**** BEGIN LOGGING AT Wed Oct 16 02:59:58 2013 Oct 16 03:11:19 shouldn't Oct 16 03:13:23 hah, really a nice script, especially the header ;-) Oct 16 03:14:48 http://maemo.cloud-7.de/maemo5/usr/local/bin/contact ;-D Oct 16 08:25:10 A wonderful splendid Prickle-Prickle Morning (UGT) to you all. Oct 16 08:47:13 Apic: lol, sie hier? Oct 16 11:22:58 Pali: then what are you still doing there. Oct 16 11:52:39 EH? Oct 16 11:55:16 sono: this sounds insulting to me Oct 16 11:55:19 hmm, debian lxde chroot terminates Oct 16 11:55:22 DocScrutinizer05: nevermind Oct 16 11:55:57 sorry, I won't. I don't like to see pali insulted or offended Oct 16 11:56:22 DocScrutinizer05: wait until he is back and ask him then. it was about a dead channel Oct 16 11:56:56 there is no need to whiteknight Oct 16 11:58:27 DocScrutinizer05: are you still actively using an n900? Oct 16 11:58:34 in fact, is anyone? Oct 16 11:58:55 I am Oct 16 11:58:55 this question is kinda silly in this channel Oct 16 11:59:14 sono: Of course. Daily and exclusively. What kind of question is that... Oct 16 11:59:19 rather ask who's NOT using a N900 and nevertheless in this channel Oct 16 11:59:30 nice. i got one a week ago. first time owner Oct 16 11:59:39 it's a very likeable device, tho i am still struggling with some things Oct 16 11:59:59 but i'll have to check them out on my own, first Oct 16 12:00:09 anyway, i love it Oct 16 12:00:21 first thing to now fall for: fremantle is NOT debian Oct 16 12:00:34 i used oldschool phones for a long time before reluctantly getting an android 2 years ago Oct 16 12:00:44 i had already forgotten about the n900 at that point Oct 16 12:01:03 wish I had more of them Oct 16 12:01:03 s/now/not/ Oct 16 12:01:06 DocScrutinizer05 meant: first thing to not fall for: fremantle is NOT debian Oct 16 12:01:06 now i got an old one from an acquaintance in the UK Oct 16 12:01:15 no, that much is clear Oct 16 12:01:42 wish i knew why the easy-chroot lxde thing doesn't open a root window, tho.. Oct 16 12:01:49 the package management is quite different Oct 16 12:01:51 it worked the first time, after reflash, it doesn't Oct 16 12:02:05 I switched from an S40 feature phone (nokia 6000-series) to n900 roughly two years ago. I still feel sorry for those n900 Oct 16 12:02:07 i might want to wipe the SD card first. Oct 16 12:02:11 i'll figure it out. Oct 16 12:02:26 wnd: sorry? Oct 16 12:02:27 s/$/s I couldn't bring home from work and that were practically doomed to be trashed/ Oct 16 12:02:28 i went N800 -> N900 Oct 16 12:02:35 wnd: oooh. Oct 16 12:02:36 with some hobby pdas between Oct 16 12:02:59 I used to use 9300i before n900 Oct 16 12:03:08 i had a nokia c3 but never used it Oct 16 12:03:14 they just make battery spares for N900 Oct 16 12:03:31 hm, i need to replace the battery too.. it only keeps for a day or so Oct 16 12:03:53 ebay, i suppose... Oct 16 12:03:59 a day of real active usage seems normal for all "smartphones" Oct 16 12:04:15 well, it's idle most of the day, of course. Oct 16 12:04:36 in my pocket, automatically switched to GSM only Oct 16 12:04:36 nevertheless it might be a software problem Oct 16 12:04:40 so it sohuld last weeks, really Oct 16 12:04:48 lol Oct 16 12:04:48 and I still haven't got around to check how long the battery would last if I switched wlan off. I now get two full days at most. Oct 16 12:04:52 DocScrutinizer05: i didn't do a lot on it.. update, remove and install some packages Oct 16 12:04:53 some apps you better steer clear of Oct 