**** BEGIN LOGGING AT Thu Jan 31 02:59:58 2013 Jan 31 03:30:02 hello i'm trying to use the skeiron extras mirror 3 repo but i'm getting a "sub-process gzip returned an error code (1)" error Jan 31 03:30:13 somebody can help? Jan 31 03:30:30 man gzio Jan 31 03:30:34 gzip Jan 31 03:30:41 what's error code 1 Jan 31 03:30:52 possibly file corrupt Jan 31 03:33:05 i can't find a list of gzip errors in the man page Jan 31 03:34:03 but doing the update again i get the same error Jan 31 03:34:48 i just flashed my nokia n900 and i need rootsh Jan 31 03:34:55 bad timing i guess... Jan 31 03:36:35 for everybody else the repos works fine? Jan 31 03:42:17 on reflection Jan 31 03:42:31 it's quite possible the repos are having an issue right now Jan 31 03:43:18 I'd have to get out of bed to check, and ... Jan 31 03:43:21 night Jan 31 03:44:58 mmh than i guess i'll just wait Jan 31 03:45:24 i should get some sleep too,it' almost 5 in the morning here... Jan 31 04:25:56 (recommended apps) I'm smelling loooong discussions what's indispensable for daily usage and what's geek shit. How about a list of *deprecated* apps instead? Jan 31 04:26:15 I already know two: powerpatch, speedpatch Jan 31 04:28:18 and honestly, any such list (recommended) is always *personal* (recommended by *whom*), and accordingly it should get represented on wiki Jan 31 04:28:23 ~jrtools Jan 31 04:28:23 methinks jrtools is http://wiki.maemo.org/User:Joerg_rw/tools Jan 31 04:30:38 a "public" and "official" wikipage always aiming for the one and only truth, and everybody will vandalize it to correct the "obvious errors" when it comes to something as individual as a list of most useful apps Jan 31 04:32:14 hey, if I wanna know which app most users think is useful, I go to ~packages and check for top of download-hitlist Jan 31 04:33:10 if I wanna know which apps wirr thinks are most useful, I go to Users:wirr Jan 31 04:35:05 please don't start such dogfight topic on a "official website", since it's not really backed up by anything or anybody. Nobody elected for those apps in any proper election, nobody appointed the editors of such page as being the ultimate experts either Jan 31 04:36:25 what do you expect the BlessN900 devels to do when you place fcam on that page? Jan 31 04:36:33 this will NOT end good Jan 31 04:37:03 ends in a lot of bickering and fights, and nobody benefiting from it in the end Jan 31 04:41:56 and I'm not even starting with more basic questions like "is busybox a good or a silly app in sense of unix spirit?" Jan 31 04:42:12 ~messybox Jan 31 04:42:12 messy... err busybox is meant for lean scripting. Regarding all the missing options and immanent limitations (see su, passwd) it's not really the interactive shell of choice. A lot of people hate busybox because a lot of system integrators don't understand the difference between busybox and a decent user interactive shell plus unix utils Jan 31 04:43:26 unix way is to have lots of small specialized though general purpose tools that you compine to your liking. Monolithic blobs like busybox are the windows way to implement things Jan 31 04:44:28 and since ls is still buggy in busybox power, we gonna invent my-acme-ls and ship that package as well then? Jan 31 04:45:08 since no sane devel interested in fixin ls will look into busybox power to do so Jan 31 04:47:10 that's a straight highway into windows hell Jan 31 04:52:14 wirr: pretty please use a filemanager and have a look into your hildon-backup files, you'll find an extremely convenient and ready-made way to deploy package/app-sets. I suggested several times somebody to simply make such a backup file with his favourite apps and deploy it for other users Jan 31 04:53:45 you're aware that hildon-backup only stores a *list of apps* to re-install on restore, it does NOT store the app binaries Jan 31 04:54:34 and, as icing on top, user can deselect particular apps during restore, so they won't get installed Jan 31 04:55:36 the only way in HAM to first select/deselect on a huge number of packages and then run the whole batch for unattended installation Jan 31 04:56:43 once such a "package" been built and published under "wirr's favorites", you bet a week later there will be a clone "kerio's favorites" Jan 31 04:56:54 or whomever it may tickle Jan 31 04:57:30 and eventually we even might come up with "cssu recommended" Jan 31 04:58:29 and I bet somebody will create "2012 most wanted" Jan 31 04:59:00 or "the spy collection", you name it Jan 31 05:04:27 hi i just got myself a N900, its currently charging up... :D where can i read about upgrading its software to the latest maemo, also i want to do some development so I want to root the device and run linux stuff on it... where can i find the howtos? Jan 31 05:04:49 haha Jan 31 05:04:54 welcome damo22 Jan 31 05:05:06 ty Jan 31 05:05:59 1st: your N900 doesn't need any "rooting", it's already open, just missing a small package called rootsh you need to install to get the commands "root" and "sudo gainroot" Jan 31 05:06:14 cool Jan 31 05:06:36 runing "linux stuff" is possible without that "root account2 generally Jan 31 05:06:42 runing "linux stuff" is possible without that "root account" generally Jan 31 05:07:00 yeah, but i want root on my own phone :D Jan 31 05:07:02 * freemangordon thinks n900 runs "linux stuff" all the time :) Jan 31 05:08:26 this is so cool :D Jan 31 05:08:30 you can install all the wealth of gnu-utils readily. And you can run all stuff like ls and ps and mc and whatever "unix stuff" you can think of, without rootsh, just some of those my need root permissions to do what you want them to do. I'm sure you're aware of this basic concept Jan 31 05:09:19 yeah, i do a bit of kernel stuff i think i know my way around *nix :) Jan 31 05:09:54 damo22: for allowing "root" and "sudo gainroot commands" open http://maemo.merlin1991.at/apt-mirror/extras/pool/fremantle-1.3/free/r/rootsh/rootsh_1.8_all.deb in microb Jan 31 05:09:56 damo22: alas you picked a very poor point in time to start with your exciting maemo adventures, since due to migration away from Nokia and towards 100% community driven, some of the resources are... flawed Jan 31 05:10:11 that one too :) Jan 31 05:10:12 but it will be fun to play with a fully fledged system that is my own phone Jan 31 05:10:33 you bet it is Jan 31 05:10:56 I'm enjoying this fun since 4 (?) years now and will never give up on it Jan 31 05:11:07 is it possible to use the small usb port as a host controller? Jan 31 05:11:16 "from my dead cold haaaaands!" Jan 31 05:11:27 damo22: you bet it is Jan 31 05:11:32 wowww Jan 31 05:11:38 yes, I made that possible, see H-E-N Jan 31 05:11:45 hehe Jan 31 05:12:05 how much power can i draw from the usb port Jan 31 05:12:10 safely Jan 31 05:12:10 200mA Jan 31 05:12:21 k Jan 31 05:12:37 DocScrutinizer05: current protection kicks in above that, ain;t? Jan 31 05:12:46 yup Jan 31 05:12:57 EOVERCURRENT Jan 31 05:13:04 cut out Jan 31 05:13:30 i was thinking of making a fully fledged audio recording device Jan 31 05:13:37 using external usb sound card Jan 31 05:13:39 GOOOD Jan 31 05:13:51 I'm actually interested in that Jan 31 05:14:01 damo22: there is a thread on talk.maemo.org about that Jan 31 05:14:07 i already wrote a kernel alsa driver quirk for this card Jan 31 05:14:20 its made by digidesign Jan 31 05:14:36 cool Jan 31 05:14:38 damo22: I guess you want to install kernel-power Jan 31 05:15:15 yeah damo22, you're speakeing with one of the two guys you need to know and make friens, for your project: freemangordon Jan 31 05:15:29 pali is the other guy and main maintainer of powerkernel Jan 31 05:15:33 sweet! Jan 31 05:16:09 and for USB stuff, ping me or pali ;-) Jan 31 05:16:15 damo22: being a kernel hacker, there is a piece of HW which sits unused in n900. A programmable filter that is Jan 31 05:16:29 ? Jan 31 05:16:40 programmable filter? Jan 31 05:16:47 as in audio filter? Jan 31 05:16:58 yes. there is a patch which enables it, but unfortunately it is incomplete Jan 31 05:16:59 ohnoes, please not "the digital filter in codec: sequel #4" Jan 31 05:17:31 DocScrutinizer05: I am talking about HW audio filter Jan 31 05:17:45 or actually, go ahead Jan 31 05:17:46 well ive already written some digital filters using biquad stuff but its cpu driven Jan 31 05:18:10 freemangordon: that's the digital parametric filter in codec I spoke of Jan 31 05:18:25 there are actually two of them Jan 31 05:18:29 damo22: the patch needs filtering bands calculated, otherwise it is functional aiui Jan 31 05:18:47 one full parametric one, plus an emphasis one Jan 31 05:18:53 ooo Jan 31 05:19:00 DocScrutinizer05: aah, ok. still having my first coffee, you know :) Jan 31 05:19:19 DocScrutinizer05: two of them? Jan 31 05:19:22 this could be fun Jan 31 05:19:23 yes Jan 31 05:19:37 HW audio filtering Jan 31 05:19:56 via alsa? Jan 31 05:20:10 damo22: now it is PA plugin which does highpass filtering ( speaker protection) Jan 31 05:20:35 damo22: I guiess there is no als plugin for that Jan 31 05:20:39 *alsa Jan 31 05:20:44 http://www.musicdsp.org/files/Audio-EQ-Cookbook.txt http://www.student.oulu.fi/~oniemita/dsp/dspstuff.txt Jan 31 05:20:55 i know that musicdsp site Jan 31 05:21:09 we implemented it in Calf plugins Jan 31 05:21:55 i made a phono RIAA filter using bilinear transform too Jan 31 05:22:01 that was fun Jan 31 05:22:14 yeah, that's what the de-emphasis filter is for Jan 31 05:22:38 the smaller one of the two (if I'm not completely making that up now) Jan 31 05:23:12 so youre saying that the N900 has a hardware codec that can do filtering but the coefficients arent calculated yet? Jan 31 05:23:23 yes Jan 31 05:23:39 what kind of curve do we need Jan 31 05:23:48 that's the question Jan 31 05:23:59 where things came to a grinding halt Jan 31 05:24:01 ;-) Jan 31 05:24:08 ahhh Jan 31 05:24:13 since evrybody wanted a graphic EQ Jan 31 05:24:20 well you probably dont need that Jan 31 05:24:30 and that's wirtually impossible with that stuff Jan 31 05:25:09 not really, i made a calculation tool that generates coefficients on the fly from analog parameters Jan 31 05:25:20 cooool Jan 31 05:25:33 yo da man Jan 31 05:26:05 damo22: https://garage.maemo.org/plugins/ggit/browse.php/?p=kernel-power;a=blob;f=kernel-power-2.6.28/debian/patches/Support-for-tlv320aic3x-codec-highpass-filter-needed.diff;h=962061affdbff324d640a6edf3b5fd17889b74e7;hb=HEAD Jan 31 05:26:10 here is the patch Jan 31 05:26:11 but we didnt end up using it, because someone else already implemented it before in Calf plugins Jan 31 05:26:15 i *think* we already have access to the coefficients Jan 31 05:26:39 its a C++ class called biquad Jan 31 05:26:55 sounds familiar Jan 31 05:27:08 ooh, "The filter can be controlled through an alsa hwdep device, via" Jan 31 05:27:11 ... Jan 31 05:27:24 freemangordon: toldya Jan 31 05:27:34 nice Jan 31 05:27:43 i *think* we already have access to the coefficients Jan 31 05:28:00 just hex number desert Jan 31 05:28:07 somewhere Jan 31 05:28:16 so, what we need, correct algo for coeffs and gstreamer plugin? Jan 31 05:28:24 yep Jan 31 05:28:30 nice Jan 31 05:28:36 afaik Jan 31 05:28:49 it's years ago since I last touched that stuff Jan 31 05:28:56 DocScrutinizer05: this is on output only, isn't it? Jan 31 05:29:08 hey cool, i can see how i can put my code in there Jan 31 05:29:08 err, probably yes Jan 31 05:29:25 but i dont know how to make the coefficients configurable Jan 31 05:29:37 see aic34(?) datasheet, and N900 schematics Jan 31 05:30:57 for example, i could add something in there that converts "#define LP_FC 200" into a set of coefficients for a lowpass filter at 200 Hz Jan 31 05:31:28 but how would anyone change the value Jan 31 05:31:36 http://maemo.cloud-7.de/hidden/ see docs/hw/tlv320aic34.pdf Jan 31 05:31:48 never post complete URL! Jan 31 05:32:02 okay Jan 31 05:32:02 anywhere Jan 31 05:32:36 damo22: did you look at the patch? Jan 31 05:32:45 freemangordon: yeah i had a glance Jan 31 05:33:09 "Added functions to manipulate the filter's coeffs and enable/disable it,..." Jan 31 05:33:19 oh geez Jan 31 05:33:28 "hwdep device called "IIR Filter" added, documentation on how to use it..." Jan 31 05:34:17 oh, there is matlab code there :D Jan 31 05:34:38 to draw the freq responce Jan 31 05:36:13 In addition to the de-emphasis filter block, the DAC digital effects processing includes a fourth-order digital IIR filter with programmable coefficients (one set per channel). This filter is implemented as a cascade of two biquad sections with frequency response given by: Jan 31 05:36:45 Nx= numerator_x Dx = denominator_x Jan 31 05:37:17 * freemangordon wonders if this can be used as an eq Jan 31 05:37:39 well page 38 of tlv320aic34.pdf looks a tiny bit more *cough* Jan 31 05:37:40 it can be used as an eq Jan 31 05:37:53 how may bands? Jan 31 05:38:17 freemangordon: a parametric EQ has no bands Jan 31 05:38:40 but do we have enough params for, lets say, 10 bands? Jan 31 05:38:50 and a biquad IIR is not even a parametric EQ even Jan 31 05:39:06 a biquad IIR doesnt work too well with bass frequencies Jan 31 05:39:41 if you try to use it in a narrow band Jan 31 05:40:10 anyway, gtg, damo22, have fun with that :) Jan 31 05:40:13 freemangordon: how many bands has y=a1*x+a2*x^2+a3*x^3+a4*x^4 ? Jan 31 05:40:16 ty Jan 31 05:40:44 its more like it can have up to 3 kinks Jan 31 05:40:46 DocScrutinizer05: hard to say, but I guess it can't be used as an EQ. Jan 31 05:41:05 this looks like a smooth function Jan 31 05:41:14 that's the point of a parametric or biquad Jan 31 05:41:19 IIR Jan 31 05:41:23 no EQ then Jan 31 05:41:25 its a polynomial quotient Jan 31 05:41:29 yep Jan 31 05:41:35 anyway, bye for now Jan 31 05:42:03 cya Jan 31 05:44:06 DocScrutinizer05: what is the benefit of the filter in hardware? would it not be more useful as a filter to protect the speakers Jan 31 05:44:25 rather than being used as an eq Jan 31 05:45:59 ie, just set a steep highpass at 20Hz and lowpass at 20kHz? Jan 31 05:46:39 sure Jan 31 05:46:40 set and forget Jan 31 05:47:00 and aiui Nokia is already using it that way Jan 31 05:47:21 :) Jan 31 05:47:24 parameters for that stuff are even on kernel's boot cmdline Jan 31 05:47:28 afaik Jan 31 05:47:37 cool Jan 31 05:48:22 take the last staement with more than one grain of salt, I only know there's "something2 on kernel cmdline related to audio Jan 31 05:48:30 :) Jan 31 05:49:03 btw 20Hz for highpass is a lot for those small speakers Jan 31 05:49:13 more like 50Hz Jan 31 05:49:19 okay Jan 31 05:49:24 maybe even 100 Jan 31 05:49:54 Nokia killed lots of speakers on their preprod protos, with that stuff ;-P Jan 31 05:49:59 so is audio currently destroying peoples speakers?> Jan 31 05:50:08 lol Jan 31 05:50:17 device has a class-D amp with no proper DC-blocking X-P Jan 31 05:50:18 * RST38h yawns Jan 31 05:50:29 hahah Jan 31 05:51:27 damo22: nope, now we got XPROT, some closed blob Nokia proprietary PA plugin Jan 31 05:51:51 aww Jan 31 05:51:58 does dynamic limiting / compression and some filtering, so the gossip Jan 31 05:52:24 kick it out and mess with ALSA dieectly and your speekers are at peril Jan 31 05:52:41 i dont really mind about that, i want to attach external sound card Jan 31 05:52:55 * DocScrutinizer05 glances at his spare speakers Jan 31 05:53:38 afaik XPROT only is in the audio pipe for speakers, not though for any other audio sink Jan 31 05:53:40 but my sound card draws 500mA Jan 31 05:53:47 eeek Jan 31 05:54:05 so i will need to power it externally Jan 31 05:54:12 :nod: Jan 31 05:54:30 Y-cable, external power plug on device, or popwered USB hub Jan 31 05:54:53 ycable would be nice Jan 31 05:55:06 * DocScrutinizer05 wonders when coffee will kick in and if that will help for typos Jan 31 05:55:23 i can get a ycable online? Jan 31 05:55:32 easy Jan 31 05:55:46 used for each >2.5" drive Jan 31 05:56:01 k Jan 31 05:56:01 there's zillion Jan 31 05:56:46 two USB-M plugs, one USB-F or mini/micro-USB-M Jan 31 05:57:15 lemme find my awesome picture for you Jan 31 05:57:51 how cool would it be to have 2 matched studio microphones hooked up to my N900 recording audio to the microSD Jan 31 05:59:13 anyway here's already the augmented booston script that gives you indications of overcurrent etc, sorry for the show effects I built in ;-P http://talk.maemo.org/showthread.php?p=1023402#post1023402 Jan 31 06:00:40 http://talk.maemo.org/showthread.php?p=921203&highlight=Y-cable#post921203 Jan 31 06:01:28 dang, the externally stored pics are a PITA. OVI fileshare vanished, so did the photo in there Jan 31 06:03:33 awesome... any idea if i can get a sim card adapter for microsim -> ordinary sim Jan 31 06:03:35 damo22: you also might want to learn about Jan 31 06:03:37 ~cssu Jan 31 06:03:37 rumour has it, cssu is http://wiki.maemo.org/Community_SSU Jan 31 06:04:01 i am selling my iphone lol Jan 31 06:04:24 those sim card adapters are rather common nowadays, never use them in slot-in SIM holders though Jan 31 06:04:33 luckily N900 has no slot-in Jan 31 06:04:42 sweet Jan 31 06:05:57 stuff like http://wiki.maemo.org/Updating_the_tablet_firmware is no news to you, I hope Jan 31 06:07:03 and, dunno if I already mentioned it, this page has some useful info, according to other users: Jan 31 06:07:07 ~jr-tools Jan 31 06:07:08 jrtools is, like, http://wiki.maemo.org/User:Joerg_rw/tools Jan 31 06:07:39 NEVER do apt-get dist-upgrade! Jan 31 06:07:47 lol ok Jan 31 06:07:49 NEVER do apt-get autoremove Jan 31 06:08:08 maemo is "different" Jan 31 06:08:35 right ok Jan 31 06:10:12 ATM the standard repos are down, due to Nokia hitting the sack Jan 31 06:10:37 so you either find your way around, or you wait for things to return to normal state Jan 31 06:10:46 ~mirrors Jan 31 06:10:46 it has been said that mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 31 06:10:50 thanks Jan 31 06:10:52 ~repo-down Jan 31 06:10:52 i heard repo-down is http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-20.log.html#t2013-01-20T02:08:42, or http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-30.log.html#t2013-01-30T00:51:04 Jan 31 06:11:04 very useful info!! Jan 31 06:11:42 I think just enough stuff to read for your first day on N900 :-) Jan 31 06:12:10 ample Jan 31 06:13:32 consider to donate! ;-) http://talk.maemo.org/showthread.php?t=88222 Jan 31 06:15:39 i will donate now, you already gave me a lot of help Jan 31 06:21:26 hey, the donation isn't for me and my help, it's for keeping your infra alive :-) Jan 31 06:21:38 well i donated Jan 31 06:22:14 all of the above links wouldn't be sustainable without community donors Jan 31 06:22:30 yeah Jan 31 06:22:34 understood Jan 31 06:24:29 thanks Doc i gtg now Jan 31 06:27:27 cya Jan 31 06:29:17 (not sustainable) not even the mirrors, which work since some guys are willing to donor their "own" server stoarage space and bandwidth, skeiron even a dedicated server just for maemo backup only Jan 31 06:30:03 s/donor/donate/ Jan 31 06:30:05 DocScrutinizer05 meant: (not sustainable) not even the mirrors, which work since some guys are willing to donate their "own" server stoarage space and bandwidth, skeiron even a dedicated server just for maemo backup only Jan 31 06:56:11 ok, quite exactly 12h of 'idle' online via 3G, with xchat logged in to some 14 channels Jan 31 06:56:42 not bad a battery standby time, eh? Jan 31 06:57:12 actually I hardly can believe it Jan 31 06:58:14 just thought "uh? what a sound? OMG I forgot to plug n900 to charger last evening" Jan 31 07:00:06 left home 12:20 ago for shopping, and switched to 3G to have a short look at maps. since then it been idling like that, since I disabled auto-switch-to-WLAN yesterday and thus it stayed on 3G even when I came back home Jan 31 07:59:58 DocScrutinizer05: enough with the "don't do autoremove", cssu-enabler fixes that Jan 31 08:01:09 every package listed as a dependency in the nokia metapackage is marked as manually installed Jan 31 08:01:24 (ty Pali) Jan 31 08:11:48 kerio: wtf is cssu-enabler ;-P Jan 31 08:15:34 hey, I have a problem with the alarm clock. The alarm has no sound :( Jan 31 08:15:44 : one that enables another to achieve an end; especially : one who enables another to persist in self-destructive behavior (as substance abuse) by providing excuses or by making it possible to avoid the consequences of such behavior Jan 31 08:15:59 ropeu: check the volume levels for