**** BEGIN LOGGING AT Sun Nov 18 02:59:59 2012 Nov 18 03:14:57 someone have compiled pastebincl ? Nov 18 03:15:09 Sc0rpius: yes few days... Nov 18 03:15:31 Sc0rpius: why wired? Nov 18 03:15:57 im now im noob with many "o" Nov 18 03:15:57 because the N900 is an obsolete phone and has been obsolete since ages Nov 18 03:16:08 why would anyone buy a N900 today? Nov 18 03:16:17 * robbiethe1st would Nov 18 03:16:33 Sc0rpius: better than djaviks ones... Nov 18 03:17:18 android looks to poor for me than clean linux... Nov 18 03:18:02 it is poor Nov 18 03:18:29 the multitasking problem is a big problem for me in new phones Nov 18 03:18:30 for example Nov 18 03:18:37 I really dig those Nokia Lumias Nov 18 03:18:45 and Windows Phone COULD be a nice OS, but it isn't Nov 18 03:18:50 it doesn't even support background sockets Nov 18 03:19:00 that means, you connect SSH to somewhere, switch application, bye bye your SSH session Nov 18 03:19:02 and that's TERRIBLE Nov 18 03:19:56 lol better i dont say anything about new nokia phones... hrhr Nov 18 03:24:07 all time like nokia phones will be top that now think all new going to be the worst of one... Nov 18 03:37:36 Drathir: ogg-support - Ogg/Vorbis/Flac support for n900 Nov 18 03:38:16 decoders-support - Support for additional video and audio decoders in the Media Player Nov 18 03:38:24 dunno which if both Nov 18 03:56:24 of* Nov 18 04:07:17 DocScrutinizer05: now looking how with apt check detailed info about packages what dependencies contain... Nov 18 04:13:02 DocScrutinizer05: ok second include first pkg Nov 18 04:21:28 DocScrutinizer05: works perfect... thanks Nov 18 04:21:55 btw noce dropbox app is... Nov 18 04:26:05 now time to ipv6 configure... Nov 18 04:30:08 sigh Nov 18 04:30:14 mer/active is still fail for mainstream stuff Nov 18 04:30:21 :) Nov 18 04:30:31 i doubt i will ever see a productive tablet with active on it Nov 18 04:30:42 stuck with win8, ios, or ad-droid forever it seems Nov 18 04:30:50 i want maemo back :( Nov 18 04:31:44 I'm hoping for enough support in Plasma Active that we could at least get a good, flagship tablet with good support for it Nov 18 04:32:02 sounds like the tragedy of the commons Nov 18 04:32:16 I mean, so far I'm not seeing anything as good as the N900 by todays standards Nov 18 04:32:26 most people like to be rape by ads Nov 18 04:32:50 sounds elitist but is true Nov 18 04:32:50 It's rare to find a phone device with greater than 800x480 res, and even rarer to find one with a HW keyboard Nov 18 04:33:04 im right with you Nov 18 04:33:09 i miss my n900 Nov 18 04:33:20 i still use my n900 tethered heh Nov 18 04:33:44 but it's awful... i would have expected at least some unofficial hackery to have gotten mer/active going on a tf101 Nov 18 04:33:54 i saw a video a long time ago .. i would have thought it would be mainstream by now Nov 18 04:50:48 guess not Nov 18 04:50:49 heh Nov 18 04:50:50 ah well Nov 18 04:50:54 ad-droid it is :-/ Nov 18 05:04:48 loud MEH on plasma Nov 18 05:05:44 I like it because it's KDE, relatively stock Nov 18 05:05:58 which means I could likely set it up to run real KDE on it Nov 18 05:06:00 plasma is about the first thing I can think of why I'd like to nuke KDE4 Nov 18 05:06:04 and/or switch between interfaces Nov 18 06:01:35 jon_y: btw adds utorrent also have adds but in options can off Nov 18 06:50:32 Drathir: utorrent started getting annoying around 3.x Nov 18 06:50:51 qbittorent is the closest Free clone Nov 18 06:51:03 it's actually not too bad Nov 18 07:12:14 jon_y: you mean transmission >:( Nov 18 07:12:26 actually maybe not, idk if there's a windows port Nov 18 07:23:08 Anyone in the community offering pre-customized images of maemo? Could be a good way to get back to a certain clean configuration. Nov 18 07:25:39 kerio: it does Nov 18 07:25:54 UI is very different from utorrent though Nov 18 07:26:19 hence qbitorrent is a clone in this regard Nov 18 07:26:39 qbittorrent kicks ass Nov 18 07:27:50 no Windows binaries for transmission Nov 18 07:27:59 wiki says is buildable though Nov 18 07:28:23 I wouldn't poke it just yet since qbitorrent is running perfectly fine on my windows box Nov 18 07:42:28 internetishard: vi__ has been fighting with fiasco-gen since yesterday trying to do exactly that Nov 18 07:56:02 N900's memory is so small, 256MB... Nov 18 08:18:52 how do i check if i can write to a file in a shell script? Nov 18 08:19:47 check if it's writable? Nov 18 08:20:14 the only things that come to my mind are ugly hacks with perl -MFile::stat or awking ls -l Nov 18 08:20:21 tadzik: plenty of ways to make a 0777 file that can't be written to Nov 18 08:20:44 such as? Nov 18 08:21:29 IO errors Nov 18 08:21:30 ACL Nov 18 08:21:31 kerio: see test(1) Nov 18 08:21:54 or [(1) Nov 18 08:22:04 jon_y: i kinda want to just open it in write mode Nov 18 08:22:10 i can't do that with a shell script though Nov 18 08:22:26 in a shell script, you just use the test command Nov 18 08:23:03 if test -w $file_writable then... fi Nov 18 08:23:17 no need to open/close Nov 18 08:24:10 if [ -w $file ]; then... is also fine Nov 18 08:24:39 could someone check how much heap hildon-ststus-menu uses? Nov 18 08:25:06 meh Nov 18 08:25:34 kerio: ls -all Nov 18 08:29:43 jon_y: ye qbittorrent is nice but i use deluge best speed and console/web acces... Nov 18 09:16:32 Drathir: why not transmission? Nov 18 09:16:41 i still found it lighter than deluge Nov 18 10:03:04 has somebody used CLI send sms seccuss, i see this thread ,but use this python script can't send sms. Nov 18 10:03:13 talk.maemo.org/showthread.php?t=36148&page=1 Nov 18 11:12:43 kerio: vi__: good to know Nov 18 11:13:01 never heard of fiasco-gen though Nov 18 11:54:39 kerio: I've confirmed it is not a maemo issue. Gentoo + blueman/bluez has the same issues. Initial PIN works fine. Reconnecting is a pure pain. Nov 18 11:54:49 I think.. Nov 18 12:01:07 Actually, it works quite well... Nov 18 12:01:12 Interesting... Nov 18 12:01:19 (quite well on my laptop) Nov 18 12:09:48 Wizzup__: hm Nov 18 12:09:54 well, freemangordon is online, so you can ask him Nov 18 12:09:59 (he's the extkbd dude) Nov 18 12:13:34 Hopefully he'll see this, since you just hl'd him :) Nov 18 14:57:21 could someone without CSSU check and report what is memory usage of /usr/bin/osso-abook-home-applet Nov 18 14:57:40 s/is/is the/ Nov 18 14:57:42 freemangordon meant: could someone without CSSU check and report what is the memory usage of /usr/bin/osso-abook-home-applet Nov 18 15:04:12 560 532"635 3796 Nov 18 15:04:36 that's Nov 18 15:05:08 swap, unique pages, proportional size, RSS Nov 18 15:05:23 k, not pages Nov 