16 12:04:56 DocScrutinizer05: powerkernel Oct 16 12:04:57 I want to see a smartphone that lasts weeks Oct 16 12:05:00 ~crappatch Oct 16 12:05:00 well, speedpatch is http://talk.maemo.org/showthread.php?p=1012405#post1012405 >>first i don't realy understand what does this patch do (that is why it is called miracle patch)<< [/quote original-author-of-speedpatch] Oct 16 12:05:28 I recharged some old tiny nokia a while ago and it idled for a good two weeks or so with an old battery Oct 16 12:05:31 joga: ahem, if i switch my samsung cheap-o android device to GSM only, it is good for a week with casual use Oct 16 12:05:52 sono, wow, never have I heard anyone claiming a week with any smartphone Oct 16 12:05:59 joga: it's the 3G modes and the display that eat power. Oct 16 12:06:00 joga: my N900 easily stays a week on standby Oct 16 12:06:05 joga: NOT active use Oct 16 12:06:11 when not logged in to GSM and WLAN Oct 16 12:06:18 joga: this n900 is dead after a day of casual use Oct 16 12:06:19 mine can probably idle for about two days or something Oct 16 12:06:41 joga: with 2G and wifi off? Oct 16 12:06:52 I actually never turn gsm off Oct 16 12:07:04 so yeah it might be better, and also useless to me ;) Oct 16 12:07:05 no, but HSPDA/UMTS Oct 16 12:07:20 anyway I've accepted that when I go home, I just plug it in charger Oct 16 12:07:21 heh Oct 16 12:07:47 should I try to watch a youtube video in metro or something the battery life would be closer to an hour than a day Oct 16 12:07:55 that's what I always did, even 15 (20?) years ago, with all mobile phones I had Oct 16 12:08:15 DocScrutinizer05: so with GSM actually off, i can get a week? Oct 16 12:08:22 DocScrutinizer05: did you ever run 2G only? Oct 16 12:08:25 continously Oct 16 12:08:32 sure Oct 16 12:08:34 how long? Oct 16 12:08:43 unclear, see above Oct 16 12:08:47 thanks =) Oct 16 12:08:58 ~power Oct 16 12:08:59 i heard power is http://wiki.maemo.org/N900_Hardware_Power_Consumption Oct 16 12:09:06 \o/ Oct 16 12:09:15 so what do you all think about jolla? Oct 16 12:09:46 just waiting for my pre-order to ship ;-) Oct 16 12:09:50 the company or the device? Oct 16 12:10:27 well, as a geek i primarly concern myself with the device, Oct 16 12:10:30 lol, that building is quite huge, that would never fit into my post office ;-) Oct 16 12:10:35 as a social being, i care about the company too =9 Oct 16 12:11:11 on a more serious note; I think the device and the company need to go hand in hand, so quite happy about all I hear so far. I care deeply about open devices Oct 16 12:11:18 about the device little is known yet, and that's already a downside of the device. and what's known is another downside for me, I don't expect to use mine as daily phone Oct 16 12:12:29 for me there are many of the FLOSS defenders from NOK there now, so I am hopeful (well and TBH what are my alternatives WRT to having a phone with an open bootloader and that has any chance in hell of booting an upstream kernel) Oct 16 12:12:50 it's pretty questionable how "open" jolla device will be Oct 16 12:13:01 indeed little is known, sailfish is also worked on behind closed doors Oct 16 12:13:07 DocScrutinizer05: that is my main concern Oct 16 12:13:12 the OS is open maybe, the device though... who knows Oct 16 12:13:12 we'll see later this year when the hardware hits the shelves if I am too much of an optimist or not Oct 16 12:15:02 anyway, thanks for keeping maemo alive everyone Oct 16 12:15:10 i'll dive deeper into it in the coming