**** BEGIN LOGGING AT Sat Feb 02 03:00:00 2013 Feb 02 06:43:12 ~botsnack Feb 02 06:43:13 DocScrutinizer06: thanks Feb 02 06:53:16 Good morning :) Feb 02 07:00:23 morning Feb 02 07:05:20 grrr, FW expired the 18h uptime it usually works semi-useful Feb 02 07:07:23 [notice]FW reboot pending. 18h uptime expired. Semi-frozen Feb 02 07:07:43 DocScrutinizer05: more like, exceeded? Feb 02 07:08:13 or that Feb 02 07:08:56 I meant "used up, no more left over of it" Feb 02 07:09:01 yup seems like it got frozen 1-2mins back Feb 02 07:09:37 Sa 2. Feb 06:57:51 CET 2013 real 0m18.894s Feb 02 07:09:39 Sa 2. Feb 07:03:10 CET 2013 real 2m27.017s Feb 02 07:09:40 Sa 2. Feb 07:10:37 CET 2013 real 0m16.212s Feb 02 07:09:54 that's Feb 02 07:10:03 Sat Feb 2 15:01:52 SGT 2013 real 0m23.019s Feb 02 07:10:04 Sat Feb 2 15:07:15 SGT 2013 real 0m11.360s Feb 02 07:10:12 that's 4mins back here :S Feb 02 07:10:29 and FW got rebooted 1min back Feb 02 07:10:38 :D Feb 02 07:11:12 Broadcast message from root (pts/0) (Sat Feb 2 09:08:04 2013): Feb 02 07:11:14 The system is going down for reboot NOW! Feb 02 07:13:37 o.O Feb 02 07:13:48 still not back Feb 02 07:15:10 GRRRRRR Feb 02 07:15:51 yikes down again? Feb 02 07:18:58 seems it recovered, just took a 10min off time ;-P Feb 02 07:19:11 yup back now Feb 02 07:19:24 it needed a botsnack too :p Feb 02 07:19:51 I'll give it a botkick in da A**e Feb 02 07:19:59 lol Feb 02 07:20:08 POS Feb 02 07:21:29 out of curiousity, is this channel related to the hildon foundation? Feb 02 07:21:30 LOL Feb 02 07:21:33 root@MaemoFirewall:~ # uptime Feb 02 07:21:35 09:21:47 up 11 min, 1 user, load average: 0.00, 0.02, 0.03 Feb 02 07:21:43 a LIIIEEE, LIIIEEEEE Feb 02 07:21:44 haha Feb 02 07:24:08 hmm guys i dont know what im doing wrong but my built in browser cant install cssu Feb 02 07:24:10 Feb 2 09:08:15 MaemoFirewall syslog-ng[2381]: syslog-ng shutting down; version='2.0.10' Feb 02 07:24:11 Feb 2 11:10:13 MaemoFirewall kernel: [ 0.000000] Linux version 2.6.32.43-57.e43.i586 (root@andrew) (gcc version 4.1.2 20070626 (Endian 4.1.2-14)) #1 SMP Wed Aug 1 Feb 02 07:24:13 :-/ Feb 02 07:25:26 damo22: you probably need some patience since the original repository.maemo.org is still sloooow Feb 02 07:26:04 pwnguin: well, kinda Feb 02 07:26:19 pwnguin: ...since HiFo is realted to maemo Feb 02 07:26:27 related even Feb 02 07:26:39 im just confirming they're not a splinter group Feb 02 07:26:52 hehehe Feb 02 07:27:05 DocScrutinizer05: they aren't a splinter group :p Feb 02 07:27:12 no, HiFo is maemo's official treasurer Feb 02 07:27:18 what is supposed to come on the screen when the cssu installation starts? Feb 02 07:27:45 i get a text window with config info about cssu repo Feb 02 07:27:58 that sounds about right Feb 02 07:28:11 so do i just wait and see what happens? Feb 02 07:28:40 yes, as long as there's no error getting thrown in that window Feb 02 07:28:47 ok Feb 02 07:29:17 with our current extremely slow repo it might take an hour or longer till stuff gets done Feb 02 07:29:40 what's going on with the site? Feb 02 07:29:54 hm? Feb 02 07:30:03 ok Feb 02 07:30:11 firewalls crashing, repos slow Feb 02 07:30:22 expiring keys Feb 02 07:30:36 firewall is a POS Feb 02 07:31:12 repo moved from a Akamai server farm with several hundred servers to a VM with a 50Mbit uplink Feb 02 07:31:32 key expired since Nokia didn't take care to renew it in time Feb 02 07:31:38 akamai server farm seems like overkill Feb 02 07:32:38 well, the farm been overkill, what we now got is too weak Feb 02 07:32:45 Doc: What do you need the firewall for in the first place? Feb 02 07:33:14 RST38h: I don't need it neither implemented it first place Feb 02 07:33:26 So, why not remove it? Feb 02 07:34:00 because rack's network config is outside our domain of administrative power Feb 02 07:34:12 is it supposed to be stuck in a web browser with [install] catalogues=community-testing.... etc Feb 02 07:34:18 Why not ask the administrator to disable the firewall? Feb 02 07:34:47 the administrator already suggested we (they) might do exactly that Feb 02 07:35:19 but that needs a grid-guy to create external IF on each VM Feb 02 07:35:35 we started with repo Feb 02 07:35:42 which is already exposed Feb 02 07:35:44 it seems to me that the .install script is not being detected as the right mime type and its just displaying the script in the web browser instead of executing it Feb 02 07:35:48 Aha Feb 02 07:36:09 damo22: that's bad. Which browser you're using? Feb 02 07:36:19 the Web thing Feb 02 07:36:24 built into nokia Feb 02 07:36:26 should work Feb 02 07:37:21 is there a way to add this thing as an apt repo and do it from the command line? Feb 02 07:37:45 damo22: sorry I'm not up to date and fluent on talking you through manual installation of cssu-s Feb 02 07:38:18 i just want to know if its possible Feb 02 07:38:26 yes, basically you can enter the info from the .install into HAMs catalogs settings under "new catalog" Feb 02 07:38:39 great Feb 02 07:38:42 i can work it out Feb 02 07:38:54 then install the app in .install via apt-get and execute it Feb 02 07:39:05 good shit Feb 02 07:39:30 the .install is just a convenience shortcut for the former actions Feb 02 07:41:42 damo22: fun stuff: it needs the exact name of the repo too Feb 02 07:42:05 damo22: it's probably better if you download the .install and open it with file manager Feb 02 07:42:43 well, it's not "need", but the enabler won't work right wrt repos, and you'll have to manually change them before doing the install Feb 02 07:43:25 DocScrutinizer05: rmo throughput? still the same? Feb 02 07:45:01 hardware question :anyone measured the current draw of an n900 while charging? mine seems to only draw 400mA which seems low... Feb 02 07:45:05 kerio: last figure from pre-boot: 23.8Mb Feb 02 07:46:12 ecc2g: that is quite normal, depending on charge state of your LiIon battery Feb 02 07:46:29 i typed it in manually, looks like its doing something good Feb 02 07:47:18 kerio: 28.5Mb Feb 02 07:47:30 about 25%... I was trying to figure out why the ac10u was 1200mA... Feb 02 07:48:50 ecc2g: that's max load Feb 02 07:49:20 LiIon charging in N900 follows a CCCV scheme though Feb 02 07:49:39 which means that you will eventually taper off with current Feb 02 07:50:08 damo22: yeah but the enabler is supposed to... nevermind, you're installing cssu stable right? Feb 02 07:50:21 actually battery never will charge with 1200mA (except when e*_ tries it) Feb 02 07:50:31 im putting on testing Feb 02 07:51:04 and 5V is step-down transformed to the battery's 3.7 avr voltage, which also means you got lower current on 5V than on battery charge Feb 02 07:51:51 damo22: I heard cssu-t has some problems with the expired key on Nokia's repos Feb 02 07:51:57 IIRC Feb 02 07:52:15 dang Feb 02 07:52:17 better install cssu-s, you can 'upgrade' to cssu-t any time later Feb 02 07:52:44 nah, it really doesn't Feb 02 07:52:45 during bulkcharge phase it should draw max current, the taper should happen above 4v and around 80+% capacity... Feb 02 07:53:00 basically yes Feb 02 07:53:04 i mean, what are the missing packages? Feb 02 07:53:43 warfare: whenever you can, .install files have the wrong MIME, they shouldn't be text Feb 02 07:53:59 kerio: I dunno. as already mentioned i'm not up to date and fluent with manual installation of CSSU Feb 02 07:54:28 kerio: WTF happened to the mime types then? Feb 02 07:54:35 it should be application/x-install-instructions Feb 02 07:54:40 DocScrutinizer05: nginx happened :) Feb 02 07:54:41 5v400mA is 2w, bl5j is 5wh, means 2.5h to charge at 100% efficiency... Feb 02 07:54:42 ok im putting stable on Feb 02 07:54:58 kerio: also, please consider warfare is our *sysiop*, not our repo/www maintainer Feb 02 07:55:15 that mime thing sucked but i worked it out manually Feb 02 07:55:29 DocScrutinizer05: who has write access to /etc/nginx.conf? Feb 02 07:55:44 where? Feb 02 07:55:47 rmo Feb 02 07:55:56 meh, that's warfare Feb 02 07:56:09 toldya :P Feb 02 07:56:19 do you say .install is on repo.m.o? Feb 02 07:56:21 that's why i said "whenever you have time" Feb 02 07:56:31 DocScrutinizer05: they are, afaik Feb 02 07:56:45 I'd like to bet against that Feb 02 07:57:03 well, dunno Feb 02 07:57:07 maybe they are Feb 02 07:57:09 * kerio wins the bet Feb 02 07:57:20 ~cssu Feb 02 07:57:20 methinks cssu is http://wiki.maemo.org/Community_SSU Feb 02 07:57:24 Doc: http://repository.maemo.org/community/community-fremantle.install Feb 02 07:57:35 I tried hooking up to a bench supply, what bothered me a bit is that the n900 seems to present a near linear resistive load... which is not what a switching supply should look like... Feb 02 07:57:40 you're right Feb 02 07:57:50 of course i am Feb 02 07:57:52 ~kerio Feb 02 07:57:52 somebody said kerio was kerio is right Feb 02 07:57:57 its a bit of a shame because tomorrow i have a chance to convince my friend to get a n900 instead of an iphone, but if the repo is sloooow its not gonna be a good demonstration Feb 02 07:58:06 lol Feb 02 07:58:07 ~useless Feb 02 07:58:08 * infobot starts crying and hides from docscrutinizer05 in the darkest corner of the room. :( Feb 02 07:58:25 haha Feb 02 07:58:46 damo22: if your friend needs a demo to choose the n900 over the iphone then he really should get the iphone Feb 02 07:58:59 she Feb 02 07:59:04 :-D Feb 02 07:59:32 how about adding an entry to your N900 /etc/hosts? ;-P Feb 02 08:00:08 (semi-serious suggestion, probably won't fly) Feb 02 08:00:24 damo22: unless your friend is looking to buy a 3 year old iphone, that comparison might be worthwhile =( Feb 02 08:01:06 * kerio feels that extensive multitasking is still in favour of the n900 Feb 02 08:01:55 true Feb 02 08:02:31 i might have replaced my n900 with a galaxy nexus, and am slightly sad at the multitasking Feb 02 08:02:32 http://talk.maemo.org/showthread.php?p=1320044#post1320044 Feb 02 08:02:47 i just swapped my iphone 4s for n900 Feb 02 08:02:57 lolwut Feb 02 08:03:35 well not literally, i still have both here im just trying to see if i can do everything i need on this Feb 02 08:03:46 just that it's not a "first step2 but rather the last step of first of three hops Feb 02 08:03:58 s/2/"/ Feb 02 08:03:58 DocScrutinizer05 meant: just that it's not a "first step" but rather the last step of first of three hops Feb 02 08:04:13 damo22: there's a compatibility layer to run angry birds Feb 02 08:04:16 you'll be fine Feb 02 08:04:24 i never used angry birds Feb 02 08:04:27 in my life Feb 02 08:04:40 :D Feb 02 08:05:08 kerio: since when does N900 need a compatibility layer to run angry birds? Feb 02 08:05:09 i just want this phone because i like to be in control of my own computing Feb 02 08:05:36 and i dont want someone telling me one day i cant make a phonecall because i havent downloaded 900Mb update Feb 02 08:05:46 :P Feb 02 08:05:50 afaik AB been a native fremantle app, to start with. Only then got migrated to other platforms Feb 02 08:06:25 DocScrutinizer05: true its been available on fremantle since the first pr-release but i don't know if it was launched here first or on iOS Feb 02 08:06:29 touchscreen seems to suck though Feb 02 08:06:53 although i havent removed the film off the front yet Feb 02 08:07:37 damo22: I always thought N900 touchscreen is the best I ever used - incl any crappy capacitive TS on those other phones Feb 02 08:07:53 damo22: so yes, remove that film, might help a lot Feb 02 08:07:57 ok Feb 02 08:08:14 be careful though, not to damage the digitizer Feb 02 08:08:56 it can't stand neither punctures nor excessive pulling force Feb 02 08:09:03 how do i uninstall cherry Feb 02 08:09:13 is it supposed to be in the list somewhere? Feb 02 08:09:24 damo22: it's not a user package Feb 02 08:09:29 apt-get? Feb 02 08:09:30 apt-get remove cherry Feb 02 08:09:34 as root Feb 02 08:10:10 