**** BEGIN LOGGING AT Thu Oct 24 02:59:58 2013 Oct 24 03:52:33 Tekk_: The Nokia N900 does not support the UMTS bands used by Telus Mobility, Bell Mobility, and their MVNOs in Canada. Telus Mobility and Bell Mobility use only CDMA2000 and UMTS, not GSM. If the Neo900 supports these UMTS bands, then the Neo900 has a significant advantage over the Nokia N900 in Canada. Oct 24 09:19:58 Hello! I'm experiencing something weird. Yesterday I was asking here about my phone not booting, but it finally booted and everything was fine. Today I was using my phone and suddenly it said that my sim card was disconnected. Ok, but after 10 seconds it told me that now it is connected and asked me for the pin code. I entered my pin code correctly two times each time receiving "invalid pin", on the third try it worked and connected Oct 24 09:19:58 correctly. Okay, but after some time this happened again... and again and again. So I decided to reboot my phone, but now it was not booting again. So I was playing around with reinserting the battery and finally it worked, but it did not manage to boot and got stuck on these loading dots. I've tried again - didn't work... So some time later I've tried it again and it worked from the third try or so. SIM card problem appeared soon after Oct 24 09:19:58 successful boot. Now it is working, but says that sim card is not there. Also if I try to charge it I get "Not Charging" message. Any ideas? Oct 24 09:39:56 Sysaxed, it sound like the issue when due to heat/the board flexing a chips solder joints have broken :( ( cant think of what to search for on the forums, but its there somewhere) Oct 24 10:11:29 hmmm, okay, anybody else? Oct 24 10:12:15 chainsawbike: thanks, but it is still weird that it happened without anything happening before Oct 24 10:12:30 I can understand if it fell and then started to act like this Oct 24 10:12:37 but maybe you're right Oct 24 10:31:43 Sysaxed: do you do any over/under clocking/volting? Oct 24 10:37:05 Hi maemo folks. Some time back one of you gave me a URL that when browsed to on the N900 would load a bunch of repos automatically. Anyone got that link mentally cached? Oct 24 10:48:52 dang Oct 24 10:49:15 for the record: http://maemo.cloud-7.de/repositories.install Oct 24 10:49:23 or simply Oct 24 10:49:30 ~jrtools Oct 24 10:49:30 rumour has it, jrtools is http://wiki.maemo.org/User:Joerg_rw/tools Oct 24 10:49:42 which has a link to the file Oct 24 10:50:24 Sysaxed: sounds like hw defect :´( Oct 24 10:51:22 when you're extremely lucky, it's just the battery that's defect Oct 24 10:52:53 Sysaxed: standard recommendation: full reflash PR1.3.1, no installation of any additional apps, test for a week in this state Oct 24 10:53:52 I know reflash sounds inconvenient, but given the alternative of instantly binning the device, it's for sure worth to give it a try Oct 24 10:54:25 excellent howto but what IS rw/tools? Oct 24 10:54:39 hmm? Oct 24 10:54:53 sorry, i just wanted to know what the outlined procedure is for Oct 24 10:55:05 whichone? Oct 24 10:55:07 becaues it sounds like something i might want to do, ... Oct 24 10:55:11 jrtools Oct 24 10:55:29 is CSSU = thumb = jrtools? Oct 24 10:55:44 jrtools is a collection of little hints I found/developed during last 4 years Oct 24 10:55:48 ah Oct 24 10:56:02 ok, CSSU = what i am using now Oct 24 10:56:07 thumb = probably somewhere in the wiki Oct 24 10:56:19 and this page = doc's personal howtos Oct 24 10:56:21 thanks Oct 24 10:56:25 is there any email client that supports gmail? Oct 24 10:56:34 not sure Oct 