IRC log of #maemo-ssu for Monday, 2014-05-05

*** kolp has quit IRC00:40
*** eugene_ has quit IRC02:28
*** okias has quit IRC03:08
*** dos1 has quit IRC03:32
*** xes has quit IRC03:49
*** jon_y has quit IRC03:52
*** LauRoman has joined #maemo-ssu04:37
*** silviof1 has joined #maemo-ssu05:01
*** silviof has quit IRC05:04
*** amiconn has quit IRC05:05
*** amiconn_ has joined #maemo-ssu05:05
*** amiconn_ is now known as amiconn05:05
*** LauRoman has quit IRC05:17
*** nox- has quit IRC06:00
*** silviof1 is now known as silviof06:14
*** amizraa4 has quit IRC06:25
*** radekp has joined #maemo-ssu06:48
*** amizraa4 has joined #maemo-ssu07:28
*** Milhouse has joined #maemo-ssu07:33
*** scoobertron has quit IRC08:32
*** luf has joined #maemo-ssu08:32
*** scoobertron has joined #maemo-ssu08:46
*** sailus has joined #maemo-ssu09:44
*** _rd has joined #maemo-ssu09:50
*** _rd has quit IRC10:10
*** _rd has joined #maemo-ssu11:34
*** _rd has quit IRC11:47
*** Milhouse has quit IRC12:02
*** Milhouse has joined #maemo-ssu12:15
*** jon_y has joined #maemo-ssu12:31
*** kolp has joined #maemo-ssu14:06
*** Milhouse has quit IRC14:06
*** freemangordon has quit IRC14:48
*** freemangordon has joined #maemo-ssu14:48
*** okias has joined #maemo-ssu14:57
*** NishanthMenon has joined #maemo-ssu15:29
*** LauRoman has joined #maemo-ssu15:31
*** sailus has quit IRC15:37
*** eugene_ has joined #maemo-ssu15:37
*** radekp has quit IRC16:05
*** eugene_ has quit IRC16:05
*** aap___ has quit IRC16:19
*** lizardo has joined #maemo-ssu16:46
*** eugene_ has joined #maemo-ssu16:47
*** Milhouse has joined #maemo-ssu16:54
freemangordonmerlin1991: needs some help with debian control files and preinst scripts16:57
freemangordon*need16:57
freemangordonfor some reason pre-depends does not work the way I was thinking16:58
freemangordonmerlin1991: I've put "Pre-Depends: kernel-cssu-flasher ( = 1:2.6.28-10cssu3 ) | kernel-power-flasher ( >= 1:2.6.28-10power51r1 )" in mp- control file, but kernel is not being flashed before the reboot17:02
freemangordonis there a way to force dpkg --configure kernel-cssu-flasher in mp- preinst script?17:03
*** eugene_ has quit IRC17:14
*** MohammadAG has quit IRC17:20
*** MohammadAG has joined #maemo-ssu17:27
*** BigBox___ has joined #maemo-ssu17:39
*** okias has quit IRC17:42
DocScrutinizer05freemangordon: I would prefer a script that in turn calls the package management17:43
freemangordonDocScrutinizer05: but how to execute that script?17:51
DocScrutinizer05well, isn't user supposed to click on the CSSU installer icon?17:51
*** eugene_ has joined #maemo-ssu17:51
freemangordonit is, but it looks hackish to me to install a krenel from cssu-enabler17:52
freemangordon*kernel17:52
DocScrutinizer05honestly? why?17:52
freemangordondunno17:52
freemangordonno particular reason besides my gut feelings :)17:53
DocScrutinizer05it is 'hackish' to force HAM into a scrupting environment17:53
DocScrutinizer05scripting*17:53
freemangordonwell, actually HAM supports scripting ootb17:54
freemangordonanyway, i'll consider cssu-enabler17:54
DocScrutinizer05unlike (some) other package managers HAM obviously doesn't support a particular sequence or priority of installation of multiple packages pending to get installed17:54
DocScrutinizer05and tbh I'd probably prefer wget and dpkg for installing kernel anyway17:55
DocScrutinizer05that's *my* guts feeling ;-)17:55
freemangordon:)17:55
*** BigBox___ has quit IRC17:55
kerioyeah17:57
*** scoobertron has quit IRC17:57
kerioshoehorning the kernel in dpkg is something nokia made up17:57
kerioand it's completely and utterly awful17:57
kerioespecially considering how it's supposed to work17:57
keriothe flasher predepends on the kernel because it'll flash the image and then delete it17:57
DocScrutinizer05:nod:17:59
keriothat's why i use uboot :317:59
keriono more flashing for me17:59
DocScrutinizer05they already did abuse .deb as script, basically18:00
