**** BEGIN LOGGING AT Sat Nov 26 02:59:57 2011 Nov 26 03:03:12 o_O gentoo on n900? Nov 26 03:03:56 тАж Nov 26 03:04:29 well, it WAS Nov 26 03:04:34 until something went and corrupted everything x.x Nov 26 03:04:40 :/ Nov 26 03:05:44 I think I need to reformat now Nov 26 03:05:46 sigh Nov 26 03:07:22 compiling on the poor little cpu doesnt sound like fun to me anyway Nov 26 03:07:53 still, crazy stuff... :) Nov 26 03:08:17 (that gentoo runs, not that yours is corrupted...) Nov 26 03:09:13 only linking Nov 26 03:09:16 I use distcc :p Nov 26 03:09:24 ah Nov 26 03:10:04 anyway, bbl Nov 26 04:48:10 luke-jr, disable fsck Nov 26 04:48:16 тАж Nov 26 04:48:19 why? Nov 26 04:48:47 Because ext* assumes you have a harddrive Nov 26 04:48:51 so? Nov 26 04:48:59 without fsck it'll certainly corrupt Nov 26 04:49:14 It actually keeps it amazingly consisten just with ext3+journal Nov 26 04:49:44 On harddrive at powerloss you lose like 512 bytes, or 4k, or the last x * 4k sectors written Nov 26 04:50:03 journal still needs fsck to recover Nov 26 04:50:11 and journal will shorten SSD lifetime Nov 26 04:50:19 On flash you lose that 512b, plus additional 128-512k randomly chosen sectors that get old data.. Nov 26 04:50:55 ew Nov 26 04:50:58 Ideally you'd have a gigantic journal :) Nov 26 04:51:06 well, the journal issues is why I use ext2, not ext3 Nov 26 04:51:16 I don't think it's an issue Nov 26 04:51:28 yes, becasue the journal assumes things get written in order Nov 26 04:52:04 and on boot the kernel rewrites whatever is in the journal, and things tend to work out better Nov 26 04:53:04 even better would be nilfs2 or LogFS.. though nilfs2 would chew through battery Nov 26 04:53:23 and would require backporting, testing, etc Nov 26 04:53:49 yeah there's that :P Nov 26 04:53:55 ext2 worked well for years Nov 26 04:54:11 which makes me suspect it's a booting issue Nov 26 04:54:17 with this new kernel Nov 26 04:54:26 I couldn't boot Maemo with it, either Nov 26 04:54:30 nfc why Nov 26 06:30:29 luke-jr: gps to laptop >> columbus Nov 26 06:30:39 ? Nov 26 06:46:11 4:11 < luke-jr> jacekowski: I want to use the N900's GPS from a full laptop. Nov 26 06:46:21 тАж Nov 26 06:46:46 ok. maybe didn't get u Nov 26 06:51:32 what i meant is that columbus is the package that does that Nov 26 06:53:30 o Nov 26 06:57:18 can't seem to find the code Nov 26 06:59:27 iirc it was developed by alterego Nov 26 07:29:27 ╠√ Nov 26 10:51:47 damnit Nov 26 10:52:05 telnetting to a port should show something if a service is running right? Nov 26 10:56:13 hello. is there a phone that has voice reconigtion, like maemo? Nov 26 11:03:23 voice recognition like maemo? wtf? Nov 26 11:03:41 Macer: depends if it's dropped/denied via firewall rules Nov 26 11:04:12 Macer: try nmap Nov 26 11:22:33 pocket sphinx ? Nov 26 13:07:00 damnit Nov 26 13:07:09 for the life of me i can't find out why smtp won't connect on this synology Nov 26 13:07:13 nor does tls imap :-P Nov 26 13:47:55 M4rtinK: yeah, or even voicetogoog Nov 26 13:48:15 well, seems that'S only on maemo6 for now Nov 26 13:49:09 anyway raptor67682 's question sounds quite upside-down Nov 26 13:51:52 ping X-Fade Nov 26 13:52:08 * DocScrutinizer idly wonders if/when Siri/theGoogle$thing will implement a pull feature, I.E. mic activation from remote, so Siri can listen whenever it thinks it's worth it ;-D Nov 26 13:52:25 * DocScrutinizer moos @ Pali Nov 26 13:52:27 ~seen X-Fade Nov 26 13:52:34 x-fade is currently on #maemo #harmattan #meego, last said: 'Can somebody remind me again how to backup your messages/conversations in harmattan? Where were they stored again?'