weeks Oct 16 12:15:16 Yeah, thanks from me too. Oct 16 12:15:23 Especially I find ModRana great! Oct 16 12:15:37 what does the hw hitting the shelves change in the status of the device regarding openness? without proper docs it's still not an open device, even when available and even when it maybe isn't tivoized Oct 16 12:16:01 DocScrutinizer05: he's going to find out when it hits the shelves, is all Oct 16 12:16:26 you find out that there are still no docs about the hw at all? Oct 16 12:16:35 I already can tell you about that Oct 16 12:16:58 open MCE is great :) Oct 16 12:17:29 said the developer of open MCE ;-P Oct 16 12:18:08 but yeah, awesome Oct 16 12:18:20 (though I had no chance to test it yet) Oct 16 12:18:47 I am running it on my own N900 with no failings yet Oct 16 12:18:55 cool Oct 16 12:19:37 jonwil: How did you manage to clone even the bugs like the one linked to in the tmo thread? :P Oct 16 12:19:53 jonwil: google-fu failure, link? Oct 16 12:19:59 j/k. I suppose they are in the meego source Oct 16 12:20:12 Its a 1:1 clone of the code, bugs included Oct 16 12:20:14 so finally we can fix the annoying indicator LED dimming, and the idiotic waste of a whole smart engine for sth silly as a 100ms ramp-up/down of kbd backlight Oct 16 12:20:38 DocScrutinizer05: someone will rip it apart, so at the latest when comercially available we'll know the used chips inside, that will already answer a lot of questions WRT to HW docs Oct 16 12:20:43 o.o i really have to install the SDK and look at the architecture soon Oct 16 12:20:46 * sono can't quite follow Oct 16 12:20:51 DocScrutinizer05: smart engine? Oct 16 12:21:23 pcfe: sure, but that doesn't qualify the device as open - at best as "ripped" or "rooted" Oct 16 12:21:43 sono: see LP5523 Oct 16 12:21:45 we prefer the term "reverse engineered" over here in corporate land Oct 16 12:21:47 ;) Oct 16 12:21:50 DocScrutinizer05: thanks Oct 16 12:21:51 has 3 mini-cpus Oct 16 12:22:01 one for each color of LED Oct 16 12:22:13 and a kernel driver for it Oct 16 12:22:26 also, http://wiki.maemo.org/N900_Hardware_LED Oct 16 12:22:33 but Nokia decided to assign one "color" aka engine exclusively to the kbd backlight Oct 16 12:22:48 DocScrutinizer05: ok Oct 16 12:23:21 yep, that link is fine, so I don't need to explain it again, since I did there already Oct 16 12:23:31 =) Oct 16 12:24:04 Aw. My Contacts in the App times out with the Throbber searching for them forever... probably Flash borked... in what Dir on FS are the Contact XML files stored? Oct 16 12:24:10 http://wiki.maemo.org/LED_patterns Oct 16 12:25:03 >>There's really no reason why you need to assign 1 of those 3 machines to this *occasional* short ramp function and break true rainbow 3 color effects on indicator LED. Sound rationale: while kbd bl ramp the CPU is awake anyway. No obvious benefit in offloading the task to a LP553 engine [except from saving 50 lines of code in mce]!)<< Oct 16 12:26:24 yea, i already wondered why they did this when i read it, and still have no answer Oct 16 12:26:39 that said, my soldering is OK... Oct 16 12:26:42 so... ;) Oct 16 12:28:55 soldering? Oct 16 12:29:19 i meant, my soldering skills are OK-ish so i don't see why i shouldn't rewire the channels.. ;p Oct 16 12:29:22 it was a joke Oct 16 12:29:22 ~/.osso does not seem to contain the Contacts Oct 16 12:29:30 Or is my whole Dir where they are stored vanished? Oct 16 12:29:48 I hope you don't think of "fixing2 this sw issue on a hw level - you can't