18 15:05:27 'smem' Nov 18 15:06:28 SpeedEvil: may I have the output of Nov 18 15:06:30 grep -A 9 heap /proc/`pidof osso-abook-home-applet`/smaps Nov 18 15:06:37 SpeedEvil: how you check that? Nov 18 15:06:52 oh... i see... Nov 18 15:07:22 smem - it's a python thing that works fine on n900 Nov 18 15:07:34 RSS is a lie Nov 18 15:07:40 moment Nov 18 15:09:22 oh neat Nov 18 15:09:28 much the same output Nov 18 15:09:47 or not Nov 18 15:10:58 grep -A 9 heap /proc/`pidof osso-abook-home-applet`/smaps|tr -d '\n' Nov 18 15:12:59 size 660, RSS 380, PSS 362, 0/20/4/356/96/240 Nov 18 15:13:22 SpeedEvil: is your address book empty? Nov 18 15:13:26 yes Nov 18 15:13:29 mostly Nov 18 15:13:35 5ish items Nov 18 15:13:47 grep -A 9 heap /proc/`pidof osso-abook-home-applet`/smaps|tr '\n' '!' Nov 18 15:13:48 SpeedEvil: yeah, looks like Nov 18 15:13:53 ^^^ even better Nov 18 15:15:37 it deffinitely looks like address book copy is kept in memory for every process using libosso-abook :( Nov 18 15:16:39 with csu 624 is big vaule? Nov 18 15:20:32 kerio: heh.. http://forum.xda-developers.com/showthread.php?p=32908157 Nov 18 15:20:38 Let's just blame the Keyboard then Nov 18 15:20:59 Wizzup__: :( Nov 18 15:21:09 silly people supporting only one brand of device Nov 18 15:21:50 i wonder if there's a bluetooth confederation that you can contact Nov 18 15:22:01 because their keyboard is definetely not a HID BT keyboard Nov 18 15:22:32 well, again, it only works the first time. May be interesting to know how ipad ``reconnects'' Nov 18 15:23:55 Wizzup__: probably by not having a good standard in security, and accepting an unauthed, unencrypted keyboard Nov 18 15:23:59 or something Nov 18 15:25:40 kerio: last night I tried modifying some hid config file on n900 to set auth and enc to none, but that also did not help Nov 18 15:25:52 I may debug more on my laptop later, but rsi is killing me, so not now Nov 18 15:49:38 Wizzup__: if you've got rsi, why do you want to use a BT keyboard? :P Nov 18 15:52:33 kerio: what is wrong with his bt kbd? Nov 18 15:53:17 vi__: it doesn't work when you connect for the second time to it Nov 18 15:53:38 hid? Nov 18 15:54:13 yep Nov 18 15:54:26 but the manufacturer said it's only for iShit devices Nov 18 15:54:56 (which I knew when I bought it, but still) Nov 18 15:58:35 guy a GK308 Nov 18 15:59:35 *buy Nov 18 15:59:39 or a Matias Nov 18 16:01:04 [51644.412200] Forced hostmode error: a high-speed device attached but not high-speed mode selected Nov 18 16:01:21 goddamned hen bullshit. Nov 18 16:01:22 the iGo stowaway is really tiny, but the layout is a bit weird and there are no dedicated number keys Nov 18 16:01:29 vi__: what is this? Nov 18 16:01:38 also just do things manually Nov 18 16:01:52 kerio: how? Nov 18 16:02:03 I am doing it manually. Nov 18 16:02:14 vi__: some wireless keyboard have to nees oryginal reciver to work.. Nov 18 16:02:33 Drathir: *bluetooth* Nov 18 16:02:45 kerio: yes Nov 18 16:02:51 bluetooth is bluetooth Nov 18 16:03:42 have own adapter with special encryption that only works with one reciver adapter conneced to pc... Nov 18 16:04:15 that would not be bluetooth Nov 18 16:04:58 with another bt adapter keyboard ward is not see... Nov 18 16:04:59 vi__: echoing the correct things to the sysfs nodes? Nov 18 16:05:20 logitech if good remember have one that... Nov 18 16:05:58 I just wanted to mount a HDD is all. It never works right. Nov 18 16:06:53 vi__: HDDs are a bitch Nov 18 16:06:57 connect a powered hub before Nov 18 16:07:14 ...and Pali's battery measurer tells me that my battery is 101% charged Nov 18 16:07:32 for fuck's sake, what's his stupid obsession with rx51-battery Nov 18 16:07:36 i don't get it Nov 18 16:08:15 kerio: the hdd is self powered. Nov 18 16:08:25 vi__: \_o_/ Nov 18 16:08:28 just keep trying a bit Nov 18 16:08:32 enumeration is still a bitch Nov 18 16:08:48 enumeration doesnt work. Nov 18 16:08:50 ffs. Nov 18 16:09:32 ask DocScrutinizer05 for help Nov 18 16:09:46 if you want to be told "rmmod those weird modules and use h-e-n" Nov 18 16:10:51 kerio: I'll look into that one Nov 18 16:11:17 Wizzup__: anyway, i thought you said something about toying with hcitool to make the keyboard work? Nov 18 16:12:09 I am using hen Nov 18 16:12:16 finally it enumerated. Nov 18 16:13:27 vi__: I'm confused who might have introduced that oddly worded "error" Nov 18 16:13:48 a) it's not an error b) it's grammatically incorrect Nov 18 16:14:06 and yeah, that's definetely pali Nov 18 16:14:13 tbh 'sounds' like pali Nov 18 16:14:24 are you sure you're using h-e-n and not usbmode(.sh)? Nov 18 16:14:35 wtf are you taliking about? Nov 18 16:14:46 kerio: I suspect pali messing with hostmode patches to KP Nov 18 16:14:49 kerio: yes, I did something, but I cannot repeat/reproduce it Nov 18 16:14:55 I am using kernel-cssu3, thumb and HEN. Nov 18 16:15:41 I still want to try a laser-ir keyboard...hehe Nov 18 16:15:44 vi__: this 'error' msg is most likely a kernel message from musb_hdrc Nov 18 16:15:50 Wizzup__: no, you don't >:c Nov 18 16:15:54 Are they terrible Nov 18 16:15:55 ? Nov 18 16:16:04 I really like the idea Nov 18 16:17:07 actually I seem to recall sth like "added error messages to hostmode" in KP changelog Nov 18 16:17:08 how the shitters does my disc have a fat32 partition on it? Nov 18 16:17:53 which already puzzled me back when, since I'm sure we had sufficient diagnostic output in original hostmode kernel Nov 18 16:17:58 Isn't MyDocs fat, vi__ ? Nov 18 16:18:21 Wizzup__: i think he's talking about his usb hdd Nov 18 16:20:04 vi__: *maybe* you mounted a superfloppy (no partition table)? Nov 18 16:20:30 vi__: check mount Nov 18 16:20:47 Wizzup__: I am partitioning gthe HDD on my router. Nov 18 16:21:15 vi__: a new HDD? Nov 18 16:21:49 kerio: yeah Nov 18 16:21:57 there was probably a partition already there Nov 18 16:22:00 kill it with fire Nov 18 16:29:26 fat32 what a joke. Nov 18 16:29:42 why we are stuck with such legacy BS? Nov 18 16:33:30 vi__: because people keep using windows xp Nov 18 16:33:47 and neither os x nor windows support ext* or btrfs Nov 18 16:34:13 is btrfs even declared stable yet Nov 18 16:35:31 Wizzup__: nope Nov 18 16:36:02 but then again, i don't particularly trust fat32 either Nov 18 16:36:41 I expect everything from my mp3 player to my server to support reiser5 Nov 18 16:38:53 you know who's not going to support reiser5? Nov 18 16:39:21 nina reiser Nov 18 16:39:26 yes, i went there Nov 18 16:39:33 *sigh* Nov 18 16:39:39 Can be make mcafee jokes too now t hen Nov 18 16:39:44 * vi__ shakes his head. Nov 18 16:39:55 ( http://gizmodo.com/5959812/john-mcafee-wanted-for-murder ) Nov 18 16:41:20 Wizzup__: bonus points for linking to a neat story, malus points for linking to gizmodo Nov 18 16:43:54 ☺ Nov 18 17:04:32 Wizzup__: oooold Nov 18 17:05:36 I know ... reiser is even older :) Nov 18 17:35:29 hello does anyone know where to find kerne power Nov 18 17:38:08 in the repositories I bet Nov 18 17:38:49 kernel-power - Flashable Linux kernel (kernel power 2.6.28.10) Nov 18 17:39:06 sorry my IRC got hung up Nov 18 17:39:14 I meant kp version 52 Nov 18 17:39:37 well if that one doesn't provide kp52, I dunno what does Nov 18 17:39:44 I don't use kp currently Nov 18 17:39:51 it's in devel, right? Nov 18 17:39:57 probably Nov 18 17:40:31 Is 52 even the latest? and does it work with thumb/does it come thumb compiled? Nov 18 17:41:09 oof, forgot to switch on devel before updating ham Nov 18 17:41:30 ~ham Nov 18 17:41:30 methinks 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 Nov 18 17:42:08 yes, that. I just installed CSSU-thumb Nov 18 17:42:17 new Firefoy just crashes on me Nov 18 17:42:23 *Firefox Nov 18 17:42:30 cssu-thumb? Nov 18 17:42:36 yes Nov 18 17:42:54 what's that? Nov 18 17:42:59 never heard of Nov 18 17:43:02 bennypr0fane: KP52 is still a prerelease Nov 18 17:43:17 FIQ|n900: it's a secret Nov 18 17:43:39 a secret that google knows about Nov 18 17:44:30 kerio pre-release, as in unstable? Nov 18 17:44:48 bennypr0fane: prerelease as in "still subject to change and not released officially" Nov 18 17:45:24 FIQ|n900: http://talk.maemo.org/showthread.php?t=84829&highlight=thumb2 Nov 18 17:45:42 Anyone in here using CSSU-thumb? Nov 18 17:48:52 I upgraded Fennec to v17 after installing thumb, but starts adn crashes right away... Nov 18 17:52:23 bennypr0fane: worksforme Nov 18 17:55:47 maybe I did something wrong... Nov 18 17:56:06 run it from xterm and see if there are error messages Nov 18 17:56:28 the command is fennec or firefox? Nov 18 17:57:06 whatever expands to a valid command on pressing Nov 18 17:58:47 (though I could figure they missed to have a proper symlink in /usr/bin and rather start the binary on /opt via FQN in .desktop - dunno) Nov 18 17:59:27 (just like alarmed...) Nov 18 18:01:27 uhm, which is the tab key? Nov 18 18:02:05 in osso-xterm you got one in the hotkey bar usually Nov 18 18:02:22 ah, ok Nov 18 18:02:31 called "Tab" Nov 18 18:02:37 ctrl-i should work too usually Nov 18 18:02:47 :nod: Nov 18 18:03:15 I wonder if I shoulda removed the previous version of fennec first Nov 18 18:03:46 is fennec still insanely slow nowdays? Nov 18 18:03:57 it was last time when I used it... Nov 18 18:04:04 (which was 2011) Nov 18 18:04:43 how are newer thumb-versions of apps handled? I mean should it be replaced the same as if it were non-thumb? Nov 18 18:04:47 last i heard been "firefox way faster than fenec" Nov 18 18:05:13 uh Nov 18 18:05:26 what? Nov 18 18:05:27 fennec gives Segmentation fault. what does it mean? Nov 18 18:05:27 firefox as in firefox-desktop...? Nov 18 18:05:28 :s Nov 18 18:05:32 also a year ago or somesuch Nov 18 18:05:42 fennec from cssu-thumb is actually usable Nov 18 18:05:50 well I'm pretty sure progress has been made since then Nov 18 18:06:01 me too Nov 18 18:06:08 but is it *usable*? Nov 18 18:06:31 I don't like microB but there's hardly better choices... or at least there was Nov 18 18:06:39 until I xwithced to thumb just now, I used Fennec 7, which seemed not slower than MicroB Nov 18 18:06:48 so yes, usable to me Nov 18 18:06:57 ok interesting Nov 18 18:07:11 and latest Thumb so Nov 18 18:07:19 *** damn typos Nov 18 18:07:26 should be a bit more so Nov 18 18:07:37 i'll not use thumb before it's deemed stable Nov 18 18:07:51 Thumb is deemed stable Nov 18 18:07:52 (currently on cssu S4.1 Nov 18 18:07:54 ) Nov 18 18:08:11 I think otherwise they wouldn't pout that in the thread title Nov 18 18:08:33 please guys what is a segmantation fault and what do I do about it? Nov 18 18:08:37 "stable thumb2" Nov 18 18:08:54 bennypr0fane: "something blew up" Nov 18 18:08:55 i.e. stable thumb2 binaries Nov 18 18:09:17 that doesn't mean that the release itself is stable over not using thumb2 at all Nov 18 18:10:01 to be fair, it *should* be the same as cssu-testing Nov 18 18:10:04 i mean, it's the same source code Nov 18 18:10:07 also it seems to require cssu testing, another thing that from what I can see is against it beins stable Nov 18 18:10:10 just recompiled Nov 18 18:10:20 *being Nov 18 18:10:26 kerio: I use cssu-stable Nov 18 18:10:26 FIQ|n900: it seems there's not much known about problems with thumb system Nov 18 18:10:40 What does that mean "stable over x" the way I see it, there is a defintion of stable, and it either fits that defintion or not Nov 18 18:10:42 err not many known problems with thumb Nov 18 18:11:11 So even if it's indeed stable, it still requires me to upgrade to cssu unstable :) Nov 18 18:11:30 bennypr0fane: fennec confimed workingforme Nov 18 18:11:37 FIQ|n900: main issue with cssu-thumb right now is probably maintenance. It's even more 'bleeding edge' and way less tested than cssu-t Nov 18 18:11:58 yeah Nov 18 18:12:24 http://talk.maemo.org/showthread.php?t=84829&highlight=thumb2 I asked these question in tmo before making the switch Nov 18 18:12:41 maybe the answers help you Nov 18 18:13:38 I guess if there haven't been any problems reported so far, you should be fine Nov 18 18:13:54 I will report in that thread if anyhting comes up Nov 18 18:14:44 bennypr0fane: bleeding edge is no problem for me manually Nov 18 18:15:16 if I could reflash properly I would happily try it Nov 18 18:16:14 ...you cna't? Nov 18 18:16:21 dat usb Nov 18 18:16:24 s/manually/normally/ Nov 18 18:16:28 or, rather, dat lack of usb Nov 18 18:16:31 :p Nov 18 18:16:40 your usb is broken? Nov 18 18:16:48 yes, currently Nov 18 18:16:49 man, that sucks Nov 18 18:17:05 i've not found anyone willing to fix it for me :) Nov 18 18:17:14 try nokia care Nov 18 18:17:19 apparently that's now a thing Nov 18 18:17:53 I believe my friend with a similar problem did Nov 18 18:17:55 huh? Nov 18 18:18:14 well if your device is less than 2 years old... Nov 18 18:18:17 They destroyed the device after calling it unfixable Nov 18 18:18:24 bennypr0fane: they usually swap Nov 18 18:18:27 with a N8 Nov 18 18:18:28 bennypr0fane: it's 3yrs old soon Nov 18 18:18:30 or something stupid like that Nov 18 18:18:38 N8 Nov 18 18:18:45 can't be serious... Nov 18 18:18:46 well technically a little over 2 Nov 18 18:18:57 but I've already had a replacement Nov 18 18:19:20 (which, believe it or not, had the same problem) Nov 18 18:19:31 I believe Nov 18 18:19:49 I seem to have lucked out with my device