**** BEGIN LOGGING AT Sat Jan 12 02:59:59 2013 Jan 12 03:00:17 its an sqlite db .. you can use the sqlite tools to dump it as text/sql Jan 12 03:00:43 or even prints it in some formatted way Jan 12 03:01:29 cehteh: you mean the phonebook data? Jan 12 03:01:37 yep Jan 12 03:01:47 my sqlite tools don't seem to like it Jan 12 03:01:55 something about closed sets Jan 12 03:02:00 mhm .. am i wrong then? Jan 12 03:02:14 no idea Jan 12 03:02:21 should be sqlite Jan 12 03:02:24 .rtcom stuff is in sqlite Jan 12 03:02:36 both should be in sqlite Jan 12 03:02:39 .osso seems to be some custom db Jan 12 03:02:49 I'll try again soon Jan 12 03:02:50 hmm Jan 12 03:03:09 for now, I will embark on a quest to find the great Nokia 3310 Jan 12 03:03:18 or anything equivalent Jan 12 03:03:39 Anyone know how much an unopened N950 is worth? Jan 12 03:03:56 1000-ish euros Jan 12 03:04:01 $1 .. i take it :) Jan 12 03:04:37 Theorbo: 1000+eur on ebay... Jan 12 03:05:22 in netto actualy nothing but 5euros... Jan 12 03:09:08 Thanks Jan 12 07:16:58 can anyone configrm is the Audio in Qt for Maemo is supported or not? Jan 12 07:18:28 confirm if* Jan 12 07:52:24 HtheB: it should work Jan 12 07:52:43 if it doesn't hit the guys over in #maemo-ssu and we might provide a fix with the next cssu update Jan 12 07:53:13 http://talk.maemo.org/showthread.php?p=1311859&highlight=ppsspp#post1311859 Jan 12 07:53:22 check the last post Jan 12 07:56:09 hm intresting, I can have a look in the qt sources somewhere next week, but you'll probably want your answer earlier :D Jan 12 07:56:37 no problem Jan 12 07:56:45 well, its to point out to pali Jan 12 07:56:53 that afaik, it should work now Jan 12 07:57:07 some guy is trying to port the psp emulator to harmattan Jan 12 07:57:20 but he gets some audio errors afaiu Jan 12 08:43:05 DocScrutinizer05: URL to announcement about it? Jan 12 09:24:41 Hoo boy. Jan 12 09:24:49 "greater good" Jan 12 09:25:27 http://hildonfoundation.org/jolla-part-2/#comment-69 Jan 12 09:26:54 so now a Councilor is just a "community member" while the BoD is above all? Wow... Jan 12 10:26:01 is there some camera application around for n900 with both sources and ability to refocus? Jan 12 10:26:40 using the standard camera subsystem that is, fcam does exist.. Jan 12 10:28:39 I suppose fcam will do fine for me, though :) Jan 12 10:29:27 is there any info in maemo system logs where did the last charging occur? Jan 12 10:30:17 seems that dmesg has something Jan 12 10:31:58 tadzik: maemo has no logs unless you install syslog Jan 12 10:32:20 I see. Thanks Jan 12 10:41:16 sup Jan 12 10:48:11 kerio: va bene Jan 12 10:48:30 ? Jan 12 10:48:41 DocScrutinizer05: ping Jan 12 10:48:43 I thought you were italian Jan 12 10:48:49 i am Jan 12 10:48:53 va bene ? Jan 12 10:49:43 if you mean "what's up" you should ask something like "come va?" Jan 12 10:50:33 ah Jan 12 10:50:44 and yeah, everything alright Jan 12 10:50:50 hehe Jan 12 10:50:54 I use va bene most of the time and people just say yeah and you ? Jan 12 10:51:10 so I assumed it to be correct Jan 12 10:51:15 f3bruary: that is more like "ok" Jan 12 10:52:09 turns out the xtube fix is not really working well Jan 12 10:52:22 it's the kind of phrase that, while being wrong or at least a bit weird, would still be perfectly understood when spoken out loud by a foreigner :) Jan 12 10:52:36 ic Jan 12 10:52:39 :D Jan 12 12:05:09 DocScrutinizer05: in RescueOS, mmc blocks for SD and mtd are swapped Jan 12 12:05:51 so, mmcblk0 SD slot, mmcblk1 system emmc Jan 12 13:03:35 Any tips before applying a screen protector? Jan 12 13:04:02 do it in a dust-free environment Jan 12 13:04:41 grummund: bathroom, run the shower with scalding hot water Jan 12 13:04:53 :nod: Jan 12 13:05:09 so the air becomes humid and the dust drops from the air Jan 12 13:05:22 exactly Jan 12 13:05:56 should i clean the screen with anything first? Jan 12 13:06:37 it's quite scratched already Jan 12 13:07:03 yes, clean it Jan 12 13:07:22 the film cames with a cleaning clith Jan 12 13:07:25 *cloth Jan 12 13:07:26 and don;t worry about the scratches, screen protector will make most of them invisible Jan 12 13:07:44 you'd better find the one that came with the device Jan 12 13:07:48 i have some anti-static foam cleaner Jan 12 13:07:54 no Jan 12 13:08:10 you rist foam to enter into device Jan 12 13:08:14 that will ruin it Jan 12 13:08:18 *risk Jan 12 13:08:46 i can apply a small amount to the cloth first Jan 12 13:08:51 how about alcohol wipes? Jan 12 13:09:52 thedead1440: moist from your breat is all that you need Jan 12 13:09:57 *breath Jan 12 13:10:09 haha Jan 12 13:10:37 heh. i have been alchol free since before xmas Jan 12 13:20:10 whoa, that's all pretty extreme Jan 12 13:21:11 I use packing tape to pick up the dust and stickify the surface for the protector Jan 12 13:24:25 just use the same type of cloths they use to clean glasses Jan 12 13:24:29 those are perfect for the job Jan 12 13:24:39 or, if you have an imac around Jan 12 13:24:45 Apple's cleaning cloths are king (no joke) Jan 12 13:24:55 All done, and snug in its new case too :) Jan 12 13:25:05 nokia case? Jan 12 13:26:21 nah, imitation leather. With a free screen protector, all for 99p. :D Jan 12 13:26:31 :X Jan 12 13:26:37 including shipping :D Jan 12 13:26:42 CP-408 is the king of n900 cases Jan 12 13:26:52 http://blog.tecnologiamovil.net/wp-content/gallery/miscellanea_singles/nokia-cp-408.jpg Jan 12 13:27:39 yebbut pricey though Jan 12 13:28:00 20euros Jan 12 13:28:47 yeah like that's 20x what i paid :p Jan 12 13:28:57 http://www.amazon.co.uk/BLACK-WALLET-LEATHER-CASE-NOKIA/dp/B0038LNLS6/ref=sr_1_1 