**** BEGIN LOGGING AT Fri Jan 25 02:59:59 2013
Jan 25 04:50:13 <jonwil> does anyone know if there is a mirror anywhere of the nokia-binaries repo?
Jan 25 04:50:17 <jonwil> the SDK one
Jan 25 05:07:47 <kerio> jonwil: that would be akin to a ssu/apps mirror
Jan 25 05:08:00 <kerio> so... probably yes, but not public
Jan 25 05:09:42 <kerio> aww, the "superbattery" calibrated to 1411
Jan 25 05:09:49 <kerio> vi____: you liar
Jan 25 05:14:53 <kerio> i mean, it's still fairly good
Jan 25 05:15:31 <kerio> but it's probably not worth it to switch from the original bl-5j at ~1300
Jan 25 05:17:55 <romaxa> freemangordon: btw you said about memory overhead... browser on N900 has it's own browserd instance, which is only sharing disk space and libxul.so loaded into memory
Jan 25 05:26:39 <jonwil> great, bugs.maemo.org is running slow
Jan 25 05:26:48 <jonwil> or not working
Jan 25 05:34:51 <jonwil> is bugs.maemo.org slow for anyone else/
Jan 25 05:34:53 <jonwil> ?
Jan 25 05:35:15 <thedead1440> jonwil: its connecting for me but after that its a never-ending wait; waited for 4mins still no luck
Jan 25 05:35:27 <jonwil> ok
Jan 25 05:37:20 <jonwil> ok, so, is there anything I can usefully do to help with browser stuff?
Jan 25 06:10:30 <chx> hi. i posted my question here http://forum.xda-developers.com/showthread.php?p=37154781#post37154781 -- basically i would like to hook an audio modem to the n900. is that ... possible?
Jan 25 06:11:21 <ds3> to do what?
Jan 25 06:16:29 <chx> so that i can send a command to the n900 to open the building door via the existing phone system :)
Jan 25 06:17:32 <ds3> so you really are looking for way to send audio progammatically over a voice call?
Jan 25 06:19:41 <chx> i want to make a voice call
Jan 25 06:19:50 <chx> a very short one, actually.
Jan 25 06:20:05 <chx> just pick up the phone, press the button six, over.
Jan 25 06:20:18 <ds3> think that is just software
Jan 25 06:20:33 <chx> well, you need to solve the "pick up the phone" problem somehow
Jan 25 06:20:54 <chx> playing the six, sure that's simple
Jan 25 06:21:25 <chx> i mean, http://jetcityorange.com/dtmf/DTMF-6.mp3 play this, done
Jan 25 06:21:28 <ds3> isn't that a bunch of dbus commands?
Jan 25 06:26:47 <chx> what...?
Jan 25 06:27:01 <chx> how do you plan to phsyically connect to an RJ11 connector?
Jan 25 06:27:34 <kerio> don't you have an internet connection at home?
Jan 25 06:28:41 <chx> I failed at communicating through what I wanted to do
Jan 25 06:28:54 <chx> Yes, I have an Internet connection
Jan 25 06:29:38 <chx> I want to use it to command a device (perhaps the n900 because that's the most open device with wifi i own) to open a door. The door is remote controlled via a traditional phone.
Jan 25 06:29:52 <kerio> and there's no other way to control the door?
Jan 25 06:32:32 <chx> nope
Jan 25 06:32:40 <chx> this is a 20 floor highrise
Jan 25 06:34:09 <kerio> does the door actually have its own phone line?
Jan 25 06:36:35 <chx> it's hooked into the building pbx
Jan 25 06:37:19 <chx> when you ring someone from outside you dial via a normal phone system
Jan 25 06:37:33 <chx> theoretically i could forward the door buzzer to my cellphone -- the only reason I cant do that is that i cancelled my landline cos it was redundant and expensive
Jan 25 06:37:45 <kerio> oh, so you have to call from outside?
Jan 25 06:37:49 <kerio> isn't that also expensive?
Jan 25 06:38:47 <chx> it uses the house pbx, so it's free
Jan 25 06:39:20 <chx> anyways, i might give this up, get a raspberry pi, a powered usb hub, plug in a wifi stick, i have plenty, buy a usb modem for a few pennies and be done
Jan 25 06:40:12 <kerio> that's what i was going to suggest, yeah
Jan 25 06:40:19 <kerio> or just a computer that you leave turned on
Jan 25 06:40:44 <chx> there's absolutely no space for that
Jan 25 06:41:06 <kerio> you have no computers at home? :o
Jan 25 06:41:30 <chx> i have four, and so what?
Jan 25 06:41:38 <chx> they are nowhere near the phone jack
Jan 25 08:10:27 <Pauly1210> hi
Jan 25 08:16:13 <jonwil> hi
Jan 25 08:24:28 <pablomobile> Hi
Jan 25 09:33:57 <kgu> How come I can't change repository url for an existing repository? (want to change it to the mirror)
Jan 25 09:43:05 <kerio> kgu: it should be made more obvious, but those catalogues are not meant to be touched by users
Jan 25 09:43:43 <kerio> disable it and add a new one
Jan 25 09:43:44 <kgu> so I will have to disable it and a new one for the mirrors?
Jan 25 09:43:49 <kerio> but it won't work properly in HAM
Jan 25 09:43:55 <kerio> for upgrades, at least
Jan 25 09:44:00 <kgu> thx :)
Jan 25 09:44:10 <kerio> unless you use an apt-mirror one
Jan 25 09:44:28 <jacekowski> i've got apt-mirror repo
Jan 25 09:45:07 <kerio> the solution is not to use apt-mirror, the solution is to stop using HAM
Jan 25 09:47:18 <kgu> really? Thought that you had to use HAM, especially for upgrades..