it Jan 31 08:16:00 (fremantle cssu, n900) Jan 31 08:16:11 DocScrutinizer05: i think that the definition of enabler fits perfectly Jan 31 08:17:29 where can I check the volume? normally the alarm always and there is no alarm in nomal profile with sound Jan 31 08:19:29 ropeu: hm, you're quite correct Jan 31 08:19:39 ropeu: which cssu? Jan 31 08:22:22 ...dammit! Jan 31 08:32:22 i was about to make him pastebin the output of profileclient -v, for the record, and check clock.alarm.enabled Jan 31 08:32:31 (literally for the record) Jan 31 08:41:38 when I got a script A that's doing >>B&<<, and I call ssh foo@bar A; what need I to do in B to make ssh A terminate instantly? Jan 31 08:42:34 I gather the question is: how do I deamonize B ? Jan 31 08:42:50 so it's detached from parent A Jan 31 08:43:23 since otherwise ssh doesn't return until B terminates Jan 31 08:43:42 or am I completely wrong? Jan 31 08:44:17 daemons aren't just put in background Jan 31 08:44:29 they disconnect stdin/stdout/stderr and ignore hangup Jan 31 08:44:40 might "exec &>/dev/null aaah ignore sighup Jan 31 08:45:02 thaaaanks Jan 31 08:45:06 so... nohup, probably Jan 31 08:45:16 meh, I hate nohup Jan 31 08:45:33 apparently you can double & Jan 31 08:45:38 like (B &) & Jan 31 08:45:49 so B detaches from the tty Jan 31 08:45:56 it does? Jan 31 08:46:06 it's a tip from linux-mag, iunno Jan 31 08:46:07 test it Jan 31 08:46:24 still working on your dead man's switch? Jan 31 08:46:36 that's a funny thing with testing, this is a live system, your maemo fw ;-) Jan 31 08:46:47 clearly you shouldn't call "reboot" then Jan 31 08:46:52 call "echo penis" Jan 31 08:47:26 o.O Jan 31 08:47:35 echo &>dev/null ? Jan 31 08:47:50 call mkdir ~doc/rebooted Jan 31 08:47:55 or touch ~doc/rebooted Jan 31 08:48:33 meh, I'll go with orthodox trap "" sighup; Jan 31 08:48:52 and exec &>/dev/null hm, is outputting to devnull the same as closing? Jan 31 08:50:06 err not exactly Jan 31 08:50:36 I seem to recall exec semantics also included closing of stdin stdout stderr Jan 31 08:50:55 forgot how to do that Jan 31 08:52:27 exec - execute commands and open, close, or copy file descriptors - thanks posix manpage Jan 31 08:54:29 exec >&0 2>&0 <&0 Jan 31 08:54:38 could have figured Jan 31 08:55:04 isn't <&0 implied? :o Jan 31 08:55:19 implied in what? Jan 31 08:55:26 isn't 0 stdin? Jan 31 08:55:45 err, you might be right there, yeah Jan 31 08:56:19 dang, I need more coffee Jan 31 08:56:29 exec >&- 2>&- <&- Jan 31 09:05:59 DocScrutinizer05: you'll love this: http://talk.maemo.org/showpost.php?p=1319363&postcount=1 User wants a multiboot with cssu and PwnPhone :D :D Jan 31 09:06:14 oh god why Jan 31 09:06:23 yes, it's doable Jan 31 09:06:35 I'll ask him if he wants pink ponies too Jan 31 09:06:43 but if you're the kind of person who'd use pwnphone, you're not good enough Jan 31 09:08:15 pink pwnie Jan 31 09:08:16 hehe Jan 31 09:08:37 Anyone with a N810 please help on instructions of how to add temp mirror: http://talk.maemo.org/showthread.php?t=88870 Jan 31 09:08:46 omg Jan 31 09:08:50 nfc Jan 31 09:08:55 and i have 2 Jan 31 09:09:06 * thedead1440 thought those holding on to N810s wouldn't have such issues Jan 31 09:09:38 I was strictly dumbass user on N810 Jan 31 09:10:00 well you aren't now so you can help him out :D Jan 31 09:10:15 no clue about maemo4 though Jan 31 09:10:19 ask ShadowJK Jan 31 09:12:13 how to fix n810: sell n810, buy n900 Jan 31 09:17:24 oops, wall is nasty Jan 31 09:17:37 hehe Jan 31 09:18:02 I entered panic mode on fw, since my *local* tests spammed the ssh screen with "warning! rebooting in 5min" Jan 31 09:18:43 lol Jan 31 09:18:52 :D Jan 31 09:20:40 jr@halebop:~> ssh newtmo ssh root@fw ./reboot1 Jan 31 09:20:41 reboot2: no process killed Jan 31 09:20:43 jr@halebop:~> Jan 31 09:20:47 that's the way I like it Jan 31 09:21:16 haha Jan 31 09:21:27 DocScrutinizer05: don't chain sshs :( Jan 31 09:21:54 kerio: why not? Jan 31 09:22:08 because the data from fw to halebop is also decrypted on newtmo Jan 31 09:22:20 sure Jan 31 09:22:22 extra security risk, extra used cpu Jan 31 09:22:35 ok no, there's no extra used cpu Jan 31 09:22:41 but it's a risk!1!1 Jan 31 09:22:48 :shrug: any other idea how to reach 10.0.0.8 behind the fw? Jan 31 09:23:24 ssh -oProxyCommand="ssh newtmo nc %h %p" root@fw Jan 31 09:23:39 (ideally you'd put that in your .ssh/config Jan 31 09:23:44 meh Jan 31 09:24:20 Host fw/HostName fw/User root/ProxyCommand ssh newtmo nc %h %p Jan 31 09:24:27 and then ssh fw Jan 31 09:24:39 ssh config is fucking sweet Jan 31 09:25:00 I'm rather happy to have *any* access to that friggin fw, after fussing around with .ssh permissions and ssh-agent and whatnot Jan 31 09:25:53 I still can't access fw from newly started xterm instances, only from a 6 weeks old one Jan 31 09:26:05 ...wat Jan 31 09:26:34 DocScrutinizer05: just copy your id_rsa.pub in .ssh/authorized_keys Jan 31 09:26:40 we have a most incredible fw :-/ Jan 31 09:27:27 kerio: yeah, and then wonder wtf one of my xterms allows acces to machine A, B, and C, while the other one only allows access to A and C Jan 31 09:27:52 DocScrutinizer05: do you have a separate identity file for the fw? Jan 31 09:28:01 no Jan 31 09:28:14 just a new general purpose rsa key Jan 31 09:28:14 oh, actually Jan 31 09:28:23 why do you have to pass through newtmo? Jan 31 09:28:36 because Jan 31 09:28:49 it's an issue of agent key passthrough Jan 31 09:29:13 which is something that was made up for people who don't properly proxy their ssh connections, and instead do ssh newtmo ssh root@fw Jan 31 09:29:32 ( >:c ) Jan 31 09:30:01 DocScrutinizer05: try ssh -oProxyCommand="ssh newtmo nc %h %p" root@fw in a new xterm Jan 31 09:30:06 dang, I never had a need for that before I touched this maemo FW a few days ago Jan 31 09:31:20 the nested ssh is ran by doc@newtmo, you can't expect *that* to have keys to the firewall Jan 31 09:31:25 what exactly is happening to the fw? vm goes down? Jan 31 09:31:30 kerio: you're my hero of the day :-) Jan 31 09:31:45 if you have a properly-configured agent, the key from doc@halebop will be passed through Jan 31 09:31:48 which is fairly neat Jan 31 09:34:59 while sleep 1800; do ssh -oProxyCommand="ssh newtmo nc %h %p" root@fw ./reboot1; done Jan 31 09:35:49 root@MaemoFirewall:~ # cat reboot1 Jan 31 09:35:51 #!/bin/sh Jan 31 09:35:52 killall reboot2 Jan 31 09:35:54 ./reboot2& Jan 31 09:36:14 :D Jan 31 09:37:51 the nice thing is: it enables and tickles the reboot2 "watchdog". So when local while loop doesn't run and calls reboot1, no bootloop will happen after first reboot on FW Jan 31 09:42:21 well Jan 31 09:42:26 that might be bad, too Jan 31 09:47:48 how might that be bad? Jan 31 09:48:37 except for the max 910s delay after a reboot until watchdog gets engaged? Jan 31 09:48:51 (I chnaged sleep in my local loop to 910) Jan 31 09:50:58 DocScrutinizer05: if your computer ends up under a bus, we won't have any reboots! D: Jan 31 09:51:52 well, til 2h ago you wouldn't have any reboots at all Jan 31 09:53:29 any with sustaining this constant semi-manual monitoring of FW, and running from "random places to my terminal at home, to give the almost dead FW a revivla boot, *I* would end under a bus in 2 days Jan 31 09:54:00 s/"/$/ Jan 31 09:54:00 DocScrutinizer05 meant: any with sustaining this constant semi-manual monitoring of FW, and running from $random places to my terminal at home, to give the almost dead FW a revivla boot, *I* would end under a bus in 2 days Jan 31 09:57:39 DocScrutinizer05: Thanks a lot for keeping Maemo alive ! :) Jan 31 09:57:49 yw Jan 31 11:29:48 DocScrutinizer05: what's the RMO situation at the moment? who are we waiting for? Jan 31 11:36:19 the "grid guy" aiui. last (related) msg from Eero been [2013-01-28 10:41:00] ok, stage now has an external vif that is configured but not up, local iptables needs still to be configured. (and [2013-01-28 10:58:32] oh, I *will* take time off when things quiet down enough for it) Jan 31 14:31:34 * ShadowJK hasn't (un/re)installed anything on N8x0 for couple of years now Jan 31 15:03:40 [sidenote] prepare for FW reboot during next 30min Jan 31 15:07:21 my body is ready Jan 31 15:20:59 time for the unfusion, kerio Jan 31 15:21:02 lay back and relax Jan 31 15:21:07 this might sting Jan 31 15:31:22 [sidenote] FW rebooted. Tomorrow we finally might get another try on repository.maemo.org Jan 31 15:41:37 \o/ Jan 31 15:41:56 courtesy warfare Jan 31 15:42:32 warfare: don't forget tomoorow council meeting ;-) Jan 31 15:42:52 1800, UTC! Jan 31 15:43:00 that's 1900CET Jan 31 16:06:17 quoting ppl sending via talk webinterface "each time m tring to connect my yappari,its always tells me connection timed out. i wil like to know wht the reason was nd what Jan 31 16:06:21 remedy can be proffer to it Jan 31 16:08:18 btw I do not fscking care if any of your scientology-fsck-apps do not work! Jan 31 16:09:12 scientology? :o Jan 31 16:11:13 Sc0rpius: ^^^ Yappari issue :D Jan 31 16:12:19 huh? Jan 31 16:19:20 chem|st: I feel with you Jan 31 16:20:26 chem|st: turn off all PM in tmo, I always thought it's useless and EVIL[TM] ;-) Jan 31 16:45:03 freemangordon, all new thumbs released in thumb repo should work OK and you've not announced new thumb version due to repo migration fiasco, or there is something broken there Jan 31 16:45:24 they should work Jan 31 16:45:26 (and/or something that require normal repos/mirrors to be set up, before updating)? Jan 31 16:45:28 thanks Jan 31 16:47:44 erm, quick question to someone knowledgeable - HAM exports its repo settings to /etc/apt/sources.list.d/hildon-application-manager.list Jan 31 16:47:55 it also affects apt-get Jan 31 16:48:31 now, editing hildon-application-manager.list works only temporaly. Is there any way to edit it (where ham reeally stores it) without using HAM itself? Jan 31 16:48:41 hildon-application-manager-config Jan 31 16:48:56 you MUST use it to change ham's configuration Jan 31 16:49:00 ~2119 Jan 31 16:49:00 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119. Jan 31 16:49:24 last thing I'm using HAM and it's slowness glory, is to keep sync between apt repos and fapman repos, so editing it directly would