apt-get pop cherry Feb 02 08:10:23 after you installed cssu, otherwise you shouldn't accept what apt-get suggests ;-) Feb 02 08:11:06 odds are it would uninstall ~70% of your system, on a stock fremantle OS Feb 02 08:11:16 DocScrutinizer05: nah Feb 02 08:11:23 nothing depends *on* the metapackage Feb 02 08:11:25 or you simply install notmynokia app Feb 02 08:12:01 DocScrutinizer05: if you install nmn, remove cherry and remove nmn, you *have* cherry and no way to uninstall it Feb 02 08:12:30 o.O ?? Feb 02 08:13:15 is it wise to update the official maemo stuff from nokia? Feb 02 08:13:34 as in "system update"? Feb 02 08:13:38 in HAM? Feb 02 08:13:53 um, not sure if i have ham yet Feb 02 08:14:02 you have HAM Feb 02 08:14:15 ok well i got a funny ! icon Feb 02 08:14:17 it's the default package manager of hildon Feb 02 08:14:37 so i touched it and it said official nokia updates are available Feb 02 08:14:40 yes, and a "system upgarde 2011.38.2" inside? Feb 02 08:14:45 for maemo 5 Feb 02 08:14:52 forget it Feb 02 08:15:04 you're supposed to have PR1.3 flashed Feb 02 08:15:15 this maemo5 update is for Pr1.3.1 Feb 02 08:15:24 CSSU will ship that for you Feb 02 08:16:04 so no, you don't need the 1.3.1 system upgrade Feb 02 08:16:08 ok Feb 02 08:16:16 ignroe it Feb 02 08:16:20 proceed Feb 02 08:16:55 i need bash, sudo, and a few tools Feb 02 08:17:12 sudo already there Feb 02 08:17:15 busybox sucks balls Feb 02 08:17:21 get the rest via apt-get Feb 02 08:17:26 ~jrtools Feb 02 08:17:27 i heard jrtools is http://wiki.maemo.org/User:Joerg_rw/tools Feb 02 08:17:36 ahh Feb 02 08:20:57 rmo : 32B/s Feb 02 08:21:11 :P Feb 02 08:22:00 GRRR http://wiki.maemo.org/N900_The_Perfect_Setup -> another staement about SR " will significantly increase battery life." without any evidence Feb 02 08:22:38 damo22: this may change any moment, some connections see 300kBps Feb 02 08:22:44 yours too, eventually Feb 02 08:22:51 nice Feb 02 08:24:12 DocScrutinizer05: kerio predicted your reaction to that page yesterday: 22:51 * kerio flees from doc's area-of-effect wrath Feb 02 08:24:42 it was for the "kernel-power is recommended" thing Feb 02 08:34:02 yes, I know Feb 02 08:34:36 I fought myself and won, to not rant on suggesting kernel-power in itself Feb 02 08:35:30 after all the page gives due credit to overclocking being deprecated Feb 02 08:36:44 I still don't see why KP is beneficial for beginners, apart form the suggested SR power savings which never been tested in RL Feb 02 08:36:59 and are deprecated by Nokia anyway, afaik Feb 02 08:37:32 since iirc there's some hw instability in at least some devices, with SR Feb 02 08:38:03 and after all SR is just a hw based way to do what fremantle stock kernel does in sw Feb 02 08:38:37 so the supposed power savings of SR are questionable at least Feb 02 08:41:13 undervolting is a completely unrelated topic. and not exactly recommended either. So what's the beginner's benefit in using PK? Feb 02 08:43:07 DocScrutinizer05: h-e-n :P Feb 02 08:43:15 damo22: any progress? Feb 02 08:43:43 kerio: for H-E-N you'll get advice which kernel to install Feb 02 08:43:57 as soon as you install the H-E-N package Feb 02 08:49:03 hm, why doesn't it also provide modules for stock? Feb 02 08:49:39 because USB is monolithic in stock kernel Feb 02 08:49:51 oic Feb 02 08:53:06 btw, rmo is sloooooooooooooow Feb 02 08:55:08 I knooooooooow Feb 02 09:09:25 kerio: I'll add the mime-type (although this should be in apache, as nginx is only the proxy for slow connections) Feb 02 09:10:19 warfare: maybe it should? i suppose that the migration changed the apache configuration and no one noticed the problem because... well, rmo was down Feb 02 09:11:48 I think so. Feb 02 09:16:36 Content Type fixed. Feb 02 09:17:03 curl -vvI http://repository.maemo.org/extras/install/extras-devel.install => Content-Type: application/x-install-instructions Feb 02 09:18:31 warfare: i might be repeating myself, but you da man Feb 02 09:18:57 ~warfare Feb 02 09:21:45 infobot: #maemo warfare is Falk Stern, maemo.org master sysop, and also DA MAN Feb 02 09:21:46 kerio: okay Feb 02 09:36:35 fasten seatbelts! warfare going to fire up the booster on repo ;:-D Feb 02 09:38:55 KILL. KILL. DESTRRROY. Feb 02 09:42:02 booster? :O Feb 02 09:42:51 kerio: we have some heavy *writing* on HDD Feb 02 09:43:00 ...WHAT? Feb 02 09:43:14 who? why? Feb 02 09:43:16 when? Feb 02 09:43:35 it makes no sense, repo. is static right now Feb 02 09:43:37 I'm installing iotop to look for the culprit. Feb 02 09:46:01 atime Feb 02 09:46:07 my suspect Feb 02 09:46:30 ooh, that's a good one Feb 02 09:46:48 i was going to suggest some excessive logging Feb 02 09:46:51 DocScrutinizer05: who keeps enabled atime? :D Feb 02 09:46:52 ...which is what atime is, pretty much Feb 02 09:47:03 we already disabled logging, didn't change Feb 02 09:47:27 DocScrutinizer05: So let's try remount with noatime ;) Feb 02 09:47:36 :nod: Feb 02 09:48:35 DocScrutinizer05: why not? Feb 02 09:49:55 I'm taking down apache for a second. Feb 02 09:51:23 WUT? from 90+%wa to <10% Feb 02 09:52:00 aah, apache been down Feb 02 09:54:11 DocScrutinizer05: mount -o remount,noatime,nodiratime ? Feb 02 09:54:31 already done that. Feb 02 09:54:59 sorry, /dev/xvda on / type ext3 (rw,noatime,nodiratime) Feb 02 09:55:11 nobarrier data=writeback! Feb 02 09:55:18 :) Feb 02 09:56:34 I'm switching from nginx to varnish. nginx keeps swapping objects to disk Feb 02 09:57:00 warfare: not lighttpd? Feb 02 09:57:17 eeek, swapping Feb 02 09:57:25 also, can't we disable the swap? Feb 02 09:57:56 anyway, wa gone done by ~30% Feb 02 09:58:01 kerio: there is no swap (at least from the kernel side) Feb 02 09:58:02 absolute Feb 02 09:58:41 kerio: I am not familiar with lighttpd, so atm I stick to the tools I know :) Feb 02 09:59:02 or so I thought Feb 02 09:59:06 fair Feb 02 09:59:58 (watch a professional at work. makes my day :-D ) Feb 02 10:00:50 oh, it's the firewall that has swap, right Feb 02 10:01:13 nope Feb 02 10:01:33 we're not looking at firewall now, since repo is exposed anyway Feb 02 10:01:40 yes, yes Feb 02 10:01:58 nevertheless FW has no swap either Feb 02 10:02:56 actually I think swap doesn't make sense on low latency systems Feb 02 10:03:20 autobuilder, ok. not on fileservers or firewalls though Feb 02 10:04:40 Should be better now Feb 02 10:05:56 Cpu(s): 1.1%us, 0.9%sy, 0.0%ni, 95.2%id, 0.1%wa, 0.0%hi, 1.4%si, 1.3%st Feb 02 10:05:57 Mem: 2049296k total, 1966308k used, 82988k free, 12960k buffers Feb 02 10:06:01 Still pushing ~26M but with no cpu load now. Feb 02 10:06:19 :nod: Feb 02 10:06:38 ~93% of all requests are served from the cache. Feb 02 10:06:38 warfare: okay Feb 02 10:06:40 quite a difference Feb 02 10:07:00 ~forget 93% of all requests Feb 02 10:07:00 i forgot 93% of all requests, DocScrutinizer05 Feb 02 10:07:38 ;-) Feb 02 10:07:42 Oh, "~" is for the bot? Feb 02 10:07:49 attention key Feb 02 10:07:52 ~warfare Feb 02 10:07:53 warfare is, like, Falk Stern, maemo.org master sysop, and also DA MAN Feb 02 10:09:23 those 28Mbps uplink might be ISP's traffic shaping actually Feb 02 10:09:28 ~botsnack Feb 02 10:09:28 kerio: thanks Feb 02 10:09:44 DocScrutinizer05: well, now that we're not cpu limited... Feb 02 10:09:51 can't we get nemein to give us more bandwidth? Feb 02 10:09:59 like, if we ask pretty, pretty please :3 Feb 02 10:10:03 we can try Feb 02 10:10:24 I need some coffee. Feb 02 10:10:35 enjoy! :-D Feb 02 10:13:15 4957 root 20 0 693m 464m 3368 S 5 23.2 1:26.78 iftop :-o Feb 02 10:13:48 464MB? Really? Feb 02 10:14:05 And we are not talking overcommit here? Feb 02 10:14:11 not Really, rather RES Feb 02 10:14:34 that's top Feb 02 10:14:42 of top ;-) Feb 02 10:15:23 I'm inklined to give it a kill Feb 02 10:16:14 PID USER PR NI VIRT RES SHR S PU %MEM TIME+ COMMAND Feb 02 10:16:16 11584 root 20 0 374m 143m 3476 S 4 7.2 0:23.79 iftop Feb 02 10:16:17 4957 root 20 0 693m 502m 3368 S 4 25.1 1:33.82 iftop Feb 02 10:16:19 12182 root 20 0 309m 83m 3476 S 4 4.2 0:13.23 iftop Feb 02 10:16:20 10033 nobody 20 0 10.6g 609m 27m S 1 30.5 0:19.18 varnishd Feb 02 10:16:30 why three? Feb 02 10:16:39 :shrug: Feb 02 10:16:46 kill them with fire Feb 02 10:16:51 we're two users there, i got one ;-) Feb 02 10:16:57 or kill them with TERM Feb 02 10:18:34 killed mine, which turned out to be the greedy one Feb 02 10:19:00 humm ...bandwidth maniacs? :) Feb 02 10:19:20 moo xes Feb 02 10:19:38 xes: plaqying with repo Feb 02 10:20:24 hello there! ...sorry for the meeting, i returned at home after 21.00 :( Feb 02 10:21:50 top porn: http://privatepaste.com/9608d64d82 Feb 02 10:22:00 the last term for migration should be last day of february..or am i wrong? Feb 02 10:22:10 yep Feb 02 10:22:17 for step 2 of 3 Feb 02 10:22:42 which is xen-grid -> maemo iron Feb 02 10:22:50 can varnish really push 26Mb/s with 3% cpu? :o Feb 02 10:23:03 step 3: maemo iron -> Hetzner/osuosl Feb 02 10:23:12 DocScrutinizer05: yay europe Feb 02 10:23:35 step 3 still under evaluation Feb 02 10:24:37 for that step3 we definitely need quite some sysop manpower Feb 02 10:25:03 for step2 it's unclear yet how much of it will be done by Nemein Feb 02 10:25:16 kerio: varnish can push around a gig with almost no cpu. Feb 02 10:25:28 ..don't forget that for sure we/they(nemein) will need at least one week of test of the new irons before transfer vm..indipendently of the virt platform Feb 02 10:25:35 DocScrutinizer05: important thing: keep the timestamps on rmo the same Feb 02 10:25:44 so we don't get flooded to hell by all the n900s Feb 02 10:26:00 :nod: Feb 02 10:26:10 kerio: with a vm transfer all will remain the same Feb 02 10:26:29 :nod: again Feb 02 10:27:05 that's why my plan says "keep the VM until end of step 3" Feb 02 10:27:58 after step3 acomplished (whatever it might be), we can start consolidation to our liking Feb 02 10:30:04 DocScrutinizer05: with our iron, the "firewall" will be the hypervisor, right? Feb 02 10:30:39 unclear yet, my take on it is 2FW? what friggin FW?!" Feb 02 10:30:51 s/2/"/ Feb 02 10:30:51 DocScrutinizer05 meant: unclear yet, my take on it is "FW? what friggin FW?!" Feb 02 10:34:28 concerning the firewall, the easy way would be to use a natted interface of the hypervisor... Easy and plain! ..In the future, some kind of fw with qos would be better Feb 02 10:35:07 xes: I guess we could add that when need arises Feb 02 10:35:21 (some kind of FW) Feb 02 10:35:40 DocScrutinizer05: for sure. The first need is to have all up and running Feb 02 10:35:58 for now Nemein's plans are to get rid of that FW since it's a PITA Feb 02 10:36:13 even in xengrid Feb 02 10:37:38 kerio, I use device with 200mhz and 16 RAM as my irc bouncer... It also shares my whole internet connection@home, servers as ssh and voip gateway, torrent client, and wireless access point :P Feb 02 10:37:53 Estel_: power usage? Feb 02 10:38:03 I think it beats your shevaplug pretty hard, especially, that it costed ~30$, and it was 3 years ago Feb 02 10:38:13 12V 100mA peak ;P Feb 02 10:38:25 well, the dreamplug does have *two* gigabit ethernets and wifi Feb 02 10:38:31 or more than 3 Feb 02 10:38:43 but it cost order of magnitude more Feb 02 10:38:55 about half an order of magnitude :P Feb 02 10:39:04 anyway, it's wrt54gl linksys router, that I'm talking about, powered by openwrt Feb 02 10:39:06 uplink steady at 27Mbps Feb 02 10:39:07 DocScrutinizer05: i think that endian is a good product, but not inside a vm. Too complicated to be managed by a team and ..almost impossible to debug in hours (rather than weeks), so f**k the fw Feb 02 10:39:08 * DocScrutinizer05 waves Feb 02 10:39:23 Estel_: makes sense - i do have a wrt810n that acts as my router Feb 02 10:39:43 but openwrt is flaky on it, and ddwrt is... eeh Feb 02 10:39:49 works, but not quite Feb 02 10:39:51 why flaky? Feb 02 10:39:58 try 