Oct 16 12:29:58 s/2/"/ Oct 16 12:29:59 DocScrutinizer05 meant: I hope you don't think of "fixing" this sw issue on a hw level - you can't Oct 16 12:30:25 Apic: hmm? Oct 16 12:30:39 Apic: i have an ~/.osso-abook/ with "db" in it? Oct 16 12:30:45 maybe you're looking for that Oct 16 12:31:12 True Oct 16 12:31:12 DocScrutinizer05: as I said, I am hopeful. I might just be an incorrectable optimist. But many of the people that went to Jolla from NOK are drinking the FLOSS cool-aid, which for me is a good thing Oct 16 12:31:30 Is there a DB-fsck> Oct 16 12:31:32 s/>/?/ Oct 16 12:31:33 Apic meant: Is there a DB-fsck? Oct 16 12:31:38 pcfe: sure Oct 16 12:31:40 DocScrutinizer05: they way i understood it, they are using the third channel for the keyboard? i should read the whole page perhaps. Oct 16 12:31:44 DocScrutinizer05: nevermind then =) Oct 16 12:31:56 i'll get there Oct 16 12:32:03 sono: the lp5523 has a muxer Oct 16 12:32:18 you don't need to solder anything ;-) Oct 16 12:32:36 DocScrutinizer05: ah ok. i got to take the time when i am off work to read the wiki a bit, just so Oct 16 12:33:19 each of the 9 LED can get miuxed to any (or none) of the 3 engines Oct 16 12:34:06 and even when not muxed to any engine, each LED still has its own PWM hardware, incl register to control brightness via I"C from APE CPU Oct 16 12:34:47 I'll try overwriting .osso-abook with .osso-abook-backup Oct 16 12:34:48 it's just that MCE doesn't allow to assign any of the 3 indicator colors to the 3rd engine Oct 16 12:35:08 the LP5523 driver though does Oct 16 12:36:08 MCE hard-muxes all kbd BL LED to engine3 Oct 16 12:36:09 Still searches forever B-( Oct 16 12:36:29 I'll check whether I have a Backup Oct 16 12:36:41 Yes, from 2013.06.20 Oct 16 12:36:43 *restore* Oct 16 12:36:54 Apic: .osso-abook Oct 16 12:37:09 Yes, found it, but restored the .osso-abook-backup and still fails Oct 16 12:37:13 Is there a .db-fsck> Oct 16 12:37:14 s/>/?/ Oct 16 12:37:17 Apic meant: Is there a .db-fsck? Oct 16 12:37:47 o.O Oct 16 12:37:54 no Oct 16 12:37:55 I'll try Backup Restore, but it says "Communication and Calendar", not "Contacts" Oct 16 12:38:01 Does it contain Contacts too? Oct 16 12:38:02 I hope so. Oct 16 12:38:03 *restore* Oct 16 12:38:12 yes, sure it does Oct 16 12:38:17 Good Oct 16 12:39:27 I think addressbook.db is a berkley db Oct 16 12:40:27 Btw, is there any compelling reason for using different db backends for contacts, calendar, events, ...? Oct 16 12:41:31 good question Oct 16 12:41:53 possibly because often the backends are FOSS Oct 16 12:42:07 and already used a certain architecture Oct 16 12:42:39 Okay restoring Backup worked \o/ Oct 16 12:45:33 \o/ Oct 16 12:46:01 tbh I got NFC about what is osso-abook-backup Oct 16 12:47:17 possibly it's just a quick copy/snapshot done by osso-backup, to compress and backup a consistent state of db Oct 16 12:47:32 or maybe not Oct 16 12:48:07 maybe the contacts app using it for implementing a sort of transactions? Oct 16 12:48:25 really no idea Oct 16 12:52:31 jonwil: shouldn't mce get a post on FPTF thread as well? Oct 16 12:52:55 DocScrutinizer05: Do You get payed by Nokia btw? Or are You completely Fan-Guy? Oct 16 12:53:16 yeah maybe it should Oct 16 12:53:25 well, I'm the moron wo got appointed to volunteer for the community Oct 16 12:53:33 ;-P Oct 16 12:53:40 http://wiki.maemo.org/Community_Council Oct 16 12:54:04 I never got paid by Nokia Oct 16 12:54:11 DocScrutinizer05, that's why we love you :P Oct 16 12:54:33 