. Nov 26 13:53:24 Pali, wish you luck Nov 26 13:53:33 indeed Nov 26 13:53:33 hello Nov 26 13:53:38 hey freemangordon Nov 26 13:53:43 hey Nov 26 13:54:39 seems no hen developers left , it is only you Nov 26 13:55:06 where is X-Fade? Nov 26 13:55:11 freemangordon: if you want to chat with H-E-N hardcore kernel hackers you should've joined ##mhd since some days. You ignored all my invites, now paul said >gtg,bbl< some 15min ago Nov 26 13:55:12 who knows Nov 26 13:55:39 sorry, saw the invitations next day Nov 26 13:55:50 i've just have changed my irc client Nov 26 13:56:20 BTW I still don't know which is the best IRC client to use Nov 26 13:56:39 irssi Nov 26 13:56:48 winoze? Nov 26 13:59:44 DocScrutinizer, what was the hen channel? Nov 26 14:13:57 DocScrutinizer: now I found some bq2415x kernel driver from Nokia: https://lkml.org/lkml/2011/3/9/176 Nov 26 14:14:18 DocScrutinizer: do you know about it? Nov 26 14:15:11 yes, I heard about it ans somewhat looked into it Nov 26 14:15:44 haven't completely analyzed it but it seems it still relies on undocumented interaction with some bme shit Nov 26 14:16:21 here is source code: https://lkml.org/lkml/2011/3/9/178 Nov 26 14:16:49 I see that this only export functions via /dev/ file Nov 26 14:16:59 Pali, did you tyr last patch i've sent? Nov 26 14:17:03 try* Nov 26 14:17:26 freemangordon: sorry not yet Nov 26 14:17:32 ok Nov 26 14:19:05 Now I'm compiling it Nov 26 14:19:28 Pali: also on a 30s glance to that patch I spot things like "bq2415x_chg_low(true)" which again make me think Nokia only did the bare minimum they thought they might find useful for their special usecase - as usual Nov 26 14:19:33 bq2415x_chg_low(true) Nov 26 14:19:43 freemangordon: did you work on ham already? Nov 26 14:19:52 no Nov 26 14:20:15 case BQ2415X_CHG_NORM: res = bq2415x_chg_low(false); Nov 26 14:20:36 i'm still fighting kp stuff Nov 26 14:21:03 friggin sh*t I don't want to choose from charge-low and charge-normal, I want to control charge current to the last digit that bq24150 allows Nov 26 14:23:36 or compare the implementation of >>bq2415x_set_current_limit(struct regulator_dev *rdev, int min_uA, int max_uA)<< with what my suggested API specifies Nov 26 14:24:26 and I bet the whole thing doesn't even touch boostmode Nov 26 14:25:56 it is not better to fix that existing nokias driver instead writing new? Nov 26 14:26:35 The existing hostmode stuff on the n900 is fragmentary. Nov 26 14:30:45 Pali: no, see: >>MODULE_DESCRIPTION("TI BQ24153/6 one-cell Li-Ion charger driver");<< -- We got bq24150 which is quite a bit different it seems Nov 26 14:31:25 though, when I think about it, I didn't find that much diffs in datasheet, still this driver bewilders me Nov 26 14:31:50 * jonwil wishes he could decide what n900 thing to work on Nov 26 14:31:58 anyway this driver is confusing, and doesn't deal *at all* with vboost Nov 26 14:34:11 I think it is more simply to add vboost to that driver instead writing all from sketch Nov 26 14:37:39 Pali: and looking through the source from start to end it appears to me the Nokia driver doesn't implement any additional intelligence so it still needs all the bme cruft like it does now. My driver spec OTOH is meant to implement an autonomous handler that only needs a ONE-TIME config e.g. during boot and then could happily run along charging under same unchanged situation for infinite time without any bme retriggering watchdog and similar Nov 26 14:37:41 cruft. The idea is to allow concurrent userland/other systems to simply set up charger to meet the situation (host powered charging, fastcharger charging, even vboost) and then leave it alone Nov 26 14:38:20 while the driver itself controls and maintains the bq24150 chip Nov 26 14:38:28 I.E. tickles watchdog etc Nov 26 14:38:39 yes, I know Nov 26 14:38:47 Nokia driver fails completely on this idea Nov 26 14:39:55 can bq chip tell us if charger or power usb cable is connected? Nov 26 14:40:10 it can't Nov 26 14:40:25 you need other kernel based and useland cruft to do so Nov 26 14:40:37 so we need additional checks (from other driver?) to check that Nov 26 14:40:49 so auto charge should be in other kernel module Nov 26 14:41:02 not in bq2415x*.ko Nov 26 14:41:06 well, it can detect VBUS 5V, it can to some degree check if PHY chip things there'S D+/- short Nov 26 14:41:56 you can set up bq24150 to start charging @ XXXmA whenever there'S VBUS 5V applied Nov 26 14:42:22 for more nifty stuff we need support from way more complex systems, and that's userland task Nov 26 14:42:44 like doing USB ENUM to learn if PC can power 500mA Nov 26 14:42:53 or if it is a PC at all Nov 26 14:43:45 or rather a dumb charger with missing D+/- short (->100mA max, according to Nokia's conservative and standards conforming policy) Nov 26 14:44:17 ok Nov 26 14:44:29 for all this other systems are involved, as charger driver can't do e.g. ENUM Nov 26 14:44:54 what is isp1704 charger? Nov 26 14:45:14 wow Nov 26 14:45:15 lol, 1704 is USB PHY detecting D+/- short Nov 26 14:45:22 in upstream kernel it is something for rx51: http://lwn.net/Articles/400910/ Nov 26 14:45:25 I'm definitely noticing a huge performance boost switching to hardfp on N900 Nov 26 14:45:45 jonwil: kernels Nov 26 14:46:56 Pali: so isp1704_charger probably means the PHY has detected D+/- short and thus the plugged in cherging device is considered to be a fastcharger Nov 26 14:47:55 Pali: bq24150 even has a hw signal line to talk to PHY directly regarding this Nov 26 14:48:14 alas the PHY is not always consistent in asserting this signal Nov 26 14:49:16 as PHY msutn't try to detect D+/- short when an active USB session is ongoing, as it would break data transfer (check for Data+ to Data- short, go figure!) Nov 26 14:49:41 ok Nov 26 14:52:44 otoh an active USB session means there's definitely no D+/- short ;-D Nov 26 14:53:18 you see charging over USB is quite entangled a bit with generic USB driver though Nov 26 14:53:55 which is why we need to get rid of BME for hostmode, basically Nov 26 14:54:59 Pali: for bq24150 proper cherging&VBOOST versatile kernel driver all this doesn't matter too much Nov 26 14:55:51 my specs specify a *sane* interface (I hope) that allows to offload the annoying bits to the driver while keeping all the nifty finegrained control to userland Nov 26 14:57:08 I even planned for raw register access for "dumb brute force access" the bme alike style Nov 26 14:59:13 so with a very simple wrapper to change direct I2C-writes/reads to /sys/bq24150/ clean reads/writes bme could still work on this driver, though would basically disable all the smartness in that driver - but well that'S what bme likes to do Nov 26 15:02:08 a "smarter bme" would just write once to things like /sys/bq24150/... USB-curlim,chterm_ena,chterm_thld,CC,CV... and then echo auto >watchdog Nov 26 15:02:27 then let kernel driver and chip do their thing Nov 26 15:03:01 you could kill smarter-bme and uninstall it and charging would continue happily, after that Nov 26 15:03:48 next time smarter-bme would be needed was on next plug-in of charger/PC Nov 26 15:04:10 to check type of "charger" and again write once