Jan 25 09:48:17 <kerio> kgu: if you're asking if you can use apt-get to do system upgrades, you can't use apt-get to do system upgrades
Jan 25 09:48:58 <kgu> ok, but otherwise you recommend using apt-get for program installations?
Jan 25 09:49:22 <kerio> kgu: if you're asking if you can use apt-get to do program installations, you can't use apt-get to do program installations
Jan 25 09:49:29 <kerio> but it should be safe
Jan 25 10:01:25 <Hurrian> kerio, IIRC you can simply "apt-get install mp-fremantle-community-pr"
Jan 25 10:01:47 <Hurrian> since that's 90% of what HAM does, minus the slowness
Jan 25 10:03:07 <Lava_Croft> The best way to upgrade is to just use HAM and be busy with something else at the same time
Jan 25 10:03:15 <Lava_Croft> That way, the slowness of HAM will elude you somewhat
Jan 25 10:17:35 <kerio> Hurrian: ofc
Jan 25 10:22:22 <Estel_> ~seen Pali
Jan 25 10:22:25 <infobot> pali <~pali@Maemo/community/contributor/Pali> was last seen on IRC in channel #maemo, 10h 30m 28s ago, saying: 'n900 has gnu userspace, so low level can be also tar -czf '.
Jan 25 10:23:01 <Estel_> freemangordon,  Pali is one to pester for bugs in bits for userspace bme replacement?
Jan 25 10:23:06 <Estel_> yep?
Jan 25 10:23:41 <Estel_> asking, because I was out of loop, and so goes on bla bla
Jan 25 10:23:52 <Estel_> = I have no idea who is working on it now
Jan 25 10:24:55 <Hurrian> Estel_, Pali is the replacement BME maintainer
Jan 25 10:25:13 <Estel_> thanks - nice to see you again Hurrian, btw :)
Jan 25 10:25:19 <Hurrian> it is hosted on his site after all
Jan 25 10:25:38 <Hurrian> Estel_, surprised to see you're back after some time out of the loop
Jan 25 10:26:08 <Estel_> I'm always back, such mere things as politics won't scare me off ;)
Jan 25 10:26:45 <Estel_> seriously though, had some serious things in rl to take care of... Methinks I need to apologize some people for lack of expected contact
Jan 25 10:27:13 <Estel_> good new is that I'm testdriving pre-pre-pre-pre alpha version of N900 body replacemebnt, since few days
Jan 25 10:27:59 <Estel_> it does have any useful things on exterior side - basically, it looks like ugly piece of alu - but, inside, it got almost all parts that need to cooperate with motherboard
Jan 25 10:28:19 <Estel_> and motherboard isn't sparking on contact anymore =D
Jan 25 10:29:16 <jonwil> BME replacement is a great thing for N900
Jan 25 10:29:35 <jonwil> now if only we could replace some other things too... :P
Jan 25 10:30:10 <jonwil> All my recent efforts at reverse engineering have so far come up with nothing useful
Jan 25 10:52:45 <user05> when to expect repository.maemo.org to be up again? Anyone knows it?
Jan 25 11:03:31 <thedead1440> user05: the /topic has the link to a TMO thread where updates are so seeing that you can see work is on-going
Jan 25 11:07:27 <user05> thedead1440: 10k u
Jan 25 11:21:59 <Aoyagi_joytop> kerio: So it was of course bogus. The phone wasn't "as good as new", it had wobbly display, warranty was 6 months, not 24. And the keyboard was Spanish. :D
Jan 25 11:22:07 <kerio> :D
Jan 25 11:50:57 <Estel_> jonwil,  and what you've tried to reverse-engineer, lately?
Jan 25 11:51:57 <jonwil> some ICD stuff, some wifi stuff, some GPRS stuff, a whole bunch of connectivity UI bits
Jan 25 11:53:09 * Estel_ nods
Jan 25 11:53:17 <Estel_> where exactly you've meet a wall?
Jan 25 11:53:34 <Estel_> lack of participation from others, probably?
Jan 25 11:54:14 <jonwil> Just cant seem to get anywhere because of undocumented libraries, undocumented APIs, unknown data structures and the fact that ARM ASM is a LOT harder to reverse engineer than x86
Jan 25 11:55:08 <jonwil> and yes, lack of interest from other people too
Jan 25 11:56:54 <Estel_> I remember your mails to mailing list
Jan 25 11:57:23 <Estel_> would love to help if not for almost complete lack of coding skills, let alone RE
Jan 25 11:57:50 <SpeedEvil> jonwil: awesome efforts.
Jan 25 11:58:00 <Estel_> Pali,  ping
Jan 25 11:58:10 <Pali> Estel_, pong
Jan 25 11:58:24 <Estel_> I've encountered strange thing when it comes to usb mode used alongside bme replacement bits
Jan 25 11:58:28 <SpeedEvil> jonwil: it seems almost as if that was one of the express design aims
Jan 25 11:58:56 <Estel_> everytime i end working with mass storage tings connected to N900 and use "periferial mode with charging button"...
Jan 25 11:59:06 <Estel_> it end up as "peripherial mode with *boost*"
Jan 25 11:59:18 <Estel_> seriously, that is what banner says, and I confirmed it in reality
Jan 25 11:59:35 <Estel_> it's peripherial mode, but boost still active, and devices connected are still light up
Jan 25 12:00:02 <Estel_> no idea why it keeps booston, though, nothing suspicious in syslog, I suspect userland bit bug
Jan 25 12:00:13 <Pali> Estel_, look at /sys/class/power_supply_bq241*/mode
Jan 25 12:00:18 <Pali> if it has really boost mode
Jan 25 12:00:22 <Estel_> as, after all, it even *announce* peripherial mode with boost, via banner
Jan 25 12:00:25 <Estel_> yes
Jan 25 12:00:28 <Pali> or only MCE LED pattern was not deactivated
Jan 25 12:00:34 <Estel_> it has, and as said, I checked it hardware-wise
Jan 25 12:00:40 <Estel_> it's still providing real boost
Jan 25 12:00:51 <Pali> Estel_, which version of usbmode do you use?