2.4 kernel Feb 02 10:40:00 xes: yep Feb 02 10:40:11 2.6 kernel on openwrt = mess, from my experience Feb 02 10:40:22 er, 610n Feb 02 10:40:47 holler if sh*t hits the fan, folks. I'm gone Feb 02 10:41:16 DocScrutinizer05: it's ok, we've got a real sysadmin now >:D Feb 02 10:41:33 root@MaemoFirewall:~ # uptime Feb 02 10:41:34 12:41:23 up 3:32, 1 user, load average: 0.00, 0.00, 0.00 Feb 02 10:41:40 kerio, I don't know for never models, but wrt54gl is quite interesting, as it works exactly as well for 12V power supply godly way, as for 5V from either battery pack or usb Feb 02 10:41:42 so another 15h left Feb 02 10:42:07 Estel_: does openwrt have a web interface? Feb 02 10:42:15 kerio: I never claimed to be any sysop Feb 02 10:42:22 pretty decent, if one want to use it (I do) Feb 02 10:42:27 LuCI Feb 02 10:42:31 DocScrutinizer05: out of curiosity - is FW stable now? Feb 02 10:42:35 kerio: system architect and project leader here Feb 02 10:42:37 Lua scripted interface Feb 02 10:42:42 freemangordon: well, it works Feb 02 10:42:51 with reboots? Feb 02 10:42:55 freemangordon: yes... for 18h Feb 02 10:42:55 Estel_: i might just try it, i suppose Feb 02 10:43:06 DocScrutinizer05: nice Feb 02 10:43:22 with constant 25 MB load? Feb 02 10:43:28 *MB/s Feb 02 10:43:28 nah Feb 02 10:43:30 Mbs Feb 02 10:43:30 kerio, only drawback - as name suggest - it's G wifi standard Feb 02 10:43:45 repo bypassing FW Feb 02 10:43:50 aah, ok Feb 02 10:43:53 I don't see a problem, as I'm allergic to N Feb 02 10:44:26 \/o Feb 02 10:44:54 http://talk.maemo.org/showpost.php?p=1320064&postcount=160 Feb 02 10:44:54 kerio: who loads RMO that badly? devices? Feb 02 10:45:10 problem logging in to package page? Feb 02 10:45:23 kerio: or it is nokia farm? Feb 02 10:45:26 problem logging in to maemo.org Feb 02 10:45:38 freemangordon: 40k active n900s (with only extras active) end up being a download every two seconds, on average Feb 02 10:45:50 known but not assigned yet, it seems Feb 02 10:45:50 kerio, btw, yesterday, some funny technician tried to convince my parents, that their wrt54gl won't support more than 17 Mb/s network access Feb 02 10:46:02 kerio: I know, that is why I ask Feb 02 10:46:34 kerio, although no one seen those 40k devices yet ;) Feb 02 10:46:36 so we are waiting server farm to stop torturing RMO? Feb 02 10:46:45 freemangordon: there's no server farm anymore Feb 02 10:46:55 but why this lload then? Feb 02 10:46:58 *load Feb 02 10:46:59 pooof, gone Feb 02 10:47:11 freemangordon: N900 Feb 02 10:47:26 so yes, *I* seen them Feb 02 10:47:51 ~5000 unique IPs served? Feb 02 10:48:01 lol Feb 02 10:48:12 what? there is enough devices to do do constant 25 MBs? Feb 02 10:48:31 *arew Feb 02 10:48:35 root@stage:~# awk '{print $1}' < /var/log/apache2/*stage_transfer | sort | uniq | wc -l Feb 02 10:48:36 31602 Feb 02 10:48:47 that's for today Feb 02 10:48:57 so we're close to 40k limit, it may end in few days Feb 02 10:48:58 * freemangordon heads to the coffee machine for another coffee :) Feb 02 10:49:08 DocScrutinizer05: we are not logging anymore ;) Feb 02 10:49:13 unless it turns out to be 80k limit:P Feb 02 10:49:14 ooh Feb 02 10:49:41 DocScrutinizer05: Doesn't make sense, as most requests don't even hit the apache anymore. Feb 02 10:49:44 the issue is that we don't actually know if the catalogue updates end or fail due to timeouts Feb 02 10:50:07 so, a shameless question - do we *know* (in terms of zhiterai *&nowing* ;) ) why it works so slow, or we only *suppose* reason? Feb 02 10:50:13 s/we/you/ Feb 02 10:50:13 Estel_ meant: so, a shameless question - do you *know* (in terms of zhiterai *&nowing* ;) ) why it works so slow, or we only *suppose* reason? Feb 02 10:50:25 s/we/you Feb 02 10:50:30 I am curious too Feb 02 10:50:43 We have 1533 established tcp connections to port 80. Feb 02 10:50:59 Divide that by the available bandwidth. Feb 02 10:51:15 (around 26mbit/s) Feb 02 10:51:32 warfare: aah, so we lack network bandwith? ok, makes sense Feb 02 10:51:33 ~26/1533 Feb 02 10:51:33 0.016960208741 Feb 02 10:51:53 hmm, does not make sense :) Feb 02 10:51:55 ~16kb Feb 02 10:52:05 Kbs :) Feb 02 10:52:08 yes Feb 02 10:52:14 I was hungry and ate s Feb 02 10:52:25 well, in fact Feb 02 10:52:25 but why web iface is slow then? Feb 02 10:52:33 ~26*1024 Feb 02 10:52:34 26624 Feb 02 10:52:46 ~26624/1533 Feb 02 10:52:46 17.367253750815 Feb 02 10:52:49 aah, it is Mbs Feb 02 10:53:13 ~17,3Kbs Feb 02 10:53:21 per head Feb 02 10:53:29 doesn't sound so slow, in practice it seems slower Feb 02 10:53:34 than that Feb 02 10:53:34 not that bad, esp for http Feb 02 10:53:41 it's probably not equally-distributed Feb 02 10:53:54 well, in practice it feels like 17 bs Feb 02 10:54:02 :nod: Feb 02 10:54:07 warfare: thought as much when you mentioned that fronend thingie, varnish? Feb 02 10:54:39 Thats probably because of the ratelimiting nemein does. This drops some packets, so tcp retransmits them. Feb 02 10:54:55 * kerio shakes his fist at nemein Feb 02 10:55:29 shameless question 2: Given known circumstances and numbers, do you think there is a chancde for it to *ever* work as intended? Feb 02 10:55:40 or is it dead-man-loop, excuse the pun? Feb 02 10:55:58 s/chancde/chance/ Feb 02 10:56:17 Estel_: we just need more bandwidth :) Feb 02 10:56:33 Estel_: we're working on it Feb 02 10:56:34 the VM is perfectly healthy, 90% idle Feb 02 10:56:38 more bandwidth doesn't fit in "known circumstances" as in "present limits we have" Feb 02 10:56:53 s/we/you/ Feb 02 10:56:54 Estel_ meant: more bandwidth doesn't fit in "known circumstances" as in "present limits you have" Feb 02 10:56:57 damn the habit Feb 02 10:57:05 Estel_: _If_ we can get more bandwidth (preferably 100Mbit/s) varnish could easily saturate that. 90% of all requests ask for the same files. Feb 02 10:57:36 warfare, seems reasonable. do You (as in organization) have money for more bandwidth? Feb 02 10:57:55 warfare: which files? extras Release, extras Packages.gz? Feb 02 10:58:07 kerio: yes. Feb 02 10:58:33 Estel_: I don't know about the monetary situation, but I will go looking for some sponsoring on monday. Feb 02 10:58:41 * Estel_ nods Feb 02 10:58:52 thanks for info... And good luck Feb 02 11:01:25 can't we ask Nemein for a temporary burst to 50mbps or something similar? Feb 02 11:01:39 i bet that once all the active n900s properly update, they won't bug us as much Feb 02 11:01:56 a HEAD is a lot faster than a GET for a 3mb file Feb 02 11:03:43 kerio: do you know how much 50mbps costs instead 20mbps? Feb 02 11:04:21 BTW do we expect that activity to cease? Feb 02 11:04:36 in short term that is Feb 02 11:06:13 oh, shit "3% [9 Packages 93261/3169kB 2%] 5763B/s 8min53s" Feb 02 11:10:44 freemangordon: my HAM update stalled at donno where, extrapolation gave me 90min ETA. 15min later it had finished Feb 02 11:11:04 * freemangordon is doing wget :) Feb 02 11:11:23 might be all the same Feb 02 11:11:43 connection badwidth per client isn't static Feb 02 11:11:49 band* Feb 02 11:12:15 I see connections with 300+kBps here Feb 02 11:12:25 others with 5Bps Feb 02 11:13:38 Top is extras/Packages.gz, followed by extras/Packages, followed by extras-devel/Packages.gz Feb 02 11:14:29 hi, just reactivated my n900 after not using it for about one year, I just started missing it ;) seems some of the old repos aren't up anymore. any recommendations what repo-list to use to get it up to date, please? Feb 02 11:14:32 who the hell is downloading extras/Packages? :s Feb 02 11:14:57 kerio: could that be me? Feb 02 11:15:02 ...why do users start using the n900 again only when the repos are down? :S Feb 02 11:15:17 topro: temporary problems caused by the migration from nokia-owned infrastructure to community-owned infrastructure Feb 02 11:15:20 come on... Feb 02 11:15:32 the repository is working, is just sloooooooooow, but we're working on it Feb 02 11:15:38 in the mean time Feb 02 11:15:38 ok, i see. no problem. i'm patient ;) Feb 02 11:15:39 ~mirrors Feb 02 11:15:40 i guess mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Feb 02 11:16:02 i am downloading the cssu-s stuff Feb 02 11:16:15 maemo-archive.wedrop.it is easy to use, just add "maemo-archive.wedrop.it/" between http:// and repository.maemo.org Feb 02 11:16:20 in your sources.list Feb 02 11:16:39 kerio: i just realized that the old repo list on my n900 contains a lot of stuff. is there a list with currently recommended repos to start over again? Feb 02 11:17:30 just the two nokia repos, extras (or extras-devel if you're feeling brave), and a cssu repo if you installed cssu Feb 02 11:18:00 ok, thanks Feb 02 11:22:00 my repo list, http://paste.debian.net/231180/ Feb 02 11:22:15 ^^ anything that needs a change? Feb 02 11:23:33 you don't need both extras and extras-testing Feb 02 11:23:46 apart from that, sounds good Feb 02 11:23:54 is this your hildon-application-manager.list? Feb 02 11:24:05 it'll be overwritten by HAM whenever you launch it :) Feb 02 11:24:27 yes, it is. where to put infos then? Feb 02 11:24:53 configure HAM Feb 02 11:25:03 or disable every repo in HAM and copy that list in /etc/apt/sources.list Feb 02 11:28:28 topro: also, HAM is a bit weird, he'll not accept upgrades from mirrors of certain repos unless you disable a certain hidden setting Feb 02 11:28:38 but that hidden setting may have UNFATHOMABLE CONSEQUENCES Feb 02 11:29:31 it might turn your n900 into a doorstop? Feb 02 11:30:05 nah Feb 02 11:30:24 ok, tell me more... Feb 02 11:30:24 there's surprisingly little that you can do in software to brick a n900 Feb 02 11:30:36 then whats with the caps Feb 02 11:30:40 :P Feb 02 11:31:13 topro: HAM has something it calls "domains", each with a certain trust level Feb 02 11:31:37 packages in repos belong to a certain domain, and installed packages from those repos also belong in that domain Feb 02 11:31:52 HAM will outright refuse to upgrade a package that comes from a domain with a certain trust level to a domain with a lower trust level Feb 02 11:32:10 repositories are put in domains based on the key used to sign them, and based on some hidden configurations in HAM Feb 02 11:32:17 those configurations also include the URL for the repo Feb 02 11:32:33 what about using FAM? any better, or even worse? Feb 02 11:32:36 so any repo *you* create as a user will not be considered belonging in a domain different than the "untrusted" one Feb 02 11:32:38 ~fam Feb 02 11:32:39 well, fapman is Faster Application Manager, a bad package manager that causes problems, don't use it, ever Feb 02 11:33:00 ok, i see. i've been away for too long ;) Feb 02 11:33:07 but considering that the nokia repos right now have 0 trust because their key is expired... i feel that there's very little you can do except disable that check in HAM Feb 02 11:33:17 enter redpill mode, go to settings, and untick "ignore packages from wrong domains" Feb 02 11:34:02 that will let HAM know that you won't be ok with its bullshit Feb 02 11:34:05 can't find this setting, is it a HAM setting or somewhere in system settings? Feb 02 11:34:19 it's a HAM setting, but it's hidden Feb 02 11:34:41 http://wiki.maemo.org/Red_Pill_Mode Feb 02 11:34:45 the I missed the red pill I think Feb 02 11:34:49 ah, ill see Feb 02 11:35:17 hmm, "There is currently no text in this page...." Feb 02 11:35:29 oh, sorry Feb 02 11:35:30 http://wiki.maemo.org/Red_Pill_mode Feb 02 11:35:35 fscking case sensitiveness Feb 02 11:37:18 kerio: I'm honestly concerned about your health! :-o Where's gone all your usual trolling? ;-D Feb 02 11:37:30 i don't even have that configuration file /home/user/.osso/hildon-application-manager Feb 02 11:37:48 nah, that's hardly possible Feb 02 11:38:04 if there's actually no such file, create it Feb 02 11:40:48 http://privatepaste.com/9b452a7c24 <- /home/user/.osso/hildon-application-manager Feb 02 11:40:54 *my* /home/user/.osso/hildon-application-manager Feb 02 11:41:40 doc: thanks Feb 02 11:42:38 kerio always claims red pill mode never really vanished from HAM. I'm not all that sure about it Feb 02 11:43:12 you'll learn if it works when (or when not) there are new settings showing up in HAM menu Feb 02 11:43:19 even with red piull mode I see no settings in HAM at all. I maus be blind Feb 02 11:43:35 just catalogs and protocol Feb 02 11:44:04 killall hildon-application-manager Feb 02 11:44:19 recheck /home/user/.osso/hildon-application-manager Feb 02 11:44:32 it might get overwritten by HAM, on quitting Feb 02 11:44:38 ok, i'll do. just have to leave for the moment. thanks a lot. bb Feb 02 11:45:04 o/ Feb 02 11:45:46 * DocScrutinizer05 suggests to install cssu HAM, it seen several fixes Feb 02 11:46:04 and should work even on stock fremantle Feb 02 11:46:15 * DocScrutinizer05 pokes kerio Feb 02 11:47:33 DocScrutinizer05: i swear it never disappeared Feb 02 11:47:39 but i might be wrong Feb 02 11:47:47 i seldom used stock :) Feb 02 11:47:59 the settings menu is on the top level of HAM iirc Feb 02 11:48:32 the matrix way to temporarily enable it was gone at some point Feb 02 11:48:50 DocScrutinizer05: i've got a cloak and everything, i can't troll like this :( Feb 02 11:49:12 aaaaw ;-P Feb 02 11:50:32 kerio: *iirc* they removed the matrix style redpill enabler at one point in time, but later on removed redpill mode completely. I could be wrong as well Feb 02 11:51:24 do I have CSSU HAM when I have CSSU stable installed ? Feb 02 11:51:36 kerio: anyway, cssu HAM should install on stock fremantle cleanly, no? via dpkg Feb 02 11:51:48 M4rtinK: yes Feb 02 11:51:51 as I don't see any settings button in the toplevel app-menu Feb 02 11:52:08 M4rtinK: you still need enable redpill mode Feb 02 11:52:36 ok Feb 02 11:52:49 though in CSSU HAM even the matrix easteregg GUI hack works afaik Feb 02 11:54:36 DocScrutinizer05: Confirmed Feb 02 11:54:40 http://wiki.maemo.org/Red_Pill_mode#Activating Feb 02 12:09:11 DocScrutinizer05: apt, too :) Feb 02 12:09:59 freemangordon: does the maemo's apt has support for http redirects? Feb 02 12:10:57 xes: the question more is: does HAM support that feature of apt? Feb 02 12:18:50 xes: no idea Feb 02 12:26:03 :( in apt 0.7.20.2 maemo13 no http redirects support .. Feb 02 12:41:15 ping lxp1 Feb 02 12:51:30 teotwaki: could you please disable mirroring of repo.m.o? thanks Feb 02 12:51:59 teotwaki: (I mean actually syncing) Feb 02 12:53:19 teotwaki: ... temporarily ... Feb 02 13:13:46 pali: is lxp1 the mantainer of apt pkg? ..http redirect is present(but disabled) in x-apt http://maemo.gitorious.org/maemo-pkg/apt?page=6 ...so, for now, road closed Feb 02 13:14:39 xes, no lxp1 wrote packet injection wl1251 driver patch Feb 02 13:15:07 xes, look here: https://gitorious.org/~pali/community-ssu/pali-apt there is fremantle apt git tree Feb 02 13:15:37 xes, if you want to patch apt, use this version ^^ Feb 02 13:20:17 Pali: ok Feb 02 13:20:54 xes, that version in maemo-pkg "master" branch is harmattan version Feb 02 13:40:39 xes: patch apt for what? Feb 02 13:44:15 oh, something like http.debian.net? Feb 02 13:44:37 kerio: i was looking if maemo's apt has alread the support for http redirects Feb 02 13:45:32 by the way, couldn't we use coralcdn for rmo? Feb 02 13:49:53 on monday warfare will check some cdn options Feb 02 13:50:07 anyway, rmo doesn't work at all for me right now Feb 02 13:51:36 seems a quarter of connections has >200kBps Feb 02 13:52:01 so i'd attribute it to bad luck for you Feb 02 13:52:31 * DocScrutinizer05 wonders when teotwaki will notice the request to stop hammering rmo Feb 02 13:53:24 my opinion is that during a migration we should apply only minimum changes to all the domain. There are already enough variables to check Feb 02 14:00:05 power management problem with packet injection wl1251 driver is solved... problem is in maemo wlancond/wl1251-cal/icd2. when driver is loaded something set link up which consume lot of power Feb 02 14:00:23 but when non packet injection driver is loaded, link is down Feb 02 14:00:47 simple solution is to open wifi network dialog, wait until it show some networks and then close it Feb 02 14:01:02 icd2/wlancond then put interface down Feb 02 14:01:38 (or another, call ifconfig wlan0 down after loading packet injection driver) Feb 02 14:06:20 xes: was your comment re cdn? Feb 02 14:07:12 DocScrutinizer05: yes Feb 02 14:07:12 Pali: nice find Feb 02 14:07:38 xes: so, we already changed that particular details, *away* from "cdn" Feb 02 14:08:05 DocScrutinizer05, I'm going to use only packet injection driver which is in kp for testing Feb 02 14:08:19 going back to cdn based repo hosting was a step back to old config Feb 02 14:08:37 now when power problem was fixed, it is possible to have it loaded instead old maemo driver Feb 02 14:09:13 Pali: when we can implement that workaround of yours as a proper fix, then yes Feb 02 14:09:41 DocScrutinizer05, I have script which loading packet injection driver Feb 02 14:09:50 and I added ifconfig wlan0 down Feb 02 14:14:15 well, that's fixing the issue once. Doesn't same problem return later on, when you use those drivers for daily usage? Feb 02 14:14:45 DocScrutinizer05, I'm going to test it Feb 02 14:15:38 seems that wifi dialog window doing some link up & link down Feb 02 14:16:09 Pali: ask jonwil to RE it, so we can fix it Feb 02 14:16:23 freemangordon, why to RE? Feb 02 14:16:33 which dialog needs to be reverse engineered? Feb 02 14:16:36 I think wifi dialog is working ok Feb 02 14:16:52 problem is when you are switching drivers Feb 02 14:16:57 Pali: seems I misunderstood what the problem is then Feb 02 14:17:06 ok Feb 02 14:17:28 freemangordon, when packet injection driver is loaded, then n900 start consume lot of power Feb 02 14:17:47 easy fix is to open wifi dialog, wait some seconds and close it Feb 02 14:17:56 or open xterm and call ifconfig wlan0 down Feb 02 14:18:08 ok, so is there anything to reverse engineer or not? :) Feb 02 14:18:19 jonwil not :-) Feb 02 14:18:20 Pali: I see Feb 02 14:18:22 ok Feb 02 14:19:16 re Feb 02 14:21:51 any of you on fosdem this year? Feb 02 14:24:08 is it ok for you if i use repository.maemo.org during migration. or does that do any harm to the migration process. that way i can go the blue-pill way despite being sloooow ;) Feb 02 14:25:52 topro_: it's probably not ok for you Feb 02 14:25:56 rmo doesn't work at all for me Feb 02 14:26:04 unless it's all you jerks stealing my bandwidth Feb 02 14:34:29 DocScrutinizer05!!! Feb 02 14:34:31 DocScrutinizer05: http://talk.maemo.org/showthread.php?t=88694 Feb 02 14:34:36 oh nvm, you've already seen it Feb 02 14:36:34 topro_: go ahead, see if it works for you Feb 02 14:37:56 for you folks too: Feb 02 14:37:59 root@stage:~# grep 'Debian APT-HTTP' /var/log/apache2/*stage_transfer|wc -l Feb 02 14:38:00 436675 Feb 02 14:38:01 root@stage:~# grep -v 'Debian APT-HTTP' /var/log/apache2/*stage_transfer|wc -l Feb 02 14:38:03 44518 Feb 02 14:38:56 goddammit mirrors Feb 02 14:39:12 stahp Feb 02 14:39:31 merlin1991: have you stopped your mirroring? Feb 02 14:39:36 root@stage:~# grep "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)" /var/log/apache2/*stage_transfer|wc -l Feb 02 14:39:37 347214 Feb 02 14:39:38 root@stage:~# grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)" /var/log/apache2/*stage_transfer|wc -l Feb 02 14:39:40 133979 Feb 02 14:39:51 kerio doc: well it works.... 1 - 200 bytes per sec. about as fast as by 2g conn so ok ;) Feb 02 14:40:04 DocScrutinizer05: damn, so many n900s without cssu :( Feb 02 14:41:51 kerio: nope, but my mirror is in sync, so it's not doing anything apart from dl the Release file (which is 4kb) Feb 02 14:44:07 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|wc -l Feb 02 14:44:08 89461 Feb 02 14:44:10 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|wc -l Feb 02 14:44:11 29095 Feb 02 14:45:26 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|wc -l Feb 02 14:45:27 20308 Feb 02 14:45:46 hold on, three versions? Feb 02 14:46:08 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|wc -l Feb 02 14:46:09 10809 Feb 02 14:46:59 mo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|grep -v "Debian APT-HTTP/1.3 (0.7.6maemo3)"|wc -l Feb 02 14:47:01 3376 Feb 02 14:47:02 what's the cssu apt? Feb 02 14:47:31 we didn't thinker with that Feb 02 14:47:37 so should be same as pr 1.3 Feb 02 14:47:50 It's terrible slow to download anything from repository.maemo.org :( Feb 02 14:47:55 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|grep -v "Debian APT-HTTP/1.3 (0.7.6maemo3)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo8)"|wc -l Feb 02 14:47:56 1710 Feb 02 14:48:06 merlin1991: 0.7.20.2maemo13.2+0cssu1 Feb 02 14:48:53 kerio: lemme rephrase we only enabled the the hidden stuff Feb 02 14:49:01 we didn't tuch the user agent Feb 02 14:49:09 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|grep -v "Debian APT-HTTP/1.3 (0.7.6maemo3)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo8)"|less Feb 02 14:49:11 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|grep -v "Debian APT-HTTP/1.3 (0.7.6maemo3)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo8)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11)"|wc -l Feb 02 14:49:12 740 Feb 02 14:49:20 oops sorry Feb 02 14:49:40 kerio: that's the whole changeset applied: http://gitorious.org/community-ssu/apt/commit/e6ae26a0e3d3018beb6b1bb0219fb1aef3128a85 Feb 02 14:50:05 hm whats the easiest way for me to install 'make' on my n900 Feb 02 14:50:23 i think i need to compile org-mode myself Feb 02 14:51:19 ludens: find a guide on installing the sdk in a chroot Feb 02 14:51:29 it's a baaaad idea to enable the sdk repo and just install the packages Feb 02 14:51:40 (rootfs will fill in seconds) Feb 02 14:52:02 fosdem meeting now Feb 02 14:52:52 merlin1991: yeah i was thinking i shouldnt need that many packages... Feb 02 14:53:21 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|grep -v "Debian APT-HTTP/1.3 (0.7.6maemo3)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo8)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11)"|grep -v "Debian APT-HTTP/1.3 Feb 02 14:53:21 but hm yeah i guess id need a bunch of stuff :/ Feb 02 14:53:22 (0.7.6maemo2)"|wc -l Feb 02 14:53:23 641 Feb 02 14:53:25 root@stage:~# grep "Debian APT-HTTP/1.3" /var/log/apache2/*stage_transfer|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13.2)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11.1)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo13)"|grep -v "Debian APT-HTTP/1.3 (0.7.6maemo3)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo8)"|grep -v "Debian APT-HTTP/1.3 (0.7.20.2maemo11)"|grep -v "Debian APT-HTTP/1.3 Feb 02 14:53:26 (0.7.6maemo2)"|grep '[0-9]maemo'|wc -l Feb 02 14:53:28 0 Feb 02 14:59:38 root@stage:~# grep "0cssu" /var/log/varnish/varnishncsa.log|wc -l Feb 02 14:59:39 916 Feb 02 14:59:41 root@stage:~# wc -l /var/log/varnish/varnishncsa.log Feb 02 14:59:42 370952 /var/log/varnish/varnishncsa.log Feb 02 15:01:34 :c Feb 02 15:02:01 DocScrutinizer05: it's to be expected i suppose, cssu users were probably aware of the downtime and switched to mirrors Feb 02 15:02:10 only means cssu users don't hammer repo with useless update requests Feb 02 15:02:22 yep Feb 02 15:02:31 speak for yourself, i hammer like crazy Feb 02 15:02:37 :P Feb 02 15:02:42 hm is it possible that someone has compiled org-mode lately and made a .deb? Feb 02 15:03:03 then itd be avilable in the repos i suppose Feb 02 15:03:20 not necessarily Feb 02 15:03:35 :) Feb 02 15:03:54 so... i need to make a chroot Feb 02 15:04:11 you're supposed to use scratchbox Feb 02 15:04:22 on a proper PC Feb 02 15:04:23 humm whats that? Feb 02 15:04:34 installing scratchbox is a