to above mentioned sysnodes Nov 26 15:04:17 * jonwil wonders why Nokia chose to ignore most of the functionality of the charger chip and write all sorts of funky code in BME instead Nov 26 15:04:32 jonwil: heritage Nov 26 15:04:35 because rapuyama used to do charging Nov 26 15:04:35 legacy Nov 26 15:04:44 yes, exactly Nov 26 15:04:47 and rapuyama has same code in it as big parts of BME Nov 26 15:04:59 even on n900 Nov 26 15:05:45 they simply ported the age-old bme code that basically was there on 6210 phones, to N900 Nov 26 15:06:10 and ignored the fact they now got a smart charger chip that could do all that nasty stuff for them Nov 26 15:07:16 probably not even Nokia did that, I suspect they hired a subcontractor that does this for them since 1999 Nov 26 15:07:37 lol Nov 26 15:07:48 well, lol but no kidding Nov 26 15:07:55 why 99? Nov 26 15:08:24 that's when I seen same shite in Nokia 6210 monitor, nder "battery" tab Nov 26 15:08:53 maybe it pedates this phone ;-D Nov 26 15:09:11 LoL :D Nov 26 15:09:13 * jonwil wonders if there is an easy way to add a "switch between 2G and 3G" option to the status bar menu Nov 26 15:09:26 there was applet for that Nov 26 15:09:34 any idea what it is called? Nov 26 15:09:42 nope Nov 26 15:09:47 i don't have n900 anymore Nov 26 15:09:48 damn :( Nov 26 15:09:52 i have N8 with symbian Nov 26 15:10:01 тАж Nov 26 15:10:02 fail Nov 26 15:10:05 If one already exists, saves me reinventing the wheel Nov 26 15:10:08 * DocScrutinizer beats jacekowski around a little Nov 26 15:10:17 but if I cant find it, it doesn't help me Nov 26 15:10:28 search for 3g Nov 26 15:10:49 3g2g-mode-selection-applet - Switch between the three network modes (3G, 2G and Dual) from the status menu Nov 26 15:10:52 2G/3G switcher applet - or sth like that Nov 26 15:11:30 I'm using it all the time Nov 26 15:11:55 found it, now I need to add it to my phone :) Nov 26 15:12:04 recently 2G(!) became abysmally poor here so I don't use it that much anymore Nov 26 15:13:50 hen a year ago I had to switch to 2G when entering my fav pub, or otherwise connection is all shaky and introduces dropouts, on 3g - now it's exactly the other way round Nov 26 15:14:40 3G sort of works now in that pub, but 2G is unbearable Nov 26 15:15:17 * DocScrutinizer wonders if they got a good price for the surplus 2G equipment they sold... fools Nov 26 15:15:48 * jonwil got a notice in the mail calling for public consultation on the erection of a new cell tower in his area Nov 26 15:16:16 duh, I read "erection" and thought "meh, SPAM" Nov 26 15:16:27 ;-P Nov 26 15:17:01 time for 12648430 тШХ Nov 26 15:17:14 HELO gentlemen Nov 26 15:17:24 hmmm so should I work on wlan stuff, h-i-m stuff, osso-backup stuff, libbmeipc/hald-addon-bme stuff, sysinfo/libsysinfo stuff, systemui stuff, connectivity ui stuff or maps stuff...? Nov 26 15:17:24 EHLO RST38h Nov 26 15:18:32 hald-addon-bme sounds like it would go nicely along with Pali 's / my bq24150 driver stuff Nov 26 15:19:33 it would basically be a 1:1 clone of hald-addon-bme plus a 1:! clone of libbmeipc and a dev package for libbmeipc Nov 26 15:19:39 libbmeipc is open Nov 26 15:19:51 no its not, not on the N900 Fremantle Nov 26 15:19:58 n900_libbme on meego is open Nov 26 15:20:08 is not same? Nov 26 15:20:10 but that's different Nov 26 15:20:13 replace hald-addon-bme, bme, and bq27200/bq24150 .ko in one pkg and we get a sane bme free kernel that kinda just works Nov 26 15:20:46 jonwil: bq24150.ko has no way to support libbmeipc Nov 26 15:20:54 hald-addon-bme is easy to write - we can use bq27200 kernel driver Nov 26 15:21:06 