Jan 25 12:00:54 <Estel_> i.e. up to 200 mA 5V
Jan 25 12:01:18 <Estel_> one related to kernel-power version from repos
Jan 25 12:01:36 <Estel_> i.e. last one that work with kp51r1
Jan 25 12:02:26 <Estel_> I'm just getting into the loop again, so sorry, if I missed something important (searched a little, but never found reports of such behavior or fixlogs)
Jan 25 12:02:34 <Estel_> (fixes in changelos*)
Jan 25 12:02:43 <Pali> Estel_ update to kp52 from TMO and usbmode to 1.0-3
Jan 25 12:04:21 <Pali> Estel_, here is last usbmode version: http://atrey.karlin.mff.cuni.cz/~pali/usbmode/
Jan 25 12:04:27 <Pali> but you need kp52 for it
Jan 25 12:07:23 <Estel_> thanks a lot
Jan 25 12:07:25 <Estel_> will do
Jan 25 12:07:31 <Estel_> then, re-check bug existence, yep?
Jan 25 12:09:27 <Estel_> Pali,  as I was able to catch you - any chances for including module I've requested (qf9700) in next KP version? I've send it via maemo-developers mailing list, including source code (afaik it just require compiling against kernel-power headers)
Jan 25 12:10:22 <Estel_> It would easy life of USB->ethernet dongle's users a lot, as every new davicom is using it, instead of old module (and, new one is free of most old davicom bugs)
Jan 25 12:10:35 <Aoyagi_joytop> Pali: I'm really curious why is it hosted on Matfyz :D
Jan 25 12:10:38 <Estel_> of course new and old module doesn't collide
Jan 25 12:10:54 <Aoyagi_joytop> What is this boost?
Jan 25 12:19:31 <kerio> how do i use more power? i'm at 475mA now
Jan 25 12:19:49 <freemangordon> disable SR
Jan 25 12:19:51 <kerio> with screen at full brightness, busy cpu, and wifi with no powersaving
Jan 25 12:19:56 <kerio> freemangordon: ooh, nice one
Jan 25 12:19:59 <kerio> how do i do that?
Jan 25 12:19:59 <freemangordon> :P
Jan 25 12:20:08 <freemangordon> echo 0 > /sys/power...
Jan 25 12:21:16 <kerio> freemangordon: and then what?
Jan 25 12:21:23 <kerio> sr_vdd*_autocomp?
Jan 25 12:21:25 <freemangordon> yep
Jan 25 12:21:29 <freemangordon> for 1 and 2
Jan 25 12:22:09 <kerio> heh, 8mA increase
Jan 25 12:22:54 <Pali> Estel_, I saw your mail, but I did not looked at it. Did you try to compile it against 2.6.28?
Jan 25 12:23:09 <Pali> Aoyagi_joytop, so what do you think? :-)
Jan 25 12:23:50 <Estel_> kerio,  use flashlight ;)
Jan 25 12:23:59 <Estel_> and bluetooth discoverable
Jan 25 12:24:23 <Estel_> instead of busy cpu, start recording HD video via camera, with flashlight on
Jan 25 12:24:29 <Aoyagi_joytop> Pali: Huh? About what? The boost thing?
Jan 25 12:24:33 <Estel_> and gps program like modrana or marble enabled
Jan 25 12:24:38 <kerio> right, the gps
Jan 25 12:24:44 <Estel_> more than 1A power usage easily achievable
Jan 25 12:24:59 <Estel_> Pali,  as per my knowledge, it compiles without problems
Jan 25 12:25:06 <Estel_> but it's second-hand knowledge
Jan 25 12:25:20 <Estel_> also, android kids are using it hapilly for more than a year
Jan 25 12:25:37 <Estel_> afaik people compiled it sucessfuly even for more ancient kernels
Jan 25 12:25:44 <Estel_> like 2.4
Jan 25 12:26:26 <Estel_> without much hassle. IIRC, 2.6.8 should be just matter of compiling against proper headers. Anyway, my mail isn't long, just check it in a second of free time, please
Jan 25 12:26:40 <Pali> Aoyagi_joytop, boost mode (in bq2415x context) provide 200mA/5V from n900 for usb port
Jan 25 12:26:50 <Pali> which is needed for usb host mdoe
Jan 25 12:26:51 <Pali> *mode
Jan 25 12:27:09 <Estel_> it's probably as much as non-coding person can do to ease your work on including, i.e. sending code, etc. It's in process of getting into mainstream linux kernel, too
Jan 25 12:27:22 <Aoyagi_joytop> Right, thanks.
Jan 25 12:27:51 <Aoyagi_joytop> Now I just need to find out what USB host is, heh.
Jan 25 12:28:07 <ShadowJK> Estel_; I'd compare busyness when running fennec on stock vs ext4, start with same .mozilla/fennec dir. swap on uSD
Jan 25 12:28:29 <Pali> Estel_, compiling some module externally against kernel power is one step and integrate it into deb package second (at deb package compile time)
Jan 25 12:28:38 <Pali> Estel_, sorry but I do not have time for second
Jan 25 12:28:58 <Pali> if you (or soembody) else can provide patch, I can look at it and include it
Jan 25 12:29:14 <Estel_> Pali,  I see
Jan 25 12:30:09 <Estel_> not that I know what is needed to be done for integrating it into kernel power modules at deb creating time :P
Jan 25 12:30:16 <ShadowJK> Estel_; in general sqlite stuff seems to suffer alot on ext3/flash, so I would expect to see benefit even in Conversations
Jan 25 12:30:23 <Estel_> ShadowJK,  hm?