Nov 18 18:20:03 never any hardware problems Nov 18 18:20:25 what the fucking fuck Nov 18 18:20:29 my N900 just shut down Nov 18 18:20:39 i was doing I/O Nov 18 18:20:45 http://youtu.be/26UA578yQ5g Nov 18 18:20:48 god fucking dammit not this again Nov 18 18:22:30 i'm just going to assume it's dsme being the usual dick Nov 18 18:27:36 32wd_to <- yep Nov 18 18:27:39 fucking dsme Nov 18 18:27:54 ShadowJK: dding a file from a mounted eMMC partition to the uSD caused this Nov 18 18:28:59 also wtf, why does cpumem-applet report 100% used cpu while top doesn't? Nov 18 18:41:49 ha, Fennec is working: "Remove /home/user/.mozilla/fennec directory and try again to run fennec. And make sure you don't run it as root" Nov 18 18:41:53 yay! Nov 18 18:41:59 and pretty fast, too Nov 18 18:42:11 lightning start, conpared to before Nov 18 18:42:27 bennypr0fane: thumb one? Nov 18 18:42:59 yes! Nov 18 18:43:13 only just made the switch Nov 18 18:43:29 no-thumb fennec 7 to thumb Nov 18 18:43:42 quite the difference i gotta say Nov 18 18:45:16 page loading not that much, but still Nov 18 18:45:26 mainly handling, responsiveness is better Nov 18 18:45:36 yeah, page loading times suck Nov 18 18:45:48 but it seems to be an upstream bug Nov 18 18:45:49 well not worse than MicroB Nov 18 18:45:56 whaat? Nov 18 18:46:04 do you have adblock? Nov 18 18:46:08 yes Nov 18 18:46:11 hehe Nov 18 18:46:13 remove it Nov 18 18:46:21 and tey microb again ;) Nov 18 18:46:25 that will make my life betteR? Nov 18 18:46:26 *try Nov 18 18:46:33 you won;t believe it Nov 18 18:46:36 yeah, adblock is *really* heavy Nov 18 18:46:39 hmmm Nov 18 18:46:55 who'd have thunk Nov 18 18:47:22 so MicroB as fast as Opera?Helium? Nov 18 18:47:40 if not faster Nov 18 18:47:47 hm, how am i supposed to make nemo boot with uboot? Nov 18 18:47:58 that should make me rather happy... Nov 18 18:48:07 kerio you use power kernel? Nov 18 18:48:15 bennypr0fane: yes Nov 18 18:48:36 ah fuck, i have to remove the battery for this :C Nov 18 18:49:03 kerio, check this: https://wiki.merproject.org/wiki/Nemo/Installing#Nokia_N900 Nov 18 18:49:45 adblock is brutal. Nov 18 18:50:11 yeah, it adds about 8010 seconds to page loading time Nov 18 18:50:16 freemangordon I aslo have fasterfox and speedDNS in MicroB. Any of that slow me down too? Nov 18 18:50:16 8-10 Nov 18 18:50:23 NFC Nov 18 18:50:25 Make a light hosts file block list and add the flashblock script to your user.css. Nov 18 18:50:40 bennypr0fane: probably yes. Nov 18 18:50:53 fasterfox is widely panned as being a POS Nov 18 18:51:10 think of it as the speedpatch of firefox. Nov 18 18:51:10 fasterfox *is* a POS Nov 18 18:51:56 what is a POS? Yes to whaT? Nov 18 18:52:13 Piece Of Shit Nov 18 18:52:45 will it suffice to deactivate Ad block or need to remove it? Nov 18 18:52:47 pile of shit. Nov 18 18:53:03 bennypr0fane: I think de-activate is ok. Nov 18 18:53:24 bennypr0fane: look into 'fanboy' host file block list. Nov 18 18:53:48 he offers a 'lite' version for lower power machines. Nov 18 18:53:54 vi___ as a replacement for ad block? Nov 18 18:54:39 what about speed DNS? Nov 18 18:54:56 bennypr0fane: yes as an ab replacement. Nov 18 18:54:56 do you have any active addon in MicroB? Nov 18 18:55:12 fuck the internet with no adblck. Nov 18 18:55:17 vi___: no need to replace me Nov 18 18:55:27 ab: hah. Nov 18 18:55:27 ab: oh you Nov 18 18:55:28 ab HAHA :-D Nov 18 18:55:36 a lurker awakens! Nov 18 18:55:53 bennypr0fane: noscript Nov 18 18:55:54 kerio you trying the latest nemo? Nov 18 18:56:18 yes, but in a weird way Nov 18 18:56:41 bennypr0fane: ghostery works Nov 18 18:56:55 vi____ what is a hosts file block list Nov 18 18:57:10 kerio weird how? Nov 18 18:57:22 bennypr0fane: look it uo. Nov 18 18:57:24 bennypr0fane: look it upp. Nov 18 18:57:25 bennypr0fane: i didn't follow the "correct" partition layout Nov 18 18:57:35 and i don't know how to specify where's the root device Nov 18 18:57:49 I think you can't Nov 18 18:57:54 BS Nov 18 18:57:58 you trying to run it off mmc? Nov 18 18:58:10 yes Nov 18 18:59:18 oh Nov 18 18:59:26 well that comlicates things Nov 18 18:59:46 apparently the only devices that the kernel sees at boot are mtdblock* Nov 18 18:59:51 guys, is there a way to load the desktop version of Facebook in FEnnec? Nov 18 19:00:56 oic Nov 18 19:00:59 vi___ the results I get are all Windows relatede Nov 18 19:01:15 for hosts file blocking Nov 18 19:01:28 are these as true for all browsers? Nov 18 19:02:35 bennypr0fane: yes Nov 18 19:05:00 I hate all sites that won't show me the desktop version if I want it Nov 18 19:06:57 FUCKING SYSTEMD NOT MAKING MY NEMO BOOT Nov 18 19:09:33 ok, fixed fstab :) Nov 18 19:13:03 (( i've not found anyone willing to fix it for me :) )) recently somebody had his N900 USB repaired at a Nokia Care point, and he actually got device back in fixed condition, paid 30GBP. It wasn't in warranty anymore, so Nokia didn't feel like keeping the N900 and shipping a crappy "replacement" N8 or whatever instead. Maybe a viable way to get yours fixed as well? Nov 18 19:14:56 hm, nemo is pretty Nov 18 19:14:56 :3 Nov 18 19:14:56 it reports my battery as empty, though Nov 18 19:17:11 kerio u find it rather usable? afaik there's still a lot missing... Nov 18 19:17:21 it's a tech demo, of course it's not usable Nov 18 19:17:22 :) Nov 18 19:20:30 and now to never use it again Nov 18 19:21:56 it is not so hard. Nov 18 19:22:15 Just give me debian+hildon+gsm on my fone. Nov 18 19:22:24 vi___: word Nov 18 19:23:26 kerio; I suspect the applet counts Wait as use Nov 18 19:25:14 or everything not "idle" as busy Nov 18 19:35:40 (( hm, nemo is pretty)) haute couture, while we all know we need an overall in daily live, not something sexy looking that doesn't even allow sitting down Nov 18 19:36:50 ^^^ this was a more generic comment about recent OS design paradigms found everywhere Nov 18 19:37:34 s/live/life/ Nov 18 19:37:55 DocScrutinizer05: but it's pretty! Nov 18 19:42:53 DocScrutinizer05: fun fact: systemd will keep trying and trying and trying if you have *something* that doesn't work in your fstab Nov 18 19:44:06 fsck systemd - yesterday I found no way to `stop akonadi-server` Nov 18 19:44:32 systemd is an undigestible POS in my book Nov 18 19:55:57 btw systemd already uses dbus, sth I thought was a joke until I've seen it mentioned in man systemd Nov 18 19:57:00 I'm rather sure there are still nasty bug sleeping in dbus. the more I appreciate a new process-#1 using dbus now Nov 18 20:02:23 oh, and of course I'd adore dbus running on all my small embedded systems like sheeevaplug, whatever, just to make systemd happy Nov 18 