DocScrutinizer05a double wrapper matroshka kernel installation scheme would be really way too much18:00
DocScrutinizer05rather unwrap the crap and rewrite/refactor it in plain shellscript. Or at least use the lowest level tool (I.E: dpkg, rather than HAM) directly from a shellscript, to install kernel and modules18:02
DocScrutinizer05with dpkg you have 100% control over sequence of installation18:03
keriowhy is the kernel even versioned like that18:04
keriodebian on sheevaplug has no on-nand kernel autoflashing, and for good reasons18:05
kerio...at least, i think so18:05
DocScrutinizer05but yeah, optimal control and same time getting rid of this abomination would be: wget the tarball, untar it to /lib/modules/<newkernel>/* and /tmp/kernel.bin, then call flasher manually for the kernel.bin, then reboot18:05
keriomeh, modules are fine like this too18:06
keriowhy do we even have to flash the kernel again?18:06
DocScrutinizer05sure18:06
keriocan't we just get ubifs support in uboot and use that?18:06
kerioor extfs18:06
DocScrutinizer05but flashing a NAND partition is sth you better do under very tight and close control18:06
kerioyeah18:06
DocScrutinizer05particularly when you first want to check if that's really what you wanna do ("check for uBoot or non standard kernel on NAND")18:07
keriowell, the pali flasher asks18:08
keriobut on deny, it just doesn't flash18:08
keriothen it goes as if it flashed18:08
DocScrutinizer05yeah, and it's a constant source of trouble18:08
freemangordonhmm, I guess I can (ab)use https://gitorious.org/community-ssu/hildon-application-manager/source/17c6a728be8dd9def69eced62880767733656713:doc/packaging.txt#L30718:08
keriopls dont18:08
kerio:c18:08
freemangordonhmm, why?18:09
DocScrutinizer05it's HAM specific aiui and will cause further crippling and obfuscation18:10
DocScrutinizer05KISS18:11
freemangordonok, cssu-enabler then18:14
*** luf has quit IRC18:29
*** eugene_ has quit IRC18:34
merlin1991freemangordon: iirc pre-depends needs no reboot19:26
merlin1991it just says installed and configured before installing this ...19:26
freemangordonmerlin1991: yes, but I wanted to exploit HAM's recovery functionality19:26
freemangordonso, if the kernel is flashed before the device reboots because of the meiiens running kernel, it will finish the installation successfuly after the reboot19:27
freemangordonI tested it, If I flash the kernel after ther reboot by hand, it finishes ok19:27
*** amizraa4 has quit IRC19:27
freemangordonbut anyway, I've already hacked into the enabler19:28
*** amizraa4 has joined #maemo-ssu19:28
*** dos1 has joined #maemo-ssu19:44
*** amizraa4 has quit IRC20:00
*** amizraa4 has joined #maemo-ssu20:00
*** M4rtinK has joined #maemo-ssu20:24
*** MohammadAG has quit IRC20:24
*** MohammadAG has joined #maemo-ssu20:31
DocScrutinizer05better that20:52
DocScrutinizer05HAM messing around with kernel is questionable anyway20:53
freemangordonwhat do you think about https://gitorious.org/community-ssu/community-ssu-enabler/commit/751c577e9998e4856f40365ba4f94a93cc825c0c?20:53
DocScrutinizer05may I start ranting?20:55
freemangordonsure :)20:55
DocScrutinizer05*never* use $? a few lines *after* where you suspect it been set20:55
freemangordonhmm?20:55
freemangordonI know that, but all of the checks are just after the line I need the reslt from20:56
DocScrutinizer0594/11020:56
freemangordonoh, shit20:56
DocScrutinizer05yeah :-P20:57
DocScrutinizer05usually you do >' if apt-get -y --force-yes install kernel-cssu kernel-cssu-modules kernel-cssu-flasher; then '<  rather than >' apt-get -y --force-yes install kernel-cssu kernel-cssu-modules kernel-cssu-flasher: if  [ "$?" = "0" ]; then '<20:59
DocScrutinizer05at least that's what I'd do21:00
freemangordonI am not exactly a scripting guru, but I guess it is a matter of coding style21:00
freemangordonand IMO "my" way is more readible21:01
DocScrutinizer05usually not21:02