oooh, and I hoped you love my for my character Oct 16 12:54:56 Fat Respect! Oct 16 12:55:35 hi moh, old arab! Oct 16 12:56:01 20's the new old? :/ Oct 16 12:56:03 hi DocScrutinizer05 :) Oct 16 12:56:51 well, for terrorists, 20 is quite old ;-P Oct 16 12:57:32 how's studying? Oct 16 12:59:54 you already got hired to write stuxnet-II ? Oct 16 13:00:33 DocScrutinizer05, painful here in Israel Oct 16 13:00:42 though in my free time, I did learn Java Oct 16 13:00:53 * MohammadAG hands DocScrutinizer05 a pistol Oct 16 13:01:48 * DocScrutinizer05 reads "Java" and shoots himself in the forehead Oct 16 13:02:29 Is there any info on what 'Update automatically' in the 'date/time' applet means exactly? Update from GSM network, I suppose? Oct 16 13:02:44 yes Oct 16 13:02:52 exactly Oct 16 13:03:04 Ok. And who is repsonsible for doing that? Oct 16 13:03:12 'Cause it's just a gconf key Oct 16 13:03:14 nowadays 99% of GSM networks support this Oct 16 13:03:59 kolp: (who's doing) there's a weird closed source time daemon, and I guess csd is also involved Oct 16 13:04:43 Ok Oct 16 13:05:05 * kolp is writing an open clone of the date/time applet, in case anyone is wondering Oct 16 13:05:35 what really really sucks is welcome daemon which asks for time/date/language while in backgrounf time and date get set by the concurrently running black magic for GSM time-adjusting Oct 16 13:06:37 so I usually boot two times in sequence after swapping battery Oct 16 13:07:03 on second boot the tinme and date got adjusted and correct from 1st boot Oct 16 13:07:11 My experience was that networks didn't support this so I have it switched off and use ntp instead Oct 16 13:07:37 depends on your carrier, you might be one of the 1% where it fails Oct 16 13:08:00 5 years ago things been worse Oct 16 13:08:32 I think I have it switched off since I got my first n900, in 2010 Oct 16 13:08:33 also roaming can introduce quite a number of really exciting effects with this automatic Oct 16 13:09:22 so if you're frequently roaming, you probably rather want to switch automatic timeadjust off Oct 16 13:09:36 since it also automatically sets TZ, etc Oct 16 13:09:52 and you have no control *when* it does that Oct 16 13:10:33 and I seem to recall that sometimes it doesn't set TZ and and clock same time Oct 16 13:10:55 rather with a few hours delay between the two Oct 16 13:11:17 Nice Oct 16 13:11:30 so first you get TZ A but with time of TZ B, and later on the TZ adjusts. Or the other way round Oct 16 13:11:39 but maybe I'm wrong on this Oct 16 13:11:52 might have been another platform Oct 16 13:13:52 this stuff is always much fun to debug, since relatively few roaming across TZ happens, and you hardly ever can do it in a controlled environment Oct 16 13:14:43 ...unless you got a CMU200 or similar, in your lab Oct 16 13:15:30 or the BB5 firmware source so you could flash a testbed frirmware that fakes such stuff from modem Oct 16 13:16:55 even with a CMU200 it helps a lot when you're e.g. Nokia so a 2 dozen carriers worldwide provide detailed data about their timestamp feature and how it's implemented, to you Oct 16 13:17:54 MohammadAG: when will you be free again? We coukd use a few brilliant heads in FPTF Oct 16 13:18:53 MohammadAG: and are you planning for a better place than this violence and idiocy infested IL? Oct 16 13:28:36 DocScrutinizer05: how is gsm time supposed to work again? Oct 16 13:29:06 kerio: hm? Oct 16 13:29:25 please ask more specifically Oct 16 13:29:36 what polls for