Jan 25 12:30:34 <Estel_> busyness of I/O or cpu time?
Jan 25 12:30:38 <Estel_> re fennec?
Jan 25 12:30:39 <ShadowJK> I/O
Jan 25 12:30:40 <Pali> Aoyagi_joytop, usb has two mode: peripheral and host: In host mode you can connect usb device and in peripheral you act as usb device
Jan 25 12:30:41 <ShadowJK> iostat
Jan 25 12:30:45 <Estel_> I see
Jan 25 12:31:37 <ShadowJK> I tried browsing with fennec, and had iostat display on another device (via ssh), and nogticed whenever fennec scrolling froze, often I/O was pegged at 100%
Jan 25 12:31:39 <Estel_> heh, conversation was always fast for me on ext4 and swap on microSD, but maybe iostat will show comparable differences
Jan 25 12:31:49 <Aoyagi_joytop> Pali: Just found a description of it, thank you.
Jan 25 12:31:58 <ShadowJK> well, ext4 in general is faster too
Jan 25 12:32:05 <Estel_> good point, will try to do
Jan 25 12:32:13 <Estel_> disabling barriers may help, too
Jan 25 12:32:19 <ShadowJK> how to compare ext4 vs ext4-stride-stripe-align, no clue
Jan 25 12:32:21 <ShadowJK> try bonnie
Jan 25 12:32:50 <ShadowJK> (bonnie++)
Jan 25 12:32:53 <Estel_> btw setting stride-stripe on mkfs time is enough, no need to use mount time stripe option?
Jan 25 12:33:08 <Estel_> and what it does? why it's stripe only in mounttime (noob mode)
Jan 25 12:33:57 <Estel_> fun fact - with journal_async_commit, partition seems to be less prone to fuck up and trigger fsck due to unexpected inconsistency
Jan 25 12:34:16 <Estel_> i.e. I feel that journal recovery is enough, more times. May be placebo, though
Jan 25 12:35:05 <Estel_> Aoyagi_joytop,  also, get in mind that HEN is older way of achieving hostmode (although, more "manual", so still work on some specific times, where usb-mode fails to enumerate)
Jan 25 12:35:12 <Estel_> generally, usb-mode is more conveinent
Jan 25 12:35:26 <Aoyagi_joytop> A lot, more, I'd think.
Jan 25 12:35:38 <ShadowJK> I've only once had ext3 fuckup, and it got fucked up by running fsck
Jan 25 12:35:39 <Estel_> so use it everyday, and if you got problems with any particular device, try all steps manually, via hen or scripts itself
Jan 25 12:36:15 <Estel_> ShadowJK,  but if fsck triggers due to unexpected inconsistency, something is already fckd up. fsck. just fckd more (i.e. whole partition)
Jan 25 12:36:47 <ShadowJK> Nokia stock setup never runs fsck
Jan 25 12:36:53 <Estel_> basically, if something is fishy, I just re-create partition and restore backup, due to inability to repair anything on flash
Jan 25 12:36:56 <ShadowJK> or maybe they added that in 1.2 or 1.3?
Jan 25 12:37:26 <Estel_> no idea, I'm doing it manually with -n (no changes to filesystem) option after unexpected reboots, if any
Jan 25 12:37:48 <Estel_> especially if device reboots during heavy operations, due to, for example, dying of low-power with bme disabled :P
Jan 25 12:37:54 <Estel_> to see if anything is fishy
Jan 25 12:38:04 <Estel_> most of the times, recovering journal is enough
Jan 25 12:38:44 <Estel_> if fsck report need for running it manually ("unexpected inconsistency"), I know that it will fckd my partition if I let it to
Jan 25 12:39:02 <kerio> well, this battery can sustain 850mA of draw
Jan 25 12:39:07 <Estel_> so either restore backup, or be in danger of running into some file screwed. No idea if important one or not
Jan 25 12:39:18 <Estel_> so i preffer to restore backup
Jan 25 12:39:29 <Estel_> kerio,  liIon can give many C of power
Jan 25 12:39:42 <Estel_> so probably no problem to draw 5A from it, for a while
Jan 25 12:39:48 <Estel_> what are you trying to achieve?
Jan 25 12:39:50 <kerio> Estel_: yeah but this is a chinese liIon
Jan 25 12:39:52 <kerio> :)
Jan 25 12:40:03 <ShadowJK> Estel_; I wouldn't generalize it like that
Jan 25 12:40:04 <Estel_> no matter, liion chemistry isn't racial
Jan 25 12:40:14 <Estel_> ShadowJK,  batteries, or partitions?
