**** BEGIN LOGGING AT Fri Jan 08 02:59:58 2016 Jan 08 03:36:53 Wizzup: on https://github.com/MerlijnWajer/uinput-mapper/blob/master/README.rst at least when watching it in browser there's a typo in the cmdline at bottom. \ missing before 1 Jan 08 03:37:41 Wizzup: sed -r 's/[^ ]+ +([^ ]+).*/"1" : 1,/' should be sed -r 's/[^ ]+ +([^ ]+).*/"1" : \1,/' Jan 08 03:38:20 and I'm not even convinced the "1" is correct Jan 08 03:44:37 jonwil: alas this generally fails for me, the time from modem isn't available in time. Jan 08 03:47:14 Yes, same for me. I either put it manually, or wait for startup to go through with wrong time, and then it auto-sets from cellular (unless it forgot the setting for auto-time, just like it sometimes forgets setting for 12-vs-24) Jan 08 03:48:02 Fixing this functionality would be _great_ . It might or not be related to backup-battery having enough power for the modem? Jan 08 03:48:47 Scratch that, loss-or-not of data (such as time, date, language, location) from memory should not affect modem, right? Jan 08 03:50:44 It's like, when startup shows correct time, I cannot be sure whether it's because it got it from modem or because it remembered it in memory. But when it does not show correct time, it's obvious that it does not manage to get time from modem. Jan 08 03:58:39 Oksana: bupbat never powers modem Jan 08 03:59:40 bupbat only powers RTC in SoC's companion chip TWL4030 aka GAIA, plus a few byte of CMOS-RAM non-volatile memory Jan 08 04:00:14 ... but it should not affect the problem since startup wizard tries to reach modem when mainbattery is already here. Unless startup modem does not actually know how to wake up modem, or to wait for connection to towers to be established Jan 08 04:00:59 it takes time to get time from network Jan 08 04:01:29 So... Would it get correct time if startup-wizard was left to rest for a while? Jan 08 04:01:47 for establishing connection as you said, and also for waiting for next timestamp telegram sent from BTS Jan 08 04:01:52 If yes, then neat, it just needs a progress bar ;-) Jan 08 04:01:53 yes Jan 08 04:02:39 actually a "sleep 15" in the initscript which launches startup wizard should already do Jan 08 04:03:01 unless startup wizard also powers up modem Jan 08 04:03:10 Yes, that's the question Jan 08 04:03:12 but I think that's done by CSD Jan 08 04:03:41 I seem to recall some users completely disabled annoying startup wizard Jan 08 04:03:52 and their modem still worked ;-) Jan 08 04:04:24 It's not annoying by itself, it's annoying when it stays on for a long time, preventing communication (like receiving calls) Jan 08 04:04:37 yep Jan 08 04:04:49 It could, for example, auto-proceed after successfully receiving time from modem? Jan 08 04:05:06 do you happen to know off top of your head where startup wizard gets invoked? Jan 08 04:06:00 AIUI, it *does* auto-proceed when RTC-time is correct on invokation Jan 08 04:06:12 Nay. There is /usr/bin/osso-startup-wizard, /usr/bin/osso-startup-wizard.launch, and /usr/bin/osso_startup_wizard/state , or something Jan 08 04:06:30 yeah, but who calls that? Jan 08 04:07:11 a grep will tell Jan 08 04:07:33 RTC-time can be correct on invocation only if battery held the data intact. It would be nice if it also auto-continued if RTC-time is off, after time gets corrected by modem Jan 08 04:08:46 For example, right now, osso-startup-wizard showed itself... with correct time. Does it mean that it got it from modem, and didn't auto-proceed without user's approval because RTC-clock was off before modem corrected it? Jan 08 04:09:25 hmmphh http://paste.opensuse.org/10325924 Jan 08 04:10:16 There is also Xsession.d/30 Jan 08 04:10:44 Oksana: no, time in RTC can also be correct when modem been fast enough (or had enough time) so it was able to set correct time in system at the moment startup-wizard gets invoked Jan 08 04:11:14 But then, it