24 10:56:37 gmail supports pop3 and imap so yes, any, probably? Oct 24 10:57:12 modest (stock mailer) supports imap and pop3 and smtp Oct 24 10:57:12 youd think so, but I get an error about the message not existimg on server Oct 24 10:57:24 hrm. "an error" Oct 24 10:57:36 maybe sniff in on that and try to find out what is going on Oct 24 10:57:44 seems imap support is kinda flawed in stock modest, use CSSU Oct 24 10:57:57 DocScrutinizer05: "modest"? Oct 24 10:58:03 the maemo mailer Oct 24 10:58:06 thanks Oct 24 10:58:19 I thought I was using CSSU Oct 24 10:58:24 bloody UI stuff never told me what it was called =) Oct 24 10:58:26 check changelog linked to on Oct 24 10:58:28 ~cssu Oct 24 10:58:28 cssu is probably http://wiki.maemo.org/Community_SSU Oct 24 10:58:56 it has fixes particularly for gmail iirc Oct 24 10:59:35 * sono finds https://wiki.maemo.org/Community_SSU/Thumb and reads Oct 24 10:59:40 gotta use that lunch break for something useful =) Oct 24 10:59:56 cehteh: no, I've not done any overclocking Oct 24 11:03:03 DocScrutinizer05: what if it is the battery, would changing the battery fix the problem? Oct 24 11:03:14 sure Oct 24 11:04:38 hmmm Oct 24 11:04:42 old worn battery LiIon cells tend to develop high effective Series Resistance (ESR), and sometimes behave quite random on this. A high ESR causes "brownout" when current bursts are drawn from battery by e.g. the modem during TX Oct 24 11:05:02 and the modem switches off earlier than the OMAP APE linux system Oct 24 11:05:23 earlier as in "higher battery voltage" Oct 24 11:05:26 DocScrutinizer05: sounds logical Oct 24 11:05:45 considering the fact that these are some random chinese batteries Oct 24 11:06:09 ok, I have another battery working perfectly with another n900 Oct 24 11:06:15 I'll give it a try soon Oct 24 11:06:18 short brownouts can cause quite random erratic behaviour of all components of a system Oct 24 11:06:41 think of it like extreme undervolting Oct 24 11:07:11 processes can segfault at random Oct 24 11:07:20 memory IO may fail Oct 24 11:07:36 make sure you try a good battery Oct 24 11:08:02 but, sorry to say this, I have only little hope it's caused by your battery Oct 24 11:08:49 try to install ksyslogd and provide a full pastebin, so I could have a look at it Oct 24 11:09:13 syslog may reveal some hints on what's going on Oct 24 14:25:10 DocScrutinizer05, hello... do you know a way to save the xchat configuration to prevent its reset after closing this app on n900? Oct 24 14:40:47 zammy: just close the app via the app's close-app button/menu Oct 24 14:41:35 alternatively there once been a "save settings" menu entry, dunno if it still exists Oct 24 14:42:23 ohthanks Oct 24 14:43:29 DocScrutinizer05, but it says something like Error key binding.. stuff Oct 24 14:44:49 hmm, then there's probably a syntax or semantics error in the key bindings Oct 24 14:46:13 can't find the "save settings" menu entry Oct 24 14:50:43 editing the keybindings is a painful exercise anyway, simply because the display is too small to show any reasonable overview or detail view for assigning a binding to a key Oct 24 14:51:31 right... for this reason I want to save the configuration DocScrutinizer05 because it's not only key binding configurable... Oct 24 14:52:08 I don't have many bindings. I think shift-space for nick-completion is basically the only one Oct 24 14:52:19 yes Oct 24 14:52:28 that's all indeed DocScrutinizer05 Oct 24 14:52:40 I edit font also Oct 24 14:52:45 and some color Oct 24 14:52:46 often I find it simpler ro edit the