20:04:12 DocScrutinizer05: i just checked, there's no systemd on my sheevaplug :3 Nov 18 20:04:17 linux more and more turns from a general purpose OS to sth tailored to fit windows-alike desktops only. Thanks to Lennart Nov 18 20:08:00 MohammadAG: PING! Nov 18 20:08:52 get this Nov 18 20:09:13 a rootfs image, with backupmenu rolled in... Nov 18 20:09:30 ...PLUS an eMMC image include inside the rootfs image. Nov 18 20:10:08 That way you only need flash the rootfs image then wipe mmc with the opt tar Nov 18 20:13:32 vi___: are you still desperately trying to figure out how fiascogen is supposed to work? Nov 18 20:13:54 I know how it is SUPPOSED to work. Nov 18 20:14:24 It just turns out no one EVER saw fit to document how the layout file should look. Nov 18 20:16:22 :-\ Nov 18 20:18:16 vi___: maybe we could ask somebody (ex-)$Nokia to shed some light on it? Nov 18 20:21:00 vi___: you meanwhile think it's definitely related to layout file? Nov 18 20:22:50 vi___: also, iirc this fiasco-gen is just some 40k and available for X86. freemangordon might be able and willing to analyze it, maybe Nov 18 20:24:17 freemangordon: you tried to unpack the non-working fiasco.bin via flasher-3.5? Nov 18 20:24:59 s/freemangordon/vi___/ Nov 18 20:24:59 DocScrutinizer05 meant: vi___: you tried to unpack the non-working fiasco.bin via flasher-3.5? Nov 18 20:25:37 DocScrutinizer05: fiasco-gen is available for x86. Nov 18 20:25:50 It has no sanity checking on the layout files. Nov 18 20:25:53 that's what I wrote Nov 18 20:26:02 So you can feed in any old shit and it will use it. Nov 18 20:26:08 sure Nov 18 20:26:17 However if you try to flash it, it will fail. Nov 18 20:26:38 I have had varying degrees of success. Nov 18 20:26:53 just when you unpack such shitty fiasco.bin via flasher, you could compare the unpacked files to the original ones you used to build it Nov 18 20:27:03 The most successful I have had is using the layout file as it appears in the header of the pr1.3 emmc. Nov 18 20:27:21 DocScrutinizer05: I have tried that. Nov 18 20:27:32 Unpacking gets you your filesystem image. Nov 18 20:27:47 It is the same as I put it in. Nov 18 20:29:04 with the nokia mmc mydocs image+the 'example' layout (from the header), flasher will accept the fiasco image, attempt to flash it then choke at 96% Nov 18 20:29:34 If I alter the header, flasher will not accept the fiasco image. Nov 18 20:29:49 vi___: flasher have any verbose mode? Nov 18 20:30:05 If I alter the filesystem image, flasher will barf at ~20%. Nov 18 20:30:47 If I compare the header of my mmc image with that of pr1.3 header I can see they are different. Nov 18 20:31:02 There seems to be an extra 'label' in the nokia one. Nov 18 20:31:32 does anyone know what exactly libedataserver.so do? Nov 18 20:31:50 evolution? Nov 18 20:31:57 yep Nov 18 20:32:14 no idea, beyond that Nov 18 20:32:15 I have a feeling that contacts are not in DB Nov 18 20:32:32 well, not in berkeley DB Nov 18 20:32:50 but in a simple file Nov 18 20:32:51 http://paste.debian.net/210523/ Nov 18 20:33:01 ^ is the header from my fiasco image Nov 18 20:33:35 http://paste.debian.net/210524/ Nov 18 20:33:48 ^ is the header from pr1.3 emmc image Nov 18 20:36:01 you will see the nokia one contains a 'RX-51_20'... line at the start. Nov 18 20:37:18 It is NOT the version string you can specify in the layout file. Nov 18 20:38:00 This page details the fiasco image format: Nov 18 20:38:04 http://www.nopcode.org/0xFFFF/?p=docu&q=fiasco Nov 18 20:41:33 and now I am stumped Nov 18 20:41:49 vi___: yours is missing "3.mmc { " Nov 18 20:42:06 DocScrutinizer05: Yes. Nov 18 20:42:14 But htf do I get it there? Nov 18 20:42:40 If I add it to the layout file, the image will not be accepted by flasher (never mind even getting to fail to flash). Nov 18 20:43:08 make sure your {} match in layout Nov 18 20:44:40 So you are suggesting I add '3.mmc { blah blah blah } to my file? Nov 18 20:44:43 vi___: see http://wiki.maemo.org/User:Joerg_rw/tools Nov 18 20:44:48 I am pretty sure I have tried that. Nov 18 20:47:05 I do not understand what you mean doc. Nov 18 20:49:35 I think it's just about format of layout file, and possibly parameter passed via >>--mmc-partition=ARG Partition for the MMC raw image (in layout)<< (which might create the 'label' you're missing) Nov 18 20:50:30 sorry, no better ideas here Nov 18 20:51:33 you are hinting at something, can you be more specific? Nov 18 21:00:15 MohammadAG: ping Nov 18 21:03:00 vi___: the structure as found on http://wiki.maemo.org/User:Joerg_rw/tools seems "wellformed" (matching { } and a label "mmc" ). your excerpt as of http://paste.debian.net/210523/ is missing that. Either it has to get included into layout file, or fiasco-gen auto-creates it when given the right parameters Nov 18 21:03:20 vi___: that's about all hints I can give. I'm largely handwaving Nov 18 21:04:52 * fil has spent the day going through old forum threads and recreating the flaining attempts of others to sort out his crappy SIP quality ... to no great effect -- I just tried Skype though, and it works fine, which makes it even more annoying Nov 18 21:05:14 s/flaining/flailing/ Nov 18 21:08:31 anyone know how to disable echo cancelation in the n900 SIP client? (just a vague theory that I want to eliminate: the freeswitch and n900 echo cancelation stuff may be getting into a fight) Nov 18 21:09:22 there's no way for two EC to get into a fight Nov 18 21:10:17 fil: wireshark has a good analysis function for RTP (SIP) streams Nov 18 21:10:36 check your inbound jitter / packet-loss Nov 18 21:18:38 fil: Limit the codecs sip can use. Nov 18 21:18:47 Limit it to the lower quality ones Nov 18 21:18:48 gsm Nov 18 21:18:53 ilbc Nov 18 21:18:55 and g729 Nov 18 21:18:58 (i think) Nov 18 21:19:32 I found that 3 most relevant sources of problems with SIP audio quality are: WiFi / router / firewall-there, Codec used (depends on far end as much as your own client), and audio card itself (PA for maemo) Nov 18 21:19:56 hey, vi___ beat me on it Nov 18 21:20:01 fil: My sip performance is utter shit as well. Nov 18 21:20:04 Performance seems to be improved by disabling the other codecs. Nov 18 21:20:05 But it is too early to say. Nov 18 21:20:18 DocScrutinizer05: Some of us are not allowed to fiddle with work wifi access points... Nov 18 21:22:20 in our case it could be other way round: the low-bandwidth codecs could just hog the N900 CPU, while 'simple' codecs with high bandwidth might work better Nov 18 21:22:43 DocScrutinizer05: Perhaps, give me a couple of days before I make a comment. Nov 18 21:22:53 Fortunatley we have wireshark on the n900! Nov 