**** BEGIN LOGGING AT Tue Jun 19 02:59:59 2012
Jun 19 06:38:55 <Macer> lol
Jun 19 06:39:09 <Macer> omg live chat tech support from nokia is a joke
Jun 19 06:39:26 <Macer> what a waste of life, time, and resources
Jun 19 09:39:11 <zeq> freemangordon: I've found I need to patch the fennec source, but the maemo fennec debian/rules seems very strange to me: unpacking the source tree during build..?!? It's too late by then to patch with quilt!
Jun 19 09:44:28 <merlin1991> zeq: quite a few maemo packages apply quilt patches within debian/rules
Jun 19 09:46:08 <zeq> merlin1991: how should it be done? Have you seen the rules for fennec.  I don't know whether it's unusual or not.
Jun 19 09:46:26 <merlin1991> I didn't see the rules for fennec
Jun 19 09:47:17 <zeq> I've tried including the quilt patchsys
Jun 19 09:47:38 <zeq> that didn't work because the source isn't unpacked
Jun 19 09:47:58 <merlin1991> is the rules file somewhere on the net hat I can have a quick look?
Jun 19 09:48:57 <zeq> it's in extras-devel
Jun 19 09:49:10 <zeq> that's the version I basing upon
Jun 19 09:49:40 <zeq> there's a Fennec 13 thread on tmo
Jun 19 09:50:07 <merlin1991> could you just pastebin the rules, I'm currently on windows
Jun 19 09:50:57 <zeq> hang on
Jun 19 09:55:30 * zeq is installing irc client on his laptop since empathy is broken
Jun 19 09:59:44 * zeq is surprised XChat takes so long to emerge...
Jun 19 10:12:41 <zeq> channel window isn't appearing in XChat on my laptop. Wonderful! 
Jun 19 10:17:56 <zeq> I've installed ircii, how do I switch to a channel I've joined??
Jun 19 10:18:42 <merlin1991> ctrl+#
Jun 19 10:20:21 <zeq> literally?
Jun 19 10:20:31 <merlin1991> # stands for number ,)
Jun 19 10:20:44 <zeq> that's what I thought
Jun 19 10:20:51 <zeq> doesn't seem to work :/
Jun 19 10:21:13 <merlin1991>  /window #
Jun 19 10:21:19 <merlin1991> I think that should always work
Jun 19 10:21:55 <zeq> doesn't seem to in ircII
Jun 19 10:22:29 <merlin1991> hm
Jun 19 10:23:18 <zeq> \/info gives info about the current window
Jun 19 10:23:49 <zeq> I'm obviously doing something wrong
Jun 19 10:24:13 <zeq> I seemed to join ^^
Jun 19 10:27:25 <fizzie> Based on very vague recollections of ircII, if you don't explicitly put your channels to different windows, your channels will not be in different windows, but just all meshed in together, with ^x switching the channel being talked to. (And /window swap <r> swaps the current window with the hidden window of that number.)
Jun 19 10:27:31 <zeq> irssi then...
Jun 19 10:31:39 <zeq_laptop> irssi works :)
Jun 19 10:34:24 <zeq_laptop> http://pastebin.com/Hyhxdhx4
Jun 19 10:35:14 <merlin1991> zeq_laptop: well then add simple call to quilt after the tar call?
Jun 19 10:36:03 <zeq_laptop> I haven't used quilt for 10 years.. 
Jun 19 10:36:45 <zeq_laptop> so how do I apply a patch series?
Jun 19 10:37:48 <zeq_laptop> the last time I did any packaging for a debian based distro was for the PS2 Linux Kit
Jun 19 10:37:55 <zeq_laptop> :)
Jun 19 10:38:49 <merlin1991> hehe
Jun 19 10:38:59 <merlin1991> you need to have the debian/patches folder
Jun 19 10:39:03 <merlin1991> the series file in there
Jun 19 10:39:09 <zeq_laptop> I've done that
Jun 19 10:39:20 <merlin1991> and after that see man quilt 
Jun 19 10:39:40 <merlin1991> (I don't remember the commandline for apply all patches)
Jun 19 10:39:56 <zeq_laptop> better install man on sb ...
Jun 19 10:41:00 <merlin1991> zeq_laptop: it's quilt push -a
Jun 19 10:41:23 <zeq_laptop> thanks!
Jun 19 10:42:16 <zeq_laptop> I'll need to add quilt to the build-deps too I guess
Jun 19 10:45:43 <merlin1991> yep
Jun 19 10:58:30 <zeq_laptop> merlin1991: what fun... quilt push -a attempts to apply the patches to the top of the build tree, not to where the source was unpacked, even if I 'cd' into it first!
Jun 19 11:03:48 <merlin1991> if you cd you need to cd somewhere && quilt push -a
Jun 19 11:04:10 <zeq_laptop> ah ok
Jun 19 11:04:51 <zeq_laptop> forgot it was a makefile for a moment ;)
Jun 19 11:14:58 <zeq_laptop> merlin1991: figured it out.  debian/patches needs to be symlinked into the unpacked source dir
Jun 19 11:16:24 <zeq_laptop> the patches I'm using assume the tree is unpacked at the top level
Jun 19 11:28:25 <freemangordon> zeq_laptop: look here https://gitorious.org/community-ssu/dosfstools/blobs/master/debian/rules
Jun 19 11:29:23 <freemangordon> I expect you only need to add include /usr/share/quilt/quilt.make
Jun 19 11:30:27 <freemangordon>  patch / unpatch to build/clean section
Jun 19 11:38:27 <zeq_laptop> freemangordon: It turns out patching it may not help afterall.  I thought I was hitting this bus: https://bugzilla.mozilla.org/show_bug.cgi?id=738416 but that's already applied to the source I'm using.  Instead it's failing at link time. (same method though)
Jun 19 11:38:28 <povbot> 04Bug 738416: was not found.
Jun 19 11:39:02 <zeq_laptop> I guess it's not in our qt lib
Jun 19 11:56:18 <StyXman> echo deb http://ubuntu.packages.gtm.p.fti.net/dop lucid-preprod pns > /var/opt/hosting/data/pns_php/etc/apt/sources.list.d/pns-preprod.list
Jun 19 11:56:22 <StyXman> duh
Jun 19 12:30:42 <freemangordon> zeq_laptop: what could not be found in link time, I am quite familiar with Qt, maybe I can help
Jun 19 12:36:26 <zeq_laptop> freemangordon: QTMLocationProvider::SetHighAccuracy()
Jun 19 12:37:36 <zeq_laptop> I've essentially reverted the patch that added it to mozilla with #if 0, trying to build again atm
Jun 19 12:38:19 <zeq_laptop> it was added for this bug: https://bugzilla.mozilla.org/show_bug.cgi?id=738416
Jun 19 12:38:20 <povbot> 04Bug 738416: was not found.
Jun 19 12:56:52 <japh> I play music with mplayer. When I slide open the keyboard fully, the sound is interupted for half a second or so. Same thing if someone sends a text. Isn't pulseaudio supposed to be able to play multiple streams at once, or is this because of something else?
Jun 19 12:59:50 <ivgalvez> japh: Probably mplayer is not configured to use pulseaudio but Alsa
Jun 19 13:00:14 <zeq_laptop> japh: do you use -ao pulse ?
Jun 19 13:00:17 <japh> ivgalvez, yes.
Jun 19 13:00:22 <japh> err. zeq_laptop - yes
Jun 19 13:00:47 <ivgalvez> then it "might be" because mplayer is not using DSP acceleration
Jun 19 13:00:52 <ivgalvez> and takes a lot of CPU
Jun 19 13:00:58 <japh> uhm
Jun 19 13:01:09 <ivgalvez> compare CPU consumption with any MAFW player
Jun 19 13:01:17 <japh> MAFW?
Jun 19 13:01:22 <ivgalvez> I mean
Jun 19 13:01:26 <ivgalvez> any regular player
Jun 19 13:01:30 <ivgalvez> stock one
Jun 19 13:01:38 <japh> the reason why I'm using mplayer is because everything else sucks
Jun 19 13:01:40 <ivgalvez> or Open Media Player for example
Jun 19 13:01:40 <japh> okay
Jun 19 13:02:13 <ivgalvez> if Mplayer consumes almost 100% of your CPU, that would explain the interruption
Jun 19 13:02:28 <japh> can I assume that if I try to replace pulse with for example oss4, things will break horribly?
Jun 19 13:02:42 <ivgalvez> use alse instead
Jun 19 13:02:57 <ivgalvez> that wouldn't allow any other application to take over your audio
Jun 19 13:03:09 <japh> wtf
Jun 19 13:03:15 <japh> why on earth didn't I think of that?
Jun 19 13:03:18 <ivgalvez> -ao alsa
Jun 19 13:03:23 <japh> that's why I have never liked alsa
Jun 19 13:03:27 <japh> and now it's my saviour
Jun 19 13:03:39 <japh> (dmix never works for me) 
Jun 19 13:03:41 <japh> thanks, ivgalvez 
Jun 19 13:03:43 <ivgalvez> anyhow, notifications are terrible
Jun 19 13:03:52 <ivgalvez> consuming a lot of resources
Jun 19 13:06:21 <japh> fwiw, mplayer uses ~15-20% of cpu, depending on if it's flac or mp3 I'm playing
Jun 19 13:07:11 <vi_> fwiw mplayer alsa output is absolute shite.
Jun 19 13:07:25 <vi_> It crashes CONSTANTLY.
Jun 19 13:07:34 <vi_> FOrcing you to restart it with ctrl+c
Jun 19 13:07:50 * japh cries
Jun 19 13:08:06 <vi_> japh: why you are not using open media player?
Jun 19 13:08:31 <japh> I don't know what it is or if I've tried it
Jun 19 13:08:45 <japh> but everything I've tried is horribly userhostile
Jun 19 13:08:52 <vi_> It is like the normal stock media player but with MOAR FEETURES!!!
Jun 19 13:09:05 <japh> I'll check it out
Jun 19 13:10:18 <japh> ... but not today it seems
Jun 19 13:10:34 <japh> I have a n950 to go checkout :D
Jun 19 13:13:01 <vi_> japh HOW??
Jun 19 13:13:03 <vi_> japh: HOW??
Jun 19 13:13:12 <vi_> Where did you get an n950?
Jun 19 13:13:20 <japh> sweden!
Jun 19 13:13:25 <vi_> o0
Jun 19 13:13:30 <petteri>  damn sweden!
Jun 19 13:13:30 <vi_> is this legitimate?
Jun 19 13:13:40 <vi_> Dont worry, I will not tell.
Jun 19 13:13:56 <japh> heh
Jun 19 13:13:56 <japh> yes
Jun 19 13:14:03 <vi_> japh: developer program?
Jun 19 13:14:15 <freemangordon> zeq_laptop: Which module is that? It is not in Qt for sure. https://bugzilla.mozilla.org/show_bug.cgi?id=679236 ?
Jun 19 13:14:16 <povbot> 04Bug 679236: was not found.
Jun 19 13:14:19 <japh> oh you meant if I got it from nokia?
Jun 19 13:14:24 <japh> that I did not
Jun 19 13:14:30 <freemangordon> povbot: shut up
Jun 19 13:14:30 <povbot> freemangordon: Error: "shut" is not a valid command.
Jun 19 13:14:40 <vi_> ...but that is the only place to get one.
Jun 19 13:14:46 <japh> 'yeah' :)
Jun 19 13:19:31 <vi_> japh: not gonna talk huh?
Jun 19 13:21:01 <japh> I happened to befriend someone who got one for dev and eventually my charm payed off
Jun 19 13:25:02 <zeq_laptop> freemangordon: qtm-location I'd guess
Jun 19 13:25:40 <japh> I didn't think they were that attractive, most people seems to like the n9 but I really can't stand not having a keyboard
Jun 19 13:34:02 <freemangordon> zeq_laptop: which qt-mobility version do you use?