Jan 12 13:29:01 1 left in stock Jan 12 13:29:12 i had a case like that at first, it died in a week Jan 12 13:29:21 not exactly the same, but also a cheap one Jan 12 13:31:02 otterbox or nothing Jan 12 13:31:41 that doesnt protect the screen iirc Jan 12 13:53:07 otterbox+screen protector Jan 12 13:53:15 lol screen protector Jan 12 13:53:23 the name suggests it protects Jan 12 13:53:31 but it protects as well as normal condoms do when having anal sex Jan 12 13:53:34 which isnt very much Jan 12 14:06:29 :O Jan 12 14:53:09 hi, i'm using maemo on n900 including the community updates (which I appreciate very much). Jan 12 14:53:51 i'd like to disable certain CA's, though, which doesn't seem to be possible using the GUI. can you help me as to find out how to disable a CA (or revoke trust)? Jan 12 14:55:01 also, i'm wondering which web browser people use / recommend. i'd like to use one which receives security updates. Jan 12 14:56:09 tomreyn: you may want to look at this post: http://talk.maemo.org/showpost.php?p=1311600&postcount=13 Jan 12 14:56:52 and on yet another topic: http://wiki.maemo.org/Community_SSU/Changelog prints a database error, revealing the database it is trying to access as well as the DB user: from within function "SiteStatsUpdate::doUpdate". Database returned error "1142: UPDATE command denied to user 'wikiuser_ro'@'wml0234' for table 'mw_site_stats' (apps.maemo.org)". Jan 12 14:57:35 ~jrtools Jan 12 14:59:10 DocScrutinizer05: infobot down Jan 12 14:59:32 thedead1440: thanks for the forums link, this should help. do you think this method will persist on CA updates? Jan 12 14:59:42 tomreyn: use cmcli to change the cert db Jan 12 15:00:18 tomreyn: the wiki is down so you need to add into your hosts file the following entry: 188.117.59.202 wiki.maemo.org Jan 12 15:00:34 s/down/being migrated/ Jan 12 15:00:52 kerio: thanks, that's basically what the script in the forums does (amongst also updating the, apparently separate, microb CA store. Jan 12 15:01:05 there's no need to change the microb CA store Jan 12 15:01:48 kerio: so it's sourced from the same backend as the one cmcli operates on? Jan 12 15:02:08 libmaemosec-certman0 diverts /usr/lib/microb-engine/libnssckbi.so away, and supplies its own Jan 12 15:02:22 tomreyn: blacklisting is probably better than deleting, fyi Jan 12 15:04:03 kerio: ssh: connect to host rikers.org port 22: Connection timed out Jan 12 15:05:02 kerio: can you explain how to blacklist? using cmcli, i assume? i'm just looking at the usage output and it doesn't meantion "disable" or "editing trust" Jan 12 15:05:17 DocScrutinizer05: I cannot connect to our new serrver... Jan 12 15:05:28 ooh Jan 12 15:05:31 sry cannot 'logon' to it Jan 12 15:05:58 I have no access yet Jan 12 15:06:03 tomreyn: you have to add it to the "blacklist" domain Jan 12 15:06:19 ask merlin1991 for details Jan 12 15:06:20 oh, got it Jan 12 15:06:46 chem|st: ping reggie, maybe he's messing around there Jan 12 15:06:46 he's the one that made the last maemosec update Jan 12 15:07:06 much appreciated that one Jan 12 15:07:14 just like diginotar's Jan 12 15:07:15 DocScrutinizer05: I tried to logon minutes after I got the ip Jan 12 15:07:17 tomreyn: cmcli -c blacklist -a pathtocert Jan 12 15:07:32 cmmcli -c common-ca -r certid Jan 12 15:07:53 merlin1991: what happens if a certificate is in both? Jan 12 15:07:57 s/cmm/cm/ Jan 12 15:08:03 retardation? Jan 12 15:08:11 :) Jan 12 15:09:18 DocScrutinizer05: I replied to rambo's mail Jan 12 15:09:27 oh, you didn't remove anything from common-ca because the intermediate CAs weren't in the common-ca list Jan 12 15:10:53 kerio: can you explain the concepts of private vs shared domains? i assume private is for internal / organisation CAs? Jan 12 15:11:11 i have no idea actually Jan 12 15:11:15 i.e. when you have your own internal PKI Jan 12 15:11:30 you could try adding a certificate with the applet Jan 12 15:11:36 to see where it ends up Jan 12 15:11:44 the applet in Settings Jan 12 15:11:55 infobo: wb Jan 12 15:11:59 right Jan 12 15:12:05 DocScrutinizer05: wat Jan 12 15:12:20 [2013-01-12 16:11:27] [Notify] infobot is online (irc.freenode.net). Jan 12 15:12:25 i see Jan 12 15:13:20 ~botsnack Jan 12 15:13:20 kerio: :) Jan 12 15:13:56 well thanks everyone. about the web browser: which one do you use and why? and is there any one which gets security updates, or updates at all? Jan 12 15:14:18 tomreyn: microB gets security updates from cssu :) Jan 12 15:14:31 it does? i never noticed Jan 12 15:14:36 second "distribution" in the world to fix the turktrust issue :) Jan 12 15:14:46 hehe Jan 12 15:14:54 er, third Jan 12 15:15:00 first and second were mozilla and m$ Jan 12 15:15:32 base don what i read Ms actually didnt revoke it yet Jan 12 15:15:57 nobody removed the turktrust CA Jan 12 15:15:58 or just provides a patch which is not shipped universally Jan 12 15:16:07 they just blacklisted the two affected intermediate CAs Jan 12 15:16:17 well i did ;) Jan 12 15:16:35 they lost my trust if i had any in them before Jan 12 15:17:41 but "revoke" is a very wrong term in this context indeed Jan 12 15:29:58 does someone use evopedia? I don't know what wikipedia-dump I have to download and evopedia cannot download it via bittorrent... Jan 12 15:32:11 i guess the idea is that you download it to your desktop using bittorrent and then move it to the maemo device using a data cable or similar Jan 12 15:32:36 the dumps' torrent files are available on their website Jan 12 15:32:52 pretty outdated though Jan 12 15:34:20 it would seem like evopedia needs a different format