save me from firing up HAM ever... Jan 31 16:49:30 kerio, thanks Jan 31 16:49:37 regular folder or gconf entry? Jan 31 16:49:48 no, no, binary Jan 31 16:50:28 oh lol Jan 31 16:50:31 /usr/bin/hildon-application-manager-config Jan 31 16:50:52 /etc/hildon-application-manager/catalogues won't work? :P Jan 31 16:51:11 not if HAM's running Jan 31 16:51:14 and/or apt-worker Jan 31 16:52:08 erm, how one is supposed to add repo by hildon-application-manager-config? Jan 31 16:52:22 usage is quite sparse Jan 31 16:52:54 beats me Jan 31 16:52:58 check the source Jan 31 16:53:09 or community-ssu-enabler's postinst/prerm Jan 31 16:54:41 * Estel_ nods Jan 31 16:55:20 Estel_: btw, have you installed the battery applet i linked you? Jan 31 16:55:55 new one without default rx51? havent seen that link yet, got 1300+ lines logged on #maemo Jan 31 16:56:01 :P Jan 31 16:56:06 so it's just assumption Jan 31 16:56:11 that it's something new Jan 31 16:56:27 Estel_: https://dl.dropbox.com/u/1405219/status-area-applet-battery_1.0-3%2Bkerio0_armel.deb Jan 31 16:56:36 or, if you don't trust the binary, https://dl.dropbox.com/u/1405219/status-area-applet-battery_1.0-3%2Bkerio0.tar.gz Jan 31 16:56:37 OK, wtf is that? Jan 31 16:56:45 Estel_: magic Jan 31 16:56:52 modified by you to get rid of rx51? Jan 31 16:57:00 pali haven't released it yet, yep? Jan 31 16:57:12 should be installed on top of bme replacement suite? Jan 31 16:57:20 it should work in every situation Jan 31 16:57:40 it's just that now rx51's design data is only used if bq27k can't provide a capacity Jan 31 16:57:50 sure, but why to edit startup scripts myself if there is thing doing it made by Pali and will be compatible with his future releases Jan 31 16:57:53 nice Jan 31 16:58:05 oh, it also has a noticeably faster startup Jan 31 16:58:24 Estel_: this is just the battery applet Jan 31 16:58:26 erm, refresh my memory, what is missing as compared to vanilla bme, re functionality? Jan 31 16:58:28 it'll work with bme too if you want Jan 31 16:58:34 no, I don't :P Jan 31 16:58:47 it's a replacement battery applet, it'll work with everything, really Jan 31 16:58:57 charging from usb port works if one limits current (or automagically, already?) Jan 31 16:59:08 I mean bme replace suite Jan 31 16:59:13 oic Jan 31 16:59:13 as whole Jan 31 16:59:28 does it shutdown device gently on user-defined voltage? Jan 31 16:59:34 not user-defined Jan 31 16:59:40 it's hardcoded at EDVF Jan 31 16:59:45 so when? and why isn't it configurable?:P Jan 31 16:59:48 which will probably be changed, because it's a bit too low Jan 31 16:59:55 erm, edvf was? 2800? Jan 31 16:59:56 accepting user configuration is a bitch Jan 31 16:59:59 or 3000? Jan 31 17:00:00 something like that, yes Jan 31 17:00:04 no, 2800 Jan 31 17:00:08 2800 is way too low Jan 31 17:00:23 that'll probably be changed in the future Jan 31 17:00:29 it's a hald-addon-bme thing afaik Jan 31 17:00:47 kerio, if you compiled applet with modified code, just go and modify it to accept user defined sysfs entry for shutdown, too :P Jan 31 17:00:56 mhm Jan 31 17:01:00 no, it's userspace Jan 31 17:01:28 if one ask me, easiest way would be to shutdown if device is continously at or below 3000mV for 15 seconds Jan 31 17:01:38 Estel_: why 15 seconds? Jan 31 17:01:49 ...which replaces all of bme funny "low voltage peaks measurement" Jan 31 17:01:53 just shut down at below 3000 Jan 31 17:01:58 nope Jan 31 17:02:04 would waste power on peaks Jan 31 17:02:11 ? Jan 31 17:02:16 even calibration uses 15 seconds on 3248 Jan 31 17:02:19 to avoid peaks Jan 31 17:02:22 hm Jan 31 17:02:38 you'd have to poll, though Jan 31 17:02:49 well, most of the time you would be 50-100 mA above real 3000mV, if you would shutdown on peak momentary 3000 Jan 31 17:02:55 hm Jan 31 17:03:09 so it would be "15 consecutive measurements below 3000" Jan 31 17:03:25 or 3, if it's measured every 5 seconds Jan 31 17:03:26 anyway, proposals are welcome, according to Pali Jan 31 17:03:36 i'm not sure how often hald-addon-bme updates Jan 31 17:03:47 anyway, test my applet pls :3 Jan 31 17:03:52 sure Jan 31 17:04:11 does it have chances o blowing up my system, or I can skip doing full backupmenu backup?:P Jan 31 17:04:59 it relies on hald-addon-bme to push charging/not charging info, so if you don't use bme or the replacement, you'll only get the "charging" notification when it polls for it, every 30s Jan 31 17:05:18 Estel_: i can almost guarantee that it'll not turn your n900 in a nuclear reactor Jan 31 17:05:22 but not quite Jan 31 17:05:51 (it should be "safe" - after installing, killall hildon-status-menu) Jan 31 17:06:07 ;) Jan 31 17:06:13 downloading Jan 31 17:06:17 btw Jan 31 17:06:29 so you got your compiling enviromnent set up, after all? Jan 31 17:06:44 i just launched the scratchbox VM Jan 31 17:07:46 I wonder if it works with virtualbox too? Jan 31 17:07:48 anyway Jan 31 17:07:52 i used virtualbox Jan 31 17:07:57 with the default ubuntu settings Jan 31 17:08:00 :) Jan 31 17:08:15 no manual steps needed to fix vm image anymore, to get rid of some bugs? Jan 31 17:08:58 thumb stuff is still in there Jan 31 17:09:00 i installed the virtualbox guest thing Jan 31 17:09:03 so you still have to clean that Jan 31 17:09:08 oh Jan 31 17:09:22 ...well, i sure hope Estel_ has a kernel with thumb support Jan 31 17:09:40 kerio: nah the gcc in the vm image is foobar Jan 31 17:09:44 regarding thumb Jan 31 17:09:55 so anything that ends up as thumb will be bad Jan 31 17:10:07 but I suppose your makefile does not check DEB_BUILD_OPTIONS for thumb? Jan 31 17:10:26 merlin1991, thanks, I vaguely remembered smth like that. No one released fixed vm image, yet? Jan 31 17:10:38 nope Jan 31 17:11:19 I would like to compile javispedro's unreleased latest version of fm radio for maemo, which allow analog passthrough without cpu... Jan 31 17:11:27 merlin1991: i don't think so, it's pali's makefile Jan 31 17:11:38 afaik it was "just compile" ready, no idea why he haven't released it... Jan 31 17:12:05 but I would like to do so with new compiler and all thumb godness. Even thinking about setting dev env. from scratch, for that, gives me headache Jan 31 17:12:15 (I'm coding/compiling noob, sadly) Jan 31 17:12:52 it's just scratchbox that's awful at being good Jan 31 17:14:25 huh? Jan 31 17:14:32 ~seen javispedro Jan 31 17:14:35 javispedro <~javier@Maemo/community/contributor/javispedro> was last seen on IRC in channel #maemo, 88d 21h 31m 46s ago, saying: 'get a macintosh!'. Jan 31 17:14:44 ough Jan 31 17:24:52 famous last words Jan 31 17:34:55 DocScrutinizer05, question to you as maemo irc specialist Jan 31 17:35:06 might be kind of mission impossible, yet... Jan 31 17:36:12 Anywhere from 2 months ago to year ago javispedro posted here - in conversation with me - link to source code of latest, unreleased version of his radio program - it was in garage, so is probably easily findable... Jan 31 17:36:37 but, in next line, he also posted link to patch, fixing some issue, patch posted "somewhere" - maybe in garage, or not Jan 31 17:36:41 now my question Jan 31 17:36:51 any ideas how to effectively search irc logs for it? :P Jan 31 17:36:59 grep Jan 31 17:37:04 having this situation description Jan 31 17:37:09 sure, but what to grep Jan 31 17:37:17 javispedro and http Jan 31 17:37:23 hm Jan 31 17:37:26 and fm maybe Jan 31 17:37:32 erm, where are logs not divided by days? Jan 31 17:37:43 searching every day separately = not fun Jan 31 17:37:47 rgrep Jan 31 17:37:55 grep -r Jan 31 17:37:56 Estel_: you'd have to use your own logs Jan 31 17:38:02 or ask someone with logs Jan 31 17:38:03 grep #maemo* Jan 31 17:38:10 the web interface is quite slow to search Jan 31 17:38:14 kerio, doesn't have from that moment Jan 31 17:38:46 cehteh, could you elaborate? I must admit I'm quite green here, what should I rgrep, as I only know weblogs of irc Jan 31 17:38:52 of this irc channel Jan 31 17:39:02 it could be at #maemo or #maemo-ssu Jan 31 17:39:04 you have personal logs? Jan 31 17:39:17 sadly, not from times in question Jan 31 17:39:26 http://mg.pov.lt/maemo-ssu-irclog/%23maemo-ssu.2012-08-03.log.html#t2012-08-03T00:58:55 Jan 31 17:39:27 ? Jan 31 17:39:30 ok then not Jan 31 17:40:08 arcean, god, how you've found it in ~30 seconds? Jan 31 17:40:12 I'm speechless Jan 31 17:40:28 google perhaps :P Jan 31 17:40:35 :P Jan 31 17:40:47 yeah, google :) Jan 31 17:41:56 well, I feel dumb :) thanks anyway Jan 31 17:42:11 in fact huge thanks Jan 31 17:48:17 kerio, see query Jan 31 17:51:27 merlin1991: where should i disable the thumbification? Jan 31 17:58:55 kerio /scratchbox/users/$somename/targets/FREMANTLE_ARMEL.enironment Jan 31 17:59:17 there should be an export line for DEB_BUILD_OPTIONS, remove the thumb part Jan 31 17:59:24 Estel_ et al.: http://db.tt/G9cHLpNy Jan 31 18:11:05 kerio is it cfm? Jan 31 18:11:16 yes Jan 31 18:11:16 with that thing defined? Jan 31 18:11:24 ~cyber-orgasm Jan 31 18:11:41 yeah, but i can't get the "analog bypass" to work Jan 31 18:12:09 oh, you always have to blow it like that? :P Jan 31 18:12:13 thanks anyway, will test it Jan 31 18:12:25 surely you defined adv audio routing on correct place? Jan 31 18:12:47 yes, yes Jan 31 18:12:54 i mean, it appears to reduce the cpu usage *slightly* Jan 31 18:12:55 ADV_AUDIO_ROUTING 1 should be responsible for it being real function, not mockup Jan 31 18:13:00 ah Jan 31 18:13:07 but not quite Jan 31 18:13:18 the question is if you will be able to blow up speakers, if yes, analog passthrough is on Jan 31 18:13:21 seriously though Jan 31 18:13:28 tune to same station on another analog radio Jan 31 18:13:35 and see if N900 sound lags behind Jan 31 18:13:40 if it lags, it's digital Jan 31 18:13:49 if it's sync'ed, it's analog passthrough Jan 31 18:14:33 when i click on "analog passthrough", it jumps ahead a bit Jan 31 18:15:02 installing, will test it in a few seconds Jan 31 18:15:09 Estel_: analog passthrough goes to the headphones anyway Jan 31 18:15:25 but it still seems at least partially controlled by pulseaudio Jan 31 18:15:49 it seems that you can define other route via options, speakers aren't connected to digital only? or are? Jan 31 18:16:12 couldn't be, javispedro was