Jun 19 13:40:07 <zeq_laptop> freemangordon: I seem to have the cssu versions installed, but this target hasn't been upgraded to CSSU.
Jun 19 13:40:45 <zeq_laptop> hmm
Jun 19 13:40:57 <freemangordon> zeq_laptop: CSSU has qtm 1.0.2, maybe you need 1.2.0, which is in extras-devel afaik
Jun 19 13:41:13 <freemangordon> but I still fail to see which symbol is missing :)
Jun 19 13:42:34 <zeq_laptop> except I've been asked to try to maintain non-CSSU compatibility... extras-devel is even less compatible...
Jun 19 13:42:40 <freemangordon> zeq_laptop: iirc I toldya to enable the thumb repo and to upgrade the target. You MUST do that, otherwise your Qt mkspecs won't be correct
Jun 19 13:43:03 <freemangordon> zeq_laptop: you're using gcc 4.6.2
Jun 19 13:43:23 <freemangordon> you're not compatible with anything but CSSU-thumb :D
Jun 19 13:43:33 <zeq_laptop> I have a target configured that way
Jun 19 13:43:43 <freemangordon> which way?
Jun 19 13:43:49 <zeq_laptop> time to start compilinng again..? :)
Jun 19 13:43:55 <freemangordon> wait
Jun 19 13:44:00 <freemangordon> let me help you
Jun 19 13:44:01 <zeq_laptop> with cssu-thumb
Jun 19 13:44:19 <zeq_laptop> I have an upgraded scratchbox target
Jun 19 13:44:25 <freemangordon> yeah, QT in CSSU is 4.7.4 
Jun 19 13:44:27 <zeq_laptop> with cssu-thumb
Jun 19 13:44:33 <freemangordon> aah, ok
Jun 19 13:44:40 <freemangordon> but qtm is still 1.0.2
Jun 19 13:44:53 <freemangordon> and AFAIK for HARM you need 1.2.0
Jun 19 13:45:06 <zeq_laptop> but I'm currently using a target which is not cssu, but has the repo in sources.list
Jun 19 13:45:19 <freemangordon> and gcc 4.2.1?
Jun 19 13:45:21 <zeq_laptop> and is up-to-date otherwise
Jun 19 13:45:25 <zeq_laptop> gcc-4.6
Jun 19 13:45:39 <freemangordon> naah, you shuld enable cssu-thumb repo
Jun 19 13:45:51 <freemangordon> and do apt-get update/upgrade
Jun 19 13:46:45 <zeq_laptop> sorry, yes cssu-thumb is in sources list, upgraded, but not with release upgrade
Jun 19 13:46:55 <zeq_laptop> other target is fully upgraded
Jun 19 13:46:59 <freemangordon> it is not needed
Jun 19 13:47:12 <freemangordon> siple upgrade should do the job
Jun 19 13:47:16 <freemangordon> *simple
Jun 19 13:47:27 <zeq_laptop> it has the cssu-thumb qt libs installed I think
Jun 19 13:47:37 <zeq_laptop> I guess they don't get held back
Jun 19 13:47:37 <freemangordon> as long as you have qt-dev-...-bla-bla-thumbN you're ok
Jun 19 13:47:55 <freemangordon> *libqt-dev...
Jun 19 13:48:35 <freemangordon> zeq_laptop: could you pastebin the exact linker erroro message?
Jun 19 13:49:31 <zeq_laptop> I don't have the build log
Jun 19 13:50:05 <zeq_laptop> it's building currently with the failing  code #if 0'd
Jun 19 13:50:19 <zeq_laptop> It will probably fail...
Jun 19 13:50:20 <freemangordon> that won't work
Jun 19 13:50:23 <freemangordon> :D
Jun 19 13:50:25 <freemangordon> yeah
Jun 19 13:51:17 <zeq_laptop> what's the qt-mobility package called?
Jun 19 13:51:39 <freemangordon> zeq_laptop: do you have that one? https://bug679236.bugzilla.mozilla.org/attachment.cgi?id=553396
Jun 19 13:51:41 <zeq_laptop> qtm
Jun 19 13:51:53 <freemangordon> let me check
Jun 19 13:52:02 <zeq_laptop> libqtm
Jun 19 13:52:28 <zeq_laptop> I don't have it installed!
Jun 19 13:52:40 <zeq_laptop> I need to add it to the build-dep for my fennec build
Jun 19 13:53:05 <zeq_laptop> you'd think the the configure would have mentioned something... :S
Jun 19 13:53:59 <freemangordon> zeq_laptop: libqtm-12
Jun 19 13:54:09 <freemangordon> that is qt-mobility 1.2
Jun 19 13:54:18 <freemangordon> but try with 1.0.2 first
Jun 19 13:54:22 <zeq_laptop> didn't have that either
Jun 19 13:54:36 <freemangordon> yeah, that is the one in extyras-devel
Jun 19 13:54:59 <zeq_laptop> I'm getting tempted to set up scratchbox on my server...
Jun 19 13:55:08 <freemangordon> :D
Jun 19 13:55:24 <freemangordon> BTW did you looak at the link i posted
Jun 19 13:55:29 <freemangordon> *look
Jun 19 13:55:59 <zeq_laptop> I'll see if it needs applying
Jun 19 13:56:31 <freemangordon> because it adds PKG_CONFIG for qtm :P
Jun 19 13:57:07 <freemangordon> and missing QtLocation LD flags :)
Jun 19 13:57:23 <zeq_laptop> it's already in
Jun 19 13:57:50 <freemangordon> so how is your configure not failing then ?!?
Jun 19 13:58:00 <zeq_laptop> good question!
Jun 19 13:58:16 <freemangordon> hmm, is there autoconf.sh?
Jun 19 13:58:33 <zeq_laptop> where?
Jun 19 13:58:49 <freemangordon> sorry, autogen.sh
Jun 19 13:58:56 <freemangordon> in main directory
Jun 19 13:59:12 <freemangordon> where did you get the source from?
Jun 19 13:59:27 <freemangordon> so I can look at debian stuff
Jun 19 13:59:31 <zeq_laptop> mozilla ftp
Jun 19 13:59:38 <freemangordon> no git? damn
Jun 19 13:59:57 <zeq_laptop> I thought I'd do better to at least work from a beta release :)
Jun 19 14:00:36 <freemangordon> zeq_laptop: seems debian/rules is screwed, is it you to make it?
Jun 19 14:00:52 <zeq_laptop> http://ftp.mozilla.org/pub/mozilla.org/mobile/candidates/14.0b7-candidates/build3/source/
Jun 19 14:01:02 <freemangordon> thanks
Jun 19 14:01:06 <zeq_laptop> I think that's the right one
Jun 19 14:01:28 <freemangordon> i need tar.bz one?
Jun 19 14:01:42 <zeq_laptop> I'm using the debian/rules from maemo
Jun 19 14:01:46 <zeq_laptop> -extras
Jun 19 14:01:54 <freemangordon> aah, ok
Jun 19 14:01:59 <zeq_laptop> it's not very nice though
Jun 19 14:02:06 <freemangordon> but it should work
Jun 19 14:04:03 <zeq_laptop>  there is this in the mozilla source: mozilla-beta/mobile/xul/installer/debian
Jun 19 14:06:26 <zeq_laptop> I did take a look at the ubuntu fennec package, it's rather dependent on ubuntu though
Jun 19 14:07:26 <freemangordon> let me check what is in maemo-extras
Jun 19 14:08:37 <freemangordon> zeq_laptop: may I have you mozconfig
Jun 19 14:08:41 <freemangordon> *your
Jun 19 14:09:03 <zeq_laptop> sure
Jun 19 14:10:23 <zeq_laptop> http://pastebin.com/sTpX6w3a
Jun 19 14:10:43 <freemangordon> 10x
Jun 19 14:12:16 <zeq_laptop> think enable-application is wrong... maybe
Jun 19 14:13:05 <freemangordon> gimme some time to download the source
Jun 19 14:13:45 <freemangordon> aah, control dependencies are for GTK build ;)
Jun 19 14:14:30 <zeq_laptop> I changed it because there only seems to be infrastructure in place for mobile/xul and mobile/android
Jun 19 14:15:27 <zeq_laptop> the old fennec was built against gtk, but that's been bit-rotting since v7
Jun 19 14:16:09 <zeq_laptop> the qt version has seen a lot of optimization since then
Jun 19 14:17:36 <freemangordon> yeah, sure, my comment was that debian/control has nothing to do with th real requirements
Jun 19 14:18:26 <zeq_laptop> yes, I changed it locally
Jun 19 14:18:57 <zeq_laptop> there's also a dependency on python-simplejson
Jun 19 14:19:02 <zeq_laptop> needed for build
Jun 19 14:19:05 <zeq_laptop> not listed
Jun 19 14:20:01 <freemangordon> thanks. BTW you're missing -O2
Jun 19 14:20:32 <zeq_laptop> the package seems to take care of -Os -O3 depending on component
Jun 19 14:20:46 <freemangordon> hmm, strange. OK
Jun 19 14:20:57 <zeq_laptop> I mean the mozilla build system itself
Jun 19 14:20:59 <freemangordon> ok, build started
Jun 19 14:21:05 <freemangordon> yeah, got it
Jun 19 14:21:27 <zeq_laptop> hope you have fast hw ;)
Jun 19 14:21:35 <freemangordon> yep
Jun 19 14:21:47 <freemangordon> 4 phenom 955 BE cores
Jun 19 14:21:54 <freemangordon> 3.2 GHz 
Jun 19 14:22:02 <zeq_laptop> similar to my server
Jun 19 14:22:08 <freemangordon> should be fast enough
Jun 19 14:22:16 <zeq_laptop> I've been building on core2duo laptop though
Jun 19 14:22:25 <freemangordon> wow
Jun 19 14:22:33 <freemangordon> that sounds slow
Jun 19 14:22:43 <zeq_laptop> definitely slower
Jun 19 14:22:59 <zeq_laptop> server is noisy when it's busy though
Jun 19 14:23:06 <zeq_laptop> gets annoying
Jun 19 14:24:18 <zeq_laptop> the laptop is nicely undervolted, so it has to be working really hard to get the fan really spinning
Jun 19 14:24:21 <freemangordon> sure, but it takes 35 minutes to build Qt here, 10 to build the kernel :). So I can live with the noise
Jun 19 14:25:31 <SpeedEvil> My fan is spinning really fast here.
Jun 19 14:25:38 <SpeedEvil> And the laptop is quite hot
Jun 19 14:25:45 <SpeedEvil> But then, it is in the bright sunshine.
Jun 19 14:25:50 <zeq_laptop> undervolting really helps
Jun 19 14:26:16 <zeq_laptop> takes time and patience to profile though
Jun 19 14:26:51 <zeq_laptop> same as n900
Jun 19 14:27:09 <zeq_laptop> except that doesn't burn my legs! :)
Jun 19 14:34:13 <teotwaki> freemangordon: I'll raise you my dual i5 2500k
Jun 19 14:34:51 <teotwaki> <fake intel fanboi> AMD SUCKS! </>
Jun 19 14:36:59 <freemangordon> teotwaki: :D. I've been AMD fan since the days I was running my 486DX2@80 at 100 MHz. 50 MHz VLB, it was flying :P
Jun 19 14:38:05 <freemangordon> zeq_laptop: it needs alsa too. but why alsa, I think pulseaudio is the way to go on n900
Jun 19 14:38:29 <zeq_laptop> freemangordon: I had an AMD 5x86 running at 150Mhz, 50MHz VLB, quite a machine :)
Jun 19 14:39:43 <freemangordon> yeah :)
Jun 19 14:40:08 <freemangordon> zeq_laptop: what about alsa?
Jun 19 14:40:26 <zeq_laptop> what about alsa?
Jun 19 14:40:38 <freemangordon> checking for alsa... Package alsa was not found in the pkg-config search path.