18 21:23:10 indeed Nov 18 21:23:16 mad useful tool Nov 18 21:24:03 even does statistics on jitter etc of RTP streams, and knows to playback them Nov 18 21:24:41 well, I've tried it in all sorts of places over the years, but since I use SIP for my work phone, clients start getting tired of dire quality pretty quickly, so I burn money on GSM even when I shouldn't need to to avoid complaints (and thus do much less testing recently) Nov 18 21:24:59 if your inbound RTP UDP stream is jitterfree and no packet loss, the problem is inside N900 Nov 18 21:25:17 I'll check that now... Nov 18 21:25:23 otherwise try low bandwidth codecs and better WiFi AP Nov 18 21:26:43 also WiFi encryption might be a source for CPU hogging Nov 18 21:28:20 alas libsofiasip isn't twinklephone, so the debugging options are not that many Nov 18 21:29:58 you however *might* try and use N900 as a WiFi network access via USB, and run twinkle (or other SIP clients of your choice) on a tethered PC Nov 18 21:30:26 fil: You can configure the codecs for sip by editing /etc/stream-engine/gstcodecs.conf Nov 18 21:30:31 so you could rule out the WiFi for source of jitter and/or packet loss Nov 18 21:30:53 fil: you can disable a codec by putting a 'id=-1' in the control block. Nov 18 21:30:57 vi___: duh! thanks for that info! :-) Nov 18 21:32:17 DocScrutinizer05: I still know 1 or 2 tricks! Nov 18 21:33:01 vi___: yeah, thanks -- I'd worked that bit out in my travels -- I've tried playing a few of the codecs, but not GSM, which I'd guess it optimised for, so worth a try Nov 18 21:33:39 my guess: sequence of those codecs in that file determines which is preferred and which less prio Nov 18 21:33:56 (although my testing recently has been simply calling the Freeswitch IVR demo, which may not be a very good test for normal calls) Nov 18 21:34:26 fil: GSM isn't optimized. Since it's done inside modem usually Nov 18 21:34:47 ah, fair enough Nov 18 21:34:49 fil: iLBC usually considered best codec to be found Nov 18 21:35:04 that didn't seem to help :-/ Nov 18 21:36:02 PCMA/PCMU are pretty simply and thus high-bandwidth codecs Nov 18 21:36:12 simple* Nov 18 21:36:29 high bandwidth low CPU load Nov 18 21:36:30 the frustrating thing is that twinkle on my laptop on the same wifi routers has never had an issue whenever I fired it up to see if the problem was external to the n900 Nov 18 21:37:01 hmm, check which codec it negotiates Nov 18 21:37:16 good idea Nov 18 21:37:17 (shows up next to call duration) Nov 18 21:37:32 during call Nov 18 21:38:02 pleased to hear somebody is using twinke still Nov 18 21:38:45 you're no by any chance using the de_DE locale? Nov 18 21:39:31 hmm, prolly not, judging by your realname Nov 18 21:39:37 nah -- en_UK (despite sitting in Berlin at this moment :-) ) Nov 18 21:40:43 I massively augmented the de_DE translation to include lots of explanations beyond what Michel originally implemented Nov 18 21:42:07 a pity my tweaks to window dimensions and icons and hotkeys etc never made it 'upstream' Nov 18 21:43:01 a few other patches did, though Nov 18 21:48:57 how do I manually re-enumerate? Nov 18 21:50:44 WTF Nov 18 21:50:59 DocScrutinizer05: hub 1-0:1.0: connect-debounce failed, port 7 disabled Nov 18 21:51:08 ^ I just got 10 of those on my PC. Nov 18 21:51:21 when I connected an n900 on backupmenu. Nov 18 21:51:32 I thought fuckit and pulled the battery. Nov 18 21:51:50 The n900 stayed on as it was connected by usb cable. Nov 18 21:51:52 that georgious sound of floopy disk... Nov 18 21:52:00 It then enumerated fine? Nov 18 21:52:25 How the hell could the removal of the battery fix the above error. That is insane. Nov 18 21:52:38 hmm, it probably can't Nov 18 21:53:00 tbh it definitely can't Nov 18 21:53:28 DocScrutinizer05: yeah, shame it's not being developed any more :-/ -- amusingly it seems that I've broken audio on the laptop too now (although I don't think that's SIP related, as youtube audio is totally fragmented at present too -- oh well) Nov 18 21:53:39 though from your description, I conclude that you're not running bme Nov 18 21:53:50 anyway, it was negociating spx-wb Nov 18 21:54:08 so I got NFC what pali's bme-alternative does to USB Nov 18 21:55:00 fil: (broken audio) I bet on PA Nov 18 21:55:12 yup Nov 18 21:55:31 (not developed anymore) say "thank you, qt!" Nov 18 21:55:31 fil: Did you run wireshark while initiating a sip call? Nov 18 21:56:03 speex-wideband, hmm Nov 18 21:56:41 not yet, will do now that I've installed it (I really ought to undo all the tweaks I've tried over the day too ... before I forget what I did) Nov 18 21:56:41 I wouldn't bet a penny on any particular of the fremantle codecs being bugfree Nov 18 21:58:11 DocScrutinizer05: heresy! Nov 18 21:59:20 did I hear SIP issues? Nov 18 21:59:27 yeah Nov 18 21:59:37 poor audio quality Nov 18 21:59:45 inbound it seems Nov 18 21:59:53 ah, I've never solved that one on the n900 Nov 18 22:00:04 this being said, make sure you use the right codecs Nov 18 22:00:12 :-D Nov 18 22:00:20 teotwaki: !!! Nov 18 22:01:01 even better: make sure you don't use the wrong codecs ;-) Nov 18 22:01:57 When your client sends an INVITE, it tells the receiving end that there's a priority list in terms of codecs, PCMU, PCMA, speex, etc. Nov 18 22:02:05 not all codecs are created equal Nov 18 22:02:21 iirc for an inbound INVITE the far end should offer all available codecs in SDP Nov 18 22:02:38 sure Nov 18 22:02:39 ~SDP Nov 18 22:02:41 IRC is known to be a form of the Streaming Documentation Protocol. Nov 18 22:02:48 hey, it's the " doc's late" day Nov 18 22:03:20 session Description Protocol afaik Nov 18 22:03:43 Does the n900 support RTCP? Nov 18 22:03:52 afaik not Nov 18 22:03:57 hmm Nov 18 22:04:02 not sure though Nov 18 22:04:20 FWIW, I've been experimenting with an in-house protocol for high-bandwidth low-latency networks. Nov 18 22:04:57 Where basically, instead of sending a beat every x seconds (usually, every 20ms) which has the same size every time Nov 18 22:05:19 I calculate the network's MTU, and use that to plug in as much data as possible Nov 18 22:05:55 hmm, interesting approach Nov 18 22:05:56 It works pretty nicely, network load went down roughly 20%, while retaining the same number of calls Nov 18 22:06:30 I gather it slightly increases latency though Nov 18 22:06:36 probably Nov 18 22:06:55 but it only works for intra-server calls, anything that talks to an audiocode has to use public protocols Nov 18 22:07:09 sure Nov 18 22:08:17 BTW, has anyone used Boost's bzip2 implementation? Nov 18 22:08:40 or rather, their istream filter? Nov 18 22:13:26 libBoost? WAAAAAAH! Nov 