There are 3 places that talk to libbmepic other than hald-addon-bme. One is nokia-closed-pulseaudio-blobs, one is dsme_thermalsurface and the last one is testserver Nov 26 15:21:12 you need to patch hald-addon-bme to directly use the sysfs nodes Nov 26 15:21:37 jonwil: :nod: Nov 26 15:21:39 we do not need bmeipc.... all applications using hal Nov 26 15:21:40 its fairly easy to make something that exposes the same outside-world interface as hald-addon-bme but talks directly to kernel Nov 26 15:22:00 nokia-pulseaudio-blobs cant be changed easily Nov 26 15:22:03 since they are closed Nov 26 15:22:11 dsme_themalsurface is open though Nov 26 15:22:19 and testserver doesnt seem to be used normally Nov 26 15:22:41 for PA we need to figure WTF it does, thermal is kinda obvious what it does and we need to do similar thing in "jrbme" or bq24150.ko, and testserver, oh my Nov 26 15:22:41 dsme thermal can be patched Nov 26 15:23:14 I posted to mailing list a while back with info on this Nov 26 15:23:20 including info on what bme call pulseaudio is making Nov 26 15:23:44 is not nokia pulseaudio open in meego? Nov 26 15:23:46 time for 12648430 тШХ Nov 26 15:23:49 grrr Nov 26 15:24:26 nope, nokia-pulseaudio-blobs are still closed in meego Nov 26 15:24:31 smaller blobs but still closed Nov 26 15:24:45 aha, my post is titled "Information on the interface to hald-addon-bme (so it can be replaced with something that bypasses BME and talks to the kernel directly)" Nov 26 15:24:49 google for that to find it :) Nov 26 15:25:05 the only thing pulseaudio cares about is the temprature value Nov 26 15:25:26 the source code for dsme-thermalobject-surface should show you exactly what bme calls are being made Nov 26 15:25:35 pulseaudio and temperature?? WTF?? Nov 26 15:25:36 and I can confirm that pulseaudio seems to only be using the same values Nov 26 15:26:03 why pulseaudio need temperature? Nov 26 15:26:06 no idea Nov 26 15:26:40 so yeah I think given the info I posted in that mailing list post, there isn't that much in hald-addon-bme or libbmeipc that is worth cloning or figuring out :) Nov 26 15:27:24 speaker protection or sth weird Nov 26 15:28:13 I could say "they need it for ultra-smart echo cancellation, as speed for sound is varying with air temperature" but LOL Nov 26 15:28:31 speed OF sound Nov 26 15:28:40 so should I work on wlan, h-i-m, osso-backup, sysinfo, systemui, connectivity ui or maps? Nov 26 15:29:03 h-i-m Nov 26 15:29:08 :) Nov 26 15:29:22 do you know what h-i-m means in this case? :P Nov 26 15:29:39 hm, hoping it is hildon-input-method Nov 26 15:29:45 yes it is Nov 26 15:29:59 what specifically do you think I should be doing on hildon-input-method? Nov 26 15:30:07 renderer Nov 26 15:30:31 hildon-input-method library is open, only plugins are closed, sure? Nov 26 15:30:47 the library is open Nov 26 15:31:00 but plugins and renderers and stuff are closed Nov 26 15:31:09 including the hildon vkb renderer widget Nov 26 15:31:27 and the library that handles the virtual keyboard layouts Nov 26 15:31:32 and more besides :( Nov 26 15:31:38 jonwil, btw there is almost working (AFAIK) example of renderer somewhere in maemo application framework Nov 26 15:31:52 freemangordon: a clean hald-addon-bme implementation that doesn't use libbmeipc but rather directly uses /sys/bq27200/* and /sys/bq24150/* would quite be worth it :-) Nov 26 15:31:59 err, Nov 26 15:32:03 jonwil: ^^^ Nov 26 15:32:07 jonwil: on http://wiki.maemo.org/Fremantle_closed_packages I see 3 packages: hildon-input-method-configurator hildon-input-method-plugins-western