freemangordoncould be, I am not competent enough on the matter to argue21:02
DocScrutinizer05you're even doing a string compare to a numeric return value, not that it matters much21:02
freemangordonyou mean the quotes? that's because I am always afraid some variable could become unset and the script will abort :)21:03
DocScrutinizer05compare = vs -eq21:04
freemangordonbetter? https://gitorious.org/community-ssu/community-ssu-enabler/commit/08749748a03ea9dcd3fb9cc011552986d7a904b521:05
DocScrutinizer05sure21:06
DocScrutinizer05jr@saturn:~/.kde4/share/config> test 0 -eq \000 ; echo $?21:08
DocScrutinizer05021:08
DocScrutinizer05jr@saturn:~/.kde4/share/config> test 0 = \000 ; echo $?21:09
DocScrutinizer05121:09
DocScrutinizer05but I don't like the whole code21:11
DocScrutinizer05if  [ "$?" = "0" ]; then A; B; C; D; if foo; then bar; fi; E; F; else echo error occured; exit 999; fi21:12
DocScrutinizer05HUH????21:12
freemangordonwtf is that?21:13
DocScrutinizer05if  [ $? -ne 0 ]; then echo error occured; exit 999;       A; B; C; D; if foo; then bar; fi; E; F;21:13
freemangordonaah21:13
freemangordonif it is c code, you can bet I'll write it like that :)21:14
*** xes has joined #maemo-ssu21:14
DocScrutinizer05ooops21:15
DocScrutinizer05if  [ $? -ne 0 ]; then echo error occured; exit 999; fi;       A; B; C; D; if foo; then bar; fi; E; F;21:15
freemangordonbut I don;t feel comfortable with ba/sh scripts so I do it just to make it work21:15
DocScrutinizer05then make that:21:16
DocScrutinizer05if  ! SomeLeeteStuffToDo op tion; then echo error occured; exit 999; fi;       A; B; C; D; if foo; then bar; fi; E; F;21:16
DocScrutinizer05now guess what's more readable21:17
DocScrutinizer05saves you from worrying about unset envs as well ;-D21:18
freemangordonI'd rather do a new -thumb release21:18
DocScrutinizer05I ,ight want to do a tiny fancy and refactor the code21:19
DocScrutinizer05for you ;)21:19
freemangordonsure, go ahead21:20
kerio"if apt-get" is way more idiomatic imo21:20
kerioespecially if it's a mode specifically meant to be used for the return value21:20
DocScrutinizer05exactly21:20
freemangordonsee, I followed the "coding style" of the already existed code21:20
DocScrutinizer05yeah, dunno who wrote it21:21
DocScrutinizer05fugly ;)21:21
DocScrutinizer05anyway, I'll eventually do a refactoring if I'm getting bored21:21
DocScrutinizer05dafaq, where's the full text instead of diff?21:22
freemangordonhttps://gitorious.org/community-ssu/community-ssu-enabler/source/afae0ce43e559fef8809ef1a266a991f169b9e8f:community-ssu-enabler21:23
DocScrutinizer05thanks, don't hold your breath, but eventually...21:23
freemangordonsure, esp that I am fine with it as it is :)21:24
DocScrutinizer05cleanup() is not trapped21:26
DocScrutinizer05indentation is foobar21:26
freemangordonyeah, identation is irritating21:26
freemangordonok, so HAM, H-D and libc6. what else?21:27
freemangordonkerio: ^^^? any other package -thumb misses?21:27
kerioidk21:27
kerioall of them21:27
freemangordonheh21:27
freemangordonI plan a rather small update until merlin1991 finally finds some time and makes the next -testing21:28
merlin1991yeah what's in scope for that?21:29
*** _rd has joined #maemo-ssu21:29
freemangordoncityifno, libc6, ham, h-d, modest, clock-ui21:30
freemangordonand few more21:30
freemangordonmerlin1991: check in http://maemo.merlin1991.at/cssu/community-devel/changes.log21:30
freemangordonooh, pulseaudio and pa-policy-enforcement21:31
freemangordonalso, statusarea-applet-battery21:31
freemangordonfmtx-middleware21:31
freemangordondunno about ke-recv, but I think it is still not ready21:32
freemangordonhave to ask Pali21:32
freemangordonhmm, 2014-03-19 04:33:22 add fremantle deb free armel ke-recv 3.19-17.121:32
freemangordonmaybe it is ready after all21:32
kerioif it wasn't thumbified, i'm running it and it's working21:33
freemangordonit is not21:33
freemangordoniirc21:33