it? Oct 16 13:29:38 is it polled? Oct 16 13:29:55 is there a way to figure out if it's working for you? Oct 16 13:30:03 kerio: it doesn't work for me xD Oct 16 13:30:17 I guess the modem either gets queried or gets asked to send an unsolicited message whenever a timestamp comes in Oct 16 13:31:05 I'm sure you could monitor sth to find out. wireshark ISI dissectors should work for sure. Dunno if dbus would show sth Oct 16 13:31:46 afaik there's no AT command to query the network if and how timestamps are supported Oct 16 13:31:59 so you just wait if it works or not Oct 16 13:32:34 I never looked into it indepth Oct 16 13:33:01 i'd rather just ntp Oct 16 13:33:08 just know the trouble we had with it in fsogsmd and relatives Oct 16 13:33:10 or mayne gps Oct 16 13:33:44 well, GPS just fails when you don't have a semi decent idea of your location and local time Oct 16 13:33:53 eeeeh Oct 16 13:33:58 it works Oct 16 13:34:00 eventually Oct 16 13:34:09 yeah, *eventually* Oct 16 13:34:19 fyi, the GSM time stuff is handled by libcsnet (aka libcsd-net.so plugin for csd) Oct 16 13:34:35 Is the gms network supposed to send timestamps on a regualr basis, if supported? Oct 16 13:34:44 jonwil to the rescue :-) Oct 16 13:34:46 It talks to the clockd daemon via dbus Oct 16 13:34:48 s/gms/gsm/ Oct 16 13:34:49 kolp meant: Is the gsm network supposed to send timestamps on a regualr basis, if supported? Oct 16 13:34:50 and the dbus call is documented :) Oct 16 13:34:56 kolp: yes Oct 16 13:35:14 aloso my network sends correct time signals Oct 16 13:35:29 hail dbus Oct 16 13:35:32 What#s the interval, approx? Oct 16 13:35:35 Any idea? Oct 16 13:35:41 when I went on holidays to different timezone (same cell network) and turned on N900 after getting off airplane, it picked up new time with no issues Oct 16 13:35:42 kolp: varies vastly Oct 16 13:35:46 new time and time zone Oct 16 13:35:58 between a few seconds and several minutes afaik Oct 16 13:36:36 Ah, ok Oct 16 13:36:47 Then there is still hope... Oct 16 13:36:48 for the dbus details, look in http://www.cncmods.net/files/dbus/net_interface.xml under get_network_time_info method and network_time_info_change signal Oct 16 13:36:50 though disclaimer: I never... Oct 16 13:37:01 thats how csd tells timed the new time Oct 16 13:37:06 time settings that is Oct 16 13:37:49 DAMN .xml files Oct 16 13:37:57 my browser just renders them Oct 16 13:38:17 Let me see who else might be using those same dbus interfaces Oct 16 13:38:35 rendering them results in blank page Oct 16 13:39:39 btw could we gett all that stuff into maemo wiki? Oct 16 13:39:53 feel free to update the wiki if you want :) Oct 16 13:39:59 I suck at doing that kind of wiki work :P Oct 16 13:41:29 * DocScrutinizer05 too Oct 16 13:41:49 maybe Sixwheeled feels inclined... Oct 16 13:41:58 http://www.cncmods.net/files/dbus/ Oct 16 13:44:26 csd_ss.h sounds braindamaged. supplementaryServices is more than just call divert Oct 16 13:44:44 csd_ss.h contains only the info I could pick up Oct 16 13:44:53 it's basically all from *0# to *99# Oct 16 13:44:57 ss.xml contains the full interface pulled from introspection Oct 16 13:44:58 aah Oct 16 13:45:08 i.e. csd_ss.h is bits from bluez etc Oct 16 13:45:56 ok, the time dbus calls are used by clockd, osso_startup_wizzard.launch and libconnui_cell.so Oct 16 13:46:10 refer to explanation in starhash Oct 16 13:46:35 well, I guess Nokia subsummed it under USSD Oct 16 13:51:22 ok, so anything else I can look up? Oct 16 13:52:32 