quite sure, that you can't even put filter there Jan 31 18:16:29 or i have no idea Jan 31 18:16:45 i assume that the analog passthrough will default to the headphones when they're connected Jan 31 18:16:49 aka every time, when using the radio Jan 31 18:17:52 I'm curious if it can be redefined during runtime Jan 31 18:18:08 javispedro mentioned smth about options for "defining audio route whatever you want" Jan 31 18:19:11 * Estel_ wonders what's the difference between /usr/bin/cfmradio and /usr/bin/cfmradio.launch Jan 31 18:19:18 both seems to be binary Jan 31 18:19:24 no sh wrapper Jan 31 18:19:36 Estel_: ls -l would've shown you :) Jan 31 18:19:41 it uses maemo-invoker Jan 31 18:19:47 which is fairly neat, not a lot of programs do that Jan 31 18:20:31 what are benefits of using maemo-launcher? Jan 31 18:20:34 invoker Jan 31 18:20:40 gtk aren't loaded again Jan 31 18:20:54 elaborate? Jan 31 18:21:07 the binary is basically loaded as a plugin by maemo-launcher Jan 31 18:22:08 neat Jan 31 18:22:23 btw what the hell "reset plugin to hf" means? Jan 31 18:22:33 i assume he meant HP as in Head Phones Jan 31 18:22:49 no, it's "reset audio to hf". Headphone? why reset to headphone? Jan 31 18:22:58 hp Jan 31 18:23:09 because you can switch the audio to the speaker Jan 31 18:23:20 or to the analog passthrough Jan 31 18:23:29 the default would be to switch back to the headphones Jan 31 18:23:40 yea, noticed that, every button have reverse one instead of toogle. no big deal Jan 31 18:24:06 no, it's just that there's more than one option for audio Jan 31 18:24:11 finally, no kinetic scrolling Jan 31 18:24:23 btw, i get no sound output if i click on analog passthrough twice Jan 31 18:24:30 lol Jan 31 18:24:32 i have to reset and reenable it Jan 31 18:24:40 it's clearly not perfect Jan 31 18:26:06 analog passthrough still works, though, sound is in perfect sync with my desktop fm radio hi-fi Jan 31 18:26:14 if analog passthrough is enabled Jan 31 18:26:17 yeah, yeah Jan 31 18:26:23 pulseaudio is still doing something though Jan 31 18:26:48 yes, but much less than before, in my case Jan 31 18:26:57 will measure power usage difference, later Jan 31 18:32:13 well, it definitelly could use some polishing - bug you've mentioned, + it tried to name stations by rds but get only part of it, resulting in funny names during auto-scan (I haven't found a way to change name of presets)... Jan 31 18:32:28 will dig if I can find configuration files for manually editting that Jan 31 18:32:54 still, lightweightness and analog passthrough as opposed to fm radio program, is amazing here Jan 31 18:33:35 + no idea who thought it's good idea to do kinetic scrolling of radio frequencies while updating them in real-time during said kinetic scroll... in FM Radio, not cfmradio, obviously Jan 31 18:35:05 after turning it off, error: Jan 31 18:35:06 Memory pool destroyed but not all memory blocks freed! 4368 remain Jan 31 18:35:18 does it mean what I think it mean? Jan 31 18:37:06 idk Jan 31 18:37:08 does it? Jan 31 18:37:23 measure the memory usage of maemo-launcher Jan 31 18:37:56 i can recompile it without maemo-launcher support i think Jan 31 18:40:25 well, every time count of left blocks is different Jan 31 18:40:36 if you just open and close it, no such message Jan 31 18:40:51 http://privatepaste.com/b6376f142f guess when I booted the FW ;-) Jan 31 18:41:04 after some usage, the loinger you use it and mess with more options, the higher number of blocks remain Jan 31 18:41:08 memory leak? :/ Jan 31 18:41:16 kerio^ Jan 31 18:42:44 Estel_: maybe, idk Jan 31 18:42:50 would it persist, is the real question Jan 31 18:43:10 hmmm Jan 31 18:43:20 Im trying to reinstall my N900, but the link to extras doesnt seem to work Jan 31 18:43:28 LinuxCode: /topic Jan 31 18:43:30 no idea why ;) Jan 31 18:43:35 ~repos-down Jan 31 18:43:40 ~repo-down Jan 31 18:43:40 ~mirrors Jan 31 18:43:40 rumour has it, repo-down is http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-20.log.html#t2013-01-20T02:08:42, or http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-30.log.html#t2013-01-30T00:51:04 Jan 31 18:43:41 from memory, mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 31 18:43:48 ~migration-fiasco Jan 31 18:43:53 Estel_: fu Jan 31 18:43:54 can't remember how it was called Jan 31 18:44:00 what? :D Jan 31 18:44:10 hmmm Jan 31 18:44:26 LinuxCode: you chose quite the fun time to reinstall your n900 Jan 31 18:44:32 but it's ok, because it means that you can also install cssu Jan 31 18:44:35 but it's still possible Jan 31 18:44:36 no kidding Jan 31 18:44:39 and quite easy to do Jan 31 18:44:54 just turn off regular repos and enable mirrors Jan 31 18:44:58 i recommend merlin1991's extras-devel-light and community-testing-light repos Jan 31 18:45:01 kerio, I resized my N900, but failed, with the rescue image it worked Jan 31 18:45:02 they're light Jan 31 18:45:09 MartinK_N9, kudos for modRana cobs repo approach Jan 31 18:45:38 youmre quite independent from migration fiasco problems (unless we're talking about dependencies, which may be in downed repo...) Jan 31 18:46:19 just defined modrana cobs in fam and it's amazingly easy and fun to use, instead of messing with repos. I just wonder how it would be handled if migration to cobs would be ever performed? Jan 31 18:46:27 separate repo entry for every package? Jan 31 18:46:30 cobs? wtf is cobs? Jan 31 18:46:31 eeeeeew Jan 31 18:46:42 kerio, MartinK_N9 created cobs for modrana Jan 31 18:46:45 even before fiasco Jan 31 18:46:51 yeah but wtf is cobs Jan 31 18:46:59 community obs Jan 31 18:47:29 aka painless approach to repositories, bulding (bye bye outdated autobuilder) and everything(tm) Jan 31 18:48:18 and thing that was meanto to be the way maemo repos will migrate, until our volounteer staff got lazy due to nokia donating too much hardware and paid worktime, top migrate depreciated autobuilder and co Jan 31 18:48:27 /rant Jan 31 18:48:44 which is why we're still in limbo, instead of using shiny cobs Jan 31 18:48:57 lol @ Merlin1973 Jan 31 18:49:36 kerio, where the fuck cfm radio keeps configs, like presets? Jan 31 18:49:40 can't seem to find it Jan 31 18:49:47 where the fuck am i supposed to know Jan 31 18:49:50 *how Jan 31 18:50:07 i suppose "where" is the source Jan 31 18:50:09 you just compiled the source code! :D Jan 31 18:50:39 now be prepared to be treated as deity and need to know answer to every question Jan 31 18:50:51 compiling is step into heaven's throne Jan 31 18:50:54 seriously though Jan 31 18:51:00 are they even saved Jan 31 18:51:30 yes, they are Jan 31 18:51:48 i gave you the tarball, you tell me Jan 31 18:51:57 :P Jan 31 18:52:14 will investigate it later, it isn't in any obvious locaion Jan 31 18:54:35 oh, gconf Jan 31 18:55:48 Estel_: /apps/maemo/cfmradio Jan 31 18:56:24 kerio, was repository.maemo.org up at all ? Jan 31 18:56:26 before Jan 31 18:56:34 before the migration? sure Jan 31 18:56:35 Estel_: stop telling BS and insulting volunteers Jan 31 18:56:41 or has it been down since the move Jan 31 18:56:42 omg Jan 31 18:56:44 k Jan 31 18:56:50 so its been down since the move Jan 31 18:57:00 ;-| Jan 31 18:58:34 Estel_: I'm tolerating zero such behaviour anymore of you Jan 31 18:58:47 it's not easy to keep a server up when akamai DDoSes it Jan 31 18:59:03 is that what is happening ? Jan 31 18:59:09 how come ?!? Jan 31 18:59:13 it was the original problem, yes Jan 31 18:59:19 because nokia used akamai Jan 31 18:59:33 and when the migration happened, akamai decided that it had to redownload the whole of rmo Jan 31 18:59:43 ohh man Jan 31 18:59:56 iptables cant fix it ? Jan 31 19:00:47 ~repo-down Jan 31 19:00:48 well, repo-down is http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-20.log.html#t2013-01-20T02:08:42, or http://mg.pov.lt/maemo-irclog/%23maemo.2013-01-30.log.html#t2013-01-30T00:51:04 Jan 31 19:02:42 the maemo packages are signed correct ? Jan 31 19:03:09 the signing of the repository isn't the problem, the problem is that the default package manager only wants packages from certain repos Jan 31 19:03:17 wants certain packages, that is Jan 31 19:03:22 no, Im asking because I have a suggestion to make Jan 31 19:03:39 how about using mirrors ? Jan 31 19:03:45 like we do in Fedora Jan 31 19:03:53 ~mirrors Jan 31 19:03:54 extra, extra, read all about it, mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/ http://talk.maemo.org/showthread.php?p=1315143#post1315143 Jan 31 19:04:25 kerio, Im talking about more than two ;-p Jan 31 19:04:34 there's three Jan 31 19:04:38 ok three Jan 31 19:04:52 we can barely afford one server, let alone one for each country Jan 31 19:05:02 Fedoras kind mirror providers have a total bw capacity of 234GB Jan 31 19:05:23 you missed the point I was making Jan 31 19:05:28 free to you Jan 31 19:05:43 anyway, the problem isn't the lack of mirrors, the problem is the lack of infrastructure, at the moment Jan 31 19:06:20 and the fact that the currently-active n900s will still hammer repository.maemo.org Jan 31 19:06:38 k Jan 31 19:06:41 for the extras repo Jan 31 19:06:52 eventually yeah, having mirrors would be good, especially if free :) Jan 31 19:06:53 just a suggestion Jan 31 19:07:00 LinuxCode: skeiron for example has a load of maybe 1% of available bandwidth Jan 31 19:07:24 so how would more mirrors help? Jan 31 19:10:26 DocScrutinizer05, if akamai keeps hammering Jan 31 19:10:35 anyway, I know nothing about all that so I just shut up Jan 31 19:10:40 was just an idea Jan 31 19:11:03 akamai servers are down, DNS stage.maemo.org is gone Jan 31 19:11:35 k Jan 31 19:11:39 tomorrow we start another effort to ramp up repo.m.o again Jan 31 19:11:45 kk Jan 31 19:13:43 btwm thank you for all your efforts Jan 31 19:13:46 -m Jan 31 19:13:52 it is really appreciated Jan 31 19:22:35 i wonder how much fun it would be to contact nokia care about the repos being down Jan 31 19:23:25 "your device is out of warranty. by the way, did you know that lumias are really great?" Jan 31 19:23:38 they already tried that with an N8 when i sent in my first n900 for repairs Jan 31 19:23:44 took me several months to get a n900 back:< Jan 31 19:25:32 Lava_Croft: I'm seriously