Jun 19 14:40:38 <freemangordon> Perhaps you should add the directory containing `alsa.pc'
Jun 19 14:40:49 <freemangordon> that is fennec configure
Jun 19 14:40:50 <zeq_laptop> didn't spot that
Jun 19 14:41:14 <freemangordon> AFAIK alsa sucks on n900, let me check where is that configured
Jun 19 14:41:45 <zeq_laptop> is it needed for the WebM support or something?
Jun 19 14:41:53 <zeq_laptop> doesn't it use gstreamer?
Jun 19 14:42:10 <freemangordon> n900? yes, it uses gstreamer
Jun 19 14:42:31 <zeq_laptop> I mean fennec/webm
Jun 19 14:42:40 <freemangordon> aah, NFC
Jun 19 14:43:56 <freemangordon> zeq_laptop: Need alsa for Ogg, Wave or WebM decoding on Linux.  Disable with --disable-ogg --disable-wave --disable-webm
Jun 19 14:44:10 <freemangordon> well, lets install it then
Jun 19 14:45:01 <zeq_laptop> yes, we definitely want those features :)
Jun 19 14:46:18 <freemangordon> the correct package is libasound2-dev
Jun 19 14:46:28 <zeq_laptop> I already had it installed
Jun 19 14:46:54 <freemangordon> yeah. BTW I am adding those to debian/control
Jun 19 14:47:07 <zeq_laptop> it's already in my control file
Jun 19 14:47:07 <freemangordon> so next time Depends: will be ok
Jun 19 14:47:12 <freemangordon> :D
Jun 19 14:47:23 <freemangordon> well, now it is in mine too :P
Jun 19 14:47:43 <freemangordon> may I have your depends:
Jun 19 14:48:38 <zeq_laptop> http://pastebin.com/4DnmgQng
Jun 19 14:48:43 <freemangordon> 10x
Jun 19 14:49:36 <zeq_laptop> I added libqtm-dev, don't know if it needs 1.2 yet as you know
Jun 19 14:50:00 <freemangordon> yeah
Jun 19 14:50:24 <freemangordon> btw you should depend only on libqt4-dev for building, without the other qt packages
Jun 19 14:50:35 <zeq_laptop> ok
Jun 19 14:51:04 <zeq_laptop> others should be in run-time deps
Jun 19 14:51:10 <zeq_laptop> don't know why i did that
Jun 19 14:51:13 <zeq_laptop> error
Jun 19 14:51:22 <freemangordon> yeah, happens
Jun 19 14:54:29 <zeq_laptop> I do think I made a mistake specifying mobile/xul, what do you have it set to?
Jun 19 14:54:49 <freemangordon> in mozconfig?
Jun 19 14:54:54 <freemangordon> I am using yours
Jun 19 14:55:02 <zeq_laptop> actually, ignore me
Jun 19 14:55:19 <zeq_laptop> it seems mobile is an alias for mobile/xul anyway from configure output
Jun 19 14:55:22 <freemangordon> let me check what are the options
Jun 19 14:55:32 <freemangordon> aah, ok
Jun 19 14:58:44 <freemangordon> ok, build process started
Jun 19 15:01:42 <freemangordon> hmm, it failed because some file missing, let me start over
Jun 19 15:02:46 <zeq_laptop> missing file?
Jun 19 15:03:02 <freemangordon> yeah, something in asm/
Jun 19 15:03:12 <zeq_laptop> strange
Jun 19 15:03:30 <freemangordon> yeah, but I tried to build it with -nc flag
Jun 19 15:03:48 <freemangordon> most probably that is the reason
Jun 19 15:07:44 <zeq_laptop> -nc is probably broken given the debian/rules unpacks during build
Jun 19 15:07:53 <freemangordon> yeah
Jun 19 15:08:05 <zeq_laptop> it's a bit yucky
Jun 19 15:08:22 <freemangordon> Ill, fix that once the build is ok
Jun 19 15:08:40 <freemangordon> s/Ill, /I'll /
Jun 19 15:08:41 <infobot> freemangordon meant: I'll fix that once the build is ok
Jun 19 15:09:11 <freemangordon> "/home/maemo/workspace/tmp/fennec-14.0b7/mozilla-beta/js/src/assembler/assembler/MacroAssemblerARM.cpp:48:26: fatal error: asm/procinfo.h: No such file or directory
Jun 19 15:09:11 <freemangordon> "
Jun 19 15:09:20 <freemangordon> WTF?
Jun 19 15:09:26 <zeq_laptop> not had that problem here
Jun 19 15:09:37 <freemangordon> linux-headers?
Jun 19 15:09:45 <zeq_laptop> probably
Jun 19 15:10:16 <zeq_laptop> I have no /usr/include/asm/procinfo.h though
Jun 19 15:10:36 <freemangordon> it is in /usr/share/source AFAIK
Jun 19 15:11:05 <zeq_laptop> I have no /usr/share/source
Jun 19 15:11:29 <freemangordon> lets ask google
Jun 19 15:12:41 <zeq_laptop> doesn't sound like it should be being included
Jun 19 15:14:34 <zeq_laptop> ah
Jun 19 15:14:36 <freemangordon> #if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,28)
Jun 19 15:14:36 <freemangordon> #include <asm/procinfo.h>
Jun 19 15:14:42 <freemangordon> funny, isn't it
Jun 19 15:14:44 <zeq_laptop> what kernel version do you have?
Jun 19 15:15:05 <zeq_laptop> why does it matter? :S
Jun 19 15:15:06 <freemangordon> Linux maemo-desktop-13 2.6.32-41-generic #90-Ubuntu SMP Tue May 22 11:31:25 UTC 2012 arm GNU/Linux
Jun 19 15:15:38 <freemangordon> yeah, should not
Jun 19 15:15:44 <zeq_laptop> wtf is causing that to happen?
Jun 19 15:16:04 <freemangordon> any clue where LINUX_VERSION_CODE is defined
Jun 19 15:16:19 <zeq_laptop> yes
Jun 19 15:16:27 <zeq_laptop> /usr/include/linux/version.h
Jun 19 15:17:03 <zeq_laptop> it's included just above the code you quoted ^^ ;(
Jun 19 15:17:06 <zeq_laptop> ;)
Jun 19 15:17:15 <freemangordon> #define LINUX_VERSION_CODE 132630
Jun 19 15:17:19 <freemangordon> WTF?
Jun 19 15:17:31 <zeq_laptop> 132636
Jun 19 15:18:07 <freemangordon> yeah, seems I have to install linux-headers-cssu
Jun 19 15:18:28 <zeq_laptop> I have "linux-kernel-headers" installed
Jun 19 15:18:38 <zeq_laptop> not sure which repo that's from
Jun 19 15:19:02 <freemangordon> cssu-thumb :P
Jun 19 15:19:13 <zeq_laptop> explains it :)
Jun 19 15:22:23 <freemangordon> hmm, for some reason apt-get thinks linux-kernel-headers in cssu-thumb are lower version than those in sdk repo :S
Jun 19 15:22:50 <zeq_laptop> why did it get work for me?
Jun 19 15:22:56 <freemangordon> NFC
Jun 19 15:23:50 <freemangordon> damn. the package name is linux-kernel-cssu-headers :(
Jun 19 15:26:15 <zeq_laptop> mine is definitely linux-kernel-headers
Jun 19 15:26:46 <freemangordon> I have that installed too, NFC why it has a lower linux version
Jun 19 15:27:13 <freemangordon> well, after installing kernel-cssu-headers it seems ok
Jun 19 15:28:22 <zeq_laptop> good :)
Jun 19 15:37:30 <freemangordon> hmm, its getting noisy here :D
Jun 19 15:40:31 <zeq_laptop> no more errors?
Jun 19 15:40:40 <freemangordon> yep
Jun 19 16:01:20 <zeq_laptop> freemangordon: is there a reason for specifying vfp in DEB_BUILD_OPTIONS over neon?  (The cortex-a8 has a really incapable vfp unit.)
Jun 19 16:02:21 <zeq_laptop> I was just thinking about the thumb SDK howto
Jun 19 16:02:36 <freemangordon> zeq_laptop: AFAIK gcc vectoriser sucks
Jun 19 16:02:49 <zeq_laptop> even on 4.6?
Jun 19 16:03:11 <freemangordon> otherwise that is a legacy from ARMEL_FREMANTLE target
Jun 19 16:03:18 <freemangordon> NFC
Jun 19 16:03:19 <zeq_laptop> I read on wikipedia the VFP is 10x slower
Jun 19 16:03:26 <freemangordon> could be
Jun 19 16:04:09 <freemangordon> but AFAIK you should enable -munsafe-math too, as NEON has some precision incompatibilities with ieee standart
Jun 19 16:04:31 <freemangordon> otherwise vectoriser wouldn't kick in
Jun 19 16:04:54 <freemangordon> It is possible I am talking bullshit too :D
Jun 19 16:05:55 <zeq_laptop> you could well be right, if so it's a rational for the default at least
Jun 19 16:06:07 <zeq_laptop> s/rational/rationale/
Jun 19 16:06:09 <infobot> zeq_laptop meant: you could well be right, if so it's a rationale for the default at least
Jun 19 16:06:44 <zeq_laptop> however, I suspect if a typical order of magnitude performance difference is typical...
Jun 19 16:07:05 <freemangordon> zeq_laptop: I is the same situation as with MMX/SSEx AFAIK, you should use asm to have some speed gain
Jun 19 16:07:12 <freemangordon> *It
Jun 19 16:07:32 <zeq_laptop> well -mfpmath=sse is std on x86_64
Jun 19 16:08:05 <freemangordon> yeah, but do you thing ARM gcc is as advanced as x86/64 is? I doubt.
Jun 19 16:08:11 <freemangordon> *think
Jun 19 16:08:38 <freemangordon> well, now we have a modern gcc version we can do some tests
Jun 19 16:08:43 <zeq_laptop> indeed
Jun 19 16:08:55 <freemangordon> just need to find a victim to be optimized
Jun 19 16:08:57 <zeq_laptop> 4.2 was pretty horrible
Jun 19 16:09:32 <freemangordon> on the other hand I think we already have neon/pld optimized memcpy in libc6
Jun 19 16:09:35 <zeq_laptop> http://en.wikipedia.org/wiki/ARM_architecture#Advanced_SIMD_.28NEON.29
Jun 19 16:09:47 <zeq_laptop> mentions mp3 as a victim
Jun 19 16:10:20 <zeq_laptop> mp3 stream decoding @10Mhz
Jun 19 16:10:31 <freemangordon> yeah, but it should be hand-coded, I don't think gcc can make it
Jun 19 16:10:49 <freemangordon> WOW, linking libxul :)
Jun 19 16:11:01 <zeq_laptop> the vectorizor was seriously improved in 4.6
Jun 19 16:11:27 <freemangordon> could be, I don't have much deeper knowledge on gcc internals
Jun 19 16:11:34 <zeq_laptop> might need -ftree-vectorize though to turn it on
Jun 19 16:11:47 <freemangordon> yeah, that one too
Jun 19 16:12:05 <freemangordon> along with -funsafe-math
Jun 19 16:12:14 <freemangordon> (or whatever the switch was)
Jun 19 16:12:23 <zeq_laptop> not relevent for fennec though because it is hand coded for NEON
Jun 19 16:12:35 <zeq_laptop> -ffast-math
Jun 19 16:12:43 <freemangordon> aah, yeah :)
Jun 19 16:12:54 <zeq_laptop> or theres -Ofast today too
Jun 19 16:13:07 <zeq_laptop> that one's fun for breaking things :)
Jun 19 16:13:42 <zeq_laptop> (My entire Gentoo is built with it btw)
Jun 19 16:14:11 <zeq_laptop> except where it's *known* to break things
Jun 19 16:15:02 <zeq_laptop> most packages actually check during configure for ieee754 compliance where they need it
Jun 19 16:15:22 <zeq_laptop> it's just those which don't...
