see other open https://garage.maemo.org/tracker/?atid=5521&group_id=1528&func=browse Mar 05 21:19:29 I have no idea if are valid or not Mar 05 21:24:24 can somebody test if uninstall menu entry for kernel-power working? (It is "Nokia Kernel") Mar 05 21:24:42 I'm using U-Boot without -flasher package Mar 05 22:03:52 Pali or freemangordon still here? Mar 05 22:04:02 I'm here Mar 05 22:04:48 Just before I start, how to check if someone is here thru some IRC command or isn't that possible? Mar 05 22:05:01 no way :) Mar 05 22:05:37 you can do something like "freemangordon, ping" Mar 05 22:06:03 Anyway, I am on the pre50 (link of freemangordon) I have no symbols missing string in dmesg Mar 05 22:06:14 ~seen MrPingu Mar 05 22:06:21 mrpingu is currently on #maemo (9m 20s). Has said a total of 3 messages. Is idling for 17s, last said: 'Anyway, I am on the pre50 (link of freemangordon) I have no symbols missing string in dmesg'. Mar 05 22:06:44 Well there was some other I thing I noticed in dmesg tho Mar 05 22:07:29 MrPingu, use ~seen command if you want to know if somebody is here :-) Mar 05 22:07:55 8.022064] smc91x: not found (-19). Mar 05 22:08:44 same error in my build too Mar 05 22:08:53 copied from dmesg =) Ah and TY Pali, IRC isn't that hard =) Mar 05 22:10:38 MrPingu, Pali, search dmesg log from here http://talk.maemo.org/showthread.php?t=51264 for smc ;) Mar 05 22:12:43 whatever this message means, it is a legacy from omap1 kernel Mar 05 22:12:53 smc91x is for smsc 91C9x ethernet device Mar 05 22:13:10 seems like for qemu (where is eth0 support) Mar 05 22:13:35 CONFIG_SMC91X=m is in config Mar 05 22:13:46 could be, anyway don't think it is a problem Mar 05 22:13:57 yes, it is not a problem Mar 05 22:15:11 smc is enabled in arch/arm/mach-omap2/board-rx51-peripherals.c Mar 05 22:15:11 damn kubuntu image, takes ages to download :( Mar 05 22:15:39 and there is some gpio request for RX51_ETHR_GPIO_IRQ Mar 05 22:16:22 have our n900 some ethernet chip? :D Mar 05 22:16:54 DocScrutinizer, see ^^^^ what funny is in nokia rx51 kernel :-) Mar 05 22:17:52 Pali, http://talk.maemo.org/showpost.php?p=1174885&postcount=459 Mar 05 22:18:17