than the one provided by wikimedia Jan 12 15:34:39 "Evopedia dump at home combines the processing power of various computers around the globe to create Wikipedia database dumps for use in the offline Wikipedia reader evopedia." Jan 12 15:35:54 does it work with this dumps? https://meta.wikimedia.org/wiki/Data_dump_torrents Jan 12 15:36:15 ok Jan 12 15:36:19 it would seem like evopedia needs a different format than the one provided by wikimedia Jan 12 15:36:27 yes Jan 12 15:38:00 based on past releases, you can probably expect the evopedia people to update the english database within the next two months. Jan 12 15:38:51 the latest release of the english wikipedia dates back to early 2012. german is more current in case you prefer this. Jan 12 15:39:13 I want both Jan 12 15:40:18 hmm german is 3.9 TB while english is just 670 MB Jan 12 15:40:37 oh thats wrong Jan 12 15:40:46 EN is 13.8 GB :) Jan 12 15:47:48 tomreyn: do you like it so far? Jan 12 15:50:35 everybody involved in KP project: what do you think about preempt-rt on N900? Will it improve anything or ruin expeperience instead? Jan 12 15:51:53 WizardNumberNext: i like what? Jan 12 15:52:41 I am going to try out RT-kernel on my server (mind its amd64) and see how it behaves (NFS and SAMBA gets slow sometimes on no preemtion). I run all my x86 (centrino and 2 AMD Phenom II) on no preemption at the time Jan 12 15:52:59 tomreyn: evopedia Jan 12 15:53:24 WizardNumberNext: i'm not using it, it takes too much space. Jan 12 15:54:03 tomreyn: if it stores all tis on your HDD, then that sucks Jan 12 15:54:19 well thats the point Jan 12 15:54:46 so you dont have to be online Jan 12 15:55:32 I could store it on my server, but after I fix my 1TB drives and they sit, ready to be fixed on shelf for something like 2 months Jan 12 15:56:14 tomreyn: that is not my issue - I have 'all you can eat data', so I do not care about data Jan 12 15:56:39 wikimedia already stores the data on their servers thanks to the many donations they receive, it's called wikipedia.org Jan 12 15:57:55 yes, wikipedia does the job, at least ost of the times Jan 12 16:26:20 ShadowJK: is that ok? http://pastebin.com/BkeUC9YZ Jan 12 16:27:04 Samsung 64GB Pro UHS-1 Jan 12 16:29:49 holy balls Jan 12 16:30:42 kerio: ? Jan 12 16:31:04 freemangordon: those random writes Jan 12 16:31:12 yeah :D Jan 12 16:31:17 btw. have you seen nokia financial numbers Jan 12 16:31:28 freemangordon: try writing 64gb of random data over it first, though Jan 12 16:31:31 to make sure it's not cheating Jan 12 16:32:01 jacekowski: btw it was you with doxbox, ain't? Jan 12 16:32:25 kerio: http://en.samaanet.com/?p=16415, I guess it is not Jan 12 16:32:48 freemangordon: ? Jan 12 16:32:49 kerio: i'd rather do random write test wint 1GB Jan 12 16:33:13 also run the same test on the n900 itself, too :) Jan 12 16:33:13 jacekowski: was it you the dosbox maintainer Jan 12 16:33:24 kerio: how? Jan 12 16:33:35 wine, ofc Jan 12 16:33:40 :P Jan 12 16:33:59 there must be an IO benchmark for linux Jan 12 16:34:14 freemangordon: no Jan 12 16:34:23 jacekowski: ok, sorry Jan 12 16:34:50 aah, it was javispedro :) Jan 12 16:35:15 kerio: there is, iozone iirc Jan 12 16:35:38 freemangordon: 64gb swap partition gogogogogo Jan 12 16:35:46 hehe Jan 12 16:49:33 kerio: btw benchmark is made with card in n900 connected via usb Jan 12 16:49:45 oh, good enough then, i suppose Jan 12 16:50:20 yep. I guess we can't go more than 10MB/s anyway. but random write is impressive Jan 12 16:50:28 *write speed Jan 12 16:54:54 I note that I got 24mbyte read using emmc and SD Jan 12 16:55:15 a leetle more if you use one nand too Jan 12 16:55:30 SpeedEvil: on device? Jan 12 16:55:39 tyres Jan 12 16:55:42 yes Jan 12 16:55:56 concurrent dad's Jan 12 16:56:00 dd's Jan 12 16:56:30 I guess it depends on io scheduler settings. I did the test with stock settings Jan 12 16:57:31 though I don't think sequental io matters much on n900 Jan 12 17:04:13 stock here too Jan 12 17:04:31 could you givme the commend? Jan 12 17:04:39 *command Jan 12 17:04:43 mum Jan 12 17:04:44 to run it here Jan 12 17:04:48 ok Jan 12 17:04:48 umm Jan 12 17:05:50 just something like time dd if=... of=/dev/null bs=131072 count=a gigs worth Jan 12 17:06:19 ok Jan 12 17:08:39 SpeedEvil: iostat reports ~15500KB/s Jan 12 17:10:30 as I understant it, 25mhz is the maximum clock Jan 12 17:10:45 so 12.5 will be the hard maximum performance Jan 12 17:11:21 afaik it is 50MHz, but could be wrong Jan 12 17:11:45 dd of=./tst if=/dev/zero bs=131072 count=65536, iostat reports between 4000 and 10000 KB/s Jan 12 17:13:03 doing the same with bs=4096 increases write speed to ~14000KB/s Jan 12 17:14:06 hmm I like that card :D Jan 12 17:14:19 freemangordon; empty card? Jan 12 17:14:23 yes Jan 12 17:15:18 ShadowJK: what is the difference? Jan 12 17:15:52 Yeah I get 200-ish IOPS on blank sandisk class 4 too, collapses to 1-2 once all pages have been touched Jan 12 17:16:33 I guess it will take a while until all 64GB of pages got touched :) Jan 12 17:16:51 ShadowJK: any idea how to test random io on the device? Jan 12 17:17:22 I was trying to compile coffeemug rebench the other day, but got stuck on libaio Jan 12 17:18:11 well, i guess i can run ubuntu installer and check what iostat reports then Jan 12 17:22:39 i'll rerun diskmark random write with 1GB file Jan 12 17:23:11 how long does the benchmark run? Jan 12 17:23:20 does it keep file around? Jan 12 17:23:28 yes, it keeps the file Jan 12 17:23:48 bot I will watch iostat in the meantime ;) Jan 12 17:24:20 finished ;) Jan 12 17:24:35 I magine if you ran it 5-10 minutes on