Jan 25 12:40:15 <kerio> Estel_: trying to calibrate this battery as fast as i can
Jan 25 12:40:37 <Estel_> kerio, but you will get lower results :P than you would get with normal heavy usage
Jan 25 12:40:50 <ShadowJK> the high energy stuff like laptop and phone batteries have max recommended draw of 2C at best :)
Jan 25 12:40:55 <Estel_> ShadowJK,  if batteries, sure, it was "mental shortcut"
Jan 25 12:41:05 <Estel_> LiFE can give more than 30C safely
Jan 25 12:41:10 <kerio> Estel_: surely only the end of the calibration matters
Jan 25 12:41:29 <Estel_> kerio,  nope, as there is no such things as absolute capacity
Jan 25 12:41:34 <Estel_> capacity is relative
Jan 25 12:41:58 <Estel_> if you measure it with 1A draw, that you never use in real life, you will get lower capacity results
Jan 25 12:42:19 <Estel_> than with using device normally until it calibrates itself
Jan 25 12:42:24 <Estel_> well, it won't hurt
Jan 25 12:42:39 <Estel_> just that you will see little less than it can really give
Jan 25 12:42:41 <ShadowJK> Estel_; there's more than chemistry involved in it :) Li-Ion comes from high energy 1C stuff to high power 60C stuff depending on the cell design.. A 60C li-ion in bl-5) size would be far lower capacity than 1200mAh though :)
Jan 25 12:42:45 <kerio> huh, sending zeros over wifi stops nicocam
Jan 25 12:42:56 <Estel_> ShadowJK,  sure
Jan 25 12:43:10 <Estel_> I just mean that he shouldn't expect it failing apart on 850 mA
Jan 25 12:43:14 <Estel_> even chinese one
Jan 25 12:43:17 <kerio> Estel_: anyway, not just for calibration, i also want to stress-test this battery a bit
Jan 25 12:44:01 <Estel_> kerio,  btw, relative capacity is why w% 1500 mAh scud run as dual-scud result in ~3100 mAh battery :P
Jan 25 12:44:30 <Estel_> (lower power draw from every cell due to halved between cells, = more relative capacity in each cell)
Jan 25 12:45:18 <kerio> Estel_: i suppose i'll run yet another calibration overnight at a constant ~200mA draw
Jan 25 12:45:23 <kerio> seems to be the most consistent
Jan 25 12:45:31 <Estel_> ShadowJK,  speaking of that, do we know design capabilities of nokia's bl-5j?
Jan 25 12:45:36 <Estel_> i.e. "how many C"?
Jan 25 12:45:43 <kerio> the highest load i've got was 950mA, now
Jan 25 12:46:00 <ShadowJK> I should test someday, say, 2A draw down to 25% followed by 100mA to empty, compared to 100mA start to finish
Jan 25 12:46:09 <Estel_> constant draw is also resulting in less capacity measured than variable one due to normal N900 usage :P
Jan 25 12:46:21 <ShadowJK> Estel_; charging 0.7C is all I remember
Jan 25 12:46:30 <Estel_> not that it matters much, but, best calibration is one regular, not hasted or slowed down
Jan 25 12:46:54 * ShadowJK has managed 1.3A draw, without using usb boost
Jan 25 12:47:04 <ShadowJK> but now I have too good 3g signal to repeat that
Jan 25 12:47:10 <Estel_> ShadowJK,  yea. BTW, I'm using 1250 mA charging (for dual scud) since I announced tesdriving it more than half year ago... or more, don't remember
Jan 25 12:47:14 <Estel_> no ill effects
Jan 25 12:47:30 * ShadowJK is doing it to single batt
Jan 25 12:47:43 <Estel_> using it on different devices, also different revisions (2101, and this 2nd less popular)
Jan 25 12:47:44 <Estel_> :P
Jan 25 12:47:48 <Estel_> poor battery
Jan 25 12:47:54 <Estel_> no ill effects, I suppose, too?
Jan 25 12:48:23 <ShadowJK> Well it kinda already was past its best perf :)
Jan 25 12:48:39 <ShadowJK> though still better than original bl-5j out of box
Jan 25 12:48:45 <Estel_> well, my wanna-be 2A 5V portable charger seems to suffer a little, as it started to give no more than 950 mA. I was forced to switch into regular N900 charger back again
Jan 25 12:48:59 <Estel_> which gives 1250 mA without overheating or complaining
Jan 25 12:49:28 <ShadowJK> My homebuilt portable charger doesn't do 1250 either
Jan 25 12:49:30 <Aoyagi_joytop> Or exploding?
Jan 25 12:49:34 <ShadowJK> on paper it should do 5A
Jan 25 12:49:51 <Estel_> I wonder about N900 charger real limits for stable work, for sure are higher than 1250 mA
Jan 25 12:49:53 <Estel_> ShadowJK,  huh
Jan 25 12:49:58 <Estel_> what went wrong?
Jan 25 12:50:03 <ShadowJK> but the usb cable I cut up was too cheap, drops too much voltage :P
Jan 25 12:50:39 <Estel_> lol? would quite overheat
Jan 25 12:50:42 <Estel_> lol? would quite overheat
Jan 25 12:50:44 <Estel_> or melt
Jan 25 12:51:03 <Estel_> from 5A -> 1000 mA loss. No?
Jan 25 12:51:42 <Estel_> Aoyagi_joytop,  N900 charger is designed for 1250 mA, so it just lives up to expectations
Jan 25 12:51:49 <ShadowJK> more like 5V -> 4.6V (so .4V * 1A = 0.4W loss?)
Jan 25 12:51:49 <Estel_> I plan to add 2 usb ports to it
Jan 25 12:52:25 <Estel_> ShadowJK,  but 0.4W seems a lotta hell capable of melting such cable
Jan 25 12:52:40 <Estel_> or I'm wrong?
Jan 25 12:52:46 <ShadowJK> well we dont know if the other bits and pieces besides the chip is designed for 1250, 
Jan 25 12:53:43 <ShadowJK> Estel_; it's equivalent to 80mA at 5V, spread across a cable?
Jan 25 12:53:47 <Estel_> no parts jumping into saturation and desoldering themselves, so it should be OK
Jan 25 12:54:13 <Estel_> ShadowJK,  I expected such voltage drop due to one part where contact is low (between internal cable wires)
Jan 25 12:54:37 <Estel_> after all, if it isn't half-broken, what would result in such voltage drop? it's made of copper, isn't it?
Jan 25 12:55:05 <Aoyagi_joytop> Uh. What parts of maemo.org are currently -not- working?
