**** BEGIN LOGGING AT Wed Jan 21 02:59:58 2015
Jan 21 07:11:29 <Maxdamantus> Hm. With newer versions of Linux (say, 3.14 and above), it looks like the kernel keeps the watchdog alive automatically, until some process opens the device or something.
Jan 21 07:11:46 <Maxdamantus> Is that not the case with the current maemo kernels? (eg, 2.6.28.10-cssu1)
Jan 21 07:17:15 * Maxdamantus wonders if the interface is different for /dev/twl4030_wdt
Jan 21 07:17:54 <Maxdamantus> I think in the newer kernels it just unifies into a single watchdog using /dev/watchdog
Jan 21 07:30:15 <Maxdamantus> Ah. There's no twl4030_wdt device in by filesystem's /dev .. that might be why busybox's `watchdog` doesn't work.
Jan 21 07:35:03 <Maxdamantus> Yay. It gets past an fsck now.
Jan 21 07:36:35 <APic> Good.
Jan 21 07:36:46 <APic> I love Filesystemchecks.  ;-)
Jan 21 07:36:52 <Maxdamantus> I don't.
Jan 21 07:37:14 <Maxdamantus> I like btrfs, but that's not easy to run on 2.6.28.
Jan 21 07:37:42 <APic> ext9* > btr
Jan 21 07:37:46 <APic> ext9* > RAZOR
Jan 21 07:37:49 <Maxdamantus> The reason people usually won't have to do an fsck in that instance is that other people will be using ubifs for their root.
Jan 21 07:37:50 <APic> (aka reis0r)
Jan 21 07:38:02 <APic> You are very right.  Sorry.
Jan 21 07:38:15 <APic> I use to do tune2fs -c 0 -i 0
Jan 21 07:38:21 <APic> When i get around to Thinking of it.
Jan 21 07:38:39 <APic> .o0(But you _cannot_ Tunafish, even though they may taste like Dolphineggs…)
Jan 21 07:38:59 <Maxdamantus> /dev/mmcblk0p2           27.4G     18.0G      8.0G  69% /home
Jan 21 07:39:13 <Maxdamantus> because a separate /home and /opt is silly.
Jan 21 07:39:42 <APic> I see
Jan 21 07:39:46 <APic> Planetary Collisions
Jan 21 07:40:07 <APic> But they oughta have thought of _something_ in the Past when introducing Partitionseparation on Muli{ces,Uni{}
Jan 21 07:40:10 <APic> FAIL
Jan 21 07:40:12 <APic> SORRY
Jan 21 07:40:13 <APic> I will shut up now.
Jan 21 07:40:14 <Maxdamantus> Making a block device available over UMS is also silly, in a relatively Linux-like environment like Maemo.
Jan 21 07:40:21 <APic> And vatch teh™ nic3 Snowdenfilm.
Jan 21 07:40:21 <APic> cya
Jan 21 07:40:23 <APic> *nute*
Jan 21 07:40:24 * APic mutes.
Jan 21 07:40:35 * Maxdamantus will go afk now.
Jan 21 07:40:41 <APic> hf
Jan 21 07:41:21 * Maxdamantus will soon see if he can kill the supposedly superfluous `watchdog` processes left from boot.
Jan 21 08:26:03 <Maxdamantus> 20:36:42 < APic> I love Filesystemchecks.  ;-)
Jan 21 08:26:32 <Maxdamantus> A filesystem having a check that's meant to be done pretty much every mount indicates it's a filesystem open to inconsistencies.
Jan 21 08:26:55 <Maxdamantus> actually, the check can fix the inconsistencies.
Jan 21 08:27:02 <Maxdamantus> but it's a filesystem open to data loss.
Jan 21 08:28:06 <Maxdamantus> unless you run it in a crazy way, eg, `-o data=journal` on ext3/4
Jan 21 08:28:52 <Maxdamantus> meh, I think inconsistency was the right description.
Jan 21 08:29:39 <Maxdamantus> with ext, you can end up with a filesystem in a state you never actually put it in, unless you use `-o data=journal`, which causes all data to be journalled, halving write performance.
Jan 21 08:30:34 <uhhimhere> so has maemo run on anything other than n900
Jan 21 08:35:25 <Maxdamantus> It runs on x86.
Jan 21 08:46:07 <uhhimhere> Maxdamantus: really?
Jan 21 08:46:15 <uhhimhere> Maxdamantus: like what
Jan 21 08:46:27 <Maxdamantus> It runs in scratchbox.
Jan 21 09:08:20 <Luke-Jr> who cares? maemo is dead.
Jan 21 09:10:24 <bencoh> noooo!
Jan 21 09:38:52 <kerio> scratchbox uses qemu for some stuff i think
Jan 21 09:49:49 <Maxdamantus> I think it uses it for building stuff.