same file it'd start declining Jan 12 17:24:44 Random Read 4KB (QD=1) : 2.407 MB/s [ 587.7 IOPS] Jan 12 17:24:44 Random Write 4KB (QD=1) : 1.086 MB/s [ 265.0 IOPS] Jan 12 17:25:14 i'll run it again Jan 12 17:25:53 oh, no, it creates new file every time :( Jan 12 17:26:19 how long does it take to run once? Jan 12 17:26:30 1-2 minutes Jan 12 17:27:21 i guess I can set the benchmark what to do, lemme check Jan 12 17:28:08 no, I can't :( Jan 12 17:28:20 ok, sekomd run, same results Jan 12 17:28:24 *second Jan 12 17:28:56 well, i'll partition the card and will put it in my primary, will benchmark again in a week or so Jan 12 17:30:54 ShadowJK: how many iop/s was your data? Jan 12 17:30:57 *adata Jan 12 17:31:13 about 20 in steady state Jan 12 17:31:50 that is I ran benchmark in one-minute chunks until numbers settled Jan 12 17:36:08 ShadowJK: or i'll create 1GB partition and will use it Jan 12 17:40:19 Would be interesting to test whether blanking stuff outside the 1G speeds the 1G Jan 12 17:42:06 hi everybody - how's the migration progressing? i've seen you published new wiki's ip addr... Jan 12 18:06:46 wirr: still waiting for the dns change from nokia afaik Jan 12 18:09:07 kerio: looks like... Tech Email:hostmaster@nokia.com Jan 12 18:09:07 Name Server:NS.NOKIA.COM Jan 12 18:09:18 (whois for maemo.org) Jan 12 18:09:50 you know how funding's proceeding? what's more needed, money or sponsoring ISPs providing hosting or housing? Jan 12 18:10:59 hm, idk Jan 12 18:15:27 two questions: n900: maximal SD(hc/xc) size? Jan 12 18:15:57 and wine? how the hell it work? does it emulates x86? Jan 12 18:17:53 wine doesn't work, really Jan 12 18:18:25 afaik there's no qemu+wine combination in the repos Jan 12 18:18:29 kerio: that is what I was thinking Jan 12 18:19:02 the "wine" in extras-devel is to run recompiled windows applications compiled against the wine libs Jan 12 18:27:25 wirr: money more needed, and particularly a professional sysop needed Jan 12 18:29:48 WizardNumberNext; well freemangordon is using a 64g card. do 128 ones exist? Jan 12 18:32:59 I'm guessing exfat doesn't work though :) Jan 12 18:57:22 ShadowJK, format it as ext3 (or if you have kernel-power as ext4) Jan 12 18:57:40 I think that the best filesystem for SD cards is ext4 Jan 12 18:58:08 On sheevaplug I use nilfs2 Jan 12 18:58:49 there is some side effect in ext4 implementation which cause better performace as ext3 for sd cards... Jan 12 18:59:44 exfat is from design step back Jan 12 19:00:16 I really do not understand why they chosed it as *default* filesystem for SDXC Jan 12 19:02:27 what else? Jan 12 19:02:55 ext2/4/... are not supported on most computers Jan 12 19:04:19 and is exfat supported? Jan 12 19:04:27 and is exfat better then other FS? Jan 12 19:04:44 and is there exfat specification? Jan 12 19:05:13 is there reference implementation (portable for other systems)? Jan 12 19:06:11 really, as replacement for old fat32 which is supported on more computer is UDF Jan 12 19:06:41 there is not stupid 4GB limit, so as filesystem for backup/storage is usable Jan 12 19:06:57 and yes, it's retarded Jan 12 19:07:12 and UDF is supported on lot of system and configurations Jan 12 19:07:19 fundamentally, Microsoft is an using its monopoly posaition,. Jan 12 19:07:22 abusing Jan 12 19:09:35 its funny that nobody written ext4 driver for windows Jan 12 19:10:09 but there exists more ext4 implementations in other non windows systems... Jan 12 19:10:17 I assume there is fine Jan 12 19:10:19 one Jan 12 19:10:27 where? Jan 12 19:10:30 I was using one back in 1995ish Jan 12 19:10:38 well. ext2 Jan 12 19:10:43 ext2 is here Jan 12 19:11:02 (and usable for ext3 without journal) Jan 12 19:11:18 but writing with ext2 driver to ext4 is not working... Jan 12 19:24:36 hm, UDF.. wonder if that would work on sd.. Jan 12 19:27:26 sure, why not Jan 12 19:27:31 well, probably not on windows Jan 12 19:41:18 I vaguely recall trying to use udf on -rw, not successful :) Jan 12 21:21:03 ~ping Jan 12 21:21:03 ~pong Jan 12 21:28:42 DocScrutinizer51: srsly, set up sasl Jan 12 21:29:01 meh Jan 12 21:29:17 why? Jan 12 21:29:23 cuz it's cool! Jan 12 21:29:37 also, * DocScrutinizer51 (~lagrange@lagrange.cloud-7.de) entra in #maemo Jan 12 21:29:53 friggin ZNC not in centOS default repos Jan 12 21:30:08 just compile from the tarball Jan 12 21:30:16 it's a nice, round version 1.0 now Jan 12 21:30:35 and the above only happens when I reboot the bouncer Jan 12 21:30:49 *when the bouncer connects to IRC Jan 12 21:30:56 this is freenode, remember Jan 12 21:32:07 DocScrutinizer05, regarding sysop, is there some info on tmo about what skills and time is required? Jan 12 21:46:44 wirr: http://talk.maemo.org/showthread.php?t=88632 Jan 12 21:52:23 freemangordon: :) I don't want to stop working on bluez and others packages ... Jan 12 21:55:10 freemangordon: I've heard you're using 64gb sd card on your n900. Do you know if it is generaly supported or is it rather matter of luck? I mean do you have any idea about addressing issues? I do not know, if SD card is more alike pure FLASH or is it block device (means interface between FLASH and SD card interface itself) Jan 12 21:56:36 SpeedEvil: I never had run into any computer, which doesn't support ext2/3/4! It is up to OS only. And I know only ONE OS, which doesn't support ext2/3/4 and I think you know the name of that crap. Jan 12 21:57:22 WizardNumberNext: you mean OS X? Jan 12 21:58:04 I would not let windows read my SD card anyway. Last it touched my HDD I lost 400GiB of data - it was last time it had access to my hardware Jan 12 21:58:28 Oh, I forgot