would have auto-proceeded, instead of waiting for user's "Save" button? Jan 08 04:12:01 I'm not sure about startupwizard actually automatically proceeding when time is correct Jan 08 04:12:33 Hn... Jan 08 04:12:44 * DocScrutinizer05 glares at >> /etc/osso-rfs-scripts/osso-startup-wizard-rfs.sh:rm -f /var/lib/osso-startup-wizard/osso_startup_wizard_state << Jan 08 04:13:52 I'm pretty sure this semaphore file is related to whether or not startup-wizard continues or blocks boot Jan 08 04:16:16 osso-startup-wizard gets started in /etc/X11/Xsession.d/30osso-startup-wizard or something Jan 08 04:18:42 Collected information to decide whether to show GUI: FirstTimeRun, DateTimeUnset, SIMpresent, CurrentSimId, ChangedSettings, ValidPin, SettingsChangedBefore, SimSeenBefore, LastValidSimId, ChangedWithoutSimBefore... It's complicated Jan 08 04:20:06 There is a possibility in osso-startup-wizard to exit if there is no need for it, but what is exactly the equation for auto-proceeding? Jan 08 04:22:00 funny enough my system has no /etc/X11/Xsession.d/30osso_startup_wizard Jan 08 04:22:12 It's from .launch file, by the way Jan 08 04:22:26 Is it /etc/X11/Xsession.d/30osso-startup-wizard ? Jan 08 04:22:39 Nokia confuses _ and - , sometimes Jan 08 04:22:51 And I do not care to remember which one is where Jan 08 04:23:04 my system has no /etc/X11/Xsession.d/30* Jan 08 04:23:51 Do you have osso-startup-wizard installed? What does dpkg -L say? Jan 08 04:24:20 And yes, I do have /etc/X11/Xsession.d/30osso_startup_wizard , short-ish as it is Jan 08 04:27:20 http://paste.opensuse.org/28699642 Jan 08 04:28:30 By the way, .launch file recognises SIM states possible as Unknown, Ok, PIN requires, SIM removed, SIM permanently blocked, SIMless Offline, PUK required, SIMlock not okay, SIM not ready (let's wait for it), SIMlock initializing, or unknown Jan 08 04:29:09 IroN900:/home/user# locate Xsession.d/30 Jan 08 04:29:10 locate: Warnung: Die Datenbank "/var/cache/locate/locatedb" ist älter als 8 Tage (Das Alter ist 135,8 Tage). Jan 08 04:29:12 /media/mmc1p2/etc/X11/Xsession.d/30xTs_Calibrate Jan 08 04:29:13 /media/mmc1p2/N900-1-archive/etc/X11/Xsession.d/30osso_startup_wizard Jan 08 04:29:44 Heh, in archive... Jan 08 04:30:08 Renamed to prevent wizard from showing up? Jan 08 04:30:24 err, I think it still shows up Jan 08 04:30:42 but maybe I actually nuked it, lemme try Jan 08 04:31:09 a pity after only 20 days of uptime Jan 08 04:32:03 Ouch... :-) Funny name does not nuke it? Wizard states are: settings-changed, SIM-seen, changed-without-pin? I am just guessing from binary file opened with vim Jan 08 04:33:11 I usually do the 'funny name' thing when I wanna get rid of files Jan 08 04:33:51 let's see, I hope it was long enough with no batt Jan 08 04:33:55 Why does binary.launch file mention scratchbox? Maybe, it considered 30 to be more important than the filenametail? Jan 08 04:35:16 hmm, dunno, either it was not long enough or this device has excellent bupbat or I actually did remove the 30 file Jan 08 04:36:02 anyway time is correct Jan 08 04:36:17 Ookay... Just don't be surprised when it boots up with wrong language and-or location ;-) Jan 08 04:36:31 and that's probably why I removed that 30* file Jan 08 04:37:05 why would it boot with wrong location or lang when I previously set it correctly? Jan 08 04:37:39 Time becomes not correct only if it forgets the time (bad bupbat) _and_ it forgets the setting for auto-updating the time from modem (which rarely ever happens, but it does) Jan 08 04:37:46 I'm however not surprised it booted with incorrect time after batt change in La Gomera Jan 08 04:37:51 ;-P Jan 08 04:38:35 Well, I haven't seen it for a long time, but I think that it can forget language-and-or-location because of bad bupbat and-or bad shutdown Jan 08 04:39:21 no, time becomes not correct after each batt removal on dead-bupbat device, until