Jan 21 09:50:08 <Maxdamantus> because crosscompiling is difficult for some reason.
Jan 21 09:51:33 * Maxdamantus thinks something like the nix package manager should solve it, but they started doing it wrong at the beginning -_-
Jan 21 10:02:49 <kerio> yeah, nokia fucked us up really hard in that regard
Jan 21 10:03:14 <Maxdamantus> It's not really Nokia's fault.
Jan 21 10:04:12 <Maxdamantus> Cross-compiling is hard because people haven't come up with the right packaging system for it yet.
Jan 21 10:04:25 <bencoh> erm ... :)
Jan 21 10:04:55 <Maxdamantus> I think nix leads the way for such a system, but they didn't take it into account when first writing nixpkgs.
Jan 21 10:05:19 <Maxdamantus> a few years later when they had already written most of the packages they realised they could've made it easy to crosscompile everything.
Jan 21 10:05:23 <bencoh> it has nothing to do with packaging system
Jan 21 10:05:38 <bencoh> it does have to do with broken build systems though
Jan 21 10:06:04 <Maxdamantus> which the packaging system can fix.
Jan 21 10:06:12 <Maxdamantus> I'm using "packaging system" broadly.
Jan 21 10:06:21 <Maxdamantus> where it would include things like portage.
Jan 21 10:06:27 <bencoh> actually that's not a "fix"
Jan 21 10:06:49 <kerio> no it wouldn't
Jan 21 10:06:52 <kerio> it would include stuff like apt
Jan 21 10:06:59 <kerio> because maemo is debian-based
Jan 21 10:07:01 <bencoh> that's more of a kludge where you trick the build system into thinking it can use a whole tree as a "native" sysroot
Jan 21 10:07:21 <ShadowJK> this is weird, my MyDocs is mounted without permissions for 'user'
Jan 21 10:07:24 <Maxdamantus> I'm not only talking about Maemo.
Jan 21 10:08:52 <Maxdamantus> An important thing with nix is that native packages shouldn't get special treatment.
Jan 21 10:10:28 <Maxdamantus> if you're building on x86_64 targetting ARM and you need to use bison, you'll address an x86_64 build of bison, which is still just contained in /nix/store/abcdefg-bison
Jan 21 10:13:05 <Maxdamantus> if the ARM program needs to run bison, it'll refer to /nix/store/gfedcba-bison
Jan 21 10:13:17 <Maxdamantus> traditionally, both will want to use /usr/bin/bison
Jan 21 11:32:52 <bencoh> hm, what happened to alopex ( http://talk.maemo.org/showthread.php?t=89506 ) ? anyone knows why it hasnt made it to extras-devel ?
Jan 21 11:34:42 <APic> Dunno.
Jan 21 11:44:07 <bencoh> qwazix: any idea (regarding alopex ^) ?
Jan 21 12:12:00 <qwazix> It died
Jan 21 12:12:52 <qwazix> It wasn't in very good state, I stopped being overly interested and qmlmozembed didn't get too far either
Jan 21 12:13:33 <qwazix> It didn't get to extras devel because it requires thumb
Jan 21 12:14:13 <qwazix> bencoh, ^
Jan 21 12:37:02 * Maxdamantus would quite like a new version of xulrunner, so he can just run conkeror.
Jan 21 12:38:34 <Maxdamantus> conkeror seems like about as reasonable a browser to use on the N900 as a normal computer.
Jan 21 12:38:41 <Maxdamantus> doesn't need much UI optimisation.
Jan 21 12:38:55 * Maxdamantus wrote some script for it that gives it kinetic scrolling anyway.
Jan 21 13:13:15 <bencoh> qwazix: oh ... how comes it needs thumb?
Jan 21 13:14:29 <kerio> it doesn't require thumb, but it requires some updated qt stuff
Jan 21 13:14:34 <kerio> that, so far, is only in cssu-thumb
Jan 21 13:14:50 <bencoh> hmm
Jan 21 17:05:44 <useretail> hello. does anyone know how to format 'fake' sd card properly to reveal it's true size?
Jan 21 17:12:38 <stryngs> useretail: Not sure i understand what u mean.  fdisk -l might be helpful?
Jan 21 20:07:24 <useretail> fake sd cards described here: https://www.youtube.com/watch?v=q6MBiEaSMkU
Jan 21 22:47:53 <bencoh> does anyone know why we have jemalloc installed (in microb-engine) but microb/browserd doesnt seem to be linked against it or load it at run time ?
Jan 21 22:48:15 <bencoh> (neither ldd nor lsof say anything about it)
Jan 21 23:12:32 <kerio> delete it and see if stuff explodes
Jan 21 23:22:04 <bencoh> :]
**** BEGIN LOGGING AT Thu Jan 22 01:55:46 2015
**** ENDING LOGGING AT Thu Jan 22 02:59:59 2015