jonwil: btw are you aware of the weird TZ integer in "get_network_time_info" ? Oct 16 13:53:28 and what's/who's_using libconnui_cell.so ? Oct 16 13:54:10 (TZ) it's sth like 15min steps +- of UTC Oct 16 13:58:26 libconnui_cell is used by connui-home-cellular (operator name widget), connui-cellular-settings (cellular settings control panel), connui-conndlgs-cellular (cellular connectivity dialogs) and connui-statusbar-cellular (cellular signal status widget) Oct 16 13:59:14 * DocScrutinizer05 ponders writing a dbuslogd, similar to syslogd Oct 16 14:00:19 wouldn't be that hard by copying dbus-monitor and changing it a bit Oct 16 14:00:25 Wouldn't that just be dbus-monitor > /var/log/dbuslog? Oct 16 14:00:39 yeah something like that in the background Oct 16 14:00:45 seems libconnui_cell.so only imports the time dbus calls just because the are part of net_interface.xml Oct 16 14:01:00 kolp: exactly Oct 16 14:01:18 though |gzip >dbuslog.gz Oct 16 14:02:04 nope, libconnui_cell actually hooks up to (and monitors) the network_time_info_change signal Oct 16 14:02:36 dbus is rather chatty Oct 16 14:02:53 not sure what it does with it Oct 16 14:03:00 the signal that is Oct 16 14:03:37 maybe it offers a lib function to get network time Oct 16 14:07:10 nope, it doesn't Oct 16 14:07:15 not sure what it does with the signal Oct 16 14:07:31 obviously nothing then Oct 16 14:07:36 it wires it up with a call to dbus_g_proxy_add_signal Oct 16 14:07:54 its possible the results it gets from the signal are sent to a registered callback Oct 16 14:08:38 hmm, then it kinda _does_ provide a function to query network time status/value Oct 16 14:08:51 I cant follow the ARM code to know for sure whats going on Oct 16 14:08:59 np Oct 16 14:09:11 not yet at least Oct 16 14:09:18 so it might be doing nothing or it might be passing it to a callback or it might be doing something else Oct 16 14:10:05 ~lart nokia for csd stack being closed Oct 16 14:10:05 * infobot gives nokia an extra strength ACME sleeping pill, sending nokia to sleep for 150 years, and awakening to seven strange dwarfs and a large apple for csd stack being closed Oct 16 14:12:05 Whats needed is not the code for CSD but the documentation for the dbus calls it makes Oct 16 14:12:15 and exposes and stuff Oct 16 14:12:17 :nod: Oct 16 14:12:51 alas that's also closed resp nonexistent, and THAT is really BAD Oct 16 14:13:13 yeah I bet even libscnet-dev wasn't supposed to be released Oct 16 14:13:52 but it was and I am for sure going to use the info it contains :) Oct 16 14:14:04 sure thing Oct 16 14:15:18 heck, I am highly surprised Nokia committed to publish info in ofono for MeeGo rather than simply keeping it secret and releasing a blob for that bit along the lines of what they did with BME and GPS and stuff Oct 16 14:16:39 I bet Intel had something to do with that though Oct 16 14:18:27 http://narf-archive.com/pix/f3366f6478d922273f64db94543488bd659802b4.jpeg Oct 16 14:22:24 its clear that nokia wanted to release as little info as possible when doing ofono stuff Oct 16 14:23:54 they don't feel confident in security and stability of their BB5 ISI interface Oct 16 14:24:28 the more low level the interface, the harder the security audit Oct 16 14:24:50 now figure what's going on with that insane shared-RAM stuff Oct 16 14:25:48 I have noticed that basically all mobile devices (even those who are notionally "open" like Google Nexus devices) have closed blobs to talk to the cellular radio Oct 16 14:26:08 