hildon-input-method-widgets Nov 26 15:32:09 freemangordon: ETAB Nov 26 15:32:20 np Nov 26 15:32:33 The info needed for such a hald-addon-bme is in my mailing list post Nov 26 15:34:06 jonwil: sure, I thought you asked what to *work on*, not what to investigate and publish docs ;-) Nov 26 15:34:23 but yeah, I got it now Nov 26 15:34:29 except that I dont want to work on kernel stuff :P Nov 26 15:34:48 hald-addon-bme is all but kernel Nov 26 15:35:14 it talks to kernel though :P Nov 26 15:35:31 I know nothing about kernel battery driver, best to let someone who does write new hald-addon-bme :P Nov 26 15:35:33 even access to bq24150/27200 is a userland plain read() Nov 26 15:36:34 jonwil, damn libhildon-im-vkbrenderer.so.3.0.3 is 33k of size, shouldn't be that hard to RE ans rewrite it Nov 26 15:36:40 fopen("/sys/platform/*/*/bq27200/whatever") Nov 26 15:36:43 and* Nov 26 15:36:51 oh and for hildon-input-method stuff, the closed packages are hildon-im-common-virtual-settings , hildon-im-fkb , hildon-im-keyboard-assistant, hildon-im-keyboard-assistant-scv , hildon-im-plugin-base-settings , hildon-im-virtual-keyboard-layouts , hildon-input-method-configurator , hildon-input-method-plugins-western , hildon-input-method-widgets Nov 26 15:36:58 plus also ezitext* packages Nov 26 15:37:07 and imengines-ezitext Nov 26 15:37:24 we don't need all that stuff for portrait keyboard Nov 26 15:37:27 ezitext are only spelling database? Nov 26 15:38:22 well as far as hildon-input-method work goes, my first target is to finish cloning western_common_vim_settings.so, hildon-input-method-configurator.bin and possibly libcptextinput.so Nov 26 15:38:42 which will mean the list of valid hwkbd layouts can be un-hard-coded Nov 26 15:38:54 as those 3 binaries are what contain the hardcoded lists Nov 26 15:39:06 as for libhildon-im-vkbrenderer, its not as simple as it looks :P Nov 26 15:39:13 hmm, how then ukeyboard adds my language? Nov 26 15:39:19 Bulgarian Nov 26 15:39:35 the hardcoded lists seem to only affect the settings Nov 26 15:39:47 h-i-m and its plugins just use whatever value is in gconf Nov 26 15:39:56 so, why then one should care? Nov 26 15:40:04 if there is a workaround Nov 26 15:40:53 because normal users dont want a hack, they want to be able to pick their new layouts from the dropdown with all the rest Nov 26 15:41:53 sure, but ukeyboard allows exactly that without any hackery Nov 26 15:42:00 AFAIK Nov 26 15:42:14 well I am probably going to clone it anyway :P Nov 26 15:42:32 it uses standart .vkb files Nov 26 15:42:42 western_common_vim_settings.so isn't rocket science although I am close to the point where I say "forget about cloning it exactly and just write something that does the same thing" Nov 26 15:43:33 jonwil: in hildon welcome (screen to choose time&date + language at boot first time) is propably hardcoded languages... I cannot see here additional my (slovak) lang, but in control panel is visible Nov 26 15:44:07 it is true that language choose box/menu is hardcoded or only bad configuration? Nov 26 15:45:14 This is really annoying that after remove and insert battery I must start phone 2 times (one to choose english and then in control panel change to other)... Nov 26 15:45:28 * jonwil looks Nov 26 15:46:12 ok, hildon-welcome.launch is not making any GTK calls Nov 26 15:46:15 its only making some X calls Nov 26 15:46:25 plus some gconf calls Nov 26 15:46:31 I mean gstreamer calls Nov 26 15:46:34 and glic Nov 26 15:46:36 glib Nov 26 15:46:38 and dbus Nov 26 15:46:55 its asking mce to turn on the display with req_display_state_on Nov 26 15:47:15 