that creator of OS X exists Jan 12 21:58:40 sorry ;) Jan 12 21:59:48 who uses OS X? only fanatics I presume. No games, no software, no nothing - what is use of that? Fancy graphics? I have that for ages on my GNU/Linux Jan 12 22:00:09 and I didn't pay 1200GBP for my desktop Jan 12 22:00:30 barely over 1/3 of that price tag Jan 12 22:00:38 WizardNumberNext: more people than Linux, actually :) Jan 12 22:01:21 because they do not know about existence of anything else then Window$ and O$ X Jan 12 22:02:48 * kerio paid 2300€ for his macbook pro Jan 12 22:02:51 haters gonna hate Jan 12 22:02:58 look - I do not have to spend whole week to install my software, not to mention drivers, do noit have tpo spend ages to pull software of the net and do not have to pay single pence for my software Jan 12 22:03:14 kerio: you must be joking me Jan 12 22:03:28 what sort of CPU and RAM do you have there? Jan 12 22:03:31 WizardNumberNext: you must be poor Jan 12 22:03:40 4x opteron 62xx? Jan 12 22:04:05 luf: ofc, CSSU needs you :) Jan 12 22:04:08 for that price tag I would have 2x Opteron 62xx + 64GiB of RAM Jan 12 22:04:37 WizardNumberNext: not with 8 hours of battery life Jan 12 22:04:37 though you may administer maemo.org in the time you take rest from bluez and such :P Jan 12 22:04:55 I am not poor, I am reasonable - I do not spend 5x more for same class CPU and 3x for same damn RAM Jan 12 22:05:11 WizardNumberNext: on the other hand, LAPTOP Jan 12 22:05:43 (besides, mine is about 5 years old now) Jan 12 22:05:50 I have almost best CPU of the line and 8GiB of RAM - price - just slightly over 400GBP Jan 12 22:06:02 i'm sure of it Jan 12 22:06:02 where PSU costed me 80GBP Jan 12 22:06:12 how long does it stay on without plugging it in? Jan 12 22:06:24 mine is dated last year Jan 12 22:06:34 desktop? Jan 12 22:06:42 wiothout pluggin in? Jan 12 22:06:52 I do not like laptops Jan 12 22:07:15 for one reason Jan 12 22:08:05 in last year I had: 5 motherboards, 2 cpu, 2/4/8GiB of RAM, 3 PSUs Jan 12 22:08:21 all of it was tested and if I didn't like it I changed it Jan 12 22:08:30 change your PSU in laptop! Jan 12 22:09:10 i already did, under warranty Jan 12 22:09:30 P.S. welcome to Apple world - everything is really integrated and you cannot change anything, and even if you can, then you either void warranty or pay for it Jan 12 22:09:31 because the plug that connects to the laptop side was broken Jan 12 22:09:44 it is not PSU of Laptop Jan 12 22:09:56 it is just general PSU Jan 12 22:10:02 I mean PSU-PSU Jan 12 22:10:11 not just single volt PSU Jan 12 22:10:21 oh, the battery then Jan 12 22:10:26 WizardNumberNext: OS X is nice, when not running on Apple hardware. Jan 12 22:10:27 you have whole PSU in laptop Jan 12 22:10:38 12v, 5v, 3.3v Jan 12 22:11:07 also, holy crap, 5 mobos and 2 cpus in a year? Jan 12 22:11:17 Hurrian: no it's not, mountain lion is a piece of crap Jan 12 22:11:28 IHurrian: I was curious for while, but, when I seen how much I need to do to run it on AMD Phenom II... Jan 12 22:11:39 freemangordon: :D Yes I still sleep several hours per day. Jan 12 22:11:49 so lion is crap as well Jan 12 22:12:02 lion is crap too, yes Jan 12 22:12:08 ah, an AMD system. Man, if that was an Intel system, that would probably bankrupt me. Jan 12 22:12:20 snow leopard is the last good one, pretty much :s Jan 12 22:12:29 luf: what a waste :D Jan 12 22:12:43 luf: disregard sleep, acquire coffee Jan 12 22:12:48 s/coffee/meth/ Jan 12 22:12:48 kerio meant: luf: disregard sleep, acquire meth Jan 12 22:12:52 kerio: I would disagree - last one was leopard - and that is for single reason Jan 12 22:12:56 it run on PPC Jan 12 22:13:16 kerio: meh, I spend most of my time in nano, Xcode, Firefox and VMWare, and launch programs using Alfred, so I avoid the horrible Launchpad Jan 12 22:13:44 the big draw to me using OS X is Mission Control, which on a single-screen system is a lifesaver Jan 12 22:13:56 except that nobody was able to make a G5 that didn't require a nuclear fusion reactor to run Jan 12 22:14:05 so that made it kinda hard to have good laptops Jan 12 22:14:38 ok, IBM never meant 74xx for laptop Jan 12 22:14:46 sorry 97x Jan 12 22:15:11 it was motorola job to make laptop ones Jan 12 22:15:24 and motorola quited AIM Jan 12 22:15:32 so onlyu AI left Jan 12 22:15:47 then Aplle kicked out IBM so only A left Jan 12 22:16:24 it's a damn shame, yeah :s Jan 12 22:16:31 that's shame they kicked them out Jan 12 22:16:43 now Apple would have Cell 9i Jan 12 22:16:56 or something even better Jan 12 22:17:06 i wonder what would've happened if apple moved to ARM instead Jan 12 22:17:23 WizardNumberNext: Cell chips need TWO nuclear fusion reactors to run! Jan 12 22:17:37 and left? Then ARM would second to best! Jan 12 22:17:53 Hurrian: who told you that Jan 12 22:18:02 full Cell 8i is in PS3 Jan 12 22:18:02 WizardNumberNext: no, i mean, instead of moving to x86 Jan 12 22:18:20 WizardNumberNext: High TDP of PS3 CPU, and it's only single-core. Jan 12 22:18:23 but i don't know what was the status of arm when they made the switch Jan 12 22:18:40 kerio: that is what I meant, Apple moving to ARM and leaving it. Jan 12 22:19:17 Hurrian: single-core? What? It is full Cell 8i - 8SPE + 1 PPE Jan 12 22:19:30 I was talking about the PPC core inside the Cell. Jan 12 22:19:32 where do you get your info from? Jan 12 22:19:43 SPE is PPC! Jan 12 22:20:17 Alright, so, go run Linux kernel process threads in it. Jan 12 22:21:22 It's impractical to do so, because of the crazy mailbox/local store structure, and the fact that they're more suited to running numerical compute