Jan 25 12:55:46 <Estel_> I mean that I can't imagine how relatively short cable without connection problems, i.e. almost-torn out continuity inside, could result in such voltage drop
Jan 25 12:55:51 <Estel_> made of iron? :P
Jan 25 12:56:02 <teotwaki> or water
Jan 25 12:56:05 <ShadowJK> Estel_; iirc there were 3 strands of copper in it. And I don't mean the thick strands of ethernet
Jan 25 12:56:05 <jonwil> well wiki.maemo.org is not working for me
Jan 25 12:56:06 <Estel_> :P
Jan 25 12:56:24 <Estel_> wiki is coming alive and dying all the time
Jan 25 12:56:38 <jonwil> I think all parts of maemo.org are going up and down right now
Jan 25 12:56:45 <Estel_> ShadowJK,  3 hairs of coppers, eh?
Jan 25 12:56:51 <ShadowJK> ya
Jan 25 12:56:54 <Estel_> definitely too cheap ;)
Jan 25 12:57:03 <Aoyagi_joytop> jon-kha: It is working for me, but it's oh so slow :D
Jan 25 12:57:24 <Estel_> well, http://maemo.org/community gave 503 error just a second ago
Jan 25 12:57:25 <Aoyagi_joytop> Browser says 7 B/s.
Jan 25 12:57:37 <Estel_> premium hosting, after all
Jan 25 12:58:18 <Estel_> add to this rogue moderators using one touch ban & clean on real people with positive thanks ratio, wandering on TMO...
Jan 25 12:58:27 <Estel_> eh, I won't get into it again, screw that
Jan 25 12:58:35 <Estel_> not my problem, anyway
Jan 25 12:58:50 <Aoyagi_joytop> :/
Jan 25 12:59:13 <Aoyagi_joytop> Is there a USB host thingy for Android too?
Jan 25 12:59:27 <Estel_> sure. Still, android sucks ;)
Jan 25 12:59:32 <Aoyagi_joytop> haha
Jan 25 12:59:52 <ShadowJK> I was tempted to reply to the people whining in alucase thread with something like "well if he hadn't been banned", but it turns out tmo has spies here :/
Jan 25 13:00:00 <Aoyagi_joytop> Yeah, I enjoy bashing it too, but I still have very little arguments for that :3
Jan 25 13:00:14 <Estel_> ShadowJK,  heh, thanks for good will anyway :) btw I replied there today
Jan 25 13:00:51 <Estel_> don't even ask how people look at me when I take alu brick to ear as mobile phone
Jan 25 13:01:13 <Estel_> (testdriving alucase with only internal finished, external things like layout non-existent yet)
Jan 25 13:01:32 <Aoyagi_joytop> Yeah, I read that before. Must be... interesting to see :D
Jan 25 13:01:42 <Estel_> (so it's raw alu brick with screen module, externally - only internals shaped, currently))
Jan 25 13:02:01 <Estel_> well, even with regular N900 with mugen cover, people took me for ticket controller in train
Jan 25 13:02:09 <jaska> heh
Jan 25 13:02:10 <Estel_> when I entered it and said "good morning"
Jan 25 13:02:17 <Estel_> having N900 in hand
Jan 25 13:02:38 <Estel_> ( was checking something on platform, stopped it when train arrived, and entered with N900 in hand)
Jan 25 13:03:10 <Estel_> they were really surprised, when I, instead of taking tickets, started to take my coat off, putting baggage on shelf etc
Jan 25 13:03:10 <ShadowJK> Estel_; check a few tickets, then say "must go check something", take ticket with you and go sit in another car. free train rides forever
Jan 25 13:03:20 <Estel_> ShadowJK,  yea
Jan 25 13:03:30 <Aoyagi_joytop> That will get you shot.
Jan 25 13:03:56 <Estel_> BTW, mbarcode is able to read barcodes from tickets here, in poland
Jan 25 13:04:36 <Estel_> so I can even scan it as ticket controller should :P
Jan 25 13:04:49 <Estel_> nothing to check validity against, though
Jan 25 13:05:09 <Estel_> need to hack into train operator database
Jan 25 13:05:13 <Estel_> ;)
Jan 25 13:07:20 <xes> Estel_: ..do you plan to add some space for a termic printer inside the alu case? ;)
Jan 25 13:07:38 <Estel_> haha
Jan 25 13:07:56 <Estel_> hostmode - cups server from easy debian
Jan 25 13:08:08 <Estel_> and usb termic printer... lovely
Jan 25 13:08:20 <xes> :) it would be nice to view some photo of the worl in progress!
Jan 25 13:08:36 <xes> *work
Jan 25 13:08:49 <Estel_> as soon as I get hands on some digital camera, beside one inside my N900 (thus inside case)
Jan 25 13:09:08 <Estel_> belive it or not, I'm using N900 as my main camera for few years
Jan 25 13:09:15 <Estel_> main and only, to be precise
Jan 25 13:09:31 <Estel_> just because fact that when I see something interesting, I never have camera with me
Jan 25 13:09:35 <Estel_> but always have N900
Jan 25 13:09:43 <Estel_> and raw photos are quite good quality
Jan 25 13:10:11 <Estel_> it resulted in me not using my regular digital camera for long time, then giving it to someone
Jan 25 13:10:13 <Estel_> BTW
Jan 25 13:10:15 <xes> I think you are the only one who tried to build something like that case...this means a great love for this device
Jan 25 13:10:24 <Estel_> tripod mounting point is already there, as only one external element
Jan 25 13:10:49 <jonwil> I prefer my Canon point & shoot (and its 12x optical zoom) to the tiny lens and sensor in the N900
Jan 25 13:10:53 <Estel_> yea, for sure. I just hope that I'll be able to finish it before only me and pre-ordering people will want to use it :P
Jan 25 13:11:09 <Estel_> it ended up in being much more work than expected, as always
Jan 25 13:11:18 <Estel_> jonwil,  understandable
Jan 25 13:11:27 <jaska> full alu? howd you deal with antennas?