thats all it does for dbus Nov 26 15:47:45 I do not know if it is app with name hildon-welcome Nov 26 15:47:55 hmmm maybe its not Nov 26 15:48:00 in fact I dont think it is Nov 26 15:49:06 jonwil, seen that http://maemo.gitorious.org/fremantle-hildon-desktop/hildon-input-method-plugins-example ? Nov 26 15:49:37 yep, I seen that, not very helpful with the stuff I am doing though because of how simple it is :P Nov 26 15:50:43 cssu stable is out: http://talk.maemo.org/showthread.php?p=1129261#post1129261 Nov 26 15:50:51 good, good Nov 26 15:50:55 * merlin1991 goes to write the maemo-dev crosspost Nov 26 15:51:19 jonwil, I am going to RE libhildon-im-vkbrenderer Nov 26 15:51:30 ok, you do that Nov 26 15:51:32 if you think you can Nov 26 15:51:42 at least I will try Nov 26 15:52:03 go ahead and use my libimlayouts header file if you like Nov 26 15:52:19 Not I am much in GTK Nov 26 15:52:30 Not that* Nov 26 15:52:45 Time to learn something new Nov 26 15:54:39 aha, osso-startup-wizzard is the app I want Nov 26 16:00:46 anyone heard of clutter? Nov 26 16:01:03 merlin1991: \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ \o/ Nov 26 16:01:17 thanks doc :) Nov 26 16:01:25 I already thought nobody cared ;) Nov 26 16:01:30 * luke-jr ponders if a decent WM is possible on N900 given its practical lack of 3D accel Nov 26 16:02:29 Compute MIPS/framebuffer-size. Nov 26 16:02:33 oh and pali, check files in /usr/share/osso-startup-wizard, those seem to be where it gets its config from Nov 26 16:02:36 It's >>>1 Nov 26 16:03:21 1.5 kIPS per pixel Nov 26 16:06:10 jonwil: thanks! Nov 26 16:07:19 bah this is too hard to follow, its using clutter (which I know nothing about), something called cogl (also unknown), plus a bunch of hard-to-follow C++ stuff like std::string and iostreams Nov 26 16:07:33 its also using libcityinfo Nov 26 16:07:44 anybody has experience with a good 1A car charger to use with N900? Nov 26 16:08:07 lots of dbus calls too Nov 26 16:08:33 * trumee only has experience with Nokia DC-10 charger which has a very short cable and useless now. Nov 26 16:09:04 its also using libi18n_locale_resolve_list_predefined Nov 26 16:09:06 whatever that is Nov 26 16:09:35 trumee: I grabbed a cheap dual-USB-port one off NewEgg Nov 26 16:09:49 luke-jr: 1A? Nov 26 16:09:53 trumee: 1A & 2A Nov 26 16:10:08 no cable, you plug your normal USB one into it Nov 26 16:10:21 luke-jr: cool. i need to buy one in UK Nov 26 16:10:51 aha, that function comes from libi18n-locale-resolver0 which is closed Nov 26 16:11:06 is there a safe way to preemptively prevent my USB port from breaking? Nov 26 16:11:19 merlin1991: nah, this will become THE BIG NEWS over the next 48h Nov 26 16:11:27 GeneralAntilles: see ^^^ Nov 26 16:11:27 luke-jr: what is the use of 2A port? Nov 26 16:11:32 GeneralAntilles: for MWKN Nov 26 16:11:45 trumee: I guess iPads use it or something. I just figured a 2nd port would allow me future expansion Nov 26 16:11:47 GeneralAntilles: that'S definitely next monday's headline Nov 26 16:13:06 lmao Doc on a newsrampage :D Nov 26 16:13:22 hmmm, the data in /usr/share/i18n_locale_resolver seems interesting (but in an undocumented, uneditable binary format) Nov 26 16:13:34 luke-jr: is it ok to plug N900 into the 2A charger? Nov 26 16:13:48 trumee: can't hurt. Nov 26 16:14:21 luke-jr: found something, http://www.amazon.co.uk/Two-Port-2-1A-USB-Charger/dp/B005JCPZ22/ref=sr_1_1?ie=UTF8&qid=1322323800&sr=8-1 Nov 26 16:15:09 aha, clutter is a drawing library Nov 26 16:15:25 i.e. its used by osso-startup-wizard because the latter is a fullscreen app and