Jun 19 16:15:31 <freemangordon> damn, what size has libxul in fennec, it is linking for the last 5 minutes :O
Jun 19 16:15:50 <zeq_laptop> it's probably big
Jun 19 16:16:03 <freemangordon> yeah, probably :)
Jun 19 16:16:37 <freemangordon> I hope I won't run out of RAM
Jun 19 16:16:53 <_PanzerSajt> hy
Jun 19 16:17:02 <zeq_laptop> with fennec 7 from extras it's 21.7M stripped
Jun 19 16:17:13 <freemangordon> hmm, I am already OOM :(
Jun 19 16:17:20 <zeq_laptop> how much do you have?
Jun 19 16:17:36 <freemangordon> It's a virtual machine, 1GB
Jun 19 16:17:43 <freemangordon> otherwise I have $
Jun 19 16:17:46 <freemangordon> 4
Jun 19 16:18:00 <zeq_laptop> I'm going to hit the same limit aren't I.../
Jun 19 16:18:18 <zeq_laptop> can we give it more?
Jun 19 16:18:29 <freemangordon> lets hope 1.5 GB of swap will be enough
Jun 19 16:19:09 <zeq_laptop> is it scratchbox that's limited? or are you running in a separate VM?
Jun 19 16:19:38 <freemangordon> I am running Nokia's wmware image
Jun 19 16:19:45 <freemangordon> *vmware
Jun 19 16:19:46 <zeq_laptop> oh right ok
Jun 19 16:19:53 <zeq_laptop> so I should be okay ... 
Jun 19 16:19:55 <freemangordon> ok, ld failed
Jun 19 16:20:07 <freemangordon> undefined reference to `MaemoLocationProvider::SetHighAccuracy(bool)'
Jun 19 16:20:08 <zeq_laptop> :(
Jun 19 16:20:12 <zeq_laptop> hmm
Jun 19 16:20:24 <zeq_laptop> OOM?
Jun 19 16:20:29 <zeq_laptop> or missing symbol?
Jun 19 16:20:34 <freemangordon> noo, it is a missing symbol
Jun 19 16:20:39 <freemangordon> the same as yours
Jun 19 16:20:43 <zeq_laptop> THE missing symbol
Jun 19 16:20:51 <freemangordon> THE missing symbol
Jun 19 16:21:03 <freemangordon> are you shure the patch from romaxa is there
Jun 19 16:21:04 <freemangordon> ?
Jun 19 16:21:11 <zeq_laptop> yes
Jun 19 16:21:35 <zeq_laptop> It might be worth trying with it reverted
Jun 19 16:21:44 <zeq_laptop> I've no idea where that comes from
Jun 19 16:21:53 <freemangordon> but why it cannot find it then. BTW I don't see any libqtm libs in ld command line, only Qt ones
Jun 19 16:21:57 <zeq_laptop> in the bugzilla entry it talks about Android
Jun 19 16:22:12 <freemangordon> hmm,let me check PKG_CONFIG
Jun 19 16:22:26 <zeq_laptop> yet my googling reveals that API is related to qtm
Jun 19 16:22:53 <freemangordon> well, that could explain why
Jun 19 16:24:03 <freemangordon> zeq_laptop: that is from config.log:
Jun 19 16:24:12 <freemangordon> "/scratchbox/compilers/arm-linux-fremantle-gcc4.6.2/bin/../lib/gcc/arm-none-linux-gnueabi/4.6.2/../../../../arm-none-linux-gnueabi/bin/ld: cannot find -lQtSensors"
Jun 19 16:26:29 <zeq_laptop> which is?
Jun 19 16:26:53 <zeq_laptop> is it related?
Jun 19 16:27:21 <freemangordon> configure:16153: checking for QtSensors QtFeedback QtLocation
Jun 19 16:27:30 <freemangordon> is it? :P
Jun 19 16:28:13 <freemangordon> could be that PKG_CONFIG cannot find libwtm package on fremantle
Jun 19 16:28:47 <zeq_laptop> okay, so what is libwtm?
Jun 19 16:29:11 <freemangordon> sorry, typo, libqtm 
Jun 19 16:29:34 <freemangordon> damn, there is no QtSensors package
Jun 19 16:29:45 <zeq_laptop> freemangordon: I have a symlink: libQtSensors.so -> libQtSensors.so.1.0.2
Jun 19 16:29:50 <zeq_laptop> but no libQtSensors.so.1.0.2
Jun 19 16:30:01 <freemangordon> but there is no pkgconfig package
Jun 19 16:30:02 <zeq_laptop> missing lib in qt
Jun 19 16:30:04 <zeq_laptop> ?
Jun 19 16:30:11 <zeq_laptop> there is no lib
Jun 19 16:30:12 <freemangordon> no, missing .pc
Jun 19 16:30:14 <zeq_laptop> just the symlink
Jun 19 16:30:34 <freemangordon> there is a library, but no .pc in -dev package
Jun 19 16:30:53 <zeq_laptop> I don't have the lib
Jun 19 16:31:11 <zeq_laptop> I have a symlink belonging to libqtm-dev
Jun 19 16:31:20 <freemangordon> hmm, me too
Jun 19 16:31:22 <zeq_laptop> the lib is missing here
Jun 19 16:31:33 <freemangordon> seems after all we need libqtm12
Jun 19 16:32:23 <freemangordon> Or I screwed it up when libqtm was build
Jun 19 16:32:31 <freemangordon> lemme check
Jun 19 16:32:58 <zeq_laptop> is it installed into /usr/lib or opt?
Jun 19 16:33:30 <freemangordon> it should be in /opt
Jun 19 16:33:32 <zeq_laptop> qtm-12 is installed into /opt/qtm12
Jun 19 16:33:46 <freemangordon> qtm is in /opt/maemo/usr/lib
Jun 19 16:33:55 <zeq_laptop> I decided to install it since you didn't, just in case
Jun 19 16:34:15 <zeq_laptop> but there's a b0rken symlink in /usr/lib
Jun 19 16:34:17 <freemangordon> but you should change which version to be used, hold on
Jun 19 16:34:22 <freemangordon> it is not so simple
Jun 19 16:35:48 <freemangordon> zeq_laptop: seems I screwed up qtm-sensors.install, lemme fix it
Jun 19 16:35:57 <zeq_laptop> libqtm-dev doesn't pull in the binaries
Jun 19 16:36:10 <freemangordon> it depends on the other packages
Jun 19 16:36:28 <freemangordon> give me 10 minutes to fix qt-mobility package
Jun 19 16:36:34 <zeq_laptop> libqtm-sensors at least isn't installed
Jun 19 16:36:42 <zeq_laptop> I have no broken deps
Jun 19 16:37:17 <freemangordon> hmm, correct
Jun 19 16:37:24 <zeq_laptop> libqtm-location is installed though
Jun 19 16:37:45 <zeq_laptop> except -dev creates symlinks to non-existant libs
Jun 19 16:37:56 <zeq_laptop> at least QtSensors
Jun 19 16:38:00 <zeq_laptop> anyway
Jun 19 16:38:01 <freemangordon> damn, seems -dev package dependencies are screwed
Jun 19 16:38:12 <freemangordon> apt-get install libqtm-sensors does the job
Jun 19 16:38:32 <zeq_laptop> is it the only missing one?
Jun 19 16:38:39 <freemangordon> I think so
Jun 19 16:39:19 <zeq_laptop> that fixed my symlink
Jun 19 16:39:28 <zeq_laptop> I wonder if it's too late for my build
Jun 19 16:39:46 <freemangordon> I am afraid yes :(
Jun 19 16:40:12 <freemangordon> as you win't have -lQtSensors
Jun 19 16:40:14 <zeq_laptop> pretty silly of configure allowing it to continue
Jun 19 16:40:21 <freemangordon> yep
Jun 19 16:40:40 <zeq_laptop> ^C
Jun 19 16:40:41 <freemangordon> I'm starting over, thanks god it takes only 30 minutes here
Jun 19 16:42:59 <freemangordon> but will increase my RAM to 2GB before that
Jun 19 16:43:10 <zeq_laptop> good idea
Jun 19 16:45:43 <zeq_laptop> I wonder if it would be worth using neon instead of vfp even for scalar math given the a8's VFPLite?
Jun 19 16:46:27 <zeq_laptop> I'm going to look into the non-conformance issue
Jun 19 16:46:34 <freemangordon> switching between neon/vfp is a very heavy op, something like 25-50 CPU ticks
Jun 19 16:47:55 <zeq_laptop> so it's better if it's used more if it all
Jun 19 16:48:17 <freemangordon> but have in mind everything else is vfp
Jun 19 16:48:56 <zeq_laptop> any idea how it interacts with the context switch buffer flush hit?
Jun 19 16:49:12 <zeq_laptop> (from the thumb bug workaround)
Jun 19 16:49:40 <freemangordon> should not be affected by it AIUI, it is BTB that is flushed
Jun 19 16:50:01 <freemangordon> and that should have nothing to do with neon state
Jun 19 16:50:23 <zeq_laptop> I still would rather see some benchmarks than reach premature conclusions
Jun 19 16:50:48 <freemangordon> or think a while that neon shares the same registers/execution pipeline with vfp
Jun 19 16:51:54 <zeq_laptop> how much is vfp really used though? what's going to be running fp heavy? serious question
Jun 19 16:52:16 <freemangordon> hildon?
Jun 19 16:52:49 <zeq_laptop> that's easy enough to NEONize though
Jun 19 16:52:51 <freemangordon> all transitions are calculated using doubles
Jun 19 16:53:09 <freemangordon> (and not floats ;) )
Jun 19 16:53:18 <freemangordon> NFC how is that compiled
Jun 19 16:53:28 <freemangordon> but suspect it is vfp
Jun 19 16:54:22 <zeq_laptop> yeah, that's a complication
Jun 19 16:55:04 <freemangordon> there is even a proposed patch for hildon-desctop replacina all sin/cos, etc with fsin,fcos, and doubles with floats
Jun 19 16:55:05 <zeq_laptop> is a double really necessary for that?
Jun 19 16:55:21 <zeq_laptop> indeed
Jun 19 16:55:24 <freemangordon> didn't make it in CSSU, don't ask me why
Jun 19 16:55:49 <freemangordon> should not be really needed
Jun 19 16:56:03 <freemangordon> I mean double precision. 
Jun 19 16:56:50 <freemangordon> But having in mind there is libclutter used, which is double precision, I wonder how much the gain will be
Jun 19 16:57:23 <zeq_laptop> worth testing I'd have thought
Jun 19 16:58:08 <freemangordon> well, join CSSU team then
Jun 19 16:58:17 <freemangordon> BTW for how long do you have your n900?
Jun 19 16:58:40 <zeq_laptop> ~1.5 years
Jun 19 16:58:54 <zeq_laptop> 3rd N900 actually (USB)
Jun 19 16:59:17 <freemangordon> hmm, why it took you so long to join the fun? Or I am missing something
Jun 19 16:59:40 <zeq_laptop> you're not missing anything
Jun 19 16:59:46 <zeq_laptop> I was busy with Gentoo
Jun 19 16:59:58 <freemangordon> aah, ok
Jun 19 17:00:03 <zeq_laptop> only so many hours in a day
Jun 19 17:00:24 <freemangordon> yeah, I know :)
Jun 19 17:01:21 <freemangordon> hmm, that looks better "Mem:       2061488    1759064     302424"
Jun 19 17:02:31 <zeq_laptop> I really should have become a full Gentoo Dev, but never took the test.
Jun 19 17:03:45 <zeq_laptop> I didn't see the need since I could fix things through bugzilla, and overlays, but lately fixes never seem to make their way from the bugzilla.