encouraging you to do exactly that, call Nokia care and ask what's up with the repos, with tablets-dev.nokia.com(!) and generally with their support for a product you still can buy *new* with 2 years warranty in the EU Jan 31 19:25:45 LinuxCode: anyway, install CSSU Jan 31 19:25:51 (if you haven't already) Jan 31 19:25:59 I had it installed before yes Jan 31 19:26:03 I already mentioned the strangeness regarding the support for the n900 Jan 31 19:26:16 they already decided it was end of life in 2010 Jan 31 19:26:21 when i sent in mine for repairs Jan 31 19:26:37 nokia care monkey did not have an answer Jan 31 19:26:44 he did mention how symbian belle would be awesome Jan 31 19:26:46 kerio, installing CSSSU fresh is just adding the repo right ? Jan 31 19:26:53 LinuxCode: not quite Jan 31 19:26:56 I seem to recall there was a package Jan 31 19:26:58 LinuxCode: you click a link Jan 31 19:27:02 on a wiki page Jan 31 19:27:05 thats all Jan 31 19:27:05 which also did the kernel ? Jan 31 19:27:07 LinuxCode: which cssu flavour do you want? Jan 31 19:27:13 ehhh Jan 31 19:27:16 no, the kernel is still untouched by cssu Jan 31 19:27:20 LinuxCode: stable or testing? Jan 31 19:27:21 LinuxCode: http://wiki.maemo.org/Community_SSU Jan 31 19:27:24 ohh stable Jan 31 19:27:26 please Jan 31 19:27:30 click Stable image Jan 31 19:27:33 read page Jan 31 19:27:36 k Jan 31 19:27:37 or the other way around:) Jan 31 19:27:40 disclaimer: cssu stable is stable as in debian stable Jan 31 19:27:42 ta Lava_Croft kerio Jan 31 19:27:45 np Jan 31 19:27:47 LinuxCode: hold on though Jan 31 19:28:01 follow the instructions, at some point it'll ask you to do a system upgrade Jan 31 19:28:14 before doing that you'll have to enable a mirror for the "community" repo Jan 31 19:28:22 k Jan 31 19:28:43 oh, and unless you've enabled a mirror for "extras", you won't be able to install backupmenu Jan 31 19:28:46 find a way to install it Jan 31 19:28:47 it is good Jan 31 19:28:48 ~seen wirr Jan 31 19:28:53 wirr <~wirr@xdsl-188-155-55-161.adslplus.ch> was last seen on IRC in channel #maemo, 19h 15m 35s ago, saying: 'cool... thanks - night'. Jan 31 19:28:55 kerio, kk Jan 31 19:29:08 seesm the wiki covers most of what you just said, thank you Jan 31 19:29:09 full on-device backups aren't common, are they Jan 31 19:29:32 the only time i used backupmenu it completely failed to put the backup back in place Jan 31 19:29:49 thats probably fixed by now thouh Jan 31 19:29:50 Lava_Croft: when did you do that? Jan 31 19:29:56 quite a while ago:) Jan 31 19:30:00 yeah, the first version was absolute crap Jan 31 19:30:07 and then it was turned into gold Jan 31 19:30:09 it completely broke Jan 31 19:30:12 yeah, i know Jan 31 19:30:13 kinda like king midas Jan 31 19:30:18 but that negative experience will always remain Jan 31 19:30:20 Lava_Croft: BACKUP YOUR N900! >:C Jan 31 19:30:22 thats the power of negative experiences Jan 31 19:30:32 he should've changed the name maybe Jan 31 19:31:09 i think i actually removed the backup i had on the device Jan 31 19:31:10 not too long ago Jan 31 19:31:21 i mean, as long as i dont touch anything and repos are down, im safe Jan 31 19:31:25 :) Jan 31 19:45:57 DocScrutinizer05, "critics are users too". As said, always, things presented by me are my personal opinion and no one is forced to agree with me Jan 31 19:46:05 I'm doing it in polite way, without demanding anything Jan 31 19:46:28 so I will keep my right to think that some decisions made by volounteers may be bad too Jan 31 19:47:20 if you're certain that anything other than eternal gratefulness to hildon foundation staff isn't permitted on this channel, kick/ban me, and I'll have to live with giving this place a laugh *shrug* Jan 31 19:47:43 not to mentiin that I'm not bombing anyone about my opinions - just dropped a comment or two when someone asked about this certain thing Jan 31 19:47:50 so, gtfo of me, kindly. thanks. Jan 31 19:47:57 s/of/from/ Jan 31 19:47:57 Estel_ meant: so, gtfo from me, kindly. thanks. Jan 31 20:15:30 Estel_: I give a shit who does or does not agree with you and what you declare your statements. Stop telling BS and don't insult people. This is an official warning Jan 31 20:17:19 to make it utterly clear to you so there will be no missunderstandings: "volounteer staff got lazy due to nokia donating too much hardware and paid worktime" Jan 31 20:17:52 will not be tolerated Jan 31 20:18:01 no matter what you declare it to be Jan 31 20:18:23 even if your next lime is "just kidding" you earn a ban on next such statement Jan 31 20:18:50 was this clear and understandable enough for you? Jan 31 20:20:14 I'm willing to elaborate even further if you still don't see why such staement is a) based on outright false propositions, a whole bunch of them, and b) an insult Jan 31 20:21:22 or rather, no. On second thought I'm not, since i'm sure you already know but would troll me to explain it to you Jan 31 20:27:18 I even admit that I'd probably not even notice such nonsense when anybody else would post it. It's just for your extremely long record of purposely using this type of twisting facts into lies and then salting them with an insult. Jan 31 21:46:47 oh noez. http://tablets-dev.nokia.com/maemo-dev-env-downloads.php seems to be down. Does anyone have the flasher and an n900 image..? Jan 31 21:47:10 i do Jan 31 21:47:25 http://maemo.jacekowski.org/binary/ Jan 31 21:47:33 jacekowski: nice. Jan 31 21:47:36 http://maemo.jacekowski.org/binary/firmware/ - firmware is here Jan 31 21:48:27 jacekowski: do you happen to have the flasher executable lying around, too? I.e. not the deb package? I don't really mind unpacking that stuff, but if you happen to have that handy, it'd be appreciated. Jan 31 21:49:22 nope Jan 31 21:49:24 just .deb Jan 31 21:49:31 or windows .exe Jan 31 21:50:47 FWIW: https://muelli.cryptobitch.de/tmp/2013-01-31-flasher-3.5 Jan 31 21:51:27 gonna be fun to run your backdoored firmware ;-) Jan 31 21:53:17 first the firmware then the eMMC, right? Jan 31 21:53:26 no Jan 31 21:55:06 ~flashing Jan 31 21:55:07 from memory, maemo-flashing is http://wiki.maemo.org/Updating_the_tablet_firmware Jan 31 21:55:28 if the system is working, you can do vanilla and then combined Jan 31 22:01:21 wirr: ping Jan 31 22:31:26 kerio, Im getting invalid signatures, key expired Jan 31 22:31:43 LinuxCode: that's from the nokia repos Jan 31 22:31:47 for the maemo-archive.wedrop.it Jan 31 22:31:48 and yes, it's a known problem Jan 31 22:31:50 ...oh Jan 31 22:31:56 wait, what? no you don't Jan 31 22:32:13 you're getting key expired for the nokia keys Jan 31 22:32:59 errrm Jan 31 22:33:13 I thought the archive had a direct rsync ? Jan 31 22:33:20 *nokia* repos Jan 31 22:33:24 aka downloads.maemo.nokia.com Jan 31 22:33:30 as opposed to repository.maemo.org Jan 31 22:33:34 I disabled all those Jan 31 22:33:58 meh, you could've used the official nokia.com servers Jan 31 22:34:01 and replaced with maemo-archive.wedrop.it Jan 31 22:34:01 but anyway Jan 31 22:34:08 yes, the nokia key is expired Jan 31 22:34:11 k /me disabled the arcdhive ones Jan 31 22:34:17 so HAM won't upgrade a package from them Jan 31 22:34:21 sec Jan 31 22:34:28 which is not a big problem, because they won't release upgrades Jan 31 22:35:23 [minor sidenote] rmo up Jan 31 22:35:40 ahh excellent Jan 31 22:36:03 yyyay Jan 31 22:36:09 persistenty? Jan 31 22:36:39 DocScrutinizer05: sloooooooooooooow Jan 31 22:36:42 since a few minutes Jan 31 22:36:49 kerio: suuuuuure Jan 31 22:36:50 everyone's dosing it now I guess Jan 31 22:37:00 exactamente Jan 31 22:37:08 20Mbit Jan 31 22:37:24 i don't get it Jan 31 22:37:27 we've got 280 people here Jan 31 22:37:35 might decay during minutes to hours Jan 31 22:37:39 these aren't n900s, they're not constantly trying to access rmo Jan 31 22:37:59 the heck knows what it is Jan 31 22:38:08 what are the user agents? Jan 31 22:43:09 no useragents available, but mostly gets on Packages.gz Jan 31 22:43:50 hm... it's quite fast now Jan 31 22:43:52 crap. my microsd card from my 5230 isn't working in my n900. grr. Jan 31 22:43:57 1026 unique ips. Jan 31 22:43:58 great, and i talked Jan 31 22:44:04 DocScrutinizer05: holy balls Jan 31 22:44:54 DocScrutinizer05: for the next migration, we should make sure that the timestamp on Release is the same Jan 31 22:45:01 1026 over how long? Jan 31 22:45:09 If the N900 is starved for updates does it retry continuously or on the next interval? **** ENDING LOGGING AT Thu Jan 31 22:46:41 2013 **** BEGIN LOGGING AT Thu Jan 31 22:48:23 2013 Jan 31 22:48:51 ah yeah we should look into reducing the bloat in the packges.gz Jan 31 22:48:58 since it is redownloaded by ham Jan 31 22:49:01 a lot Jan 31 22:49:08 kerio: :nod: Jan 31 22:49:12 merlin1991: i don't think ham supports pdiffs Jan 31 22:49:14 does it? Jan 31 22:49:19 kerio: I'm going to check it Jan 31 22:49:41 but even if it doesn't support pdiffs, the repos can be brought in shape again Jan 31 22:49:57 some packages have a bazillion versions in the repo even though they should only have 3 at max Jan 31 22:50:13 i assume the only bottleneck is extras Jan 31 22:50:31 nah extras-devel Jan 31 22:50:35 because it's the only one enabled by default in pr1.3 Jan 31 22:50:36 extras is small compared Jan 31 22:50:43 extras is peanuts Jan 31 22:50:51 merlin1991: but devel isn't enabled by default Jan 31 22:51:10 when I run apt-get I get waiting for headers, then that seems to time out Jan 31 22:51:17 hehe Jan 31 22:51:22 guess it is overloaded Jan 31 22:51:42 kerio: extras 3.1m, devel is 8.8m Jan 31 22:51:51 when everything is settled we need to resync maemo.org/downloads with extras Jan 31 22:52:00 LinuxCode: and that's why it's been mostly down since the migration Jan 31 22:52:13 it's out of sync well before the migration Jan 31 22:52:59 * LinuxCode disabled extras-devel Jan 31 22:53:03 -d+s Jan 31 22:53:16 a load of packages could be removed due to being obsolete/broken Jan 31 22:54:33 merlin1991: extras-devel-light is like 7mb Jan 31 22:54:37 it's not a lot less Jan 31 22:56:27 down to 17MBit Jan 31 22:56:27 well the biggest part in the packages file is the package icon Jan 31 22:57:14 which seldom changes, hence is easily