Jan 25 13:11:34 <Estel_> it's just that I never have such things with me when needing it, murphy law
Jan 25 13:11:42 <jonwil> yeah
Jan 25 13:11:51 <Estel_> two concepts for antennas, testdriving one of them for wifi
Jan 25 13:11:52 <jonwil> I use my N900 camera if I need to take photos and dont have my normal camera with me
Jan 25 13:12:14 <Estel_> 1. separated alu part of backcover acting as antenna booster itself, shaped accordingly
Jan 25 13:12:30 <Estel_> hopes for increasing signal power over vanilla N900
Jan 25 13:13:10 <Estel_> 2. non-alu parts where antennas are, like in N950, struggling to keep vanilla signal not weakened
Jan 25 13:13:24 <Estel_> I preffer 1st option, as I have some experience in designing antennas
Jan 25 13:13:59 <Estel_> wifi seems to work ok, quite hard to determine if better in noisy enviromnent, but for sure no worse
Jan 25 13:14:06 <Estel_> no GSM antennas, yet
Jan 25 13:14:09 <Estel_> next step
Jan 25 13:14:28 <Estel_> when I finish all antennas, I'll start designing and test manufacturing outer case shape
Jan 25 13:14:34 <Estel_> with all reequired holes etc
Jan 25 13:14:54 <Estel_> and of course connector for hirose u.fl n board
Jan 25 13:15:22 <Estel_> connected to single standarized mini antenna out on case... Will love to connect it to some of my custom antennas :P
Jan 25 13:15:55 <kerio> Estel_: just use an even number of mirrors to take your picture
Jan 25 13:15:55 <Estel_> I'm also tempted to use FMTX testpoints, to allow connecting external antenna to it, too
Jan 25 13:16:02 <Estel_> kerio, haha
Jan 25 13:16:09 * kerio is only slightly kidding
Jan 25 13:16:51 <Estel_> well, kind of idea, but I'll, probably, just lend digital camera from someone
Jan 25 13:17:06 <Estel_> BTW I though, once, about crazy attachement for N900
Jan 25 13:34:08 <yana666> @freenode.net
Jan 25 13:37:21 <yana666> @freemangordon
Jan 25 13:39:38 <ccxCZ> since maemo.org is down, is there some location I can install rootsh from?
Jan 25 13:39:54 <Pali> ccxCZ use some mirror
Jan 25 13:40:26 <Pali> ccxCZ: http://maemo.merlin1991.at/apt-mirror/extras/pool/fremantle-1.3/free/r/rootsh/
Jan 25 13:40:29 <ccxCZ> such as?
Jan 25 13:40:34 <ccxCZ> kthnx
Jan 25 13:43:29 <freemangordon> ~mirrors
Jan 25 13:43:29 <infobot> i guess mirror is http://maemo-archive.wedrop.it/ http://maemo.merlin1991.at/apt-mirror/   http://talk.maemo.org/showthread.php?p=1315143#post1315143
Jan 25 13:43:34 <freemangordon> yana666: ^^^
Jan 25 13:46:28 <ccxCZ> is there a way to connect to network via usb on stock N900?
Jan 25 13:46:35 <kerio> yes
Jan 25 13:46:46 <kerio> but really, just send the .deb via bluetooth or something
Jan 25 13:46:52 <kerio> if you have no wifi
Jan 25 13:47:03 <kerio> or, hell, use usb mass storage mode
Jan 25 13:48:36 <ccxCZ> doesn't installing deb require root?
Jan 25 13:48:51 <Pali> ccxCZ, you can install deb package via file manager
Jan 25 13:49:07 <Pali> store it to MyDocs via usb mass storage
Jan 25 13:49:32 <Pali> file manager open deb package in application manager which is running as root
Jan 25 13:49:43 <ccxCZ> working on it now, thanks
Jan 25 13:49:55 <Pali> (only apt-worker backend running as root)
Jan 25 13:53:15 <Estel_> sorry, got phone
Jan 25 13:53:27 <Estel_> as for crazy N900 accesory I thought about, once:
Jan 25 13:53:51 <Estel_> small adjustable mirror, allowing to use back camera for voip call, yet, still see person you're talking too, on screen :P
Jan 25 13:54:08 <Estel_> of course it was just crazy idea, never took any steps to build it
Jan 25 13:54:43 <Estel_> would look funny, though, with some arm reaching out of N900's back, with small mirror on it
Jan 25 13:56:54 <freemangordon> merlin1991: does your rmo mirror still give hash error?
Jan 25 13:58:13 <kerio> how can it give a hash error? have you changed the keys?
Jan 25 13:59:00 <FIQ> it gave a hash error for me too when I tried, actually
Jan 25 13:59:17 <FIQ> though atm I don't really need the repos so cba to try again
Jan 25 13:59:29 <merlin1991> freemangordon: yep :/
Jan 25 13:59:37 <freemangordon> merlin1991: :(
Jan 25 13:59:42 <merlin1991> kerio: the release file is foobar
Jan 25 13:59:46 <FIQ> (it went up yesterday though for a little, and I took that possibility to download picodrive :3)
Jan 25 13:59:47 <freemangordon> do you know which repo works in HAM?
Jan 25 13:59:49 <merlin1991> all hashes are wrong
Jan 25 14:00:16 <merlin1991> freemangordon: http://extras-devel.merlin1991.at/
Jan 25 14:00:30 <merlin1991> that one is the minimal one
Jan 25 14:00:30 <FIQ> I don't know how apt works, but can't you just.. I don't know, disable hash checking, or re-do that file?