Jun 19 17:03:57 <zeq_laptop> OT though
Jun 19 17:04:32 <freemangordon> zeq_laptop: I was not joking about CSSU, we really need more devs there, think about it
Jun 19 17:05:11 <freemangordon> and have in mind the real fun is starting now, as I see it
Jun 19 17:05:13 <zeq_laptop> Really need to brush up on my Debain-fu
Jun 19 17:06:10 <zeq_laptop> I agree, although I'm somewhat concerned about divergence from mer
Jun 19 17:07:05 <freemangordon> well, SB and git is a piece of cake, I am not linux developer only(because of my job), n900 was my first contact with debian, my knowledge was on fedora
Jun 19 17:07:36 <zeq_laptop> git is no problem, and I'm getting the hang of SB
Jun 19 17:08:07 <zeq_laptop> I just forgot a lot about Debain packaging over the years
Jun 19 17:08:22 <zeq_laptop> like the various patch systems
Jun 19 17:08:48 <freemangordon> well, to be honest all of the packages I saw because of n900 have different debian/rules :D
Jun 19 17:09:08 <freemangordon> and it seems only quilt is used
Jun 19 17:09:54 <freemangordon> anyway, it is up to you, but feel welcomed
Jun 19 17:10:04 <zeq_laptop> I read earlier that quilt is now integrated in format3 rules
Jun 19 17:10:12 <freemangordon> (hmm, welcomed? is there such a word in english?)
Jun 19 17:10:22 <zeq_laptop> thanks, I probably will :)
Jun 19 17:10:30 <zeq_laptop> yes
Jun 19 17:10:36 <freemangordon> aah, thanks
Jun 19 17:10:46 <zeq_laptop> but not that usage ;)
Jun 19 17:10:51 <zeq_laptop> only kidding
Jun 19 17:10:52 <freemangordon> :D
Jun 19 17:11:21 <zeq_laptop> you should feel welcome to join/or welcomed aboard
Jun 19 17:11:43 <freemangordon> aaah, yeah, thanks
Jun 19 17:12:33 <freemangordon> hmm, still not hitting swap, thats good
Jun 19 17:13:22 <zeq_laptop> more more errors in config.log?
Jun 19 17:13:48 <freemangordon> zeq_laptop: BTW have in mind kernel-cssu which is currently a dependency (i.e. is installed on your n900) has a bug
Jun 19 17:14:12 <zeq_laptop> oh?
Jun 19 17:14:18 <freemangordon> thumb-2 related, there is a newer version in the repo, but it breaks mp-fremantle-community-pr
Jun 19 17:14:46 <zeq_laptop> I should wait then?
Jun 19 17:14:52 <freemangordon> I am wating to see if X crashes are really fixed before making a new release
Jun 19 17:15:53 <freemangordon> zeq_laptop: well, most probably no, the bug is: when a thumb2-compiled signal handler is called, it is possible some fun things to happen
Jun 19 17:16:27 <zeq_laptop> like corruption?
Jun 19 17:17:06 <freemangordon> like whatever you can imagine if first 1-4 instructions of the functions are not executed :D
Jun 19 17:17:13 <freemangordon> *function
Jun 19 17:17:19 <freemangordon> i.e. signal handler
Jun 19 17:17:24 <zeq_laptop> LOL
Jun 19 17:17:27 <zeq_laptop> that's not good!
Jun 19 17:17:37 <freemangordon> who said it is 
Jun 19 17:17:51 <freemangordon> the good think is there is a fix
Jun 19 17:18:04 <freemangordon> and I compiled a kernel with that fix
Jun 19 17:18:38 <freemangordon> so far it seems pretty OK, running on my primary device (the only phone I am using)
Jun 19 17:19:14 <freemangordon> WTF?
Jun 19 17:19:18 <freemangordon> "/home/maemo/workspace/tmp/fennec-14.0b7/mozilla-beta/dom/system/unix/QTMLocationProvider.h:40:34: fatal error: QGeoPositionInfoSource: No such file or directory"
Jun 19 17:19:22 <zeq_laptop> I'm running the version with the bug on my primary device, it's obiously not hit too often
Jun 19 17:19:36 <zeq_laptop> fun
Jun 19 17:19:38 <freemangordon> do you hit it at all?
Jun 19 17:19:42 <zeq_laptop> not yet
Jun 19 17:19:57 <freemangordon> yeah, that is a tricky one
Jun 19 17:20:06 <zeq_laptop> although I'm cooking dinner atm and on battery which slows the laptop down
Jun 19 17:20:29 <freemangordon> zeq_laptop: http://git.et.redhat.com/?p=kernel-kraxel.git;a=commitdiff_plain;h=d71e1352e240dea32d481ad8d662e8de4406ac7e
Jun 19 17:20:35 <zeq_laptop> (BIOS override with media bay battery - don't ask)
Jun 19 17:21:18 <freemangordon> That is the kernel bug I hit when thumb-compiled X
Jun 19 17:22:24 <freemangordon> nice, isn't it :)
Jun 19 17:22:38 <zeq_laptop> our kernel's that old now...?
Jun 19 17:22:40 <zeq_laptop> :(
Jun 19 17:22:54 <freemangordon> well, there are lots of backports
Jun 19 17:22:56 <zeq_laptop> I forgot how old it is
Jun 19 17:23:40 <zeq_laptop> even more backports in KP/CSSU
Jun 19 17:23:47 <zeq_laptop> but still
Jun 19 17:23:56 <freemangordon> this IS KP
Jun 19 17:24:02 <zeq_laptop> yes, I know
Jun 19 17:24:17 <freemangordon> v -pre51
Jun 19 17:24:44 <zeq_laptop> I just meant that even though the kernel has a lot of backported fixes, it's still a very old kernel by now
Jun 19 17:25:00 <freemangordon> yeah, but there is no better option for fremantle so far
Jun 19 17:25:01 <zeq_laptop> some fixes get missed
Jun 19 17:26:28 <zeq_laptop> is there any plan to eventually transition to mer+hildon?
Jun 19 17:27:12 <zeq_laptop> or is there not sufficient coverage of non-distributable binaries for a fully functional device?
Jun 19 17:27:30 <freemangordon> none afaik. And TBH there are still too many pieces missing
Jun 19 17:27:48 <freemangordon> zeq_laptop: GF came, have to go :)
Jun 19 17:27:59 <freemangordon> happy fennec-building
Jun 19 17:28:41 <zeq_laptop> okay see you later
Jun 19 18:09:13 <zeq_laptop> freemangordon: when you get back, I've determined the reason for the build failure: QTDIR is unset which results in various paths being from "/" instead of "/usr", specificly in the case of the build failure: $QTDIR/include/QtLocation
Jun 19 18:54:10 <freemangordon> zeq_laptop: I am semi-back, but thanks. Anyway I won't have time today to continue fennec building, hope you succeed
Jun 19 19:02:30 <zeq_laptop> freemangordon: that's no problem, thanks for your help with it so far.  Hit the next failure now; no QFeedbackEffect
Jun 19 19:24:33 <zeq_laptop> freemangordon: It does need at least qtm-11 for QtFeedback, but qtm-11-dev doesn't install the pkgconfig files.
Jun 19 20:18:47 <kaizerslawten> Hello Guys
Jun 19 20:40:41 <Raimu> freemangordon: Why's the new kernel break the mp?
Jun 19 21:04:02 <freemangordon> Raimu: because mp depends on the exact kernel version 
Jun 19 21:04:13 <freemangordon> not on ( >= )
Jun 19 21:04:26 <Raimu> Drat.
Jun 19 21:05:52 <freemangordon> Raimu: well, take it easy, if there are still no crashes in the next 12 hours (i.e. till tomorrow evening) I will made a new update, everything is in place but mp- thingie
Jun 19 21:08:58 <freemangordon> Raimu: I don't want to distribute something I am not convinced it is stable. And it is really hard to hit that bug, seems noone but me (because of X) has trigger it.
Jun 19 21:11:03 <Raimu> Yep, I understand.. I only did wonder why put it in early.
Jun 19 21:13:00 <freemangordon> Raimu: because it takes some time to upload about 100 MBs to the repo, I prefer to have everything uplodaed in advance as I cannot plan my free time ;)
Jun 19 21:19:26 <esaym153> gdb is not in any repos for the n810?
Jun 19 21:24:57 <Raimu> freemangordon, touché :)
Jun 19 21:57:34 <vi__> stop everything.
Jun 19 21:57:39 <vi__> we have a winner.
Jun 19 21:58:01 <vi__> "Thumb2, better than speedpatch."
Jun 19 21:59:06 <Woody14619> lol!
Jun 19 21:59:38 * Woody14619 awards vi__ a trophy for his great find.
Jun 19 22:00:02 <Woody14619> I almost want to print that out and hang it on my cube wall...
Jun 19 22:03:36 <vi__> Woody14619: so is maemo.org toast?
Jun 19 22:04:38 <vi__> Woody14619: I have space on a webserver if that can help?
Jun 19 22:05:02 <Woody14619> I don't think so... We have some internal fighters still.  Nothing can be said officially, but I suspect we'll get the 6 months promissed at least, if not 12 or so.
Jun 19 22:05:08 <vi__> could probably run the forums from it and hold the wiki.
Jun 19 22:05:41 <Woody14619> And Reggie has offered a complete dump of the forum, as that's all his systems/servers/content....
Jun 19 22:06:13 <vi__> Woody14619: at the very least I WILL be hosting the firmware images and flasher program.
Jun 19 22:06:41 <Woody14619> What I worry about really is the garages and repos.  But I susspect that there are at least one or two mirrors out there already, and those will have lots of time to get stuff from before they close, if they're hit.
Jun 19 22:07:37 <vi__> I could probably hold a mirror as well.
Jun 19 22:08:07 <vi__> although I have NFC how to set that up.
Jun 19 22:08:10 <Woody14619> And, there's a rush on to get OBS up, running, and maybe now even packaged so individuals can get it (and the blobs needed from Nokia) to run their own setup.
Jun 19 22:09:07 <Woody14619> So, No... I don't think the end is neigh.  But even if it is, we're pretty close to being able to limp along on our own as a community if it happens.
Jun 19 22:10:08 <vi__> somehoe I do not see TMO becomming a subforum in another forum.
Jun 19 22:10:23 <Woody14619> The problems with mirrors are: Some of the blobs are not something you can redist (or even fetch for that matter), as their proprietary Nokia.  And even if you did, the bandwidth would kill just getting it, yet alone offering it to others.
Jun 19 22:10:25 <vi__> ^somehow
Jun 19 22:10:56 <Raimu> Chuckling at "Better than Speedpatch"
Jun 19 22:11:01 <Raimu> Seeing that'd kill me.
Jun 19 22:11:26 <vi__> freemangordon can stop now.
Jun 19 22:11:33 <Woody14619> Like OVI store... If Nokia goes away, those apps do too.  (Save the few that crafty folks may have figured out how to re-package)
Jun 19 22:11:35 <vi__> he has reached his zenith.
Jun 19 22:12:00 <DocScrutinizer06> fsck thumb
Jun 19 22:12:03 <vi__> Woody14619: fortunaltley there was nothing but shit in the ovi store.
Jun 19 22:12:20 <vi__> ^fortunatley.
Jun 19 22:12:20 <Woody14619> But vi__ it thumb-2 will never compete with the next version of speedpatch.  (Due out tomorrow, right? ;)
Jun 19 22:12:33 <DocScrutinizer06> 97% risk for 15% savings
Jun 19 22:12:43 * Woody14619 nods, a little...  There is Angry Birds in OVI store, after all...
Jun 19 22:13:00 * Woody14619 needs his Angrey Birds fix...
Jun 19 22:14:08 <Woody14619> I've seen worse compromises Doc. ;)
Jun 19 22:14:14 <vi__> DocScrutinizer06: risk of what?
Jun 19 22:14:33 <vi__> Woody14619: 14619 is your zipcode?
Jun 19 22:14:47 <Woody14619> And I think the 97% risk is a tad high... Since so far the patch has worked 100%.