raw config files in ~/.xchat2/* Oct 24 14:53:04 does it affect? Oct 24 14:53:12 I tried... it seemed not Oct 24 14:53:31 you mustn't have xchat started when editing those files Oct 24 14:53:42 yes I know Oct 24 14:53:50 but anyway did not effect Oct 24 14:53:58 strange Oct 24 14:54:34 I couldn't figure where else xchat stores settings, afaik they are all in /home/user/.xchat2/ Oct 24 14:55:13 and several users said they were happy with my files from Oct 24 14:55:18 ~jrxchat Oct 24 14:55:18 from memory, jrxchat is http://wiki.maemo.org/User:Joerg_rw/xchat/ Oct 24 14:55:44 which implies it had effect to "edit" aka change the content of those files Oct 24 14:55:45 but most of all weird thing on xchat is that this can handle its own tray icon settings and conf but no real output of this trayicon Oct 24 14:56:24 there are a few options that are meaningless under maemo/hildon Oct 24 15:03:44 I hate xchat... not editable at all Oct 24 15:08:29 xchat trayicon worked in Maemo4 Oct 24 15:11:03 * RST38h does remember he had to kick xchat to enable that Oct 24 15:20:47 hello Oct 24 15:24:48 alright DocScrutinizer05 I was doing one of mines again... I mixed the hexchat of my linux on pc with this n900 xchat and that was messing up all Oct 24 15:25:22 now xchat loads naturally its own configuration Oct 24 15:25:43 but still I would minimize it in tray Oct 24 15:26:31 minimizing in tray would mean you need a systemmenu entry to un-minimize it Oct 24 15:27:21 in hildon there are no dedicated popup menues from particular systray icons when clicking them Oct 24 15:30:19 so? no way at all? Oct 24 15:30:46 just keeping this as an open chat as all task window Oct 24 15:37:40 what alternative to having xchat in taskswitcher would you suggest to un-iconify it? Oct 24 15:38:05 I can't think of any, and I don't know of any that xchat offers Oct 24 15:38:12 maybe screen and some tips of screen it self Oct 24 15:38:31 hmm? Oct 24 15:38:49 trying to answer not being developer Oct 24 15:38:56 sorry Oct 24 15:39:50 I'm asking for a method from user perspective, to un-iconify xchat when minimized to systray Oct 24 15:42:29 by the same part idk how do you call it which you see battery and clock? Oct 24 15:43:06 that's user friendly side for trayicon Oct 24 15:44:58 otherwise the telepathy plugin for irc should be more developed Oct 24 15:45:06 but idk Oct 24 15:46:07 telepathy irc plugin is not going to give you proper experience no matter how you develop it Oct 24 15:46:23 simply because it has to use a completely alien ui Oct 24 15:47:13 right that's the of idk was minding to Oct 24 15:54:07 if this telepathy thing-room could remain open Oct 24 15:56:27 remote host, mosh, tmux, irssi, bitlbee Oct 24 15:56:50 what's this Oct 24 15:56:50 bit of effort and learning curve, but *so* worth it Oct 24 15:56:58 is there any reason not to use the fimger for pressing the softkeys in terminal? Oct 24 15:57:12 irssi I already know Oct 24 15:57:29 but are samely likely xchat Oct 24 15:58:05 I would iconify the client Oct 24 15:58:12 samn900: first you need an always-on system to connect to, at home (pi, openwrt, ... or an entire rootserver or shell e.g. uberspace.de) Oct 24 15:58:16 whichever client Oct 24 15:58:47 samn900: then you run mosh on that. configure ssh with key auth for easy access Oct 24 15:59:13 samn900: run a tmux session there, where you keep irssi open Oct 24 15:59:36 maybe a bounce in tcl of my eggdrop on my vps do you mean? Oct 24 15:59:43 samn900: connect irssi to bitlbee for other IM clients Oct 24 16:00:09 nah bncs are