compressed Jan 31 22:58:02 uncompressed it's 15 mb vs 26 Mb Jan 31 22:58:33 merlin1991: btw, e-d-l feature request: strip useless fields from packages' control Jan 31 22:58:48 at least in Packages Jan 31 23:01:17 Stripping the icons from control and replacing them with url which cssu ham would lazyload would be the right thing Jan 31 23:01:33 It would break non-cssu ham icon display though Jan 31 23:03:36 1679 unique IPs Jan 31 23:03:58 17Mbit Jan 31 23:04:10 *.m.o stable it seems Jan 31 23:04:27 should, since we're bypassing FW now Jan 31 23:07:12 oh repo is up? or what? :o Jan 31 23:07:20 DocScrutinizer05: let's hope the situation normalizes over time Jan 31 23:07:26 or we're screwed Jan 31 23:10:06 I guess it will decrease over time, when the ones waiting since eternities all have downloaded the packages.gz and further requsts are statistically spread and most just checking for updates rather than actually doing full packages.gz download Jan 31 23:10:07 quite obviously something is trafficshaping/limiting our repo upstream at ~17Mb Jan 31 23:10:38 DocScrutinizer05, when I download packages it sits there for a while Jan 31 23:11:50 kerio: why we're screwed? each client seems to leach with ~10kb right now Jan 31 23:12:10 update takes a bit longer than usual ;-D Jan 31 23:12:13 now its dl Jan 31 23:12:21 ohh wait, no, Jan 31 23:13:41 now Jan 31 23:14:07 once connected it seemed fast Jan 31 23:14:20 what did you say is the estimated download size for one HAM doing update check? Jan 31 23:16:22 2124 IPs Jan 31 23:16:46 it works though Jan 31 23:16:50 Im happy Jan 31 23:16:56 so a big thank you Jan 31 23:17:22 all the thanks and presents go to warfare Jan 31 23:17:52 what about topic? Jan 31 23:18:08 our incredible new maemo sysop wizzard Jan 31 23:18:46 FIQ: well, I'll change the topic when we seen stuff working for a day... or two Jan 31 23:18:57 ok :P Jan 31 23:19:14 then the minute after change, everything goes down Jan 31 23:20:22 hmm Jan 31 23:20:45 Ign https://downloads.maemo.nokia.com ./ Release Jan 31 23:20:49 this host Jan 31 23:20:56 is it still online? Jan 31 23:21:02 DocScrutinizer05: are we to assume that extras is the main culprit? Jan 31 23:21:18 if so, ~3mb for each download Jan 31 23:21:25 FIQ: yes Jan 31 23:21:36 i thought they turned stuff off Jan 31 23:21:38 but ok Jan 31 23:22:24 kerio: dunno, what about -devel? Jan 31 23:22:49 DocScrutinizer05: well the question is more hits on which repo? Jan 31 23:23:33 the question is if 2 seconds of 17Mb will service one N900 Jan 31 23:24:05 since statisticaly we get one update every 2 seconds Jan 31 23:24:31 obviously they won't, even for extras Jan 31 23:24:31 one update is 7mb for extras, 8.8 for -devel Jan 31 23:24:40 feels like apt has froze on me Jan 31 23:24:47 guess it's just slow Jan 31 23:24:49 though Jan 31 23:26:41 I guess we should reject so many connects until the bandwidth starts to drop below max Jan 31 23:26:53 I'll kill r.m.o for a few seconds, I have to do a restart. Jan 31 23:26:59 (only of apache) Jan 31 23:27:06 would effectively put a random selection of N900 on wait Jan 31 23:27:15 like mine Jan 31 23:27:26 @ W: Failed to fetch http://repository.maemo.org/extras/dists/fremantle-1.3/free/binary-armel/Packages Unable to connect to repository.maemo.org http: Jan 31 23:27:27 :P Jan 31 23:27:59 * merlin1991 switched from xfce xdm4 to awesome today, feels funky Jan 31 23:28:05 warfare: <3 Jan 31 23:28:24 warfare: ohh that explains the above Jan 31 23:28:45 kerio: so far I didn't find any support for pdiff in ham :/ Jan 31 23:29:50 Back again and fast. Apache can only handle 256 clients. As most connections are slow, Apache slots fill very fast. I put an nginx proxy in front of the Apache (too lazy to migrate the config) which can handle around 8k connections with ease. This frees slots from apache and makes everything faster. Jan 31 23:30:05 :D Jan 31 23:30:26 hehe proxy black magic Jan 31 23:30:34 btw why does the repo even run on apache? o_O Jan 31 23:30:37 weirdly-coded russian programs with little documentation save the day! Jan 31 23:30:47 maybe Jan 31 23:30:51 * FIQ borrowed the server for 10s to update his stuff Jan 31 23:30:58 it's a basic serve files setup, not exactly the workload you need an extensible super configureable http server for Jan 31 23:31:09 merlin1991: There are some PHP scripts served. Jan 31 23:31:26 well, it seems to be quite fast now Jan 31 23:31:29 might just be me but it felt faster than before the migrate lol (took 10s to update) Jan 31 23:31:31 I guess lighttpd might be up to the task, but who cares Jan 31 23:31:50 warfare: well, there are other httpds which work with php I believe Jan 31 23:31:54 ~warfare Jan 31 23:32:01 * merlin1991 has no data about stability of lighttpd with high load Jan 31 23:32:05 infobot: warfare is DA MAN Jan 31 23:32:05 peak: 21.0Mb Jan 31 23:32:06 okay, kerio Jan 31 23:32:12 nginx Jan 31 23:32:14 now a download stared immediately Jan 31 23:32:17 looks good Jan 31 23:32:19 you can put PHP on nginx as well, using some fastcgi helper Jan 31 23:32:33 praise nginx! Jan 31 23:32:33 like the one stolen from lighttpd :P Jan 31 23:32:34 "too lazy to migrate the config" Jan 31 23:32:36 fucking apache Jan 31 23:32:43 valid reason Jan 31 23:32:44 warfare: oh right Jan 31 23:33:07 infobot: no, warfare is too lazy to migrate the config Jan 31 23:33:08 okay, kerio Jan 31 23:33:09 Also, I'm sitting in a british hotel and my internet is crap. All praise mobile-shell. Jan 31 23:33:37 no seriously, rmo is blazing fast now Jan 31 23:33:56 tadzik: did they realy steal the fastcgi stuff from lighttpd? :D Jan 31 23:34:30 praise the warfare! no, wait Jan 31 23:34:56 give him a hotel upgrade Jan 31 23:35:10 I hate Hotels that charge for crap wifi, or charge at all Jan 31 23:35:20 At least the don't charge for it. Jan 31 23:35:35 wow, that is rare in the UK Jan 31 23:35:41 something included Jan 31 23:35:42 ;-p Jan 31 23:35:55 what do you mean by mobile-shell? Jan 31 23:35:55 mosh Jan 31 23:35:56 kerio: mosh.mit.edu Jan 31 23:36:32 someone needs to build that for maemo. Jan 31 23:37:52 warfare: well, it's a commandline program Jan 31 23:38:09 you probably just need to build the debian tarball in scratchbox Jan 31 23:38:17 and it'll work Jan 31 23:38:25 kerio: actually ham uses libapt-pkg to download the index files, so I need to check there for pdiff support Jan 31 23:38:50 kerio: please stop abusing nick factoids Jan 31 23:38:50 ~forget warfare Jan 31 23:38:50 DocScrutinizer05: i forgot warfare Jan 31 23:38:50 ~lock warfare Jan 31 23:38:50 warfare: point me to the source an I'll build it for you ;) Jan 31 23:39:45 merlin1991: not really, but the general way to handle php and this sort of crap in nginx is to use the lighttpd tool of some sort Jan 31 23:40:56 [00:33:37] no seriously, rmo is blazing fast now Jan 31 23:41:05 spawn-fcgi, this is Jan 31 23:41:16 it took 10s to update all repos (devel activated) for me :P Jan 31 23:41:27 merlin1991: http://ftp.de.debian.org/debian/pool/main/m/mosh/ ;) Jan 31 23:41:35 instead of like 3min Jan 31 23:41:43 ooh mosh Jan 31 23:41:58 that thing is awesome, too bad there isn't a build for n900 Jan 31 23:42:35 merlin1991 offered to build it. Jan 31 23:42:46 merlin1991: :D Jan 31 23:42:59 yeah sometimes I have insane moments where I create way too much work for me Jan 31 23:44:34 bah scratchbox dpkg-source can't handle that Jan 31 23:45:02 merlin1991: Such insane moments happen to all of us. Jan 31 23:45:21 kerio: my HAM will take another ~60min to actualize the catalogs, I'd say Jan 31 23:46:05 well... Jan 31 23:46:07 ~ham Jan 31 23:46:27 kerio: well our libapt-pkg at least supports pdiffs Jan 31 23:46:41 kerio: I didn't check if ham calls it correctly to use it though Jan 31 23:48:10 warfare: it has quite a few build dependencies that are not yet in the maemo repos :/ Jan 31 23:48:40 merlin1991: Ok, thanks for looking. Jan 31 23:49:05 not saying it's impossible though, just wont be done today :) Jan 31 23:50:47 warfare: due to constant keep-alive packet pnging back and forth, mosh is a battery killer on N900 Jan 31 23:50:47 particularly on 3G Jan 31 23:50:49 and I seem to recall somebody trying to build mosh a year or so ago. Faced some problems, I guess Jan 31 23:50:49 well, ham is Hildon Application Manager. Hazardous Application Manager. Slow like molasses. anticipate you got time for a beer or two until it finishes whatever it does Jan 31 23:51:12 well it uses some weird protocoll to code converter as a build step Jan 31 23:51:22 and debhelper7 Jan 31 23:51:43 I found a .deb: http://blog.villekangas.com/?p=133 Jan 31 23:51:44 I see that someone made infobot reply to ham in HAM speed Jan 31 23:51:46 how fitting Jan 31 23:52:30 bah does not have the source packge to the deb Jan 31 23:52:31 evil Jan 31 23:52:53 FIQ: server hickup Jan 31 23:53:49 50:45] well... 50:47] not saying it Jan 31 23:54:08 DocScrutinizer05: mhm ok Jan 31 23:56:00 hm my scratchbox currently dls with 400B/s Jan 31 23:56:05 we're getting faster and faster, 23.7Mb Jan 31 23:56:22 DocScrutinizer05, it seems much smoother Jan 31 23:56:33 looks like you have more hits though, my dl from here sucsk Jan 31 23:56:54 and HAM fisnished: (no updates available) Jan 31 23:57:15 ~630 concurrent connections Jan 31 23:59:28 DocScrutinizer05, so ban me already, and save me from reading your twisted morality and bs statements. Deal? Feb 01 00:00:32 /ban #maemo *!