Jun 19 22:14:55 * Woody14619 nods to vi__
Jun 19 22:15:27 <Woody14619> You know now where I (and ~35K other people) live.
Jun 19 22:16:28 <vi__> the odds in favour of your anonimity just got a whole shitload slimmer.
Jun 19 22:16:39 <Woody14619> .oO(If you had the +4, you could come pee on my lawn...)
Jun 19 22:17:22 <Woody14619> Well, considering my city/state is in my profile, and there are only 5 zips in the city... not by much. :)
Jun 19 22:18:07 <vi__> your foolishness with personal privacy has allowed me to rule out the entire southern hemisphere.
Jun 19 22:18:17 <vi__> you cannot hide forever woody.
Jun 19 22:18:53 <DocScrutinizer06> vi__: risk of 69% users fsckd
Jun 19 22:19:00 <Woody14619> Pleh... You'd have better luck tracking my IP back through IRC... that makes me one of 300 people working for the company.
Jun 19 22:19:34 <Woody14619> How so, DocScrutinizer06?
Jun 19 22:19:51 <MrPingu> What's wrong with thumb2? I have been using it without any problems, so far
Jun 19 22:19:58 <MrPingu> Note: I started fresh
Jun 19 22:20:15 <DocScrutinizer06> the patches are not verified to work against "normal apps"
Jun 19 22:20:33 <MrPingu> Define "normal apps"
Jun 19 22:20:42 <DocScrutinizer06> BlessN900
Jun 19 22:20:43 <Woody14619> That's being done now, AFAIK... And really, lots of folks have been using "normal apps" for some time now.
Jun 19 22:20:59 <Woody14619> I'd hardly use BlessN900 as an example of "normal"
Jun 19 22:21:02 <DocScrutinizer06> MyACMEapp
Jun 19 22:21:23 <Woody14619> Considering it broke itself each time Nokiia released a new PR...
Jun 19 22:21:27 <DocScrutinizer06> Woody14619: though it is
Jun 19 22:21:38 <DocScrutinizer06> we all live in a bubble
Jun 19 22:21:57 <MrPingu> Is it tested on a fresh device?
Jun 19 22:22:01 <DocScrutinizer06> there are probably 2 mio sold N900
Jun 19 22:22:16 <Woody14619> Besides, it' not like this is going into Extras... It's going into a sub-repository behind another sub-repository.
Jun 19 22:22:27 <DocScrutinizer06> users in this chan: 278
Jun 19 22:22:37 <Woody14619> You have to enable CSSU, then enable CSSU-testing.
Jun 19 22:22:49 <DocScrutinizer06> Woody14619: and I'm all fine with that
Jun 19 22:22:57 <Woody14619> 90% of the world hasn't even turned on Extras, as it's off by default.
Jun 19 22:23:09 <Jaffa> Woody14619: No, it's on by default in Fremantle
Jun 19 22:23:14 <DocScrutinizer06> incorrect, since PR1.2
Jun 19 22:23:14 <Woody14619> Maybe 1% have CSSU installed.
Jun 19 22:23:24 <vi__> Woody14619: and how many fucking spacker mongs have installed T2 because they think it is just the next upgrade release...
Jun 19 22:23:27 <vi__> ?
Jun 19 22:23:42 <vi__> there be a shit storm a brewin'
Jun 19 22:23:44 <Woody14619> Only as many as complain on TMO... so.. 4. :)
Jun 19 22:23:51 <MrPingu> Pick me
Jun 19 22:23:57 <Woody14619> Again, this assumes it breaks things.
Jun 19 22:24:09 <Woody14619> Which so far it has not.
Jun 19 22:24:10 <MrPingu> No, I exactly knew it's testing
Jun 19 22:24:20 <DocScrutinizer06> vi__: no way
Jun 19 22:24:31 <Woody14619> And it's being tested by plenty of folks right now...
Jun 19 22:25:04 <DocScrutinizer06> plenty by which metrics?
Jun 19 22:25:21 <MrPingu> Really I haven't had any problem so far :)
Jun 19 22:26:00 <DocScrutinizer06> WFM is no argument when it comes to Millions of deployed systems
Jun 19 22:26:27 <Woody14619> So, yes, will someone enable CSSU, CSSU-Testing, and then go grab zeebra-pron-0m5-thumb2 just because they can?  Sure.  But they'll already have speedpatch installed, so it's not like they'll notice if something else breaks.
Jun 19 22:26:45 <DocScrutinizer06> thimb is fine for those who want to gamble with it. It's no way ever an alternative for CSSU
Jun 19 22:27:03 <Woody14619> Doc, as I recall you used this same argument against CSSU and KP 6 months ago, no?
Jun 19 22:27:28 <DocScrutinizer06> I never used any argument against CSSU
Jun 19 22:27:46 <Woody14619> hmmm.. maybe it was chemist. :P
Jun 19 22:28:07 <DocScrutinizer06> KP however is breaking module compatibility and thus never qualifies for an recommended upgrade
Jun 19 22:28:21 <zeq_laptop> freemangordon: how did libqtm11/12 ever build? multimedia.pro contanis -march=armv7a instead of armv7-a, fixing that and there's a build failure at: qgraphicsvideoitem_maemo5.cpp
Jun 19 22:28:22 <Woody14619> There we go...
Jun 19 22:28:57 <DocScrutinizer06> BlessN900 (and nnn other commercial apps) need kernel module compatibility
Jun 19 22:29:21 <vi__> brb
Jun 19 22:29:28 <DocScrutinizer06> we're not doing CSSU for the 278 in here
Jun 19 22:29:30 <Woody14619> Then if you want BlessN900, don't install KP.  Simple.
Jun 19 22:29:40 <DocScrutinizer06> indeed
Jun 19 22:30:00 <DocScrutinizer06> you're however allowed and encouraged to install CSSU
Jun 19 22:30:21 <Woody14619> No, not any more than you are KP.
Jun 19 22:30:22 <DocScrutinizer06> and that's why CSSU will never include a new kernel
Jun 19 22:30:28 <vi__> DocScrutinizer06: no one else who is left still using N900 care about cssu etc.
Jun 19 22:30:36 <Woody14619> I know far more people that overclock via KP than use CSSU... Just saying.
Jun 19 22:30:51 <DocScrutinizer06> doesn't matter
Jun 19 22:30:56 <vi__> ^that is not already on TMO etc.
Jun 19 22:31:02 <vi__> ~tmo
Jun 19 22:31:03 <infobot> [tmo] http://en.wikipedia.org/wiki/TMO, or http://talk.maemo.org, or http://de.wikipedia.org/wiki/Terrestrial_Trunked_Radio#TMO. It's *not* T-MO (see ~T-MO) or trolls, morons, oxen.
Jun 19 22:31:05 <DocScrutinizer06> CSSU is official SSU 
Jun 19 22:31:13 <DocScrutinizer06> by community
Jun 19 22:31:23 <DocScrutinizer06> for *everybody*
Jun 19 22:31:42 <jacekowski> ~t-mo
Jun 19 22:31:43 <infobot> well, t-mo is T-Mobile
Jun 19 22:31:53 <jacekowski> ~maemo
Jun 19 22:31:55 <Woody14619> I'm still trying to see your point here Doc.   Thumb, AFAIK, is going to be in a separate sub-repo, requiring KP >=50
Jun 19 22:32:11 <vi__> DocScrutinizer06: but your *everybody* includes joe shitmuncher who doesnt care about kp.  He does not exist anymore.
Jun 19 22:32:11 <DocScrutinizer06> no problem with me
Jun 19 22:32:50 <DocScrutinizer06> we're discussing null issues
Jun 19 22:32:55 <DocScrutinizer06> n8
Jun 19 22:33:17 <Woody14619> So in that reguard, it's not going to be *in* CSSU, nor require it as far as I can see.
Jun 19 22:34:05 <zeq_laptop> DocScrutinizer06: how's your c++?
Jun 19 22:34:25 <Woody14619> Eventually, someone will make zeebra-pron-0m5-thumb2 and make it only for thumb though... which will pull in the scruffy nerf hearders, like it always does.  But again, they have to jump through two flaming hoops to get there... so... If they burn, it's on em.
Jun 19 22:34:44 <DocScrutinizer06> Woody14619: ack
Jun 19 22:35:04 <Woody14619> .oO(That's what backupmenu is for, right?)
Jun 19 22:35:21 <DocScrutinizer06> not exactly
Jun 19 22:35:44 <DocScrutinizer06> BM can roll back, but nor create alternative branches
Jun 19 22:36:50 <zeq_laptop> attempting to rebuild libqtm11 due to missing pkgconfigs, hits this after fixing above mentioned -march=armv7a error: http://pastebin.com/yCbY1Xu7
Jun 19 22:36:50 <Woody14619> True, but that roll-back is essentially a re-flash, minus firmware and kernel (for now), so... 
Jun 19 22:36:58 <DocScrutinizer06> Woody14619: the basic axiom is: pick up as many users as possible where they are. With all their needs and restrictions they face
Jun 19 22:37:25 <Woody14619> For CSSU, yes... 
Jun 19 22:37:33 <DocScrutinizer06> yes, only for CSSU
Jun 19 22:37:45 <zeq_laptop> code is viewable here: http://qt.gitorious.org/qt-mobility/qt-mobility/blobs/2a2b7c24c714ea65c6cf732175fc60a1867dae08/src/multimedia/qgraphicsvideoitem_maemo5.cpp
Jun 19 22:37:46 <DocScrutinizer06> for PK we go bleeding edge
Jun 19 22:37:53 <DocScrutinizer06> for thumb
Jun 19 22:38:21 <MrPingu> That's why we have cssu-stable
Jun 19 22:38:24 <Woody14619> For KP and thumb, it's more about giving functionality to an older device for the ones that are going to stay on it when the new shiny comes out and the rest of the nerfs throw away their old device for a new one.
Jun 19 22:38:34 <DocScrutinizer06> MrPingu: stable and testing
Jun 19 22:38:56 <DocScrutinizer06> Woody14619: yes
Jun 19 22:39:13 <MrPingu> IMO: testing can be bleeding edge too
Jun 19 22:39:13 <zeq_laptop> I can't make out what's wrong with the code given the error. (if *anybody* can help)
Jun 19 22:39:27 <Woody14619> And frankly, with the age of the N900, lots of nerfs have left.  The community is quickly approaching non-nerf status.
Jun 19 22:39:28 <DocScrutinizer06> we of course should break new ground. But not on CSSU
Jun 19 22:40:10 <DocScrutinizer06> CSSU is the genuine continuation of Nokia SSU
Jun 19 22:40:25 <Woody14619> Why not?  And again, this is not something, last I knew, that was being rolled into CSSU.
Jun 19 22:40:31 <MrPingu> but I agree thumb2 supported kernel should be in PK, cssu-kernel doesn't belong in ssu
Jun 19 22:40:37 <Woody14619> It's reliant on KP50, not CSSU.
Jun 19 22:40:51 <MrPingu> Just for the very reason you mentioned earlier
Jun 19 22:41:14 <MrPingu> With PK v50 in extras
Jun 19 22:41:28 <Woody14619> zeq_laptop, and whats the error you're getting again?
Jun 19 22:41:28 <zeq_laptop> I'm with DocScrutinizer06, and I'm working on thumb related things.  There's no problem having both streams.
Jun 19 22:41:56 <zeq_laptop> http://pastebin.com/yCbY1Xu7
Jun 19 22:42:04 <MrPingu> Can extras-devel be used for a v51 with thumb ;)
Jun 19 22:42:09 <Woody14619> I agrees.. that's why I'm trying to figure out why Doc is so down on thumb?
Jun 19 22:42:42 <DocScrutinizer06> I'm just disappointed nobody verifies THUMB
Jun 19 22:42:52 <zeq_laptop> verifies?