the correct time gets fetched from cellular network. Which works great in DE-telefonica network, but obviously not so good in La Gomera Jan 08 04:40:04 no, never seen it forgetting lang or (timezone)location Jan 08 04:40:49 timezone otoh is supposed to get auto-updated on roaming - if the cellular network provides all the needed data correctly Jan 08 04:41:03 I know, I just mean that sometimes I have seen the device forgetting both auto-update-from-modem-setting and 12-vs-24 setting. Not so sure about language-and-or-location, it is not that obvious (English-USA vs English-UK) Jan 08 04:41:29 never happened to me so far Jan 08 04:41:55 but then I almost never shutdown hard Jan 08 04:42:27 * Oksana winces - having "Recharge battery" shutdown is not rare for me Jan 08 04:42:48 well, that's a graceful shutdown Jan 08 04:43:43 Well, I also have moments when GUI freezes, and the only way to do anything is to shut-down it with power button. Is it graceful, too? Jan 08 04:43:55 I may note that I'm not running the BME-replacements. All stock on my system Jan 08 04:44:17 yes, that should also be almost graceful Jan 08 04:44:58 Because I do not have cases (well, haven't had them for a very long time) when I would have to take the battery out without shutting down first Jan 08 04:45:20 To be fair, LED does not work, so taking battery out is always case of wait-and-guess Jan 08 04:45:32 well, when startup-wizard never runs, those config files never are opened in write-mode and thus no buffers may get lost during hard shutdown ;-) Jan 08 04:46:47 oh that's a pity Jan 08 04:47:18 I should just replace flex-ribbon, front-camera doesn't work either Jan 08 04:47:26 :-/ Jan 08 04:48:00 for "wait and guess" a count up to 40 is not unreasonable Jan 08 04:58:02 btw /etc/osso-cud-scripts/ seem to be CleanUpDevice Jan 08 04:58:32 lemme guess - rfs then means ResetFactoryState Jan 08 04:59:05 as in /etc/osso-rfs-scripts/ Jan 08 05:00:12 rtcom-eventlogger-clean.sh :-)) Jan 08 05:00:39 now after 5 years the *official* method to handle bloated eventlog-db Jan 08 05:02:44 :-)) http://paste.opensuse.org/26394682 :-)) Jan 08 05:02:52 Hmm, wouldn't cleaning eventlog remove all SMS or something? Jan 08 05:03:00 yes, exactly Jan 08 05:03:16 all SMS and all call-log Jan 08 05:03:18 ~rescueos Jan 08 05:03:19 i guess rescueos is http://n900.quitesimple.org/rescueOS/ Jan 08 05:03:20 * Oksana is rather attached even to such useless things Jan 08 05:04:56 It's a case of "my cellular provider knows when I send and receive SMS, why shouldn't I know it, too?" Somewhat similar to lifelogging concept Jan 08 05:08:57 well, once your $HOME/.rtcom-eventlogger/el-v1.db grows to several MB size, you really want to backup it and then run /etc/osso-cud-scripts/rtcom-eventlogger-clean.sh Jan 08 05:10:47 Does the Conversations become too sluggish? Jan 08 05:12:48 everything comes to a grinding halt when an inbound call pulls in the complete 30MB el-v1.db into RAM cache Jan 08 05:14:06 it's a rdbms file, and you know how those are handled, right? Jan 08 05:14:34 Hrm, wouldn't it be a question of separating database into monthly chunks-files, and only handling one or two files at a time? Jan 08 05:16:57 It's not like Photos doesn't have an option already to display images grouped by months, so it would fit the them Jan 08 05:17:01 theme* Jan 08 05:20:18 Or, just have separate databases for separate contacts. Like, when you want to add something to database, you already know the phone-number-or-whatever, and you can pull up appropriate database. But, number of months in a lifetime is likely larger than number of contacts, and the spread is more uniform, so it would make more sense to go with months, as long as you know time correctly Jan 08 05:23:40 Oksana: you noticed that missed calls adds the number of not yet seen missed calls from same number and lists them only once, with