*@* Feb 01 00:01:39 you can rest assured that no one will care (too much) and, in practice, you may do whatever you want here, including enforcing your totally childish views on freedom of speech as "what irritates me is bad". I won't give away freedom of expressing my opinions (politely, without calling names etcj, just because some funny guy try to shake his ban-hat at me. That's about it, "methinks" Feb 01 00:02:21 Estel, is this really necessary ? Feb 01 00:02:29 trolololololol Feb 01 00:02:38 doesn't sound very constructive to me Feb 01 00:02:52 FIQ: well, yeah Feb 01 00:03:19 hm 1400B/s, this is never going to finish Feb 01 00:03:54 merlin1991: the index download was fast, but the package download is very slow Feb 01 00:04:08 M4rtinK: it's an index dl Feb 01 00:04:11 yeh I estimate 10-20kb Feb 01 00:05:54 so, about constructive things - I think we all ove warfare a big beer Feb 01 00:06:08 actually, a barrel of beer or wine, up to him Feb 01 00:06:23 a truck of beer Feb 01 00:07:12 I've followed some of his reports and discussions (sparse, as he was quite busy) for last few days, and I'm astoned by his knowledge on how to shape this mess into usable organism Feb 01 00:07:17 a panamax tanker maybe ? Feb 01 00:07:42 do any of you guys recall the applet that changes the time to time and date ? Feb 01 00:07:51 the name of the package, that is Feb 01 00:07:52 the fact that it works given the circumstances is true miracle, and (almost) one man show. Much kudos to warfare! Feb 01 00:08:35 Estel_: you're again starting to tell BS, but astonishingly enough without insult Feb 01 00:08:36 LinuxCode, advanced clock replacement and some plugin for it? Feb 01 00:08:49 advanced clock sounds familair Feb 01 00:08:59 I will check for that Feb 01 00:09:03 thank you Feb 01 00:11:04 DocScrutinizer05, actually, I can mess you with mud complementing you, at the same time - just don't see it as necessary or feasible ;) Funny enough, the older I'm, the more straightforward in expressing things I become. And fearless of "consequences" ;) Feb 01 00:11:11 LinuxCode, no problem Feb 01 00:11:51 Estel_, are you a pensioner ? Feb 01 00:12:02 am i supposed to understand that weird sentence? Feb 01 00:12:04 pensioners do that.... they stop thinking first and just talk ;-p Feb 01 00:12:18 not always the best thing to do though Feb 01 00:12:44 DocScrutinizer05, in fact, and frankly, I haven't expected you to get it, at least yet ;) Feb 01 00:13:13 LinuxCode, whatever pensioner is ;) Feb 01 00:13:25 a retired person Feb 01 00:13:44 65+ Feb 01 00:13:53 Estel_: in fact I haven't expected any reasonable staement from you, ever Feb 01 00:14:13 seriously though, it's rather about seeing less and less need to obfuscate things into unnecessary wrappings Feb 01 00:15:09 LinuxCode, you haven't lived in Poland then - we already have 67 as min limit of going into retirement, and in few years, it will be extended further ;) Feb 01 00:15:34 I will never have a pension, or a retirement Feb 01 00:15:42 I'm quite certtain, that today's tweenties will work 'til death, no matter of official statements Feb 01 00:16:33 * LinuxCode stares at dl progress meter Feb 01 00:16:36 ;-} Feb 01 00:16:49 frankly, I would still use mirror Feb 01 00:17:07 for some reason it wouldnt work for me Feb 01 00:17:11 a) faster b) one less person to share download at official = others have it faster Feb 01 00:17:13 25.6Mb Feb 01 00:17:15 kept complaining about binutils Feb 01 00:17:20 it would, it's just a matter of method Feb 01 00:17:50 anyway, this works nicely Feb 01 00:17:55 mhm, merlin's repos via fapman seems ok without any hash errors (ignored), and binutils should be available via bb-p? Feb 01 00:17:57 I dont care if it is slow or not Feb 01 00:17:59 around 32kbit/s per connection. Feb 01 00:18:12 what generates so many IP? Feb 01 00:18:43 all non-TMO'ed devices, of absent users, just started to do auto-update check? Feb 01 00:18:50 RIPE Feb 01 00:18:57 (I mean hundreds of non-power user's devices) Feb 01 00:19:09 Estel_: even the power ones Feb 01 00:19:13 almost all user-agents are "Debian APT-HTTP/1.3 (0.7.20.2maemo13.1)" Feb 01 00:19:17 didnt they sell 800k of these N900 units ? Feb 01 00:19:28 or more Feb 01 00:19:58 funny when you think about it - one small switch, and thousand of devices all over the world start "waking up" Feb 01 00:20:11 (in terms of update) Feb 01 00:20:53 1375 unique ips. Feb 01 00:21:12 half get dropped, or DocScrutinizer05 was wong about last number? Feb 01 00:21:15 LinuxCode: X-Fade mentioned there are some 30k to 40k active, according to Akamai repo server farm statistics Feb 01 00:21:32 interesting, not that many then Feb 01 00:22:07 * LinuxCode will be cheeky now and enable extra-devel Feb 01 00:22:11 I definitely don't want to be around when 30k devices starts, at once to down anythingp, even smallest file, due to date differences Feb 01 00:23:25 so 30k devices and only 1k reading tmo... Feb 01 00:23:38 LinuxCode: in the long rin they shall spread equaly over the day, resulting in one update every 2 seconds. Since alarmd has no sync to time-of-day for the update period Feb 01 00:23:50 freemangordon, for inclusion in thumb repo, is ready-for-compile tarball (autobuilder standard) enough, or you need compiled binaries too? Feb 01 00:23:51 hundreds of new users each month btw Feb 01 00:24:15 DocScrutinizer05, yeh Im sure it will calm down Feb 01 00:26:30 reszing the 2GB partition made a huge difference Feb 01 00:26:44 30k devices is a pityful amount Feb 01 00:26:45 mine was chocker block, and the whole N900 slowed down to a crawl Feb 01 00:26:50 anyone have any idea how many were produced Feb 01 00:27:43 I think in the million+ Feb 01 00:29:13 so that leaves 960.000 devices at least Feb 01 00:29:28 considering HW faults, a lot of them probably died Feb 01 00:29:33 I seem to recall Nokia been taken by surprise by the preorders and sales during first 4 weeks, so the wirst batch of several 100k was way too small to satisfy the demand Feb 01 00:30:00 i got mine early june 2010, so the initial MUST HAVE IT rage was mostly over then Feb 01 00:30:04 s/wirst/first/. Feb 01 00:30:40 still, 30k devices that are actively used is not bad Feb 01 00:30:47 40k tops Feb 01 00:30:55 enough to hurt a lot of people if shit goes awry:< Feb 01 00:33:30 does apt download the catalogue each time it runs ? Feb 01 00:33:40 no Feb 01 00:33:42 it keeps a local cache Feb 01 00:33:44 k good Feb 01 00:33:52 usually it should just check for date of file Feb 01 00:34:01 k good Feb 01 00:34:44 which is why I think in a few day we will laugh about bandwidth problems with repo Feb 01 00:35:37 certainly wont have me downloading loads of stuff Feb 01 00:35:39 hehe Feb 01 00:37:22 I'm off. l8ters. Feb 01 00:37:30 nn ;-} Feb 01 00:37:55 I see downloads with 255kb Feb 01 00:38:01 am I right to assume, that the RSS feed thing , is still a battery life destroyer Feb 01 00:39:19 n8 warfare Feb 01 00:39:58 please do not say N8 Feb 01 00:40:16 i still have nightmares of trying to make nokia careless take back that shit n8 and give me back my n900 Feb 01 00:40:55 hehe Feb 01 00:42:55 Fr 1. Feb 01:26:58 CET 2013 real 0m0.949s Feb 01 00:42:57 Fr 1. Feb 01:31:59 CET 2013 real 0m1.098s Feb 01 00:42:58 Fr 1. Feb 01:37:01 CET 2013 real 0m0.925s Feb 01 00:42:59 Fr 1. Feb 01:42:01 CET 2013 real 0m2.766s Feb 01 00:43:01 :-D Feb 01 00:43:49 DocScrutinizer05, dl the extras-devel def takes time Feb 01 00:43:58 so that is probably the culprit Feb 01 00:44:13 (thats my *.m.o monitor using "wget http://wiki.maemo.org/Updating_the_tablet_firmware" Feb 01 00:44:15 ) Feb 01 00:54:07 2770 clients Feb 01 01:01:26 hmmm Feb 01 01:01:30 chmod: cannot access `/var/repository/staging/extras-devel/incoming-builder/fremantle/*': No such file or directory Feb 01 01:01:31 chmod: cannot access `/var/repository/staging/extras-devel/incoming-builder/diablo/*': No such file or directory Feb 01 01:01:33 chmod: cannot access `/var/repository/staging/extras-devel/incoming-nonfree/diablo/*': No such file or directory Feb 01 01:01:59 * DocScrutinizer05 suspects a missing NFS mount Feb 01 01:07:19 or not, since in the job throwing that error, a >>pid_skip /var/tmp/rsync_to_live.pid<< is showing up Feb 01 01:09:43 ~seen pali Feb 01 01:09:45 pali <~pali@Maemo/community/contributor/Pali> was last seen on IRC in channel #maemo-ssu, 2d 33m 54s ago, saying: 'merlin1991, freemangordon can you compile it for cssu-devel?'. Feb 01 01:09:55 ~seen Pali Feb 01 01:09:56 pali <~pali@Maemo/community/contributor/Pali> was last seen on IRC in channel #maemo-ssu, 2d 34m 6s ago, saying: 'merlin1991, freemangordon can you compile it for cssu-devel?'. Feb 01 01:22:13 hmmm trying to install maegios, there seem to be missing packages Feb 01 01:22:19 but it wont tell me which Feb 01 01:32:19 Go to the command line Feb 01 01:32:36 I was in apt Feb 01 01:33:12 What's it say? Feb 01 01:33:40 apt should say something useful Feb 01 01:34:03 Hash Sum mismatch Feb 01 01:34:50 Can you pastebin the entire thing, please? Feb 01 01:35:28 yeah, that's some thing to check, if the hashsum madness is still valid on the repos Feb 01 01:36:07 this only seems to happen when I enable extras-devel Feb 01 01:39:06 * LinuxCode disables it again Feb 01 01:39:34 might explain why I had issues on an alternate mirror too Feb 01 01:40:02 is there a way I can force this one ? Feb 01 01:57:11 LinuxCode, if you've got the .deb downloaded Feb 01 01:57:27 you can find it in /var/cache/apt/archives Feb 01 01:57:38 and install with dpkg and the --force=all option Feb 01 01:57:43 k Feb 01 01:57:46 or is it --force all Feb 01 01:57:49 I will try that in a minute Feb 01 02:03:50 good stuff Feb 01 02:03:56 package was fine, no need to force Feb 01 02:05:57 thank you very much robbiethe1st Feb 01 02:06:04 and everyone else Feb 01 02:06:13 off to bed, I am sure I be back tomorrow ;-} Feb 01 02:06:17 good night Feb 01 02:24:52 I've searched maemo.org but can't find anything relevant- is there any way to check which 3G band my N900 is using? (2.1GHz or 900MHz) Feb 01 02:27:41 Um, it's the one that T-mobile uses but AT&T doesn't Feb 01 02:28:44 ? Feb 01 02:29:39 Here in Australia carriers use 850MHz, 900MHz, and 2.1GHz for 3G Feb 01 02:31:41 hm, looking at the "featureset" of systemd, I'm expecting some embeded distribution that just ships systemd, busybox and 2 or 3 service deamons Feb 01 02:33:24 ping DocScrutinizer05 ? Feb 01 02:33:40 woooooody! Feb 01 02:33:44 :-D Feb 01 02:33:45 :) Feb 01 02:33:57 So... Just posted to the forums... Feb 01 02:33:58 long time no see! Feb 01 02:34:07 lists.maemo.org is well.. dead. :( Feb 01 02:34:19 probably related to turning on the repos? Feb 01 02:34:45 Yeah.. been a busy bee at work... (though I spent some time in the forums today while waiting for tests to run. ;) Feb 01 02:34:59 ugh, lists is dead? Feb 01 02:35:10 nah, should be unrelated to repo **** ENDING LOGGING AT Fri Feb 01 02:59:58 2013