Jun 19 22:43:24 <MrPingu> Idk, however "it's working fine for me said by Mrpingu (normal user #999)" insnt a scientific proof
Jun 19 22:43:25 <zeq_laptop> if you mean for stablility, we're doing that through active testing.
Jun 19 22:43:26 <Woody14619> Again, I think it's being verified.... 
Jun 19 22:44:12 <merlin1991> freemangordon: you killed the thumb repo :
Jun 19 22:44:15 <merlin1991> *:P*
Jun 19 22:44:16 <DocScrutinizer06> verifies as in "deliver *proof* is's a fix"
Jun 19 22:44:17 <MrPingu> OTOH, if there are many working fines we can render it tested...
Jun 19 22:44:31 <DocScrutinizer06> WFM is no proof
Jun 19 22:44:39 <Woody14619> There are "normal" apps in the OVI store that break constantly that obviously got less "verification" than thumb is already.
Jun 19 22:44:40 <merlin1991> reprepro by default removes older verions
Jun 19 22:44:50 <merlin1991> and your kernel update you uploaded removed the old one xD
Jun 19 22:44:52 <DocScrutinizer06> too many savvy people dispised thumb
Jun 19 22:45:22 <DocScrutinizer06> you're not rehabilaizing it with "WFM"
Jun 19 22:45:31 <Woody14619> Doc: many savy people also dispised overclocking.... Yet it's a pretty common practice there days...
Jun 19 22:45:39 <zeq_laptop> DocScrutinizer06: that's why thumb2 was developed
Jun 19 22:45:55 <zeq_laptop> original thumb was god awful
Jun 19 22:46:04 <DocScrutinizer06> and we seen pretty a number of fuckedup devices
Jun 19 22:46:20 <MrPingu> Of OC?
Jun 19 22:46:28 <DocScrutinizer06> with NONe OF THEM been caused by OC
Jun 19 22:46:37 <DocScrutinizer06> ;-P
Jun 19 22:47:03 <MrPingu> I guess it's more because user fault, not knowing what they are doing
Jun 19 22:47:18 <MrPingu> While OC can be stable if done right ;)
Jun 19 22:47:27 <DocScrutinizer06> 95% of users silently binned the N900 when it went apeshit due to OC
Jun 19 22:47:42 <zeq_laptop> DocScrutinizer06: evidence?
Jun 19 22:48:01 <DocScrutinizer06> MrPingu: nope, OC is like running your car engine with two times the RPM it's designed for
Jun 19 22:48:06 <MrPingu> Now SR is fixed  it's even easier for end-users to use good OC-settings
Jun 19 22:48:10 <DocScrutinizer06> it will wear out
Jun 19 22:48:15 <Woody14619> Doc, I've had my device overclocked since before Titan put out kernels... No damage yet.
Jun 19 22:48:50 <MrPingu> I run however for 2 weeks without OC and it's doing fine actually :P
Jun 19 22:48:52 <zeq_laptop> underclocking and mild overclocking doesn't wear it out any quicker
Jun 19 22:48:53 <DocScrutinizer06> Woody14619: I run my car around teh block at 22800RPM once a week since 5 yreas
Jun 19 22:49:05 <zeq_laptop> I meant *unvolting*
Jun 19 22:49:11 <Woody14619> And yes, it's like over reving your engine.  But then I see little Honda Fit's every day with a custom huge engine and jacked up tires...  
Jun 19 22:49:30 <DocScrutinizer06> zeq_laptop: aha, and you know that because you designed the OMAP chip?
Jun 19 22:49:33 <zeq_laptop> it's the current and heat that kills
Jun 19 22:50:05 <zeq_laptop> I know that from the principle of electron migration
Jun 19 22:50:15 <DocScrutinizer06> aha, you again know that because you invented integrated circuits?
Jun 19 22:50:38 <Woody14619> Doc: The point being, if YOU don't want that, then by all means, don't YOU install it.  But to poo-poo people doing work to make life better for those that want to take that (far less than 1%) risk... seems a tad silly.
Jun 19 22:51:03 <MrPingu> Anyway I think OC is not that bad as you propose it is...
Jun 19 22:51:19 <zeq_laptop> It is bad if you just ramp up the clockspeed/voltage
Jun 19 22:51:21 <DocScrutinizer06> Woody14619: I'm not doing any of that
Jun 19 22:51:39 <DocScrutinizer06> I'm just not taking shit from amateurs
Jun 19 22:51:43 <Woody14619> Especially when it's not being tied to or injected into anything "stable" or main-stream.
Jun 19 22:51:55 <MrPingu> I won't discuss it anyfurther as I can't come up with good arguments ;)
Jun 19 22:52:09 <MrPingu> My knowledge is just too low :P
Jun 19 22:52:31 <DocScrutinizer06> that's the point
Jun 19 22:53:05 <MrPingu> That's why I said: I Think..
Jun 19 22:53:08 <DocScrutinizer06> I'm building chips for bagles now
Jun 19 22:53:26 <Woody14619> Doc: Galileo was considered and amateur astrologer in his day.  People poo-pooed his theories all the time... but turns out he was right in the end on most of it.
Jun 19 22:53:39 <DocScrutinizer06> I see the results of test our engineers at chipfab do
Jun 19 22:53:44 <Woody14619> Being and ameture != always wrong.
Jun 19 22:54:10 <zeq_laptop> DocScrutinizer06: and what do they tell you?
Jun 19 22:54:19 <MrPingu> Woody14619 you give a bit of confidence haha :P
Jun 19 22:54:31 <DocScrutinizer06> they tell me it's 49-dimesional problem
Jun 19 22:54:51 <Woody14619> Sounds reasonable to me. :)
Jun 19 22:54:54 <DocScrutinizer06> and things like UV and temperaturee are just 2 of those dimensions
Jun 19 22:55:57 <DocScrutinizer06> you goddomn dunno what's gonna happen when you OC, but odds are it WILL hirt your chip
Jun 19 22:56:07 <Woody14619> And yet, they can make a chip that runs at 500Mhz with a reasonable MTBF...  And you believe them when they say that, but not when someone say, if we up this a little, we can get it to run at 600 without much life loss.
Jun 19 22:56:13 <zeq_laptop> yep, current, cosmic rays, quantum events ...
Jun 19 22:56:48 <DocScrutinizer06> Woody14619: EM is a e^x function
Jun 19 22:56:54 <Woody14619> I don't think anyone argues that OC doesn't have potential to lower the life of the chip..
Jun 19 22:57:05 <MrPingu> How many of these fscked device were caused ONLY by OC?
Jun 19 22:57:13 <zeq_laptop> as undervolting increases it
Jun 19 22:57:25 <DocScrutinizer06> kicks in at 528, at 533 it hits the roof, at 600 it skyrockets
Jun 19 22:57:51 <Woody14619> It clearly does.  But the failure rate on the chip is ~5 years.  If OC lessens that life by a year, but lets you actually continue to USE the device another 2 vs getting a faster one when you "need" it, what's lost?
Jun 19 22:58:18 <DocScrutinizer06> omap has EOL of 100k@500 and 10k@600
Jun 19 22:58:29 <DocScrutinizer06> guess what it has at 700
Jun 19 22:58:48 <Woody14619> Would you sooner not have the OC option, and have people dump the device like the GTA02 when something faster comes along?
Jun 19 22:59:13 <Woody14619> Or have the option to OC and have that device be viable for another year, vs in a land-fill.
Jun 19 22:59:15 <zeq_laptop> DocScrutinizer06: they are mean failure times
Jun 19 22:59:27 <DocScrutinizer06> I just hate dudes that think they can argue from their primary scool math
Jun 19 22:59:32 <zeq_laptop> it could be *worse* ;)
Jun 19 22:59:42 <DocScrutinizer06> zeq_laptop: yep
Jun 19 23:00:25 <DocScrutinizer06> actually it's 15 of 1000 assumed dead after 10k hours @ 600
Jun 19 23:00:37 <DocScrutinizer06> at extended temp range
Jun 19 23:00:53 <zeq_laptop> do they have the stats for undervolting?
Jun 19 23:01:04 <DocScrutinizer06> NOBODY of you all did those testes
Jun 19 23:01:06 <Woody14619> And 10K @600 takes how long for a typical user?  Considering it runs 95% of it's life at 0...
Jun 19 23:01:09 <zeq_laptop> I am genuinely curious
Jun 19 23:01:09 <DocScrutinizer06> or sis anybody?
Jun 19 23:01:22 <DocScrutinizer06> did*
Jun 19 23:02:05 <Woody14619> Doc: yes... TI did.  Which is why they sold them as 500Mhz devices and not 700Mhz.  Their ramps are in the chip spec.  And are not that dramatic.
Jun 19 23:02:29 <DocScrutinizer06> Woody14619: you can cosider whatever you like. Your neighbour, watching lots of vidoes via mplayer w/o DSP accel, may have different results. WERY different
Jun 19 23:02:31 <Woody14619> And Titan DID project out that curve, which is why his suggestion was not to go over 850.
Jun 19 23:02:53 <zeq_laptop> 805 I believe
Jun 19 23:03:00 <MrPingu> Is our chip in N900 originally clocked at 500 and overclocked to 600 by Nokia?
Jun 19 23:03:25 <Woody14619> Doc: That same neighbor playing videos on stock is going to get the same life out of it, if not less, when he can't watch 720p and goes and gets a new Android device after a year.
Jun 19 23:03:51 <MrPingu> Or wat that a random example, Woody14619?
Jun 19 23:03:58 <DocScrutinizer06> MrPingu: yep
Jun 19 23:04:37 <DocScrutinizer06> MrPingu: TI recommendations: 100kh @OPP1..4 15kH@OPP5
Jun 19 23:04:40 <DocScrutinizer06> iirc
Jun 19 23:04:42 <Woody14619> Again... if the *usable* lifetime of a device can be extended via OC, you've lost nothing.
Jun 19 23:05:36 <Woody14619> If the option is get a new device or OC the one you have to keep up, one makes more land fill.  Who cares if it lessens the life of the device which you would have otherwise thrown away?
Jun 19 23:05:44 <DocScrutinizer06> Woody14619: then you need to argue on THAT basis, and you'll not see any post from me
Jun 19 23:06:06 <Woody14619> At > 2 years post launch, I think that's pretty much the given at this point.
Jun 19 23:06:16 <DocScrutinizer06> arguing "OC is safe as we don't know for sure"... MEH
Jun 19 23:06:39 <MrPingu> Besides shorter life time of chip are there more considerable negative effects?
Jun 19 23:06:47 <zeq_laptop> This reminds me, the very laptop I'm on a Dell Lattitude, from Dell comes with BIOS voltages above std values for the chip in order to reduce initial failure rates from marginally defective silicon!
Jun 19 23:06:48 <DocScrutinizer06> no
Jun 19 23:06:56 <DocScrutinizer06> MrPingu: ^^^
Jun 19 23:07:16 <DocScrutinizer06> shorter battery standby time maybe
Jun 19 23:07:46 <Woody14619> .oO(Now you're opening the door for "race to finish" debates... ;)
Jun 19 23:07:51 <DocScrutinizer06> MrPingu: however chip usually doesn't cease to work, it just "segfaults" every more often
Jun 19 23:08:10 <DocScrutinizer06> Woody14619: not really
Jun 19 23:08:15 <Woody14619> And/or nukes your flash data.
Jun 19 23:08:24 <DocScrutinizer06> yep, that too
Jun 19 23:08:24 <ds3> are there currently any reliable source for replacement digitizers on the N900?
Jun 19 23:08:40 <DocScrutinizer06> saremi
Jun 19 23:08:47 <Woody14619> some would say OC would allow longer standby, as what work is being done finishes faster.
Jun 19 23:08:53 <ds3> saremi is a store/site/person?