kernels a la CUDA Jan 12 22:21:53 it is like pure AltiVec core Jan 12 22:22:38 Hurrian: show me Linux kernel, which is able to run on more then one core at once! Jan 12 22:23:03 it is monolitic kernel and it runs on one core Jan 12 22:23:10 Indeed. IMO, Sony should've rode the tide with the PS3, developed a quad-PPC core chip, and not bothered with an SPE Jan 12 22:23:17 then you do not need more, then one core for it Jan 12 22:24:16 Sony had no voice about design of Cell! It is IBM creation! Jan 12 22:24:27 They ordered the chips. Jan 12 22:24:43 thats right, they ordered chips, not designed them Jan 12 22:25:05 they ordered 10M chips Jan 12 22:25:25 they could've asked for whatever they wanted, really Jan 12 22:25:29 nubmer of them doesn't make any frickin difference Jan 12 22:25:45 it still would be designed by IBM Jan 12 22:25:57 yes, but you can ask for specific features if you're sony Jan 12 22:26:06 and for game console you need a lot of SPEs Jan 12 22:26:14 yes, they did Jan 12 22:26:15 In Linux 2.2 there was the "Big Kernel Lock", which made kernel on all other cores/processors stop if one cpu accessed any complex data structure. In 2.4, 2.6 and 3.x it's pretty darn fine grained though Jan 12 22:26:20 they asked for 7 SPEs Jan 12 22:27:03 ShadowJK: as far as I remember, they removed BKL totaly Jan 12 22:27:42 which doesn't mean, that plain kernel thread could be run on more then one core Jan 12 22:27:54 it is all about helpers Jan 12 22:28:07 but pure kernel is unsplitable Jan 12 22:28:25 even macro-kernel is unsplitable Jan 12 22:28:26 WizardNumberNext: IIRC, from the PS3 prototypes, 8 SPEs were planned, but to cut costs on the already ~$1000 PS3 they accepted partially defective ones Jan 12 22:28:50 A single thread rarely runs on more than one cpu :-) Jan 12 22:29:07 ShadowJK: you know what I meant Jan 12 22:29:31 nowadays the newer PS3s usually have the Cell chips perfectly manufactured, and they can probably be safely enabled. Jan 12 22:29:37 fs, drivers and so on can run wherever Jan 12 22:30:33 Hurrian it is valuable info, maybe I would get one and set up some 'super computer' on it Jan 12 22:30:53 WizardNumberNext, and what's left after you remove fs, drivers etc Jan 12 22:30:55 but wait RAM is crap there - it is too small for anything sewrious Jan 12 22:31:01 wirr: there is no info on tmo about details for master sysop of maemo.org. consider of it as 2 dedi servers running ~12 VMs Jan 12 22:31:36 WizardNumberNext: I run kernel 3.7 on mine, the SPE guides on IBM are a good start Jan 12 22:31:44 memory handling, process management, ... Jan 12 22:31:50 And regarding the RAM, just swap on the VRAM block device Jan 12 22:31:59 do not remember all at the moment, but there are 4 Jan 12 22:32:07 wirr: needed time is sth the sysop needs to tell me, not I tell sysop Jan 12 22:32:28 It's XDR, so it's quite a lot faster than DDR3. You're still limited to 512MB RAM though. Jan 12 22:32:35 Hurrian: knowing what RAM is supported, which won't be easy to find out Jan 12 22:33:13 I guess you'll need 30min/day minimum, watchin nagios/munin/whatever, syslogs, CERT alerts etc Jan 12 22:33:37 WizardNumberNext: the PPC CPU has access to 256MB of RAM, and you can swap on the 256MB of VRAM (since there is no accelerated 3D video driver for the PS3) Jan 12 22:33:40 if you know your business Jan 12 22:34:13 you can swap it as long as you know what is supported Jan 12 22:34:33 there is bit more then plain mbits per chip Jan 12 22:34:37 WizardNumberNext: Now that gives you 512MB of fast memory, outside of that you'll have to swap to HDD, which is incredibly, unbearably, slow on the PS3 because of the HDD encryption/decryption chip Jan 12 22:34:43 wirr: initially you'll need more time setting things up and making familiar with stuff Jan 12 22:35:07 DocScrutinizer05, as far I get from the councils post, the goal is to have a full team of sysops, right? Jan 12 22:35:10 WizardNumberNext, ignoring the helper threads, the kernel isn't some process that's always running. It's more like event-driven pieces of code running with the highest priviledges. When a page fault happens, the cpu core on which it happened jumps into kernel code and starts executing the memory management stuff, and schedules another process to be run on that core. When an interrupt happens, the core that interrupt was routed to gets to handle it. The pr Jan 12 22:35:10 like bitness of RAM, like its organization and its low-level organization Jan 12 22:35:16 ocess doesn't get moved to core 0 or anything Jan 12 22:35:19 brb Jan 12 22:36:17 wirr: yes, the plan is to have one professional master sysop, and find as many maintainers for services as possible to delegate stuff Jan 12 22:37:12 I know ShadowJK, I know that for quite long time, it is purely easy answer to SPE problem - you cannot run kernel there, but who cares - kernel doesn't run all the time anyway Jan 12 22:37:41 you can't run stuff with a crew of 30 peer sysops Jan 12 22:37:46 btw interrupt handling is another one Jan 12 22:37:50 DocScrutinizer51: will sysop be paid? Jan 12 22:37:51 Well it's also that the thing is deprived of any meaningful I/O to manage :-) Jan 12 22:38:07 will take more time for coordination overhead than anything else Jan 12 22:38:18 freemangordon: dependsa Jan 12 22:39:07 and I/O handling Jan 12 22:39:10 we probably prefer to pay a small amount, and have proper responsibility and liability Jan 12 22:39:28 afk, sorry Jan 12 22:39:49 DocScrutinizer51: what have to have sysop? Jan 12 22:40:37 or otherwise, how much I have learn? Jan 12 22:40:52 DocScrutinizer05, I agree, the