keriothen i have no clue21:33
freemangordonhmm, it is21:33
freemangordonke-recv (>= 3.19-15.7+thumb0),21:33
kerioi have thumb over devel21:34
freemangordonwell, check your installed version21:34
keriobut my phone is all the way over *there*21:34
DocScrutinizer05freemangordon: are you sure this script shouldn't start with something like >' if [ "$(id -u)" != "0" ]; then exec sudo "$@"; fi; '<  ?21:38
freemangordonit will fail anyways if started as user21:39
freemangordonbut maybe a check is really needed21:39
DocScrutinizer05and it sure like hell witll start as user when user clicks icon21:39
freemangordonno, it is started with sudo, from the .desktop file21:40
DocScrutinizer05aaah, ok21:40
DocScrutinizer05I'd rather make the script aito-sudo itself, but that's for sure a matter of style21:40
DocScrutinizer05auto*21:40
*** _rd has quit IRC21:41
DocScrutinizer05consider setups where sudo egenrally needs root password entry, like mine21:41
DocScrutinizer05generally*21:42
kerioDocScrutinizer05: if it's well-done it'll add a sudoers file21:43
freemangordonit adds21:43
xeswho takes care about pulseaudio? Someone has ideas to fix the random  "Master sink "sink.hw0" not found" with wrong audio routing during calls?21:46
kerioomg it's xes!21:47
xeshello there!21:47
freemangordonxes: usually Pali is the PA man21:48
DocScrutinizer05hi xes21:49
DocScrutinizer05that PA error report is unique afaik21:50
DocScrutinizer05anyway CND21:50
freemangordon~cnd21:51
DocScrutinizer05maybe related to RE'ed policy enforcer LISP files?21:51
DocScrutinizer05Can Not duplicate21:51
freemangordonI doubt , those are in "production" for really long time21:51
DocScrutinizer05:nod:21:51
DocScrutinizer05thus I wonder what's wrong with xes' system21:51
DocScrutinizer05hw failure?21:52
freemangordonxes: anything in dmesg?21:52
DocScrutinizer05cmt SSI pin popper?21:52
freemangordonsmells like21:52
freemangordon(hw failure)21:52
DocScrutinizer05:nod:21:52
DocScrutinizer05sink.hw0 sounds like SSI interface to cmt21:52
DocScrutinizer05McBSP<foobar>21:53
xesthe error happens on 50% of reboots21:53
freemangordondid you try a reflash?21:53
DocScrutinizer05rrright21:53
xestried all the fixes found in tmo about miissing/optified libs21:54
DocScrutinizer05please check with stock fremantle21:54
freemangordon:nod:21:54
DocScrutinizer05BM is your friend21:54
xeswhen is does not happen the phone runs well for weeks untill ...reboot21:54
DocScrutinizer05hmmmmmm21:54
DocScrutinizer05weiiiiird21:54
DocScrutinizer05wasn't there another similar report about something with audio crapping pants during boot recently?21:55
DocScrutinizer05somewhere on tmo maybe?21:55
xesi will prepare a dmesg log to post. Also the process start order is the same between working/not working situation21:55
DocScrutinizer05mem allocation might differ nevertheless21:56
*** nox- has joined #maemo-ssu21:56
DocScrutinizer05also race conditions are a nightmare to debug in boot/sysinit21:56
DocScrutinizer05at OM we had a kernel that worked great as long as we had debug output enabled21:57
DocScrutinizer05as soon as we switched off debug it crashed hard21:57
freemangordonxes: are you on -thumb?21:57
DocScrutinizer05nobody ever found the reason21:57
DocScrutinizer05yes, that was a report about thumb crashing audio or sth21:58
DocScrutinizer05on tmo iirc21:58
freemangordonwell, assuming that nowadays most run -thumb, I am not surprised :)21:58
freemangordondo you have the link?21:58
DocScrutinizer05sorry nope21:59
DocScrutinizer05iirc was in one of "your" threads22:00
freemangordonhmm, I usually read them regularly, but might have missed that one as I was veryy busy for the last 2 months or so22:00
DocScrutinizer05I wouldn't feel surprised to learn that PA or gstreamer or even ALSA doesn't like thombification22:00
freemangordonI doubt, too many people are on -thumb22:01
freemangordonit is rather something wrong with xes' device, either HW or SW22:01