18 22:13:36 hm? Nov 18 22:14:11 DocScrutinizer05: do not criticise boost Nov 18 22:14:12 twinkle used regex from libBoost, so eventually I tried to compile that monster Nov 18 22:14:48 compiling Qt or openoffice is a walk in the park compared to libBoost Nov 18 22:15:00 Boost is piss easy Nov 18 22:15:06 first of all, half of it doesn't need to be comiled Nov 18 22:15:14 it's piss fsckg HUGE Nov 18 22:15:34 the other half just requires jam. cd boost; ./bjam Nov 18 22:16:15 also, DocScrutinizer05, the world has helped us get package managers, you know, things like aptitude and yum Nov 18 22:16:17 Use them. Nov 18 22:16:43 back when I had a laptop with a 300MHz P-II and a HDD that for sure wouldn't suffice for Boost even if completely empty Nov 18 22:17:11 or recent gcc, for that matter Nov 18 22:18:48 well, I just thought it's a bit of overkill for a silly regex, and I definitely never managed to compile it back when Nov 18 22:19:16 prebuilt .so didn't help - wrong version Nov 18 22:22:54 sorry, lapto overheated Nov 18 22:23:21 hehe, like did mine when I tried to get libBoost built Nov 18 22:23:24 ;-P Nov 18 22:23:42 DocScrutinizer05: Boost is probably one of the greatest things in the C++ world. Nov 18 22:23:58 s/greatest/hugest/ Nov 18 22:24:23 most of it is template-only, which means there is no dependency when distributing it. It's peer reviewed, which means the quality and optimisation level is just crazy. Nov 18 22:24:52 yeah sure Nov 18 22:25:09 it just gave me one week of swearing Nov 18 22:25:39 not being able to build twinkle, due to a silly regex from a 5000GB lib Nov 18 22:32:16 -devel:~50MB, -doc-html:~120MB Nov 18 22:32:56 not much by today's metrics, a friggin lot back when Nov 18 22:50:50 I see what you mean about wireshark & SIP -- been a while since I played with wireshark (it was probably called ethereal at the time ;-) ) -- I don't think it had all this cleverness then Nov 18 22:54:33 it's not very pleased with the RTP n900-->Freeswitch. seems that there are several jumps in the packet sequences -- whereas Freeswitch-->n900 it seems pretty cheerful about Nov 18 22:56:35 that suggests that CPU on N900 is overloaded Nov 18 22:57:16 though N.B. that wireshark doesn't guarantee lossless capture Nov 18 22:58:26 ideally you'd capture the traffic somewhere later in the nework stack, e.g on a PC working as AP Nov 18 22:59:34 or (maybe even more simple) at the far end on inbound traffic Nov 18 23:01:15 you *might* try and renice -5 wireshark Nov 18 23:02:03 resp use tshark or pcap to capture, and not hog CPU with full wireshark UI in realtime Nov 18 23:02:31 I'm using tshark, and looking at it with wireshark on my PC Nov 18 23:02:31 DocScrutinizer05: ++ Nov 18 23:02:42 fil: 1 step ahead! Nov 18 23:03:21 yeah, didn't seem wise to clog everything up with X-over-ssh Nov 18 23:04:58 so was dumping to a file on /tmp, then scp-ing it over afterwards Nov 18 23:06:22 I'll try a few more tests (with tshark un-niced) to see if that makes any odds Nov 18 23:08:19 if it changes stuff to either the better or the worse, we're on something. If no effect at all, we probably can rule out CPU load as a culprit Nov 18 23:11:44 well, I just start to wonder if jumps in packet sequence might also be a result of clogging outbound WiFi, so eventually the RTP stream process detects there's a backlog on packets to send, at skips some packets to keep up with time pace Nov 18 23:12:05 erhm, no way. Nov 18 23:12:36 well, already good enough :-) Nov 18 23:12:38 thanks Nov 18 23:13:16 PCMU is 96Kbit/s tops, 64Kbit/s avg Nov 18 23:13:30 :nod: Nov 18 23:13:40 how's this related though? Nov 18 23:13:41 if that clogs your wifi, well, there's your VoIP quality problem. Nov 18 23:14:27 depending on WiFi situation, there might be low bandwidth, or collisions, or packet retransmissions, or whatever Nov 18 23:14:47 and if his wifi really can't handle 96kbit/s Nov 18 23:14:54 then why is he even trying voip calls? Nov 18 23:14:55 :P Nov 18 23:15:22 I don't think the RTP send process should or would fill up a send buffer with an unlimited number of packages pending to make their way thru WiFi Nov 18 23:16:00 well, one packet every 20ms Nov 18 23:16:17 os 500 packets per second Nov 18 23:16:41 yeah, but what if channel is occupied by some other burst transmission between AP and random client for 1.5s Nov 18 23:17:10 RTP sender should drop packets in this case Nov 18 23:17:45 since packet sequence # is entagled to timebase afaik Nov 18 23:20:38 right, with nice -5 we seem to get all packets, and there was one missing packet inbound, none outbound, but the playback quality is dire (this listening to the IVR menus, since it's past midnight here, and sleeping wife & child next door precludes making phones ring and muttering to myself ) Nov 18 23:26:00 so your audio quality issues are inbound or outbound? Nov 18 23:26:17 what's up with microB occasionally acting as if I run a desktop WM? Nov 18 23:26:22 @ its' dialogs Nov 18 23:27:25 i mean, sure, some stuff in extras/etc looks like it does on the desktop, but a preinstalled browser? Nov 18 23:30:15 generally people tell me that they can hear me OK -- the problem is on the inbound channel, which after (guessing) 30s (somethimes longer) starts breaking up, with the appearance of the sound being somewhat shuffled -- then after getting completely incomprehensible (and I'm getting quite good at understanding the garbled stuff) it goes dead for a while, and then iether gives up completely, or recovers for a bit Nov 18 23:33:52 as for wifi bandwidth, I can ssh from phone to laptop at > 500MB/s, and that's with both devices on the wifi (so colliding), and probably limited by the speed of the flash the file's coming off of Nov 18 23:34:04 s/ssh/scp/ Nov 18 23:34:05 fil meant: as for wifi bandwidth, I can scp from phone to laptop at > 500MB/s, and that's with both devices on the wifi (so colliding), and probably limited by the speed of the flash the file's coming off of Nov 18 23:35:17 the weird thing: I've seen exactly same behaviour with twinkle and 300MHz PC attached to internet via a WEP AP Nov 18 23:36:01 initially fine, after 15..30s gets worse, even blocks, but then usually recovered Nov 18 23:36:08 and stayed fine Nov 18 23:37:07 inbound sound Nov 18 23:38:05 so I tend to blame WiFi usually, when SIP audio fails in quality Nov 18 23:38:22 if it doesn't work at all, it's usually firewall/NAT Nov 18 23:38:58 you might test with USB networking to access internet Nov 18 23:39:39 yup, I can try that -- and I'll hard-wire the laptop while I'm at it Nov 18 23:41:05 also make sure your router/NAT has no