totally 90's imho Oct 24 16:00:09 I knewed about it but never started to configure a try it Oct 24 16:01:24 but maybe it's time to try at least Oct 24 16:02:15 take one step at a time Oct 24 16:02:27 first comes the host and ssh Oct 24 16:02:36 I have a vps Oct 24 16:02:38 the hardest part perhaps Oct 24 16:02:43 ok Oct 24 16:02:55 I have an eggdrop but not connected to this server freenode Oct 24 16:03:03 to another one Oct 24 16:04:34 I never understood how the bounce script works actually so I never tried it Oct 24 16:04:58 you dont need it in the described setup Oct 24 16:07:39 I prefer IRC proxies/bouncers because that gives me the option to freely choose my client (and run it locally, if desired) and still stay online Oct 24 16:08:47 i prefer having a central instance Oct 24 16:08:57 yes that's what I knewed... wnd so how this happens and what makes the bouncer Oct 24 16:09:27 it makes managing channels etc. SO much easier Oct 24 16:09:43 cool Oct 24 16:10:11 as well as offering all the other advantages of a remote shell Oct 24 16:10:27 can I use the same eggdrop to connect at two different servers with two different nicks? Oct 24 16:10:48 nowadays shells are actually commercially available, cheap products.. Oct 24 16:11:10 even root VMs are dirt cheap Oct 24 16:11:27 last bnc i ran was 1998 or so Oct 24 16:12:19 mosh is nonsense Oct 24 16:13:10 and why you need tmux when you already run mosh? Oct 24 16:13:46 no, and yes. Oct 24 16:13:55 mosh does not replace tmux Oct 24 16:14:01 unless I'm mistaken, mosh doesn't transfer (cleanly) from one host running the client to another Oct 24 16:14:06 and atm i am on the bus Oct 24 16:14:22 if you ever used ssh while moving Oct 24 16:14:32 I do that all the time Oct 24 16:14:40 you won't claim mosh is nonsense Oct 24 16:14:51 mosh IS nonsense Oct 24 16:15:00 if yoy say so Oct 24 16:15:16 DocScrutinizer05: it seems like my phone works correctly with another battery Oct 24 16:15:55 calling an app anything "mobile" that sends a ping every 5 sec or whatever, that is the definition of nonsense Oct 24 16:16:07 Sysaxed: great Oct 24 16:16:08 ok Oct 24 16:16:32 you could as well call ping a battery-saver Oct 24 16:16:59 (mosh is the next best thing after butter btw. already lost link twice this bus ride) Oct 24 16:17:22 *shrug* Oct 24 16:17:33 doc. you close it when you dont actively use it. it really is that siimple Oct 24 16:17:45 that's bullshit Oct 24 16:18:00 ok Oct 24 16:18:30 I could as well call VNC a "desktop optimized for mobile" Oct 24 16:18:35 no Oct 24 16:18:38 yes Oct 24 16:18:41 no Oct 24 16:18:46 =P Oct 24 16:18:59 I just did it, so yes Oct 24 16:19:06 =) Oct 24 16:19:38 that's the number one reason I don't want to use mosh: to run it on my n900 without killing the battery, I'd have to close it before idling Oct 24 16:20:02 closing it takes a click Oct 24 16:20:10 so what? Oct 24 16:20:14 new session into tmux takes a click Oct 24 16:20:20 indeed, but creating another connection takes much more than that Oct 24 16:20:21 so what? Oct 24 16:20:29 no Oct 24 16:20:46 tmux: run byobu Oct 24 16:20:57 why the heck do you need to "create connections"?? Oct 24 16:20:58 ssh: public key Oct 24 16:21:16 mosh-client: oneliner Oct 24 16:21:24 pfff Oct 24 16:21:40 wow I am trying sygic on n900 Oct 24 16:21:42 it's cool Oct 24 16:22:19 I can't trust a mobile device to have a passphraseless ssh key that allows me to log in to any relevant host, and using an agent doesn't really solve anything here Oct 24 16:23:28 thats why i use this host that only has crap