a integer showing how many calls you missed? This will need the db crawling through all records, even when they are separated into many files. Also a rdbms doesn't work like "let's chunk it up into files" Jan 08 05:24:41 In fact, most parts of the system would not notice if they were fed only last-month part of database. Like, user would not need to know about calls missed previous month, or the month before that. But, user would need a file-defensive script for moving files around when the next month comes. Jan 08 05:25:33 anyway rtcom-call-ui rtcom-messaging-ui osso-addressbook all three access this el-v1.db and it seems to me we can't change those three programs easily Jan 08 05:26:45 But, scrolling down the conversations, and into earlier messages (be it in main-list or in separate-conversation) would be tricky. And calling up communication history with a contact. And yes, all these programs are not easily changable. Jan 08 05:38:35 * DocScrutinizer05 should actually clean out his el-v1.db - it's again 1.3MB Jan 08 05:40:23 also when you clean resp delete the el-v1.db, don't forget to again add the stored procedures to calculate call duration Jan 08 05:41:33 * DocScrutinizer05 glares at recaller, which doesn't record SIP Jan 08 05:45:01 What about orecchiette? Does it record SIP? Jan 08 05:45:24 Albeit, it has to have its GUI started up, to be able to auto-record calls Jan 08 06:15:23 I'm sure patching recaller would be rather easy. The author simply didn't think of SIP when defining the dbus messages to listen for Jan 08 06:16:43 actually _I_ wouldn't listen for the call messages but for the audio policy changes Jan 08 06:17:21 well, you want to listen for the call messages anyway, for the telefone numbers Jan 08 07:37:18 KDE custom shortcut, trigger: "Search" (((searchkey on my mouse))) Action: Command/URL: /usr/bin/kfmclient openProfile webbrowsing "https://ixquick.com/do/metasearch.pl?query=`qdbus org.kde.klipper /klipper getClipboardContents`" Jan 08 08:21:55 techporn: http://i.imgur.com/UOv9RUD.gif?_You_guys_clearly_aren__039_t_fucking_appreciating_all_this_shit. Jan 08 09:01:39 DocScrutinizer05: you're right, let me look at that... Jan 08 09:09:57 Wizzup: I get a list like http://paste.opensuse.org/6742143 even with the fixed \ Jan 08 09:10:06 is this how it's supposed to look? Jan 08 09:11:18 no Jan 08 09:11:25 thought as much :-) Jan 08 09:11:44 http://sprunge.us/RHSC Jan 08 09:12:26 yeah, gathered that. And I suspect the first part should be a "\2" maybe? instead of the "1" Jan 08 09:12:59 lemme try if I can fix it by guessing Jan 08 09:14:31 actually nope, \1 needs to be \2 Jan 08 09:14:46 what gcc version do you have Jan 08 09:16:41 it works for me, oddly enough Jan 08 09:17:50 sed -r 's/[^ ]+ +([^ ]+)(.*)/"\1" : \2,/' ? Jan 08 09:19:02 after sed it looks like http://paste.opensuse.org/71983028 Jan 08 09:20:55 http://paste.opensuse.org/96477214 is the result of gcc -E -dM /usr/include/linux/input.h | egrep ' (EV|SYN|KEY|BTN|REL|ABS|MSC|LED|SND|REP|SW)_[A-Za-z0-9_]+' | ( echo "#include " ; echo "input_constants_dict = {" ; sed -r 's/[^ ]+ +([^ ]+)(.*)/"\1" : \2,/' ; echo "}" ) | gcc -E -o /dev/stdout - | grep 'input_constants_dict = {' -A 100000 > uinputmapper/uinput_gen.py jr@saturn:~> less uinputmapper/uinput_gen.py Jan 08 09:21:34 does this look ok-ish to you? Jan 08 09:22:27 well, it's the head -n 46 Jan 08 09:22:45 Yeah, that looks good Jan 08 09:22:48 I think :) Jan 08 09:24:07 http://paste.opensuse.org/75886370 full Jan 08 09:25:02 cmd: http://paste.opensuse.org/98031835 Jan 08 09:25:16 don't ask me what the second gcc does Jan 08 09:25:32 I fixed the whole thing by mere guessing Jan 08 09:26:03 ok - I would like to work on it more now, but must dash to work - back later :) Jan 08 09:26:11 :-) Jan 08 09:26:32 I'll pester you how to actually install that thing, then Jan 08 09:26:33 It presumably expands the macro labels that aren't inside string literals. Jan 08 09:31:40 gcc -E -dM /usr/include/linux/input.h | sed 's/^.