ApplicationLevelGateway-SIP enabled Nov 18 23:41:24 they are known to be utter crap most of them Nov 18 23:42:42 Vigor for example didn't work at all with ALG-SIP enabled Nov 18 23:43:49 generally speaking, a NAT can introduce quite some jitter weirdness, depending on grunt of embedded MCU used in that NAT device Nov 18 23:45:17 if there's even a firewall enabled on same NAT, possibly with packet inspection, you might face problems regarding performance Nov 18 23:45:34 * RST38h moos evilly Nov 18 23:45:58 RST38h: you must do nightshift Nov 18 23:46:26 isn't it like 2 o clock in moscow? Nov 18 23:47:23 I am in the US Nov 18 23:47:30 aaah that's why Nov 18 23:47:59 Land of McDonaldses, etc Nov 18 23:49:46 Aren't there McDonalds everywhere these days? Nov 18 23:50:34 Yeah, but everywhere else has OTHER STUFF too Nov 18 23:53:20 Uh Nov 18 23:53:20 OK Nov 18 23:53:33 well, seems to me there's no other place with more 'other stuff' dealing with burgers, than USA Nov 18 23:53:48 * RST38h did not really mean burgers Nov 18 23:54:00 * RST38h never tasted a burger.Scary. Nov 18 23:54:24 well McDonals==burger Nov 18 23:54:38 so 'other stuff' also == burger, no? Nov 18 23:54:51 They also have apple pies, which are relatively safe to eat because they come there packaged. Nov 18 23:55:01 You've never had a burger? . . . Nov 18 23:55:11 GA: Yeah I know. Nov 18 23:55:21 That's slightly surreal to think about. Nov 18 23:55:27 I kinda can understand that Nov 18 23:55:55 GA: Well, to be fair I did taste all the individual components :) Nov 18 23:55:57 though I'd have a hard time thinking of somebody I'd know, execept vegetarians Nov 18 23:57:54 lol i run qmpdclient from server throught ssh... Nov 18 23:58:06 anyway here we got McDo and BurgerKing, and that's it basically re burgers. There's also an occasional KFC Nov 18 23:58:26 resolution looks great... Nov 18 23:58:28 in USA you got ~20 such frenchise burger firers Nov 18 23:58:43 fryers* Nov 18 23:58:46 Not really Nov 18 23:58:47 I mean Nov 18 23:58:51 There are some regional ones Nov 18 23:59:06 But for national chains it's basically Burger King and McDonalds Nov 18 23:59:10 DocScrutinizer05: royal dinner? Nov 18 23:59:32 GA: Do not forget Wendy's and KFC Nov 18 23:59:37 hmm, so that changed since mid 80s then Nov 18 23:59:45 wendy's yeah Nov 19 00:00:20 kfc > mcdonalds Nov 19 00:00:32 we even had a wendy's burger here for maybe 1 year, in the late eighties Nov 19 00:00:45 Ah, yeah, Wendy's. Nov 19 00:00:47 Forgot about them Nov 19 00:00:50 KFC isn't burger. Nov 19 00:01:06 I haven't had fast food in a while. Nov 19 00:01:23 KFC does burgers all right Nov 19 00:01:29 OMG I'm trying to get such a friggin burger since 48+ hours now Nov 19 00:01:30 * RST38h had to google though, never been to KFC Nov 19 00:01:34 Uh Nov 19 00:01:36 No they don't. Nov 19 00:01:39 They do fried chicken Nov 19 00:01:47 and there's a barbecue sandwich thing Nov 19 00:01:48 I'll probably end like that korean guy Nov 19 00:01:48 twister rulez... Nov 19 00:01:49 No burgers. Nov 19 00:01:59 not cheap but tasty.... Nov 19 00:02:02 Ah! Chicken burger! Nov 19 00:02:02 Just looks similarly Nov 19 00:02:56 I remember when I could eat Chick-fil-A. Nov 19 00:02:59 Those were the days. Nov 19 00:03:42 * DocScrutinizer05 glares at that frozen 300g steak Nov 19 00:04:11 * DocScrutinizer05 notices some icecrean next to it Nov 19 00:04:31 Doc: Grind, mix with paper, add glutamate and artificial coloring, etc Nov 19 00:05:15 xclock also works throught ssh... Nov 19 00:05:28 why shouldn't it? Nov 19 00:06:20 DocScrutinizer05: i never eat thats big high steak which only see3in films... Nov 19 00:07:20 * DocScrutinizer05 eats icecream, hopes for enough energy from it to head out and have a walk to next BurgerKing Nov 19 00:07:50 RST38h: i dont think thal nice lokks with ssh -X ... Nov 19 00:09:22 DocScrutinizer05: bet things with someone take 1l of icecreams switch on the tv and watching good film... Nov 19 00:09:30 best* Nov 19 00:10:16 only if you're not hungry Nov 19 00:11:31 I cook burgers from frozen all the time. Nov 19 00:11:44 works well, and gives time for the onions to brown Nov 19 00:13:12 burgers yes, 300g argentinian steak hardly Nov 19 00:14:16 bought this for BBQ, couldn't think of any sane way to prepare it indoors, with the tools available here Nov 19 00:15:09 Frozen burgers work really well. Nov 19 00:15:47 I was advised that frozen fish fillets wrapped in aluminum foil cook to perfection in a George Foreman grill. Nov 19 00:15:50 I need to investigate this. Nov 19 00:16:27 I should investigate preparing steak in the microwave Nov 19 00:16:55 could use my lighter to "BBQ" it when done with cooking Nov 19 00:17:18 Hehe Nov 19 00:17:27 I'm going to build a sous vide one of these days. Nov 19 00:18:36 GeneralAntilles: fresh fish put into allu foil with onion and other spices on grill with not much fire... Nov 19 00:19:09 Grill doesn't fit on my countertop and fresh fish doesn't last for months in the freezer. :P Nov 19 00:19:19 fresh - best fishing and catched by you... Nov 19 00:20:43 that make was really aromatic and soft... Nov 19 00:21:04 Yeah, don't have time to catch fresh fish. Nov 19 00:21:14 but always the best is potato from fire place... Nov 19 00:21:18 DocScrutinizer05: chop it up, which is easy part frozen, and stir fry Nov 19 00:22:05 a lot of fan making them... and also little dirty works looks on hands but worth that... Nov 19 00:22:42 Doc: Better construct a way to create plasma in themicrowave and prepare steak in plasma! Nov 19 00:23:26 SpeedEvil: is big difference betwen fresh prepared by you and from shop, sad but true... Nov 19 00:23:59 yeah. mines better. Nov 19 00:26:12 im little hungry now... Nov 19 00:41:59 ok, so how does one persuade the n900 that it's on the network (which it is) enough so that the dialer believes it and lets one dial via SIP Nov 19 00:49:03 ah, sussed the dummy stuff Nov 19 00:54:10 how depressing -- seems much better over USB, which makes it look like the whole problem is wifi. Meanwhile, wifi is th only way that the VoIP would be properly useful :-/ Nov 19 00:55:14 RST38h: good plan Nov 19 00:55:47 RST38h: found a can of fish in tomato though, I guess it will taste even without bread Nov 19 00:57:21 Doc: Do check that it is not cat food Nov 19 01:03:33 vi__: I heard you were doing something similar to my question earlier (which was) "Anyone in the community offering pre-customized images of maemo? Could be a good way to get back to a certain clean configuration." Nov 19 02:34:57 RST38h: ooh, you think there's a difference? **** ENDING LOGGING AT Mon Nov 19 02:59:59 2012