ridiculous process Feb 02 15:04:43 aha ok Feb 02 15:04:44 use the vm Feb 02 15:04:52 http://talk.maemo.org/showthread.php?p=1318174#post1318174 Feb 02 15:04:57 :c Feb 02 15:05:57 ludens: http://skeiron.org/tablets-dev/maemo_dev_env_downloads/ the vm image here Feb 02 15:06:00 the giant one Feb 02 15:06:24 alright can i just run that in qemu? Feb 02 15:06:34 http://wiki.maemo.org/Documentation/Maemo_5_Final_SDK Feb 02 15:06:50 it's a vmware vm, but it runs fine with the usual ubuntu settings in virtualbox Feb 02 15:07:26 okay.. Feb 02 15:07:51 http://wiki.maemo.org/Documentation/Maemo_5_Final_SDK_Installation Feb 02 15:08:11 DocScrutinizer05: is it irresponsible to use skeiron to download the vm image? Feb 02 15:08:13 it's like 1.5gb Feb 02 15:08:22 meh Feb 02 15:08:26 and then i will need to create a .deb to install the binary on my n900? Feb 02 15:08:57 ludens: chances are, you'll just have to get the debian tarball and compile it Feb 02 15:08:57 rrrright Feb 02 15:09:04 with dpkg-buildpackage Feb 02 15:09:46 hm oh Feb 02 15:10:28 kerio: seems last month you guys used 8% of traffic limit on skeiron Feb 02 15:10:38 there's some headroom still ;-) Feb 02 15:10:38 BTW I reflash and now I see no updates (I'm on version 20.2010.36-2). I think there is 38 at least ... Feb 02 15:11:49 luf: there's 2011.38 Feb 02 15:11:53 from nokia Feb 02 15:12:13 but HAM is a jerk Feb 02 15:12:16 kerio: Tell it to the HAM ;) Feb 02 15:12:22 no, it's a known thing Feb 02 15:12:30 it won't let you update because of the key thing Feb 02 15:12:44 enter redpill mode and disable "ignore packages from wrong domains" Feb 02 15:12:44 nokia dropped the ball Feb 02 15:12:56 ~redpill Feb 02 15:13:23 nokia also dropped the bass Feb 02 15:13:24 http://www.youtube.com/watch?v=kA2pjVAswek Feb 02 15:14:14 luf: "red-pill 1" "red-pill-permanent 1" in ~user/.osso/hildon-application-manager Feb 02 15:14:18 two lines Feb 02 15:14:23 then access the settings menu Feb 02 15:14:31 http://wiki.maemo.org/Red_Pill_mode Feb 02 15:14:47 infobot: #maemo redpill is http://wiki.maemo.org/Red_Pill_mode Feb 02 15:14:48 kerio: okay Feb 02 15:14:50 ~listkeys redpill Feb 02 15:14:53 Factoid search of 'redpill' by key (1): #maemo redpill. Feb 02 15:14:55 :D Feb 02 15:15:03 pff Feb 02 15:15:34 ~listvalues Red_Pill_mode Feb 02 15:15:36 Factoid search of 'Red_Pill_mode' by value returned no results. Feb 02 15:15:45 grrr Feb 02 15:16:00 ~listvalues http://wiki.maemo.org/Red_Pill_mode Feb 02 15:16:02 Factoid search of 'http://wiki.maemo.org/Red_Pill_mode' by value returned no results. Feb 02 15:16:25 kerio: thanks you're faster than google :D Feb 02 15:16:35 o.O Feb 02 15:16:39 ~redpill Feb 02 15:16:40 redpill is, like, http://wiki.maemo.org/Red_Pill_mode Feb 02 15:16:52 DocScrutinizer05: thanks. Feb 02 15:17:11 hm "Recent Emacs distributions include a packaging system which lets you install Elisp libraries. You can install Org with M-x package-install RET org." i wonder if that would work for me? Feb 02 15:17:18 ludens: try it Feb 02 15:17:33 ~listvalues wiki.maemo.org Feb 02 15:17:35 Factoid search of 'wiki.maemo.org' by value (18 of 76): #maemo closed ;; #maemo closed_packages ;; #maemo power ;; #maemo redpill ;; becomeroot ;; bluetooth-pan ;; bme ;; broken-maemo ;; extras ;; extras-product ;; extras-upload ;; extras-uploading ;; flash-partitioning ;; free ;; jrtools ;; maemo-flashing ;; maemo-qt ;; mafw. Feb 02 15:17:46 ~listvalues wiki.maemo.org/ Feb 02 15:17:47 Factoid search of 'wiki.maemo.org/' by value (18 of 76): #maemo power ;; accelerometer ;; bluetooth-pan ;; bluetooth-profiles ;; boot-from-sd ;; boot-sd ;; broken-maemo ;; categories ;; cssu ;; ctl-sh-p ;; extras-devel ;; extras-upload ;; flasher ;; hildon-extras ;; jebbadbus ;; jrxchat ;; maemosdk ;; n900-power. Feb 02 15:17:55 ~listvalues wiki.maemo.org/Red Feb 02 15:17:57 Factoid search of 'wiki.maemo.org/Red' by value (2): #maemo redpill ;; red-pill. Feb 02 15:18:07 infobot: what the hell man Feb 02 15:18:07 Tem and Tain are the man. Feb 02 15:18:10 ... Feb 02 15:18:13 ok Feb 02 15:18:19 ~factinfo red-pill Feb 02 15:18:19 red-pill -- created by GeneralAntilles at Wed Nov 19 18:27:28 2008 (1535 days); it has been requested 4 times, last by Stskeeps at Thu Feb 5 01:44:00 2009. Feb 02 15:18:32 what? Feb 02 15:18:33 kerio: def will. would be a lot easier Feb 02 15:18:35 ~red-pill Feb 02 15:18:36 extra, extra, read all about it, red-pill is https://wiki.maemo.org/Red_Pill_mode Feb 02 15:18:40 hrmpf Feb 02 15:18:46 :-P Feb 02 15:18:49 infobot: no, red-pill is see redpill Feb 02 15:18:50 kerio: okay Feb 02 15:19:07 it's all part of my plan to slowly gain ownership of all the factoids Feb 02 15:19:08 kerio: that's definitely rude Feb 02 15:19:21 DocScrutinizer05: https? really? Feb 02 15:19:30 ooh, k Feb 02 15:19:57 once i have the copyright on all the factoids, i'll charge 19.99$/month to access them Feb 02 15:20:03 nfc why some guys use https with wiki. Pali as well Feb 02 15:20:13 mwahahahahahaha Feb 02 15:21:03 password login protection? Feb 02 15:21:20 https everywhere? Feb 02 15:21:21 * DocScrutinizer05 redirects all https://wiki.maemo.org pages to "this is not your bank account!" Feb 02 15:21:40 fix your browser Feb 02 15:21:43 hah Feb 02 15:21:51 this is bug in your brower Feb 02 15:22:03 Pali: it's not even fully encrypted anyway Feb 02 15:22:14 i can do a MITM through static.maemo.org Feb 02 15:22:15 https has at least same protection as http Feb 02 15:22:43 so why to not use https? Feb 02 15:22:50 and at least *some* overhead on serverside Feb 02 15:23:23 if there is problem on server, turn https off Feb 02 15:23:34 I'm tempted ;-P Feb 02 15:23:41 the fuck Feb 02 15:23:52 lol Feb 02 15:24:53 sup, Skry Feb 02 15:24:57 Small trouble with redpill ... I have no .osso/h-a-m file and when I created I lost all repositories ... Feb 02 15:25:06 you're lying Feb 02 15:25:15 anyway, close ham Feb 02 15:25:21 wait for apt-worker to stop dicking around Feb 02 15:25:28 then create that file with the two lines Feb 02 15:25:36 sometimes HAM decides to not show any catalogue Feb 02 15:26:13 I hate HAM Feb 02 15:26:18 indeed Feb 02 15:26:36 Fucking sometimes ... Feb 02 15:26:38 luf, you need to wait until apt-worker will parse all lists files Feb 02 15:27:14 also when you start new HAM instance before apt-worker from old instance exited you have problem... Feb 02 15:27:15 Pali: it didn't try to parse ... Feb 02 15:27:19 But now it's ok. Feb 02 15:27:33 Pali: maybe I was too fast :( Feb 02 15:27:36 luf: there should be a Settings menu on the toplevel Feb 02 15:27:43 luf, when you start HAM it spawn apt-worked and it parsing all packages from repositories Feb 02 15:28:00 this take about 2-10 minutes on my n900 :-( Feb 02 15:28:06 luf: it's hard to *not* be too fast for HAM Feb 02 15:28:34 when you install/uninstall/update any package HAM again reparsing packages Feb 02 15:28:39 and again it took lot of time Feb 02 15:29:16 Pali: I know about that too long. But I can't understand why it starts when another apt-worked is running. Feb 02 15:29:20 if HAM doesn't work, you're too fast. Waaaaaait! Feb 02 15:29:45 luf, HAM spawning apt-worker with 4 names pipes on /tmp/ Feb 02 15:29:59 after apt worker is spawned files are removed Feb 02 15:30:16 so any new application cannot connect to apt-worker Feb 02 15:30:38 Pali: ok so HAM should report the error in that case. Feb 02 15:30:48 yes should Feb 02 15:30:59 apt-get does ;-P Feb 02 15:31:22 I already implemented some code in CSSU HAM which tell apt-worked to immediately quit when turning HAM off Feb 02 15:31:41 and I still wonder if that's a good idea Feb 02 15:32:15 DocScrutinizer05, I implemented only new command via pipe protocol Feb 02 15:32:45 so if apt-worker does not wait for something from pipe, it does not know that HAM was exited Feb 02 15:33:25 well, as long as apt-worker has no filehandles on modify/write Feb 02 15:33:44 https://gitorious.org/community-ssu/hildon-application-manager/commit/b3b3567563c30d47b8a02fe6dc38a59d1a63a5ea Feb 02 15:33:48 https://gitorious.org/community-ssu/hildon-application-manager/commit/1125cbbd2ca3a0644f684ec6ba2eaa44df5dad53 Feb 02 15:34:07 only this was implemented ^^ Feb 02 15:34:20 I'd just hate apt-worker corrupting my files Feb 02 15:35:15 Even with redpill I see no ssu update :( Feb 02 15:35:37 first patch checking for cancel command in operation "parse packages" --> this is only read only operation Feb 02 15:35:42 luf: do you see the settings menu? Feb 02 15:35:42 you need to config it Feb 02 15:36:15 second patch adding new command quit, so HAM can tell apt-worked to quit, when apt-worked doing nothing Feb 02 15:36:18 Yes I unset wrong domains. Feb 02 15:36:31 luf: are you *sure* Feb 02 15:36:32 I'm trying it once again. Feb 02 15:36:38 i'd do another refresh Feb 02 15:36:46 kerio: sure. I checked it. Feb 02 15:37:02 Yes. I restarted HAM once again and I'm waiting now. Feb 02 15:37:54 waiting is good for you ;-) Feb 02 15:38:02 I see more now. Feb 02 15:38:21 DocScrutinizer05: trust me. It's not good for me :D Feb 02 15:39:55 ok, giving up. rmo is really broken Feb 02 15:40:21 going red-pill as suggested Feb 02 15:40:49 topro_: it's the only way :( Feb 02 15:43:01 luf: =D Feb 02 15:43:33 DocScrutinizer05: couldn't we configure skeiron or maemo-archive to offer the appropriate paths when asked for repository.maemo.org? Feb 02 15:43:44 so you could change /etc/hosts and HAM would be happy Feb 02 15:43:57 Can't some CDN sponsore maemo? Feb 02 15:44:09 luf: coralcdn is free anyway Feb 02 15:44:59 I think that repos should be cached there ... Feb 02 15:45:20 kerio: already duely noted by brkn Feb 02 15:45:22 BTW update Maemo 5 -> It's not from Nokia do you want to continue? :D :D :D Feb 02 15:45:47 hmm, i get hash sums mismatch on one file when using the suggested mirror Feb 02 15:46:24 (cdn) yes, warfare will check our options on monday Feb 02 15:46:33 * DocScrutinizer05 enters parrot mode Feb 02 15:46:38 DocScrutinizer05: great ... Feb 02 15:46:49 DocScrutinizer05: are we paying warfare? Feb 02 15:46:55 no Feb 02 15:47:05 then let's make him work on sunday too! :D Feb 02 15:47:19 we never could afford such a wizard Feb 02 15:47:56 overtime pay is 1.5 or 2 times the usual rate, isn't it Feb 02 15:48:33 wow the ELPA is prett amazing. compiled the latest org-mode build inside emacs Feb 02 15:49:01 ? Feb 02 15:49:56 is there any way I can get the cmt-firmware-rx51 .deb file or the /boot/cmtImage-10w08 file? Feb 02 15:50:40 eddyb: it's in the nokia ssu repo Feb 02 15:51:24 eddyb: you'll need the user/pass that are stored in /etc/apt/auth Feb 02 15:51:31 ...or maybe you could look in some mirrors Feb 02 15:51:41 but nobody told you that Feb 02 15:51:44 kerio: how can I use the device to download the .deb file? Feb 02 15:52:01 eddyb: isn't it already installed? Feb 02 15:52:25 oh, it's one of those "install, flash and remove" Feb 02 15:52:30 mhmm Feb 02 15:52:37 the file in /boot is missing Feb 02 15:52:46 it's supposed to be missing Feb 02 15:52:48 it's flashed Feb 02 15:53:02 yeah, that's why I want to get the .deb file Feb 02 15:53:05 oic Feb 02 15:53:08 eddyb: sudo apt-get install --reinstall --download-only cmt-firmware-rx51 Feb 02 15:53:16 heh thanks :D Feb 02 15:53:17 and then grab the .deb from /var/cache/apt/archives/ Feb 02 15:53:30 that's exactly what I was looking for Feb 02 15:53:42 also, do you happen to know if anyone tried reverse engineering that image? Feb 02 15:54:00 haha Feb 02 15:54:30 ~mirrors Feb 02 15:54:30 [mirror] http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Feb 02 15:54:32 do you have datasheets of BB5? Feb 02 15:55:11 and even if you had, you can't flash patched or altered images to cmt/BB5 Feb 02 15:55:45 which renders all RE kinda useless academic effort Feb 02 15:56:12 unless you're actually just