*\(\(EV\|SYN\|KEY\|BTN\|REL\|ABS\|MSC\|LED\|SND\|REP\|SW\)_[A-Za-z0-9_]\+\).*$/"\1": \1/te; d; :e' | gcc -E -o - - Jan 08 09:32:57 http://paste.opensuse.org/19514918 Jan 08 09:35:01 gcc -E -dM /usr/include/linux/input.h | (echo '#include ' && sed 's/^.*\(\(EV\|SYN\|KEY\|BTN\|REL\|ABS\|MSC\|LED\|SND\|REP\|SW\)_[A-Za-z0-9_]\+\).*$/"\1": \1/te; d; :e') | gcc -E -o - - Jan 08 09:35:28 Meh. Jan 08 09:35:44 So that's what the final grep is for. Jan 08 09:38:52 Missed a semicolon too, before the `te` Jan 08 12:24:31 anyone know what SIP User Agent that n900/maemo uses (as seen from a sip server) ? Jan 08 12:25:03 I know Symbian devices would use things like 'Nokia RM-244 400.34.0.11' - but with n900 being linux I'm not sure Jan 08 12:27:49 hmm i think it mightbe Telepathy-SofiaSIP/0.5.18.1 Jan 08 13:11:14 would make sense since it uses telepathy Jan 08 15:17:57 stevenm: yep Jan 08 15:18:03 sth like that Jan 08 15:18:11 sth!? Jan 08 15:18:17 oh something Jan 08 15:19:04 can it be changed? the platform i'm trying to connect to checks useragents and allows s60 devices though - but wouldn't allow that through Jan 08 15:19:05 Telepathy-SofiaSIP Jan 08 15:19:31 duh, it checks the type of client? Jan 08 15:19:36 yup Jan 08 15:19:59 I'm afraid this string is prolly hardcoded into the SIP lib Jan 08 15:20:27 which means you can change it with a hexeditor Jan 08 15:20:32 why would they do that by the way? Jan 08 15:20:53 Sicelo, you mean 'why have they done that by the way?' Jan 08 15:20:56 do what? check client name? Jan 08 15:21:02 because they're BT and they're shit Jan 08 15:21:17 what's BT? Jan 08 15:21:21 https://www.btwholesale.com/pages/static/products-services/ip-voice-services.htm Jan 08 15:21:45 their SBC checks user agents so only approved devices which they've checked for compatibility can connect Jan 08 15:21:48 so are you sure only S60? Jan 08 15:21:51 hmm Jan 08 15:22:14 no lots of others too - but I have an asterisk server I managed to connect (which I can change user agents on) so I can test which user agents get past and register... and which don't Jan 08 15:22:18 s60 devices are approved anyway Jan 08 15:23:26 i just wonder if the N900 is not failing something else (not necessarily UA).. for example, N900 can't connect to Ekiga servers Jan 08 15:23:49 who said I'd even tried it? Jan 08 15:24:24 if the useragent is Telepathy-SofiaSIP then then I've tried that on my test asterisk install and it failed - where as a Nokia RM-?? user agent works Jan 08 15:24:38 ok Jan 08 15:25:13 don't worry about it anyway - lots of stuff doesn't work with their crappy platform Jan 08 15:25:42 thats why i've got asterisk in the first place to act as a proxy... although i'd much rather use kamailio but i'm not sure how to yet Jan 08 15:31:40 Telepathy-SofiaSIP/0.6.2 in /usr/lib/telepathy/telepathy-sofiasip Jan 08 15:41:39 a ~> cp /usr/lib/telepathy/telepathy-sofiasip /usr/lib/telepathy/telepathy-sofiasip_BACKUP; sed -i s#Telepathy-SofiaSIP/0.6.2#S60 # /usr/lib/telepathy/telepathy-sofiasip ;# should do Jan 08 15:48:14 though it seems you have an old version 0.5.18.1, so you should find the string by yourself (strings /usr/lib/telepathy/telepathy-sofiasip|grep SofiaSIP) and make sure you replace it by exactly same number of chars via sed Jan 08 21:14:59 hahaha!! Jan 08 21:15:22 got OWA access via own method Jan 08 21:15:58 I'll close the ticket with IT Dept. Jan 08 21:17:24 https://github.com/iSECPartners/jailbreak helped .. to export a non-exportable cert. N900 happily accepted it Jan 08 21:20:57 ? Jan 08 21:27:15 for context, http://mg.pov.lt/maemo-irclog/%23maemo.2015-12-21.log.html#t2015-12-21T10:43:02 Jan 08 22:04:37 !erlehmann **** BEGIN LOGGING AT Sat Jan 09 00:45:36 2016 **** ENDING LOGGING AT Sat Jan 09 02:59:58 2016