team has to be small enough to stay agile Jan 12 22:41:00 s/have learn/have to learn/ Jan 12 22:41:01 WizardNumberNext meant: or otherwise, how much I have to learn? Jan 12 22:41:20 but still large enough to allow people to do their tasks reliable besides their day jobs Jan 12 22:41:29 s/reliable/reliably/ Jan 12 22:41:29 wirr meant: but still large enough to allow people to do their tasks reliably besides their day jobs Jan 12 22:42:15 going for some water Jan 12 22:45:21 WizardNumberNext: I bought the card today and it is still in my devel device Jan 12 22:45:31 DocScrutinizer05, is a full resume expected with the application? Jan 12 22:46:23 can someone tell me why i can't create a new page on the new wiki? Jan 12 22:46:28 WizardNumberNext: but afaik there should be no problem with UHS cards(besides the speed will not be more than lats say 20MB/s) Jan 12 22:48:19 freemangordon, WizardNumberNext: I'm getting seq. read of up to 18MB/s and seq. write of approx 8.5MB/s with UHS-1 Jan 12 22:49:15 freemangordon: 20MB/s I would love it. At the moment I cannot seam to get decent speed of neither SD nor MMC. Barely 1.5MiB/s. But I am using data=journal, which efectively halfs speed Jan 12 22:50:55 data=writeback nobarrier or you're a wuss Jan 12 22:51:30 kerio data=journal - it minimizes data lost Jan 12 22:51:59 as i said, wuss :P Jan 12 22:52:02 if you would loose 400GiB of DATA you would do everything to prevent it from happening again Jan 12 22:52:18 you can't lose important data Jan 12 22:52:24 if it was important, you'd have backups Jan 12 22:52:35 now I have Jan 12 22:52:42 data=journal isn't a backup! Jan 12 22:52:49 but I am not makeing backup of n900 everyday Jan 12 22:52:58 wirr: 15 MB/s here, depending on the block size(dd that is). but that card impresses with random i/o, at least so far Jan 12 22:53:06 every day is a bit too much, but once every three or four days is recommended Jan 12 22:53:12 it is not backup, but it helps to keep data Jan 12 22:53:16 it takes so little effort Jan 12 22:53:25 as data is written first Jan 12 22:53:54 WizardNumberNext: I am talking about raw speed, not fs Jan 12 22:54:00 I think I might import my backup script Jan 12 22:54:29 WizardNumberNext: just install backupmenu Jan 12 22:54:30 freemangordon, then I should expect something close to 20mb/s Jan 12 22:54:37 I have backupmenu Jan 12 22:54:47 live backups won't really work Jan 12 22:55:00 but backupmenu doesn't make backup automaticaly to all my server drives at once Jan 12 22:55:10 oh i thought you meant on the n900 Jan 12 22:55:20 I meant on my n900 Jan 12 22:55:28 oh, *to* the servers Jan 12 22:55:32 I run NAS on server Jan 12 22:55:47 backupmenu then usb mass storage and copy the tarballs Jan 12 22:56:00 so you can bzip2 on a computer with a real cpu Jan 12 22:56:09 yes, to server, as 10 drives is 10 times bigger data security, then one SD card Jan 12 22:56:12 WizardNumberNext, try increase the commit time to something crazy like 300 seconds Jan 12 22:56:47 In general though, I'd say that all sorts of data integrity cleverness is completely foiled by SD behaviour... to the point that for example running fsck on SD can make matters far worse than they are Jan 12 22:56:56 ShadowJK: that is actually better, then data=journal Jan 12 22:57:28 It's commit=1 in Maemo, to maximize interactivity Jan 12 22:57:34 shit, I have ext4 on SD Jan 12 22:57:54 ext4 does much better than ext3, which does much better than ext2, as far as speed goes Jan 12 22:58:20 oh, speed-wise ext4 is a king Jan 12 22:58:33 Something about its access patterns being different Jan 12 22:59:38 \I was running resierfs3 for years and years, few weeks after ext4 went into kernel I started to use ext4 and it is far faster, then reiserfs and doesn't have problem with reiserfs loopback on reiserfs parition Jan 12 23:00:39 ext4 lacks in spouse-killing, though Jan 12 23:00:41 nilfs2 would probably be fastest fs for SD right now Jan 12 23:00:56 that was horrible Jan 12 23:01:02 spouse killing :( Jan 12 23:01:02 need to compare features though Jan 12 23:07:21 trying to broadcast music with fm transmitter. I can hear beep sound every 5 seconds and then it turns off after a minute. I'm not using the default music player, I'm using cmus. It thinks that there's no audio playing, that's why it beeps and then turns off. How can I make it accept cmus sounds? Jan 12 23:11:52 how come the sound from default music player is different from the sound from cmus? Jan 12 23:12:13 you need libplayback or something like that Jan 12 23:13:56 Sysaxed, yes you need to use libplayback library for that Jan 12 23:14:19 uhhh, I can simply install it? Jan 12 23:14:48 that library is closed and poor documented Jan 12 23:14:58 look at OMP code how is using Jan 12 23:15:26 Sysaxed, you need to add libplayback support to your sound application Jan 12 23:15:31 Pali: did 3.8 boot? Jan 12 23:15:40 Pali, well, cmus is not my application :)) Jan 12 23:15:41 seriously guys, how can i create a new page on the wiki? there seems to be no way to register for a new account... Jan 12 23:15:50 freemangordon, rescueOS yes Jan 12 23:15:57 wow, great Jan 12 23:16:02 with fbcon? Jan 12 23:16:06 yes Jan 12 23:16:08 :D Jan 12 23:16:15 also usb network woking Jan 12 23:16:20 that's good Jan 12 23:16:22 * freemangordon guesses Pali don;t expect fremantle to boot :D Jan 12 23:16:23 what about wifi? Jan 12 23:16:30 or... hm Jan 12 23:16:33 the uSD? Jan 12 23:16:37 and emmc Jan 12 23:16:54 Pali is there any hack? Maybe I can fake some sound playing? Jan 12 23:17:27 ~interim-dns Jan 12 23:17:27 Sysaxed: play 4'33" on