on it Oct 24 16:24:02 my "mosh": while ! ssh -o ServerAliveInterval=300 -p22022 78.46.9287.666 screen -R; do sleep 1; done Oct 24 16:25:48 to make it more like your mosh, use ServerAliveInterval=3 Oct 24 16:26:44 hhahaha Oct 24 16:27:11 and honestly, fix your APN config, so you don't get a new DHCP lease each time you roam from one cell to another Oct 24 16:27:48 my ssh sessions stay for a week Oct 24 16:27:57 sgood for you Oct 24 16:28:07 actually for 24h-1s Oct 24 16:28:13 some people use consumer carriers Oct 24 16:28:50 some people do what? Oct 24 16:30:18 german vodafone, good luck with that apn config Oct 24 16:33:12 hmm? Oct 24 16:33:22 what's wrng with vodafone? Oct 24 16:33:28 wrong even Oct 24 16:35:42 only thing I could figure is: your carrier has a sparse coverage and a national roaming agreement with another carrier network, so each time your device is roaming from one network to the other it most likely would get a new DHCP lease and thus IP-addr Oct 24 16:36:52 while obviously discarding the old lease locally. So when roaming back from network B to A, it can't resume the lease it had in A either, but again would request a new DHCP lease Oct 24 16:37:23 this *could* get handled in dnsmasq/ICD, I guess Oct 24 16:38:44 anyway a connectivity dropout of a few (tens of) seconds never may justify a new DHCP lease Oct 24 16:39:22 look, i can paste all that to their hotline if you like Oct 24 16:39:41 I don't care Oct 24 16:39:44 if they're feeling funny they might escalate the ticket Oct 24 16:39:50 it's your crappy connection Oct 24 16:39:55 and send us an amusing response Oct 24 16:40:23 I'm just telling you that your problems are not exactly normal Oct 24 16:40:39 they pretty much are, here Oct 24 16:43:22 either way i relax while you frantically hit ~. Oct 24 16:43:39 * sono shrugs Oct 24 16:43:48 let's let it pass Oct 24 16:45:05 I'm listing the possible causes so you could investigate further, but I'm not interested in reports of the "it's like this here, and I found a way to automate the 50 keypresses I need to do twice per minute to cope with it" class Oct 24 16:46:39 there just isn't anything to investigate except where you live so ican move there and have a week old tcp connection on my mobile, too. Oct 24 16:46:44 when a carrier network doesn't allow DHCP leases that stay longer than 30s, then there's sth fundamentally wrong with the network. I can't believe anybody would be that stupid. Oct 24 16:47:48 sono: ok, then please spare us your ranting about your problems, when you don't feel like investigation what cause the have Oct 24 16:47:57 wo.o Oct 24 16:48:04 i was ranting? Oct 24 16:48:32 even worse, you suggested mosh to a user that for sure doesn't have same problems you got Oct 24 16:48:34 you went "bullshit" all over me without a second thought Oct 24 16:48:59 ok you know what? i just shut up, Oct 24 17:28:47 DocScrutinizer05: hey :D Oct 24 17:29:06 ah Oct 24 17:29:08 maybe not.... Oct 24 17:29:15 Sysaxed: hey! :-) glad a new battery fixed it Oct 24 17:29:34 maybe not? Oct 24 17:29:47 well, I'm not sure, but it seems like I've fixed my old battery Oct 24 17:30:07 seems like contacts on it were pushed to sides from wear Oct 24 17:30:24 so I picked them with a needle and bent towards the center Oct 24 17:30:36 and it seems like everything works great now Oct 24 17:32:18 yeah, sounds pretty plausible. Particularly when you maybe connected/contacted the battery with other devices or chargers or whatever Oct 24 17:32:43 cleaning the device contact blades also might help Oct 24 