yes, exactly for that reason Oct 16 14:26:47 security by obscurity Oct 16 14:27:51 Yah Oct 16 14:28:42 the modem OS usually has very low inherent security - there's no memory protection, no users not to think of permissions, no proper preemptive multitasking, zilch Oct 16 14:29:34 so every little exploit in a e.g. uart driver or HSI driver instantly compromises the complete modem system Oct 16 14:30:57 think of a modem OS as kernel-only, with all the processes implemented as kernel modules Oct 16 14:31:57 real fun when *something* corrupts your process' local variables memory area Oct 16 14:32:22 you can't even say which of the 190 processes running on the modem been the culprit Oct 16 14:32:33 kinda like windows2 Oct 16 14:33:06 ok, maybe windows3.1 Oct 16 14:37:19 you dont want to know about some of the stuff that I saw when I worked for Motorola on a 6 moth student internship (back when the RAZR V3 flip phone was the best thing since sliced bread) Oct 16 14:37:59 I worked on ST-E NovaThor modem until a year ago Oct 16 14:38:09 I can only concur Oct 16 14:39:54 though one thing I actually would like to share, alas I mustn't diclose the original internal memos: verification and certification of a radio stack is a process that nobody who hasn't seen it can figure how much work it involves Oct 16 14:41:49 messages like "meanwhile the UMTS validation reached 30min uninterrupted calls even with cell handover in same band" are quite usual Oct 16 14:43:10 Its not just about the phone, its about all the 100s of different devices the phone could talk to over Bluetooth or WiFi or cellular or whatever Oct 16 14:45:41 well, that too Oct 16 14:45:53 but not when you're just working on the modem Oct 16 14:46:42 this been a different department in ST-E, sitting in a different location several 1000km away, and probably also several 100 people Oct 16 14:48:04 definitely developing the modem software alone is a way huger project than e.g. developing a system like maemo Oct 16 14:49:41 maybe the bb5 cmt-firmware is not as large (in MB) like the rootfs image, but that doesn't mean it's less complex Oct 16 14:58:11 if only the cellular standards bodies could make their damn standards less complex (look at WiFi or even WiMax compared to UMTS/LTE) Oct 16 14:58:45 I suspect the #1 reason they are so complex has to do with patents Oct 16 14:58:57 and every one wanting to throw in their latest patent into the ring Oct 16 14:59:09 to make more money from every phone sold Oct 16 14:59:21 eeh Oct 16 14:59:29 sure, a big-ish part may be that Oct 16 14:59:36 but it's still really complex stuff Oct 16 15:01:11 anyway, i need a better home router Oct 16 15:01:21 what should i buy? Oct 16 15:01:28 why? got a D-Link? ;-P Oct 16 15:01:34 linksys wrt610n Oct 16 15:01:42 dd-wrt crashes like a mofo Oct 16 15:01:49 openwrt doesn't support it Oct 16 15:02:07 i have to use my isps Oct 16 15:02:13 mashina: hah Oct 16 15:02:16 what ISP is that? Oct 16 15:02:19 it isnt bad but the webui only works in ie and webkit Oct 16 15:02:25 probably some crappy ISP Oct 16 15:02:25 i have a dsl modem of my own Oct 16 15:02:30 it uses pppoe :3 Oct 16 15:02:35 ftth actually Oct 16 15:02:43 oh ok, thats great :) Oct 16 15:02:51 i can see the ONT in the basement Oct 16 15:02:53 doesn't ftth usually give you an ethernet cable with dhcp? Oct 16 15:03:05 yeah but they do some magic for iptv Oct 16 15:03:13 vlan and dhco abuse Oct 16 15:03:30 disregard their reality and make up your own Oct 16 15:03:31