hunting for info/knowledge Feb 02 15:56:24 isnt cmt image signed? Feb 02 15:56:28 yep Feb 02 15:56:34 DocScrutinizer05: I'm mostly curious Feb 02 15:56:36 that's why Feb 02 15:56:44 with downgrade protection too? Feb 02 15:56:46 ask jacekowski then Feb 02 15:56:56 Pali: iirc yes Feb 02 15:57:30 DocScrutinizer05: until we find a bug in the signature verification :) Feb 02 15:57:34 or some kind of attack vector Feb 02 15:58:01 signed as in public/private key-pair, and the private key is in a buried 10m thick steel safe, somewhere in Finland :D? Feb 02 15:58:03 we had that issue with leaked PR1.2-asia which updated cmt and downgrading the rootfs created an incompatibility Feb 02 15:58:26 eddyb: maybe Feb 02 15:58:34 not maybe, exactly Feb 02 15:58:44 DocScrutinizer05: we don't actually know if it's nokia that holds the private key Feb 02 15:59:01 unless we know that because they're required to by some regulations Feb 02 15:59:02 err, yes we do Feb 02 15:59:16 maybe it was subcontractors! Feb 02 15:59:29 since the guys were able to create updates Feb 02 15:59:57 -rw-r--r-- root/root 6567893 2010-03-02 11:10 ./boot/cmtImage-10w08 Feb 02 16:00:04 6MB? Feb 02 16:00:14 sounds pretty small Feb 02 16:00:15 RE not possible... Feb 02 16:00:56 it's probably only the filesystem with a few configs Feb 02 16:01:25 what? Feb 02 16:01:42 filesystem? Feb 02 16:01:46 I remember looking at the wiki, why is there no actual info on the GSM/3G hardware stuff? Feb 02 16:01:47 in that firmware? Feb 02 16:02:19 eddyb: you didn't find it Feb 02 16:02:31 /boot/cmtImage-10w08 is FIASCO image Feb 02 16:02:50 DocScrutinizer05: I blame google trying to be a smartass Feb 02 16:02:54 amazingly small Feb 02 16:02:55 contains parts: cmt-2nd cmt-algo cmt-mcusw Feb 02 16:03:26 cmt-2nd, version BB5_09.36, size 81408 bytes Feb 02 16:03:34 pfff Feb 02 16:03:35 still get hash mismatch when using maemo-archive.wedrop.it mirror extras repo Feb 02 16:03:42 cmt-algo, version BB5_09.36, size 519808 bytes Feb 02 16:03:45 DocScrutinizer05: this? http://wiki.maemo.org/N900_Hardware_Schematic#Page_13_RF_Part Feb 02 16:03:50 that's basically BB5's NOLO Feb 02 16:04:02 cmt-mcusw, version rx51_ICPR82_10w08, size 5966464 bytes Feb 02 16:04:21 DocScrutinizer05, why there are 3 parts of cmt firmware? Feb 02 16:04:36 that might be an extremely downsized radio stack Feb 02 16:06:02 Pali: I already told you what's 2nd and algo Feb 02 16:06:29 but why are separated files? Feb 02 16:06:31 mcusw is probably MicroControllerUnit Feb 02 16:06:54 SoftWare Feb 02 16:07:01 because that are separate cpu cores Feb 02 16:08:17 Pali: BB5 has an ARM, a DSP, and some auxiliary MCU cores Feb 02 16:08:34 DocScrutinizer05: I'm only seeing "Aura - Vapaus, ... Cellphone components." Feb 02 16:09:30 what are you looking for? Feb 02 16:09:35 rapuyama? Feb 02 16:09:59 your description of BB5 sounds promising, but that's about it Feb 02 16:10:23 there IS NO datasheet about BB5 Feb 02 16:10:33 DocScrutinizer05: except I have no clue what Rapuyama means/is Feb 02 16:10:46 rapuyama == BB5 Feb 02 16:10:51 == cmt Feb 02 16:10:52 oh Feb 02 16:11:12 see, if you hadn't told me that, there was no way to find out from the wiki, at least in the places I've looked Feb 02 16:11:56 http://www.gsmrapid.com/showthread.php?t=4957 this looks like a complete SoC Feb 02 16:12:37 like it was designed to be the only SoC in a phone or something Feb 02 16:13:00 so, uhm, the firmware has ARM code? Feb 02 16:13:25 http://wiki.maemo.org/N900_Hardware_Phone Feb 02 16:13:59 eddyb: yep, it's designed to be the SoC for a dumbphone Feb 02 16:14:06 eddyb: yes, that page is about what you can find regarding BB5 internals Feb 02 16:14:12 the n900 is not a phone, it's a computer with an embedded phone :D Feb 02 16:14:44 ~cmt Feb 02 16:14:56 ~useless Feb 02 16:14:57 * infobot starts crying and hides from kerio in the darkest corner of the room. :( Feb 02 16:15:24 DocScrutinizer05: yeah, I read that page, and the only interesting info there is "V ICPR82_10w08" Feb 02 16:21:04 ICPR82_10w08 is version of rx51 cmt-mcusw image Feb 02 16:23:12 eddyb: you can find specs of ISI interface to BB5. Everything else is mostly useless since you neither can reasonably disassemble nor patch and flash Feb 02 16:23:27 why can't I disassemble? Feb 02 16:23:42 is the image compressed/encrypted? Feb 02 16:23:57 also, where is signature verification done? resident firmware in that SoC? Feb 02 16:24:05 you can, but what's the use of seeing things like "move 0xFF43 to reg 167" ? Feb 02 16:24:41 you seen that block schematics on gsmrapid? you noticed the block labeled "security"? Feb 02 16:24:53 hehe Feb 02 16:25:17 and you can learn a lot of things from register writes Feb 02 16:25:37 also ask freemangordon about general ARM security concept Feb 02 16:25:47 who would bother to obfuscate hardware accesses? Feb 02 16:26:03 s/writes/writes and/or reads/ Feb 02 16:26:06 sorry, I'm EE and that starts boring me Feb 02 16:26:55 you don't learn *anything* if you don't know the hw you're talking to Feb 02 16:27:08 UART read/write routines are somewhat easy to spot Feb 02 16:27:21 *shrug* Feb 02 16:27:35 DocScrutinizer05: you assume opaque code Feb 02 16:28:03 but code is written by humans, and only slightly modified by machines Feb 02 16:28:09 and what are you assuming? comments? Feb 02 16:28:40 I am assuming I will continue working on my decompiler Feb 02 16:28:55 you have NFC how the custom DSP in rapu works Feb 02 16:29:36 so UART might actually be the only code you're able to decifer Feb 02 16:30:31 MIPI-HSI already you're lacking proper datasheets for Feb 02 16:31:11 and on radio stack you're utterly lost Feb 02 16:32:02 maybe you could tackle the GPS chip control Feb 02 16:32:12 *might* be feasible Feb 02 16:32:26 partially Feb 02 16:34:25 * DocScrutinizer05 never gets the catch why people are interested in RE of a complex radio stack like in BB5. When they could start googling for osmocom Feb 02 16:34:57 DocScrutinizer05, so bb5 is arm soc chip? Feb 02 16:35:05 integrated into n900? Feb 02 16:35:09 yes Feb 02 16:35:30 hmmmm Feb 02 16:36:05 and omap soc can communicate with bb5 via SSI? Feb 02 16:36:07 so my archiver opened the firmware as text, so I took a quick look Feb 02 16:36:15 Max RLC window size: %d receiving window size: %d, configurationOK: FALSE Feb 02 16:36:22 McBSP but yeah Feb 02 16:36:24 Simultaneous reception of SCCPCH and DPCH is not supported - configurationOK: FALSE Feb 02 16:36:34 talk about comments, hehe Feb 02 16:36:55 BMC info trace (RRC) : Output bit set - repetition of requested message Feb 02 16:37:27 cmt-mcusw_rx51_ICPR82_10w08 image is ARM code? Feb 02 16:37:51 there is some header Elf2flash 08.37.001 Feb 02 16:37:53 now excite me an dpretend you can RE a whole WCDMA/GSM radio stack with those diagnostic output strings Feb 02 16:38:27 there is also header PRIMAPP and PAPUBKEYS Feb 02 16:38:38 which is X-Loader format certificate Feb 02 16:38:48 with signature Feb 02 16:38:54 DocScrutinizer05: nah, but it helps to have a ton of debug strings left in :P Feb 02 16:38:59 it's a standard ARM system, so what? Feb 02 16:39:18 of course it's using X-Loader Feb 02 16:40:09 DocScrutinizer05, there is still no exploit for omap HS devices? Feb 02 16:40:19 no, not afaik Feb 02 16:40:34 the security guys did a decent job Feb 02 16:40:56 btw BB6 most likely isn't omap Feb 02 16:41:04 BB5* Feb 02 16:41:34 it's ARM BB5, not ARM OMAP Feb 02 16:41:43 I did not belive that nokia did decent job - wrote bootloader X-loader & NOLO without security errors Feb 02 16:42:31 we can see lot of problems in maemo and harmattan... Feb 02 16:43:22 ROMBL and probably even x-loader isn't written by Nokia Feb 02 16:43:45 Pali: you're using 0xFFFF for unpack that image, right? I see you own the gitorious repo Feb 02 16:43:46 even NOLO is a perverted uBoot Feb 02 16:44:08 eddyb, yes Feb 02 16:44:21 DocScrutinizer05, NOLO is not uBoot Feb 02 16:44:31 there is noting common in binary Feb 02 16:44:48 and it is Nokia X-Loader Feb 02 16:45:45 * DocScrutinizer05 is searching an empty spot on his wall full of of tally charts, to add another scratch for another useless "we wanna RE the GSM radiostack" discussion Feb 02 16:47:37 Pali: (nothing common) you bet there's othing, otherwise my friend Harald would come onto them. That's why I said "*perverted* uBoot" Feb 02 16:48:35 and x-loader code is always based on that refernce implementation by TI(?) Feb 02 16:48:56 or even ARM Feb 02 16:48:58 reference implementation by TI? Feb 02 16:49:23 so this could be something closed... Feb 02 16:49:34 I did not see it yet Feb 02 16:50:22 it's maybe not exactly open source Feb 02 16:50:43 but all chip manufs get it and hand it to their customers Feb 02 16:51:08 it's probably hardly ever changed much Feb 02 16:51:28 ask jacekowski Feb 02 16:51:32 DocScrutinizer05, is NOLO doing someting with cmt? Feb 02 16:51:38 sure Feb 02 16:51:47 ping jacekowski Feb 02 16:52:41 I'm thinking if will replace NOLO with uboot if cellular modem will be working or not Feb 02 16:52:58 Pali: the proper syntax is: >> : ping Feb 02 16:53:06 if kernel/userspace has enought code to start up and configure it Feb 02 16:53:22 jacekowski: ping Feb 02 16:53:50 Pali: it probably will just delay cmt bootup Feb 02 16:54:16 usually in uBoot/Nolo and userland/kernel initialization gets duplicated Feb 02 16:55:17 I'd be surprised if not already NOLO would power up cmt Feb 02 16:55:37 and set up the interfaces to it Feb 02 16:55:47 DocScrutinizer05, in nolo console there is command for cmt bootup Feb 02 16:55:57 /nokia/sa_nmp/groups/dss/BB5_flash_algos-next/BB5_flash_algos/common_services/flash_msg.c Feb 02 16:55:59 yep, thought as much Feb 02 16:56:03 so this is reason why I asked if it will be possible Feb 02 16:56:22 that's fun, file names Feb 02 16:56:42 so what's fun in that? Feb 02 16:56:53 you want proper coredumps at least Feb 02 16:59:19 ~wiki osmocom Feb 02 16:59:22 I couldn't find a matching article in wikipedia, look for yerselves: http://en.wikipedia.org/wiki/Special:Search?search=osmocom&go=Go Feb 02 16:59:53 hmm thought as much Feb 02 16:59:56 ~wiki osmocombb Feb 02 16:59:59 At http://en.wikipedia.org/wiki/OsmocomBB (URL), Wikipedia explains: "{{cleanup|date=May 2012 | reason = grammar}} {{Refimprove|date=May 2012}} {{Infobox Software | Logo = | Screenshot = | Beschreibung = | Maintainer = | Hersteller = | AktuelleVersion = | AktuelleVersionFreigabeDatum = | Betriebssystem = | Written in = C | Kategorie = | Lizenz = GPL (Free software) | Website = [http://bb.osmocom.org/ bb.osmocom.org] }} 'OsmocomBB' ('O'pen 'S'ource ... Feb 02 17:00:25 haha Feb 02 17:00:35 there's C code in mcusw Feb 02 17:00:47 probably from some kind of assert Feb 02 17:01:02 ~useless Feb 02 17:01:03 * infobot starts crying and hides from kerio in the darkest corner of the room. :( Feb 02 17:01:57 ~attack kerio Feb 02 17:01:58 * infobot grabs a pen, screams like she's possessed, and begins chasing kerio Feb 02 17:02:45 (list1 != NULL && list2 != NULL) && ((list1->no_of_freqs > 0 && list1->freqs != NULL) || list1->no_of_freqs == 0) && ((list2->no_of_freqs > 0 && list2->freqs != NULL) || list2->no_of_freqs == 0) Feb 02 17:02:46 meh, got some serious stuff to do Feb 02 17:03:25 you can probably make up partial structure definition just from all these strings Feb 02 17:03:52 emphasis on 'make up' Feb 02 17:05:28 * DocScrutinizer05 seems to recall compilers sometimes doing that: fill unititialized areas with random cruft which often is sourcecode Feb 02 17:07:03 o/ Feb 02 17:07:08 l8rs Feb 02 17:07:16 nah, this is from a stringifing macro Feb 02 17:07:21 ASSERT usually does that Feb 02 17:09:49 all the code looks like asserted expressions, but you can correlate the actual code checking the expression and the code in the string Feb 02 17:11:49 