DocScrutinizer05pointer to 2nowhere"22:01
DocScrutinizer05pointig to a location for absolute addrs of var that's outside own code segment when "short" thumb code used22:02
DocScrutinizer05there's a number of nasty optimizations in PA and ALSA22:02
DocScrutinizer05goto * + 4 //jump to end of routine, 2 opcodes down from here22:03
DocScrutinizer05stuff like that (I made up this one)22:04
freemangordonI don;t think there is handcrafted assembly in PA or ALSA22:04
DocScrutinizer05really? ;-P22:04
* DocScrutinizer05 ponders about the bet22:04
xessorry for the delay.  I'm fighiting with a server... So, i compared also alsa dump, volumes, mixers & co... nothing changes22:04
dos1wouldn't handcrafted assembly be unaffected by gcc using thumb or not?22:04
freemangordonand gcc usually complains if there is something it doesn't like22:04
xesfreemangordon: anyway, i'm on -thumb22:05
freemangordondos1: :nod:22:05
DocScrutinizer05dos1: yes, but the code it's linked to wpould change "ABI"22:05
freemangordonthere is no change in the ABI22:05
dos1DocScrutinizer05: it would be a bit risky to code like that, as any different compiler optimalization could also "change ABI"22:06
DocScrutinizer05see above example. for thumb two intructions are 2 bytes, not 4. (err words)22:06
freemangordonyeah, bytes22:06
freemangordonso, it is just an offset22:06
xesfreemangordon: the final user behavior is that the audio of all the gsm calls is routed to the speaker and there is no way to address it22:06
freemangordonDocScrutinizer05: no sane developer will hardcode the instruction len in a jump22:06
DocScrutinizer05I'm rather sure nasty stuff similar to this really brainfscked made up example can be found in ALSA and particularly PA22:07
freemangordonxes: do you have syslog installed?22:07
xesthis is the only similar report i have found so far: http://talk.maemo.org/showthread.php?t=6685522:07
xesfreemangordon: sure22:07
freemangordonmay I have one? :)22:08
DocScrutinizer05audio is all about efficiency and don't-care about any "reasonable coding"22:08
*** _rd has joined #maemo-ssu22:09
DocScrutinizer05xes: http://talk.maemo.org/showthread.php?t=66855&page=3 sounds like age-old fscked up optification that breaks PA when trying to play shake-hands sound while /opt not available22:10
DocScrutinizer05it never recovers from that22:11
DocScrutinizer05we had *lots* of fun with that one22:11
xesDocScrutinizer05: i know but:  1 A pulseaudio restart should fix    2 every fix suggested does not help22:11
DocScrutinizer05since it only happened to those who got gstreamer-codecs-bad (or whatever) installed, which optified out the flac/whatever plugin22:12
freemangordonxes: did you try to list alsa controls and compare?22:12
DocScrutinizer05yes, he did22:12
xesfreemangordon: diff says they are equal22:13
DocScrutinizer05ooh and he checked for all optification uissues. sorry22:13
DocScrutinizer05hmmmmmmm22:14
DocScrutinizer05ffs22:14
DocScrutinizer05FIRST: find out what the heck is hw:022:14
kerioit's probably the 0th hardware22:22
xessyslog --> https://etherpad.mozilla.org/zOLk5gIADx22:24
xesmost of times, to fix it i have to shudown and remove battery for a few seconds. It seems like a peripheral that remains in an hanged/wrong state. Sometimes, a fortunate reboot fixes it22:26
freemangordonhmm, why "Linux version 2.6.28.10-power52+git"?22:28
xeslatest pali's kp22:28
freemangordondid you try kp52 from the repos?22:29
kerio:o22:29
kerioi want that22:29
keriogimme that22:29
xesfreemangordon: yep - same situation22:31
freemangordonxes: what about "Jan  1 01:00:09 N900 ohmd[946]: GLIB ERROR ** default - module rule_engine failed to load but listed in require aborting..."22:31
freemangordonand...22:32
freemangordonJan  1 01:00:09 N900 init: ohmd main process (946) killed by ABRT signal22:32
freemangordonJan  1 01:00:09 N900 init: ohmd main process ended, respawning22:32
*** _rd has quit IRC22:33
xesfreemangordon: no idea22:33