Jun 19 23:09:03 <MrPingu> Hmm, don't buy from phf  (germany)
Jun 19 23:09:15 <zeq_laptop> Woody14619: that there is actually some research on
Jun 19 23:09:25 <DocScrutinizer06> Woody14619: I know "race to finish" concept
Jun 19 23:09:45 <DocScrutinizer06> MrPingu: shop
Jun 19 23:09:51 <Woody14619> As do I.  At least 2 tests were done (one by vi__, one by Titus) that shows the "optimal" speed near 500Mhz.
Jun 19 23:10:09 <MrPingu> Yet, they are offering to pay back shippingcosts twice
Jun 19 23:10:14 <Woody14619> So, that's a good argument for a min-clock of 500.  Though not one for overclock per-se. ;)
Jun 19 23:10:48 <MrPingu> Though I have lost a bit of trust...
Jun 19 23:11:16 <DocScrutinizer06> Woody14619: 500? surprise surprise ;-P
Jun 19 23:11:29 <Woody14619> Almost like it was designed to go that fast? ;)
Jun 19 23:11:56 <ds3> DocScrutinizer: http://www.saremi-mobilfunk.de/ <--- is this the site you are suggesting?
Jun 19 23:12:18 <DocScrutinizer06> Woody14619: there's another aspect: thermal design
Jun 19 23:12:36 <MrPingu> Thinking of just ignoring and buy it from saremi, like you suggested, doc
Jun 19 23:12:36 <MrPingu> Or just hope I get a genuine one after this time...
Jun 19 23:12:42 <DocScrutinizer06> all Nokia hw engineers warned us to lock CPU at even 600
Jun 19 23:12:44 <Woody14619> Again, we're in violent agreement on this... :P
Jun 19 23:12:47 <MrPingu> Or send it back, and ask for refund and then buy from saremi :P
Jun 19 23:13:20 <MrPingu> Choices, choices... :P
Jun 19 23:13:48 <Woody14619> Except that, I say, if people want to OC, let them.  Where you seem to say, don't bother with OC (or thumb, or other misc micro-enhacement) as it won't be used by the masses.  
Jun 19 23:14:18 <ds3> whoa... $120
Jun 19 23:14:18 <Woody14619> If that were the case, items like the N900 (and the GTA* series, and smart cars, and ....) wouldn't exist.
Jun 19 23:14:35 <DocScrutinizer06> no, I hate it as I feel bad buying used N900
Jun 19 23:15:33 <ds3> was there 2 entirely different releases of the N900? the 'original' N900 stuff do not match what I got
Jun 19 23:15:34 <DocScrutinizer06> aI'm perfectly happy with somebody deliberatly frying his N900 @ 1.2GHz for numer crunching, as he needs his PIN for credit card
Jun 19 23:16:15 <Woody14619> I picked up a refurb from a Chinese distributer a couple months ago for about that...   Old guts, new kb, case, stylus, battery, etc.
Jun 19 23:16:28 <Woody14619> No ds3... If the stats are not the same then it's a knockoff.
Jun 19 23:16:29 <DocScrutinizer06> I hate tose 'experts' suggesting Nokia and TI don't know shit and "WFM, since 7 months now!!!1!1!!1eleven"
Jun 19 23:16:59 <Woody14619> I don't think anyone actually involved in OC have ever said that...
Jun 19 23:17:24 <MrPingu> Ds3, be aware there ARE fake digitizers
Jun 19 23:17:25 <DocScrutinizer06> if you're aware and accepting you're frying your CPU by OC, go on do it if you think it's worth it
Jun 19 23:17:33 <ds3> Woody14619: the original USB cables I got had a different overmold then pictured... also I didn't get a micro to old style charger adapter... mine came from Dell so I wonder if the US had different stuff
Jun 19 23:17:35 <Woody14619> Why blame them for something a nerf says after locking their device on1100Mhz?
Jun 19 23:17:36 <zeq_laptop> Am I the only one who thinks undervolting is more significant anyway?
Jun 19 23:17:42 <ds3> MrPingu: sigh.
Jun 19 23:17:55 <ds3> I should just go learn to manufacture my own 
Jun 19 23:17:56 <DocScrutinizer06> ~oc
Jun 19 23:17:58 <infobot> from memory, oc is an Optical Carrier, An OC1 has 672 channels or 44.736 Mbit/s
Jun 19 23:18:09 <DocScrutinizer06> ~omap-oc
Jun 19 23:18:10 <infobot> extra, extra, read all about it, omap-oc is http://mg.pov.lt/maemo-irclog/%23maemo.2010-08-01.log.html#t2010-08-01T22:16:05  read that!, or and this http://mg.pov.lt/maemo-irclog/%23maemo.2011-03-11.log.html#t2011-03-11T03:04:11
Jun 19 23:18:27 <Woody14619> zeq_laptop: No... I agree, especially enabling SR1/SR2, which Nokia just didn't want to put the effort into to optimize.
Jun 19 23:18:31 <DocScrutinizer06> ~factinfo oc
Jun 19 23:18:31 <infobot> oc -- created by jameswf-home <n=james@ip72-204-221-181.ph.ph.cox.net> at Mon Feb 25 00:03:20 2008 (1576 days); it has been requested 13 times, last by DocScrutinizer06, 34s ago.
Jun 19 23:18:45 <DocScrutinizer06> ~overclocking
Jun 19 23:18:46 <infobot> "OK, listen up.  This is your CPU."  apt drops the CPU into a hot frying pan.  "This is your CPU on overclocking.  Any questions?"
Jun 19 23:19:03 <Woody14619> ds3: yes, the US one came with the converter (from Nokia USA anyway)
Jun 19 23:19:41 <DocScrutinizer06> thanks apt!
Jun 19 23:19:49 <MrPingu> As much as I like to hang aroun any longer here, I have to get some sleep
Jun 19 23:19:52 <MrPingu> Bye!
Jun 19 23:19:54 <DocScrutinizer06> ~botsnack
Jun 19 23:19:55 <infobot> DocScrutinizer06: :)
Jun 19 23:20:23 <zeq_laptop> MrPingu: you're not alone ... goodnight
Jun 19 23:20:35 <DocScrutinizer06> n8, now for real
Jun 19 23:20:48 <Woody14619> doc: I think really, since the package to install warns of just that, that it's a known....
Jun 19 23:21:19 <DocScrutinizer06> you'd start to wheep if you'd seen what I've, in this chan
Jun 19 23:21:50 <zeq_laptop> Woody14619: I would agree, except why the hell does it go upto 1150?!?
Jun 19 23:22:13 <DocScrutinizer06> don't assume any degree of stupidity at $user, you're underestimating
Jun 19 23:22:40 <zeq_laptop> it really should be limited to something like 805.
Jun 19 23:22:45 <ds3> Hmmm
Jun 19 23:22:50 <Woody14619> again... If I put out the super-gizmo-app that lets you make your system X times faster at the cost of 1/X of it's life, and state that in the installer...  It seems wrong to punish or rant becuase someone make X=99 for their device and then whines.
Jun 19 23:23:24 <DocScrutinizer06> nobody rants
Jun 19 23:23:30 <Woody14619> zeq: And who decides that?  You?  Council?  Technically, Nokia did decide that... at 600.
Jun 19 23:23:41 <Woody14619> doc: I disagree... scroll up... ;)
Jun 19 23:23:43 <DocScrutinizer06> [20.06.2012 01:17:23] <DocScrutinizer06> if you're aware and accepting you're frying your CPU by OC, go on do it if you think it's worth it
Jun 19 23:24:16 <zeq_laptop> Woody14619: it's so far out of spec for any random omap
Jun 19 23:24:20 <Woody14619> And again, it's part of the package install... it pops up a user UI that you must click OK on to install.
Jun 19 23:24:39 <Woody14619> zeq_laptop: And yet some can handle it...
Jun 19 23:24:47 <zeq_laptop> for a while
Jun 19 23:24:51 <Woody14619> yup.
Jun 19 23:25:09 <Woody14619> Again.... X times the speed, 1/X times the life.
Jun 19 23:25:09 <DocScrutinizer06> ok, I (eben *I*) click OK so many times with just wondering what I did *after* I did
Jun 19 23:25:32 <DocScrutinizer06> but I agree we can't do any better
Jun 19 23:25:49 <DocScrutinizer06> and no, I don't think it should be limited to XXX
Jun 19 23:26:14 <Woody14619> Trust me... I regular TMO.... I've seen the wailing when someone installs CSSU and wants out, not realizing they need to reflash to go back.
Jun 19 23:26:43 <DocScrutinizer06> I just hope everybody clearly mentions 2it *will* shorten your CPU lifetime by a probably remarkable timespan, but otherise it seems to work"
Jun 19 23:26:48 <Woody14619> Despite it being in the repo description, and the main package update, and the wiki, and...
Jun 19 23:27:17 <zeq_laptop> Woody14619: that's because it isn't from Nokia, so how can it be *official*
Jun 19 23:27:47 <zeq_laptop> regarding CSSU
Jun 19 23:28:13 <DocScrutinizer06> there's just too many fools out there that use primary scool math and physics to claim they *know* OC is SAFE
Jun 19 23:28:17 <Woody14619> And all the documentation pretty much does, as does the installer.  Now, abill_uk, he'll argue till the death that it's 100%, but he also thinks Elop has cameras setup in his house spying on him to get new ideas for the new Maemo7 prototype due out next month...
Jun 19 23:28:54 * Woody14619 nods
Jun 19 23:29:19 * Woody14619 shrieks:  "There's someone on the internet, and their WRONG!"
Jun 19 23:29:26 <Woody14619> they're
Jun 19 23:29:26 <DocScrutinizer06> and me is a character that powers up the ICBM and MTHELs when reading that
Jun 19 23:29:37 <zeq_laptop> Woody14619: better the first time :)
Jun 19 23:30:05 <Woody14619> http://xkcd.com/386/
Jun 19 23:30:10 <zeq_laptop> DocScrutinizer06: yet they won't listen, because they *know*
Jun 19 23:30:32 <DocScrutinizer06> zeq_laptop: *sigh* I learnt you're right
Jun 19 23:31:09 <DocScrutinizer06> ~xkcd
Jun 19 23:31:10 <infobot> rumour has it, xkcd is http://xkcd.com/ or only the best webcomic of romance, sarcasm, math and language.
Jun 19 23:31:36 <Woody14619> And on that note.... I'm off.  Have a nice night/morning/whatever-damn-time-it-is-where-you're-at.
Jun 19 23:31:46 <zeq_laptop> bye Woody14619 
Jun 19 23:31:55 <DocScrutinizer06> nite fellas, I gotta design your next modem tomorrow
Jun 19 23:32:11 <zeq_laptop> goodnight DocScrutinizer06 
Jun 19 23:32:15 <Woody14619> .oO(and by gone I mean away... because really, to close the IRC window.. such a bother...)
Jun 19 23:32:45 * zeq_laptop closes the lid
Jun 19 23:33:35 <DocScrutinizer06> me mumbles "google U8500, novathor, lte"
Jun 19 23:38:06 <DocScrutinizer06> http://www.stericsson.com/products/L8540.jsp
Jun 20 00:14:45 <brzys> hey guyz
Jun 20 00:15:11 <brzys> im trying to mount img on my n900 but im getting an error: "mount: mounting /dev/loop0 on /mnt failed: Invalid argument", any ideas?
Jun 20 00:25:07 <cehteh> how do you try that?
Jun 20 00:25:26 <cehteh> filesystem not in kernel/ not loaded?
Jun 20 00:27:15 <brzys> mount -o loop,offset=3225 slitaz.img .slitaz
Jun 20 00:57:21 <cehteh> brzys: give the filesystem type
Jun 20 00:57:43 <cehteh> and modprobe the filesystem if you dont have so
Jun 20 00:58:24 <cehteh> autodetection may work to but iirc you need -t auto at least then (and it can be unreliable in few cases)
**** ENDING LOGGING AT Wed Jun 20 02:59:59 2012