19:11:18 up 10:01, Feb 02 17:12:06 shall I give FW another reboot before I leave? Feb 02 17:12:32 why leave? :( Feb 02 17:12:34 magic limit is 18h uptime Feb 02 17:13:03 kerio: because I actually have a life, though most don't believe that Feb 02 17:13:11 * kerio doesn't believe that Feb 02 17:13:50 and that BB% talk kinda bored me a little Feb 02 17:14:23 leads nowhere, always Feb 02 17:14:51 so, shall I? Feb 02 17:16:54 meh, I did Feb 02 17:16:57 o/ Feb 02 17:17:32 \o Feb 02 17:27:02 * ShadowJK wonders how many entries conntracking table has at 17hrs 50m Feb 02 17:51:20 Pali: how do you make sense of the header in mcusw? Feb 02 17:52:07 eddyb, find on inernet X-Loader image format Feb 02 17:52:32 you will see that mcusw has similar/same header Feb 02 17:52:46 please don't tell me google didn't gave me the right results because I didn't use a hyphen Feb 02 17:53:52 it is somewhere on and-developers.com Feb 02 17:54:00 motorola using omap3 too Feb 02 17:54:09 so needs same x-loader image headers Feb 02 17:54:21 not this? https://github.com/bradfa/x-loader Feb 02 17:57:07 no Feb 02 17:57:24 Omap HS X-Loader specification is not public Feb 02 17:57:51 information about certificates & signatures for HS devices are not public Feb 02 17:58:02 what I found was some RE info on and-developers.com Feb 02 17:58:11 i've got signing software Feb 02 17:58:22 ha, here it is: http://and-developers.com/partitions:isw Feb 02 17:59:06 https://maemo.jacekowski.org/tools/ Feb 02 17:59:08 jacekowski, I think that it is possible to download signing SW from TI after registration Feb 02 17:59:13 Pali: thanks, that's much better Feb 02 17:59:13 this thing can sign the software Feb 02 17:59:20 but there is no infomration about formats Feb 02 17:59:24 but you need the keys Feb 02 17:59:33 no information about HS devices registers... Feb 02 18:00:18 let's ship something in cssu that silently bruteforces the signing key for cmt Feb 02 18:00:54 kerio, better will be signing key for X-Loader Feb 02 18:01:06 kerio: hihihi Feb 02 18:01:21 but we have nothing which can check if signature is correct Feb 02 18:01:38 we do not have public key (which is burned into cpu) Feb 02 18:01:57 what we have is signature glued into X-Loader binary Feb 02 18:02:00 clearly, we'll have to constantly try to flash the signed firmwares :D Feb 02 18:02:29 Pali: oh Feb 02 18:03:41 nolo does some checking Feb 02 18:03:54 but i don't remember if it was just checking if checksum is correct or signature Feb 02 18:18:06 jacekowski, what is NOLO doing with cmt? Feb 02 18:18:17 it is something which is in kernel/userspace too? Feb 02 18:24:30 Pali: I'm confused... the file is 0x5B0A80 bytes long, but the last entry ends at about 0x901B22 Feb 02 18:31:00 Pali: and the first one starts at 0x5F000, so there should be 0x8A2B22 bytes in the file Feb 02 18:35:37 hi Feb 02 18:36:46 i bought a used n900, flashed RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin and RX-51_2009SE_2.2009.51-1_PR_COMBINED_MR0_ARM.bin, enabled rd-mode, tried to run apt-get dist-upgrade, phone rebooted and now i get "no initfs" Feb 02 18:36:53 and i faile to flash again Feb 02 18:36:57 what to do now? Feb 02 18:37:33 nusse: haven't you noticed that dist-upgrade asked you to delete ~260 packages? Feb 02 18:37:37 did you think that it was normal? Feb 02 18:37:53 no initfs is normal, but the hell if you're going to be able to boot Feb 02 18:37:57 flash again Feb 02 18:38:42 btw, you flashed the wrong one Feb 02 18:38:46 kerio: kerio it hasnt installed anything, was still downloading, it wasnt deleting 260 packages i think Feb 02 18:38:55 did i? please point me to the correct images Feb 02 18:39:15 RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin Latest Maemo 5 Global Feb 02 18:39:15 right now i can't make the flasher recognize the phone, it just sits in "no initfs" Feb 02 18:39:37 and RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin Latest Vanilla version of the eMMC Feb 02 18:39:49 so yeah, you've got the right VANILLA Feb 02 18:40:02 nusse: hm, i doubt you managed to fuck up the bootloader Feb 02 18:40:18 are you absolutely sure that the flasher doesn't recognize the n900 anymore? Feb 02 18:40:27 no Feb 02 18:40:31 of course not Feb 02 18:40:34 remove the battery, connect the usb cable, launch the flasher with the appropriate command, hold u, put the battery in Feb 02 18:40:57 a usb device is showing up Feb 02 18:41:06 New USB device found, idVendor=0421, idProduct=0106 Feb 02 18:42:05 yay Feb 02 18:42:18 remember to follow the flashing guide quite closely Feb 02 18:42:36 remove the battery between each flashing step Feb 02 18:43:25 that doesnt mean the flasher recognizes anything ;) Feb 02 18:45:43 nusse: not even dist-upgrade can fuck a n900 up enough to require coldflashing Feb 02 18:45:46 http://ix.io/4fx Feb 02 18:46:02 are you using linux? blacklist cdc_phonet Feb 02 18:46:07 try another usb cable Feb 02 18:46:26 no such module loaded Feb 02 18:46:36 yeah but blacklist it Feb 02 18:46:46 autoloaders have a tendency to fuck this kind of thing up Feb 02 18:49:07 i havent even compiled it Feb 02 18:49:42 so it recognizes something and puts it into flash mode but then nothing happens Feb 02 18:52:13 try as root Feb 02 18:52:32 I had quite restrictive permissions on Arch by default Feb 02 18:53:31 i'm root Feb 02 18:53:44 i did successfully flash it 4h ago Feb 02 18:56:37 nusse, try to use usb1.0 or usb1.1 port Feb 02 18:57:12 I have big problems with my usb3.0 ports and sometimes usb2.0 port does not working too Feb 02 18:57:14 i'll try another computer Feb 02 18:57:45 nusse, and also: for eMMC flashing you need to have working rootfs Feb 02 18:58:00 is your phone bootable? Feb 02 18:58:15 if not, then rootfs is not working Feb 02 18:58:23 so you cannot flash eMMC image Feb 02 18:58:57 Pali: I'm probably late. but I found something interesting at 0x87AF Feb 02 18:59:36 Pali: no it is not booting Feb 02 18:59:54 nusse: do: first flash combined image, reboot - but do not boot!, then flash eMMC image, reboot and then flash combined image again Feb 02 19:00:04 Pali: also something similar at 0xFFE1 Feb 02 19:00:33 eddyb: really, I'm not interesting in cmt firmware Feb 02 19:00:44 it is something which I do not want to replace on my n900 Feb 02 19:01:11 it is also not possible due to strong signing key Feb 02 19:01:31 Pali: I'm trying to make sense of the values listed in the header Feb 02 19:02:17 eddyb, I think that mcusw file is: some mcusw header, then x-loader header and then data Feb 02 19:02:25 pali: yep, its working :) Feb 02 19:02:41 eddyb, but information is useless Feb 02 19:02:41 Pali: how can the initfs geht lost? Feb 02 19:03:08 nusse: nokia decided to not use initfs anymore Feb 02 19:03:10 shall i boot between flashing or not? Feb 02 19:03:19 Pali: k Feb 02 19:03:23 Pali: yepp, and the x-loader header has triplets (offset, length, name). but I'm not sure how the offset works Feb 02 19:03:26 no you should not boot between flashing Feb 02 19:03:42 so i wait some seconds and remove the battery? Feb 02 19:03:58 if you boot phone between reboots you break phone again... Feb 02 19:04:24 ok Feb 02 19:04:28 nusse, after eMMC flashing is done, try reboot command from flasher Feb 02 19:04:40 (if reboot command will not work, remove battery) Feb 02 19:04:41 using -R? Feb 02 19:04:54 but i flash the combined after that? Feb 02 19:05:11 correct order is: combined, emmc, combined Feb 02 19:05:39 yeah, understood that Feb 02 19:07:23 haha i just tried to do 'apt-get dist-upgrade' Feb 02 19:07:43 it wanted remove all installed packages and install 5 new ones Feb 02 19:09:34 so it seems to be booting, how do i upgrade now? Feb 02 19:09:52 nusse: now you have PR1.3 installed Feb 02 19:10:03 gaaah Feb 02 19:10:16 nusse: you decided to do this in the midst of a bigass server migration Feb 02 19:10:46 so you have to set some mirrors up Feb 02 19:10:47 there's random NOLO strings throughout the first few KBs of the image. it's like "YOLO" staring at me Feb 02 19:10:52 i tried to use a mirror Feb 02 19:11:01 nusse: HAM won't like that, though Feb 02 19:11:05 so you have to tell it to stfu Feb 02 19:11:07 i used apt-get Feb 02 19:11:20 yeah but you weren't supposed to use dist-upgrade :) Feb 02 19:11:27 nusse, what do you want to upgrade? Feb 02 19:11:29 upgrading is not possible btw without mount --bind something from mydocs to /var/cache/apt/archives Feb 02 19:11:30 you have now last version Feb 02 19:11:43 nusse: /var/cache/apt is already in /opt Feb 02 19:11:51 and you don't need to upgrade much, really Feb 02 19:11:53 it'll fit Feb 02 19:12:12 really? Feb 02 19:12:29 i guess since 2010 there were a lot of bugs Feb 02 19:12:39 yes, already fixed Feb 02 19:12:59 so i do want to upgrade? Feb 02 19:13:09 what do you want to upgrade? Feb 02 19:13:17 nusse: the only upgrade from nokia is just a security fix that's also included in cssu Feb 02 19:13:30 nusse: only upgrade is to cssu Feb 02 19:13:41 ~cssu Feb 02 19:13:42 i guess cssu is http://wiki.maemo.org/Community_SSU Feb 02 19:15:05 ah okay, so i want cssu Feb 02 19:15:09 yep Feb 02 19:15:10 thx Feb 02 19:15:34 nusse: follow the installation steps closely Feb 02 19:15:37 i've been just using the old n900 for two years not reinstalling anything, so i have no clue Feb 02 19:15:38 especially the installation of backupmenu Feb 02 19:16:28 do i need backups? i could just reflash Feb 02 19:16:37 restoring a backup is a lot faster :) Feb 02 19:16:48 ah no, flashing took maybe 5mins Feb 02 19:17:03 and you can make one after installing cssu, too Feb 02 19:17:09 and before tinkering with the system Feb 02 19:17:13 and after tinkering with the system Feb 02 19:17:24 no, i'm lazy Feb 02 19:17:36 but repository.maemo.org seems still down Feb 02 19:17:46 it's up, but so slow that it seems down Feb 02 19:17:50 which is possibly even worse Feb 02 19:18:22 add a mirror repo for extras Feb 02 19:18:29 install backupmenu Feb 02 19:18:46 you could do it in your sources.list, so you don't have to wait for HAM to do its thing Feb 02 19:24:44 what do i install after adding the extras repo? Feb 02 19:25:25 "backupmenu" Feb 02 19:25:33 seriously, it's good Feb 02 19:25:52 install it, shut down, boot with the keyboard open, select backupmenu Feb 02 19:26:09 (always backup both rootfs and optfs) Feb 02 19:30:12 community-ssu-enabler :) Feb 02 19:30:55 nusse: to install cssu, enable a cssu mirror, install community-ssu-enabler, run community-ssu-enabler, update with HAM Feb 02 19:31:01 or update with apt-get and then reboot Feb 02 19:35:57 kerio: thx Feb 02 19:42:15 hey Feb 02 19:43:33 hey Feb 02 19:48:54 so community-ssu-enabler removed the mirror i added and added maemo.org again Feb 02 19:49:40 community-ssu-enabler adding its own repository Feb 02 19:49:50 into "trusted" database Feb 02 19:49:58 hm, it shouldn't remove the mirror... oh, did you add it to the file in sources.list.d? Feb 02 19:50:04 that file gets overwritten by HAM all the time Feb 02 19:50:23 nusse: HAM won't let you update from a mirror unless you tell it to do so Feb 02 19:50:56 kerio, it removing only "http://repository.maemo.org/community/" repositories Feb 02 19:53:25 can i install cssu without that crappy enabler? :) Feb 02 19:56:42 nusse: you've ran the enabler Feb 02 19:56:45 it's ok Feb 02 19:56:54 now just add a mirror Feb 02 19:56:58 and upgrade Feb 02 19:57:55 either do it with apt, but it's unsupported Feb 02 19:58:22 or enable HAM's redpill mode and disable "ignore packages from wrong domains" but it's unsupported too Feb 02 19:58:34 or wait until warfare sets up cloudflare for rmo Feb 02 20:00:50 kerio: don't spread fud Feb 02 20:01:27 DocScrutinizer05: see #maemo-ssu backscroll Feb 02 20:01:44 also i thought you were away and you had a life and all sorts of stuff like that :P Feb 