*** eugene_ has joined #maemo-ssu22:34
freemangordonis ohmd running?22:37
xesfreemangordon: the ohmd error happens also when everything works22:37
freemangordonI don't think it is normal22:37
xesit seems you have found something. ohm-agent is running but all the failed sessions i have saved don't have ohmd running22:39
freemangordonafaik ohmd is the one that applies various policies22:40
freemangordonmaybe you should try to start it by hand and see why it aborts22:41
freemangordonxes: though you should really consider a reflash22:42
xesfreemangordon: hey... it's not windows ;) and a random issue is something that could happen again22:43
freemangordonyeah, it is not :)22:44
xesanyway, thank you ;)22:45
*** LauRoman has quit IRC22:47
freemangordonxes: please, teel me how it went once you find why ohmd crashes22:51
freemangordonit could still be a bug which deservers cssu attention22:51
freemangordondeserves even22:51
xesfreemangordon: i'm already comparing with saved backups (since 2011 ;) ) i will report as soon as i find something or... i will need some other suggestion :)22:52
freemangordonxes: I am almost sure it is ohmd22:52
freemangordonxes: http://www.gossamer-threads.com/lists/maemo/developers/5501922:53
freemangordon"It handles also audio policies and tries to make..."22:54
*** xes has quit IRC22:59
*** xes has joined #maemo-ssu23:12
*** eugene_ has quit IRC23:17
DocScrutinizer05http://www.gossamer-threads.com/lists/maemo/developers/55019 is EPIC! Recommended read for every devel23:21
*** NishanthMenon has quit IRC23:24
*** eugene_ has joined #maemo-ssu23:24
freemangordonDocScrutinizer05: in what regard? for educational purposes on what omhd is?23:33
DocScrutinizer05yes, among other aspects. Also about why maemo is superior to any random OS adapted to 'just run' on embedded23:34
freemangordonyeah23:34
DocScrutinizer05freemangordon: is ohmd freed by now?23:35
DocScrutinizer05or is it still a blob? or never been?23:35
freemangordonat least some version https://gitorious.org/maemo-multimedia/ohm/source/a683b973722553fce95f260e4829ce4decb26d3e:23:35
DocScrutinizer05cool23:36
DocScrutinizer05good for FPTF23:36
freemangordonI hope fremantle version is there too23:36
DocScrutinizer05it's those closed blobs that made Nokia look really nasty23:37
xeshmmm ...myfaultmyfaultmyfault: ohmd's rule engine uses:  /usr/lib/swipl-5.6.50 that in an unhappy moment of "zero disk space" i moved and symlinked into opt.. grrr23:37
xesfreemangordon: thanks!23:37
freemangordon:)23:38
freemangordonis it fixed now?23:38
xes3 reboots without the problem.. sounds (really!) good23:39
freemangordoncool23:39
freemangordonDocScrutinizer05: according to gtk we have "ohm (0.1.2-0maemo33) unstable; urgency=low"23:41
freemangordon*gitk23:41
freemangordonthat is the version I have on my device :D23:41
DocScrutinizer05xes: haha :-D23:42
xesfreemangordon: another issue for you is about really fast sd + swap on sdcard (like your samsung 64GB sdxc) i bough the same, but since now the device boots really faster hildon-status starts before mount optified mount points. I fixed with:     sleep 5 && /usr/sbin/dsmetool -t /usr/bin/hildon-status-menu     in /etc/X11/Xsession.post/15hildon-status-menu23:42
DocScrutinizer05ugh23:43
freemangordonxes: not here at least, but I didn't messed with startup scripts, unlike you :P23:43
xesDocScrutinizer05: :P23:43
freemangordon*mess23:43
DocScrutinizer05hildon definitely shouldn't start up before /opt mounted, eh?23:43
freemangordonafaik23:44
DocScrutinizer05actually xsession shouldn't start before /opt available23:44
xesanyway with swap on that fast sd N900 seems another device23:44
DocScrutinizer05o.O23:45
freemangordonyep23:45
*** okias has joined #maemo-ssu23:55
gypsyis the best idea to do swap striping or to remove the emmc swap altogether?23:55

Generated by irclog2html.py 2.15.1 by Marius Gedminas - find it at mg.pov.lt!