17:34:04 I'm not sure if it keeps working(last time it started to fail again on another day), but if I don't come here crying that my n900 is dead, then it is fixed :) Oct 24 17:34:30 install syslog Oct 24 17:36:22 a bq27200.sh 5 >>/media/mmc1/bat.log also comes in handy Oct 24 17:36:50 mmm, what for? Oct 24 17:37:04 to check how battery behaves Oct 24 17:37:15 oh Oct 24 17:37:28 well, doesn't really matter for me while it works Oct 24 17:37:54 If only I get the same problem once again... Oct 24 17:38:13 then you'll miss reference data from the period when it worked Oct 24 17:39:23 maybe don't use 5 (period time) but 30, to not consume too much space and power Oct 24 17:41:54 good point Oct 24 19:34:21 how do I activate sygic map? Oct 24 20:32:19 zemmy: IIRC sygic send a key in an e-mail on purchase. Oct 24 20:32:32 shhh Oct 24 20:32:49 I know... but yet solved Oct 24 20:33:03 anyway thanks Oct 24 20:34:40 solved how? Oct 24 20:36:14 freemangordon: your maemo extras package xkb-data-extkbd eating 2MB on rootfs, what about optification? Oct 24 20:40:05 secret Oct 24 20:40:24 cook secret XP Oct 24 20:45:18 mmm, I would be interested if I reflash my device with SMM10 on, if I could re-install it again. Oct 24 20:52:58 freemangordon: look at this autobuilder error: https://garage.maemo.org/builder/fremantle/gucharmap_2.28.2-1nix1/armel.root.log.FAILED.txt Oct 24 20:53:14 autobuilder failed to prepare build env Oct 24 20:58:53 Pali: weird Oct 24 21:02:57 Pali: seems lynx has never made it to stable Oct 24 21:03:16 So I guess the package is broken Oct 24 21:03:37 why would any package need lynx to get built? Oct 24 21:03:46 NFC Oct 24 21:04:06 freemangordon: what I did was: downloaded source package from extras, increased version, added debian/optify and sent it to autobuilder Oct 24 21:04:18 nothing more, I did not added any dependency Oct 24 21:04:20 me neither, probably autobuilder too Oct 24 21:05:02 Pali: well, SB in autobuilder is hathor, not aphophis as used to be before the migration Oct 24 21:05:11 *apophis Oct 24 21:05:30 that might be the reason Oct 24 21:05:54 Setting up lynx (2.8.6-2.1maemo3) ... Oct 24 21:05:54 update-alternatives: error: alternative path /usr/bin/lynx doesn't exist. Oct 24 21:05:54 dpkg: error processing lynx (--configure): Oct 24 21:06:22 so lynx should be fixed first IMO Oct 24 21:07:05 Pali: I don;t dare to downgrade autobuilder SB :) Oct 24 21:07:32 the gucharmap package should get checked what for it allegedly needs lynx to support make and/or gcc Oct 24 21:08:23 I mean, it#s absolute nogo to try and download sth from internet to autobbuilder, during package build Oct 24 21:08:24 so problem is in lynx package? Oct 24 21:08:30 I will look at it then... Oct 24 21:08:44 and I hope it would fail anyway Oct 24 21:09:08 I pushed more packages to extras-devel which needed optification Oct 24 21:09:22 for the life of mine I can't think of any reasonable use for lynx during building a package Oct 24 21:09:24 to make rootfs more free Oct 24 21:09:45 gtk using lynx when building documentation Oct 24 21:09:59 generate text files from html (or similar) Oct 24 21:10:01 Pali: I hope you thoroughly checked for each single one if it actually may get optified Oct 24 21:10:22 Pali: look at this http://pastebin.com/6wzPMw3j, according to google it is from 2010 Oct 24 21:10:23 DocScrutinizer05: I checked that which are installed on my n900 Oct 24 21:11:25 and I have my own script which sort packages by sum of size of files stored on ubifs rootds Oct 24 21:11:37 Pali: unfortunately