02 20:06:38 very strange thing: I cannot charge, while n900 is on-on. It happily charge, while it is off and while bateery is almost full, bur not anywhere below 85% Feb 02 20:07:17 that is strange, indeed Feb 02 20:07:26 it is third time it is happening Feb 02 20:07:33 WizardNumberNext: try using a different charger Feb 02 20:07:51 maybe the cable is busted and can't sustain a high enough current Feb 02 20:07:54 two "different" chargers Feb 02 20:08:00 I have two of each Feb 02 20:08:05 AC-10X Feb 02 20:08:23 AC-3X AC-8X and desktop Feb 02 20:08:36 and obviously computer Feb 02 20:08:44 would start with computer and see Feb 02 20:08:47 how do you know that it's not charging? Feb 02 20:08:57 AC-3X is too weak and doesn't move it Feb 02 20:09:10 checking bq27200 Feb 02 20:09:26 not to mention pernament green led Feb 02 20:09:50 the green led is bme Feb 02 20:09:54 or, rather, hald-addon-bme Feb 02 20:09:54 I have changed my led-pattern for chraging to red-orange-yellow-green Feb 02 20:10:20 I know - I think it might be bme fault, but... Feb 02 20:10:47 while I plug charger OS is getting up (till some point) anyway or at least it looks like Feb 02 20:11:00 but in this state it charges Feb 02 20:11:37 are the ntfs drivers included with kernel-power51? Feb 02 20:11:48 lokimotive: wtf are ntfs drivers? Feb 02 20:12:05 ntfs is included Feb 02 20:12:06 he meant ntfs fs driver Feb 02 20:12:09 rather module Feb 02 20:12:10 there's no such thing, or at least there should be no such thing Feb 02 20:12:17 there's FUSE Feb 02 20:12:27 Pali: ew, the read-only one? Feb 02 20:12:29 yes there is kernel driver Feb 02 20:12:35 kerie: too later - it started to charge now Feb 02 20:12:38 hm Feb 02 20:12:41 by default yes Feb 02 20:12:43 crap Feb 02 20:12:51 or at least was trying to Feb 02 20:12:55 but you can mount it into rw mode too Feb 02 20:13:07 Pali: no, i mean, the one that's heavily suggested to be used read-only Feb 02 20:13:22 yes that Feb 02 20:13:47 kerio: so what's to see on #maemo-ssu? That a silly service provider has bizzare ideas about domain ownership? Feb 02 20:14:01 kerio: 168 upgraded, 0 newly installed, 0 to remove and 5 not upgraded Feb 02 20:14:16 nusse: is this with an "apt-get upgrade"? Feb 02 20:14:17 kerio: on computer it doesn't charge as well Feb 02 20:14:21 kerio: yes Feb 02 20:14:23 you should really use HAM for this Feb 02 20:14:32 but i'm not sure wether cssu is really enabled Feb 02 20:14:42 nusse: apt-cache policy mp-fremantle-community-pr Feb 02 20:14:49 it should be installed with version 0 or something Feb 02 20:14:57 i'm not sure you ran cssu-enabler Feb 02 20:15:50 [2013-02-02 20:53:25] can i install cssu without that crappy enabler? :) Feb 02 20:15:51 [2013-02-02 20:56:42] nusse: you've ran the enabler Feb 02 20:15:53 [2013-02-02 20:56:44] it's ok Feb 02 20:16:05 DocScrutinizer05: he doesn't have the repos anymore! Feb 02 20:16:12 ...except that maybe i was dead wrong Feb 02 20:16:17 the repos are added by postinst Feb 02 20:16:40 installing cssu can be installed only by enabler Feb 02 20:16:43 nusse: run "sudo community-ssu-enabler" Feb 02 20:16:59 enabler doing all needed Feb 02 20:17:11 Pali: except setting up mirrors and telling HAM to ignore domains Feb 02 20:17:15 gpg key, ham catalogue Feb 02 20:17:31 metapackage Feb 02 20:17:46 Pali: are you sure? i thought that repos and domains were added by the postinst, and the enabler binary did the metapackage thing Feb 02 20:18:01 I mean enabler package Feb 02 20:18:18 btw, why nokia still haven't signed their repo with admin@maemo.research.nokia.com? Feb 02 20:18:38 kerio: "already enabled" Feb 02 20:18:43 wirr: because they're something that can't be expressed in a public, publicly-logged channel Feb 02 20:18:54 wirr, because some manager in nokia must tell to developers: you can do that Feb 02 20:19:07 nusse: you'll have to do this with HAM, i'm afraid Feb 02 20:19:18 i think i dont like ham Feb 02 20:19:54 nusse: put "red-pill 1" "red-pill-ignore-wrong-domains 0" "red-pill-permanent 1" in three lines in ~/.osso/hildon-application-manager Feb 02 20:20:14 * wirr wonders why he even asked... Feb 02 20:20:15 (without quotes) Feb 02 20:20:18 application manager show an cssu update Feb 02 20:20:27 nusse: ...does it? :o Feb 02 20:20:33 is r.m.o down?? Feb 02 20:20:46 Pali: rmo is slow enough to be somewhat useless right now Feb 02 20:20:47 at least for me Feb 02 20:21:52 nokia pc suite, wtf Feb 02 20:22:00 nusse: indeed Feb 02 20:22:05 "to update your device to this version you must use nokia pc suite" Feb 02 20:22:06 nusse: close ham Feb 02 20:22:11 wait for it to actually close Feb 02 20:22:25 then create ~/.osso/hildon-application-manager as i said Feb 02 20:22:29 ok Feb 02 20:23:27 (i assume you want to install community-testing) Feb 02 20:25:10 yes :) Feb 02 20:25:32 then you'll have to also add an extras-testing or extras-devel mirror Feb 02 20:25:50 because there's a problem in the current cssu-testing Feb 02 20:26:12 could it be that there are problems with cssu-thumb and usb-mode? Feb 02 20:29:01 add an extras-testing mirror, add a community-testing mirror, disable the standard community-testing, disable that extras mirror, wait 20 to 30 minutes for HAM to do whatever it thinks it should be doing, refresh, do system upgrade Feb 02 20:33:18 lxp1: ping Feb 02 20:36:22 kerio: i should be able to do all that also using vi and apt-get Feb 02 20:38:48 where is openssh package for n900, its not being found by my repos Feb 02 21:10:11 Nice :D Feb 02 21:10:13 Get:28 http://repository.maemo.org fremantle-1.3/free Packages [26,9MB] Feb 02 21:10:13 11% [28 Packages 3842283/26,9MB 14%] 241B/s 1d 13h16min30s Feb 02 21:10:25 hahaha Feb 02 21:10:30 I think kerio adviced to be patient and calm :D Feb 02 21:25:07 hah Feb 02 21:25:13 what's this? http://www.52rd.com/bbs/Dispbbs.asp?BoardID=8&ID=132730 Feb 02 21:25:18 luf: you are a lot further then me Feb 02 21:26:01 WizardNumberNext: I'm just waiting 2 hours ... :D Feb 02 21:26:30 I chisgarged my battery twice Feb 02 21:26:50 I'm using net over usb :D Feb 02 21:26:56 s/chisgarged/discharged/ Feb 02 21:26:57 WizardNumberNext meant: I discharged my battery twice Feb 02 21:27:16 might an idea Feb 02 21:27:28 s/might/might be/ Feb 02 21:27:29 WizardNumberNext meant: might be an idea Feb 02 21:28:16 sorry my brain is dead whole day and even 2 litres of extra-strong green tea doesn't help Feb 02 21:28:51 * WizardNumberNext might be becoming zombie Feb 02 21:29:03 WizardNumberNext: :D Nothing helps better than sleeping :D Feb 02 21:29:20 luf: an hour after waking up? Feb 02 21:30:16 ok, now it is 12 hours, but my brain should be pretty much alive for next 2 days on that amount of green tea Feb 02 21:30:17 s//x/ Feb 02 21:30:34 :( Feb 02 21:30:42 lokimotive: what the hell are you trying to do? Feb 02 21:30:53 sorry i just had to try Feb 02 21:31:09 exchange nothin for somthing - I wish i could do that! Feb 02 21:31:21 would be very nice Feb 02 21:31:26 I would be rich Feb 02 21:31:32 s//money/ Feb 02 21:32:18 lokimotive - I have to try it out on real sed Feb 02 21:32:36 but I m afraid I would run out of paper soon Feb 02 21:32:57 ink - I have for another 750 sheets Feb 02 21:34:17 exaclty 750 sheets? Feb 02 21:34:26 s/exaclty/exactly/ Feb 02 21:34:27 lokimotive meant: exactly 750 sheets? Feb 02 21:34:32 tehee Feb 02 21:34:53 WizardNumberNext: it's the worst one hour after waking up :D Feb 02 21:34:58 not sure - I have put something like 300-350 sheets into drawer Feb 02 21:38:03 I had 500, how much i have now? No idea, but, when I would use all this, then I still would have ink for another 500 sheets. I have office monster Feb 02 21:38:53 eddyb: what's that? Feb 02 21:39:06 hm i need a laser printer Feb 02 21:39:23 kerio: nothing interesting, sadly Feb 02 21:39:49 lokimotive: mine is HP OJ PRO 8500A Plus (910N) Feb 02 21:39:57 sorry 910A Feb 02 21:40:42 atm i only have an hp deskjet and every time i set it up either only scanning works or only printing Feb 02 21:41:55 mine does everything Feb 02 21:42:11 but I have spend more then 100 after finding it on street Feb 02 21:43:50 was worth every single pence I spend on it - new from hp is 280gbp Feb 02 21:44:31 :) Feb 02 22:05:36 Spam kills http://www.news.com.au/technology/suicide-bomber-blown-up-prematurely-by-spam-text/story-e6frfro0-1225997374717 Feb 02 22:24:33 is the extras-devel repo reachable? Feb 02 22:24:47 ~mirrors Feb 02 22:24:48 i heard mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Feb 02 22:25:59 hm my problem is that if i try to install cssu i get an error that packages are missing Feb 02 22:26:19 i have the mirror the list though Feb 02 22:26:30 extras and extras-devel that i Feb 02 22:26:37 s Feb 02 22:42:59 lokimotive: a working extrAh crap Feb 02 22:43:18 Pali: this shit right here is the reason we should get rid of HAM domains Feb 02 22:44:07 kerio, and what will solve this? Feb 02 22:44:09 lokimotive: enable redpill mode, disable "ignore packages from wrong domains", do it again Feb 02 22:44:12 installing cssu not! Feb 02 22:44:22 speaking like yoda, you are Feb 02 22:44:40 Pali: it's for the *next* time rmo goes down Feb 02 22:44:52 besides, you'd still get a confirmation message Feb 02 22:45:45 kerio, instead enable red pill mode, write info how to edit ham domain file Feb 02 22:46:43 Pali: i suppose that those mirrors should provide a package rhat adds the repo officially Feb 02 22:47:17 kerio, so write this to mirror admins Feb 02 22:48:02 or maybe HAM could stop being annoying and ask the user Feb 02 22:48:15 these mirrors were provided for free, ffs Feb 02 22:51:01 kerio, write patch for HAM Feb 02 22:54:05 let me guess, ham is closed source? Feb 02 22:55:02 nope Feb 02 22:55:32 you can enjoy its source in its whole horrible beauty :) Feb 02 22:55:56 hehe Feb 02 23:00:44 lokimotive, HAM is fully open source, and reading source code causing hair loss :D Feb 02 23:04:08 Pali: haha, this will be much easier than i thought Feb 02 23:04:27 src/settings.cc, line 54 Feb 02 23:04:37 s/true/false/ Feb 02 23:10:04 * kerio wrote the patch Feb 02 23:29:15 callate Feb 02 23:38:40 After several hours: Err http://repository.maemo.org fremantle-1.3/free Packages Feb 02 23:38:41 503 Service Unavailable Feb 02 23:39:46 bendo blak berry Feb 02 23:42:02 what Feb 02 23:42:19 hello Feb 02 23:42:20 anthony1 are you trying to say "selling blackberry" ? Feb 02 23:42:40 yes Feb 02 23:43:04 i'm speek englis Feb 02 23:43:25 i from venezuelan Feb 02 23:57:27 Evening Feb 02 23:59:11 I have a weird issue: headphones not recognized in higher-level sw, but dmesg spits out gpio 177 connected/disconnected Feb 03 00:01:39 hoooooooooooooooooolllllllllllllllllllllllllaaaaaaaaaaaaaaaaaaaaaaaaaaa Feb 03 00:02:03 Hi? Feb 03 00:02:15 anthony1 do you know what #maemo is? Feb 03 00:07:28 quien ES MI amigo Feb 03 00:09:15 right.... Feb 03 00:13:14 damn, r.m.o is still running slow :( Feb 03 00:14:32 hmmm https://bugs.maemo.org/show_bug.cgi?id=6360, does not give me any info :( Feb 03 00:14:33 04Bug 6360: Headset problems Feb 03 00:17:13 * MrPingu guesses he needs to sacrifice his uptime for some music Feb 03 00:19:14 q Feb 03 00:20:27 was that guy.. real? Feb 03 00:21:01 Reboot fixed it. hmpf Feb 03 00:22:54 MrPingu: uptime R.I.P. :D Feb 03 00:23:48 luf: yeah, but how can that bug be invalid, then... Feb 03 00:24:37 Almost got a month, and freemangordon promised me to release a new thumb after I got an uptime of a month :P Feb 03 00:24:38 MrPingu: no idea Feb 03 00:25:07 Ain't gonna happen with these weird bugs Feb 03 00:25:11 MrPingu: so freemangordon can relax. Good news. Feb 03 00:26:01 :) Feb 03 00:27:35 Probably kerio won't like this *evil laugh* Feb 03 00:28:44 Nah, freemangordon you are not bound to keep your promise ;) **** ENDING LOGGING AT Sun Feb 03 02:59:59 2013