doesnt need to use GTK Nov 26 16:15:52 although clutter still has a dep on gtk Nov 26 16:19:11 avoiding GTK is impossible now :/ Nov 26 16:19:19 even BlueZ and pkg-config use GTK Nov 26 16:22:55 bluez only requires it for the pin-input widget, which is just ridiculous... Nov 26 16:24:53 r00t|home: wtf? srsly? Nov 26 16:24:58 why doesn't it use pinentry? Nov 26 16:25:13 at least that's the issue i had with bluez on notebooks Nov 26 16:25:50 there is a cli-tool included, but debian only delivers the source, no binary of it, and the package depends on the whole gui cruft Nov 26 16:26:00 x.x Nov 26 16:26:03 I use Gentoo anyhow Nov 26 16:26:39 you seem wrong, nor bluez nor bluez-tools depend on gtk Nov 26 16:26:50 bluez-gnome does, obviously Nov 26 16:27:34 me? maybe they changed it by now, but it used to be that way Nov 26 16:29:09 and pkg-config only requires glib Nov 26 16:42:20 DocScrutinizer, I was going to stick it on the Miscellaneous page. Nov 26 17:07:53 GeneralAntilles: well, it's finally an "official" community repo for maemo recommended to *everyone* Nov 26 17:08:03 :P Nov 26 17:09:27 basically we could even expect from Nokia to include concerns about compatibilty to CSSU-S for any further (obviously never coming) update Nov 26 17:10:41 for N900 users today is a historic date, a quantum leap (though I hate the term as it's basically wrong) Nov 26 17:11:40 Err - I missed most of todays discussion - wut? Nov 26 17:12:19 CSSU-Stable is officially out in the wild Nov 26 17:12:58 Ah! Nov 26 17:18:25 freemangordon offline?? DANG! Nov 26 17:22:57 yawn Nov 26 17:23:08 as if CSSU has any potential Nov 26 17:23:10 too much closed Nov 26 17:23:23 you can't maintain something without source Nov 26 17:37:13 luke-jr: maybe it is like quake... they are just waiting a little while before releasing it all Nov 26 17:37:23 nonsense Nov 26 17:37:57 as it stands though regardless of all that... maemo is probably still the best portable linux distro you can use on a device with a phone built in Nov 26 17:38:33 the hildon closed bits are probably the worst thing nokia has closed right now. but i'm no dev so i don't know Nov 26 17:39:52 i mean if you can show me another linux distro on a device which allows for some pretty good comm i'd love to see it Nov 26 17:40:02 but it is obviously not gentoo :) Nov 26 17:59:32 luke-jr: defeatist! we got some fixes to "closed" stuff in CSSU, we even got or are about to get true rewrites of other closed bits, for the mere benefit of openness, as being closed is a bug in itself as well Nov 26 18:01:17 Wish somebody'd port Swipe Nov 26 18:03:36 swipe? swype? Nov 26 18:03:40 Fuck that Nov 26 18:03:41 Oh wait, swype or swipe? Nov 26 18:03:50 swype is horrible, i tried it once Nov 26 18:03:51 swipe is harmattan's ui Nov 26 18:03:52 _once_ Nov 26 18:04:14 It's a nice concept, and ou Nov 26 18:04:22 You do type faster Nov 26 18:04:30 But you also spend more time correcting mistakes Nov 26 18:04:38 how does swipe improve typing speed? Nov 26 18:04:56 swype Nov 26 18:04:59 It's easier to reach a radius than the letter itself Nov 26 19:02:05 jonwil, ping Nov 26 19:02:47 ~seen jonwil Nov 26 19:02:54 jonwil <~jonwil@27-33-137-199.static.tpgi.com.au> was last seen on IRC in channel #maemo, 2h 47m ago, saying: 'although clutter still has a dep on gtk'. Nov 26 19:06:09 http://www.mobilebusinessbriefing.com/articles/nokia-shares-slip-on-weak-windows-phone-estimates/19742/ Nov 26 19:06:17 omg... 500K? Nov 26 19:06:26 they're going under :) Nov 26 19:07:51 yeah.. Nov 26 19:08:08 freemangordon: paul is back in #mhd Nov 26 19:08:21