mediaplayer Jan 12 23:17:28 i heard interim-dns is "http://mwkn.net/2013/01/community.html and http://maemo.cloud-7.de/migration-project/DNS-TBD-.etc.hosts" Jan 12 23:17:30 :) Jan 12 23:17:34 wirr ^^^^ Jan 12 23:18:04 kerio, wifi was not tested, but wl1251 driver was autoloaded Jan 12 23:18:18 eMMC was not detected Jan 12 23:18:29 is omap_hsmmc not in mainline? Jan 12 23:18:30 s/eMMC/internal slot/ Jan 12 23:18:32 Pali meant: internal slot was not detected Jan 12 23:18:54 and external slot was detected, but inserted SD card not :-( Jan 12 23:18:59 kerio: it should be, I guess some configuration Jan 12 23:19:05 omap hs mmc driver is in upstream Jan 12 23:19:36 hm, does the upstream driver deal with the backcover sensor too? and how? Jan 12 23:19:49 Sysaxed, libplayback is needed Jan 12 23:19:59 there is no other good solution for that Jan 12 23:20:18 kerio, yes Jan 12 23:20:25 it show me that cover is closed Jan 12 23:20:34 Pali: maybe some program that does a dummy playback Jan 12 23:20:51 so you can then do whatever you want with non-libplayback-aware programs Jan 12 23:21:04 kerio, yeah, that's what I'm asking about Jan 12 23:21:07 kerio, not possible Jan 12 23:21:13 why>? Jan 12 23:21:23 sound from system (like events) are high priority Jan 12 23:21:29 so what? Jan 12 23:21:39 Pali: his issue is that fmtx turns off after a bit if there's no playback Jan 12 23:21:58 Sysaxed: as i said, play 4'33" in loop with Mediaplayer Jan 12 23:22:01 :D Jan 12 23:22:02 but there is a playback... a perfect playback! Jan 12 23:22:09 kerio, yeah, that's what I'm going to try Jan 12 23:22:14 maybe it will work Jan 12 23:22:21 try Jan 12 23:22:34 maemo auto system is very bad Jan 12 23:23:15 libplayback, patched pulseadio, ohm, ohm plugins, libprolog, prolog plugins, pulseaudio policy, alsa policy, alsaped, .... Jan 12 23:23:36 do not forget for cmtspeech, fmtx or bluez, ... Jan 12 23:24:07 also dres Jan 12 23:25:06 freemangordon, 76 kernel patches for 3.8-rc3 are here: https://atrey.karlin.mff.cuni.cz/~pali/linux-n900/ Jan 12 23:33:51 wirr: I don't thin k a full resume is needed, rather some info about your qualification and your relation to maemo, plus a statement what you think you can do and for how much Jan 12 23:35:02 Pali: thanks, I was already using the new IP for wiki.maemo.org. still i can't find the possibility to register a new account or create a new page :-( Jan 12 23:36:53 DocScrutinizer05, ATM I'm rather thinking of donating some time and money :) I already have a payed daytime job... so i'll contact the council by email Jan 12 23:36:59 wirr, when I'm logged in I can create new pages the usual way. Haven't tried to register new account, for obvious reasons. Jan 12 23:39:35 you need to register new account on garage.maemo.org Jan 12 23:39:41 kerio, default music player blocks music from cmus Jan 12 23:39:55 I think that wiki using same login database Jan 12 23:40:05 Pali: ah, thats the point, thanks Jan 12 23:40:13 so playing silence helps Jan 12 23:40:14 wtf Jan 12 23:40:17 hold on... Jan 12 23:40:17 DocScrutinizer05 ^^ correct? Jan 12 23:40:49 wtf just happened Jan 12 23:40:54 Pali: might be, I actually can't recall anymore Jan 12 23:41:22 DocScrutinizer05, do you know what exactly is testserver doing in n900? Jan 12 23:41:44 not exactly, might be jumpserver for login Jan 12 23:41:44 /usr/bin/testserver Jan 12 23:41:50 oooh Jan 12 23:41:54 on N900 Jan 12 23:42:02 DocScrutinizer05 Jan 12 23:42:06 it's errrr, NFC really Jan 12 23:42:07 http://maemo.org/packages/view/testserver/ Jan 12 23:42:27 probably for preproduction or whatever Jan 12 23:42:38 that daemon is starting in LOCAL or TEST mode Jan 12 23:42:50 so is server application for phoenix? Jan 12 23:43:01 TEST mode is only via test mode battery afaik Jan 12 23:43:09 yup Jan 12 23:43:11 no Jan 12 23:43:16 basically, sth like that Jan 12 23:43:19 phoenix can boot to test mode too Jan 12 23:43:25 maybe Jan 12 23:43:26 you have button for that Jan 12 23:43:31 OK Jan 12 23:43:37 oh no Jan 12 23:43:38 :C Jan 12 23:43:38 I honestly never cared too much Jan 12 23:43:47 can remove it? :-) Jan 12 23:43:58 some free space on rootfs Jan 12 23:44:06 on your own system you're for sure free to remove it Jan 12 23:44:13 ;-) Jan 12 23:44:40 I'd be reluctant to remove it in e.g. CSSU unless we understand exactly why we don't need and don't want it anymore Jan 12 23:45:45 btw, is there (on irc or tmo) somebody (ex)nokian which know maemo system? Jan 12 23:46:02 hah, good question Jan 12 23:46:29 I guess nobody living or dead meets that spec ;-D Jan 12 23:47:13 it actually would be interesting who's been the system architects of maemo Jan 12 23:47:45 yes, it should be interesting Jan 12 23:48:24 I'd consider everything not written in the docs as lost knowledge Jan 12 23:48:25 ari jaaksi? Jan 12 23:48:40 rings a bell here Jan 12 23:49:08 but I don't know if it's the system-architect bell it rings Jan 12 23:51:23 Pali: (testserver) it *might* also be involved in NOLO flashing MyDocs Jan 12 23:51:46 no, for flashing eMMC is used softupd Jan 12 23:51:55 in UPDATE mode Jan 12 23:51:59 among others, yes Jan 12 23:52:11 mhm Jan 12 23:52:57 I guess all that is a rather white area on maemo's map Jan 12 23:53:23 only those who been there know there be dragons Jan 13 00:23:00 freemangordon, heh, crystaldiskmark claims 157.7 iops 4k rnd write for the adata card Jan 13 00:42:28 Pali: do you have initrd for maemo for 3.8-rc3 as well? If so, can you share both kernel and initrd? Jan 13 00:44:25 * WizardNumberNext is late with last question **** ENDING LOGGING AT Sun Jan 13 02:59:58 2013