google cannot find any references to that pastebin Oct 24 21:11:52 ??? Oct 24 21:11:55 but you're not the first to hit that lynx problem Oct 24 21:13:19 Pali: packages need check if they are needed during boot before they may get optified Oct 24 21:14:04 DocScrutinizer05: do you think packages in maemo-extras could be needed for booting? Oct 24 21:14:48 no those packages should not be needed at boot - before rcS-late (which mounting /opt/) Oct 24 21:14:59 I seem to recall some fool optified PA codecs, which caused epic audio breakdown Oct 24 21:15:24 DocScrutinizer05: those packages are in extras, not in SDK ;) Oct 24 21:15:33 I optified only packages which are not installed by default on clean n900 Oct 24 21:15:35 ( Pali's packages) Oct 24 21:15:50 Pali: that's safe Oct 24 21:15:55 thanks Oct 24 21:16:28 there is one very big, part of ssu (which I did not touched!): posix-locales Oct 24 21:16:29 (at least as long as the packages don't replace other packages that are installed by default) Oct 24 21:16:38 eating 23MB on rootfs Oct 24 21:17:00 yeah, so? Oct 24 21:17:14 maybe we should check if that package is needed and when... Oct 24 21:17:15 known failwale Oct 24 21:17:20 Pali: you can;t optify that, it is part of libc afaik Oct 24 21:17:30 then, ok Oct 24 21:17:31 the package IS needed, even during boot Oct 24 21:17:50 DocScrutinizer05: I do not know that, so this is why I'm asking... Oct 24 21:18:02 it's all locales Oct 24 21:18:09 Pali: maybe ask aapo, he played a lot with libc Oct 24 21:18:38 but i doubt locales can be optified Oct 24 21:18:38 core maemo system component Oct 24 21:18:47 no they can't Oct 24 21:19:01 freemangordon: I see dosbox in cssu-thumb, can you optify it? Oct 24 21:20:02 Pali: me?!? it is there just for the test purposes Oct 24 21:20:15 I just get the source from the repos and recompiled Oct 24 21:20:36 Pali: BTW are you sure it is not optified? Oct 24 21:22:08 yes, dosbox has 1,7MB on rootfs Oct 24 21:22:29 hmm, I guess I can optify the .deb Oct 24 21:23:21 * DocScrutinizer05 blames Nokia's bad-to-the-bones rotten mindset to not FOSS locales Oct 24 21:23:33 which gives us this 23MB abomination Oct 24 21:24:13 Pali: dosbox will have to wait, there are more thing I have to do I don;t have time for Oct 24 21:24:22 ok :-) Oct 24 21:25:00 I mean, it's *TEXT*!!! Oct 24 21:25:25 first of them to report OMAPFB CMA failures to CMA maintainer, but I don;t think I'll have time before Sunday :( Oct 24 21:28:10 jonwil could RE this friggin blob locale.cache 23MB, and split it into original locale text files one for each locale Oct 24 21:28:52 DocScrutinizer05: it is part of libc Oct 24 21:28:57 afaik Oct 24 21:29:06 shouldn't be too hard to write a simple disassembler based on awk or whatever Oct 24 21:29:49 freemangordon: sure, the file format is completely documented Oct 24 21:30:44 it's basically just like shipping text in a tarball for which no untar exists since nobody ever needed such tool Oct 24 21:31:22 the "tar" however exists in source FOSS Oct 24 21:32:17 Nokia "tarred up" the 39some locales into that locales.cache and never published the original files from which that locales.cache got built Oct 24 21:32:23 afaik Oct 24 21:32:25 looks like package is open: http://repository.maemo.org/pool/maemo5.0/free/p/posix-locales/ Oct 24 21:32:33 in maemo SDK Oct 24 21:33:16 posix-locales_2.7.2-1maemo7-binonly0+0m5.tar.gz doesn't exactly look like source package :) Oct 24 21:33:47 but older versions are Oct 24 21:33:47 https://maemo.gitorious.org/maemo-af/posix-locales Oct 24 21:34:41 we're missing the latest version Oct 24 21:35:00 but changes should't be that hard to add Oct 24 21:35:12 Pali: BTW why do you ned more space on rootfs? Oct 24 21:35:17 *need Oct 24 21:35:21 that's not all the plain textx that show up instead of "tana-fi-modest-error-imap" Oct 24 21:35:28 because I had 10MB :-) Now I have 20MB Oct 24 21:35:40 Pali: rm your syslog :P Oct 24 21:36:59 this 23MB blob contains ALL texts of ALL language versions of ALL core apps Oct 24 21:37:29 anyway, /me is out, night guys Oct 24 21:37:39 * DocScrutinizer05 too Oct 24 21:42:57 freemangordon: it looks like posix-locales 2.7.2-1maemo7-binonly0+0m5 is based on git version 2.7.2-1maemo6 Oct 24 21:43:16 Pali: I mean the compressed(?) cache is 23MB in size. The original set of language specific text definitions must be significantly more than this - this is one of the reasons Nokia didn't ship them on device I guess Oct 24 21:43:54 I'm generating that binary files from source files from git Oct 24 21:44:09 and will run diff on binary files if are different... Oct 24 21:44:31 which binary files? Oct 24 21:44:51 which source files? Oct 24 21:44:55 binary from http://repository.maemo.org/pool/maemo5.0/free/p/posix-locales/ Oct 24 21:45:07 and source from https://maemo.gitorious.org/maemo-af/posix-locales Oct 24 21:45:17 again: I don't see any of the textx of any of the apps in those Oct 24 21:45:30 texts* Oct 24 21:45:40 on http://repository.maemo.org/pool/maemo5.0/free/p/posix-locales/ is version 2.7.2-1maemo7-binonly0+0m5 Oct 24 21:45:48 so what? Oct 24 21:45:54 and on https://maemo.gitorious.org/maemo-af/posix-locales 2.7.2-1maemo6 Oct 24 21:46:06 again, how's that related? Oct 24 21:46:24 find source files of package posix-locales Oct 24 21:46:31 for version 2.7.2-1maemo7-binonly0+0m5 Oct 24 21:46:40 look at http://repository.maemo.org/pool/maemo5.0/free/p/posix-locales/posix-locales_2.7.2-1maemo7-binonly0%2b0m5.tar.gz Oct 24 21:46:51 specialy at script update-from-git.sh Oct 24 21:46:58 where are the texts that show up e.g. on microb-menu? Oct 24 21:47:12 these are posix locales Oct 24 21:47:17 not microb locales Oct 24 21:47:34 posix locales are used in coreutils, gnu apps, etc. Oct 24 21:47:35 * DocScrutinizer05 headdesks Oct 24 21:47:51 (maybe in busybox too, do not know) Oct 24 21:48:25 and there are info for LC_NUMERIC Oct 24 21:48:28 LC_TIME Oct 24 21:48:43 and all other info used also on linux desktop... Oct 24 21:48:43 http://repository.maemo.org/pool/maemo5.0/free/p/posix-locales/posix-locales_2.7.2-1maemo7-binonly0+0m5.tar.gz has zilch Oct 24 21:49:09 just the 23MB blob Oct 24 21:49:26 look how that blob is generated Oct 24 21:49:46 from https://maemo.gitorious.org/maemo-af/posix-locales by script update-from-git.sh Oct 24 21:50:07 aha Oct 24 21:51:00 https://maemo.gitorious.org/maemo-af/posix-locales/source/94d56d5480188e718bcace41ebe27f83b00f5a04:locales/de_DE looks pretty sparse Oct 24 21:51:17 I can't find *any* of the texts I'd expect in there Oct 24 21:51:52 it's the plain locale format definitions, no texts Oct 24 21:58:02 /usr/share/hildon/filemanager.desktop: Name=sfil_ap_name. sfil_ap_name is not defined in that maemo.gitorious.org aiui Oct 24 21:58:27 it however comes from /user/lib/locale/locale-archive Oct 24 22:00:37 all those tana_fi_homebutton and sfil_ap_name and what_not_elsefoobar come from those 23MB blob, and I never seen the source for all those definitions. timeless once published his own version of texts, since he wasn't too happy with the job done by Nokia's translators Oct 24 22:01:57