Jan 25 14:00:45 <kerio> merlin1991: only because extras is the only sanctioned repo, and nobody uses it
Jan 25 14:00:58 <freemangordon> well, there is a lady on the channel I am trying to help to set-up mirror repos, I am not sure I want her to enable extras-devel :)
Jan 25 14:01:10 <kerio> technically if you install something from extras, and then enable extras-devel, you won't be able to upgrade it from extras-devel
Jan 25 14:01:11 <ccxCZ> yay, worked
Jan 25 14:01:35 <merlin1991> freemangordon: I guess you're right :D
Jan 25 14:01:51 <ccxCZ> now back up he old system before screwing up.. just dd | gzip the block devices?
Jan 25 14:02:03 <kerio> freemangordon: extras devel is cool
Jan 25 14:02:07 <kerio> ccxCZ: nope
Jan 25 14:02:12 <freemangordon> well, lets try with http://maemo-archive.wedrop.it/repository.maemo.org/extras/
Jan 25 14:02:26 <kerio> ccxCZ: you MUST NOT directly read or write a mtd
Jan 25 14:02:28 <kerio> ~2119
Jan 25 14:02:29 <infobot> The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED",  "MAY", and "OPTIONAL" in this document are to be interpreted as described in RFC 2119.
Jan 25 14:02:30 <ccxCZ> what is the recommended procedure then?
Jan 25 14:02:44 <FIQ> that one gave bogus bzip files for me @ freemangordon
Jan 25 14:03:04 <kerio> ccxCZ: install backupmenu, reboot with open keyboard slider, follow UI
Jan 25 14:03:20 <kerio> but be wary, backupmenu doesn't backup your kernel
Jan 25 14:03:56 <FIQ> and thus, restoring to a backup with another kernel mess up the system
Jan 25 14:04:00 <FIQ> which I found out the hard way
Jan 25 14:04:07 <kerio> "mess up"
Jan 25 14:04:19 <kerio> just flash the old kernel from a n900
Jan 25 14:04:24 <freemangordon> FIQ: what
Jan 25 14:04:26 <freemangordon> ?
Jan 25 14:04:37 <FIQ> freemangordon: ?
Jan 25 14:04:43 <kerio> ccxCZ: there's (almost) nothing you can do via software that will permanently damage your n900
Jan 25 14:04:53 <freemangordon> <FIQ> that one gave bogus bzip files for me @ freemangordon
Jan 25 14:05:29 <FIQ> freemangordon: ah, when I tried out the mirrors, one gave hash errors (merlin1991's), wedrop also gave hash errors, and the third one which name I've forgot gave bogus bzip files
Jan 25 14:05:50 <FIQ> (trying to update with that repo just gave error "bzip returned with an error code (1)"
Jan 25 14:05:52 <FIQ> )
Jan 25 14:05:57 <freemangordon> so all of them don;t work?
Jan 25 14:06:07 <FIQ> Well that's my own experience
Jan 25 14:06:12 <freemangordon> but what was that reflash one?
Jan 25 14:06:23 <FIQ> Perhaps errors has been fixed, but when I tried, it simply didn't work
Jan 25 14:06:36 <freemangordon> ok
Jan 25 14:06:39 <ccxCZ> Linux Nokia-N900 2.6.28-omap1 #1 PREEMPT Fri Aug 6 11:50:00 EEST 2010 armv71 unknown
Jan 25 14:06:47 <ccxCZ> that's what I got here now ↑
Jan 25 14:07:31 <ccxCZ> so I guess I should get corresponding images for it if I'd want to restore it later? and tar up all the files...
Jan 25 14:08:55 <jonwil> ok, found a nokia-binaries mirror, still doesn't give me any nice targets to reverse engineer though...
Jan 25 14:10:53 <Pali> jonwil, can you look at libdevlock and libcodelockui binaries?
Jan 25 14:10:58 <ccxCZ> where do I get backupmenu?
Jan 25 14:11:21 <Pali> jonwil, libdevlock is needed for mce for devicelock, and there is missing header file
Jan 25 14:11:28 <Pali> but there is i386 version
Jan 25 14:11:35 <Pali> ccxCZ, from extras-devel
Jan 25 14:13:55 <ccxCZ> got it, thanks
Jan 25 14:15:38 <jonwil> Will see what I can do to produce libdevlock info
Jan 25 14:15:52 <jonwil> libcodelockui already has -dev package so no need to reverse engineer that really
Jan 25 14:19:10 <freemangordon> seems like ttp://skeiron.org/repo/repository.maemo.org/extras/ is ok
Jan 25 14:20:30 <ccxCZ> what is the most convenient way to install custom chroot? can I reformat the mmc to say ext2 instead of vfat?
Jan 25 14:20:58 <SpeedEvil> yes
Jan 25 14:21:12 <ShadowJK> iirc it wont get automounted
Jan 25 14:21:18 <SpeedEvil> that is true
Jan 25 14:21:25 <ShadowJK> I'd use microsd
Jan 25 14:21:32 <ShadowJK> can manually mount htat on boot
Jan 25 14:21:51 <SpeedEvil> I left the fat as 8g, and put a new 16 or so in the free space
Jan 25 14:23:50 * ShadowJK has two new 16g microSD cards on his desk to test
Jan 25 14:24:19 <kerio> ShadowJK: the cssu ke-recv will automount everything
Jan 25 14:24:53 <ShadowJK> If anyone wants a "Samsung Essential Class 6 16GB" uSD card, I'll send it for just cost of post :P
Jan 25 14:24:59 <ShadowJK> (it's rubbish)
Jan 25 14:26:17 <Skry> ShadowJK: deal!
Jan 25 14:27:07 <Skry> Finland?
Jan 25 14:27:14 <ccxCZ> any specific ones you'd recommend?
Jan 25 14:27:34 <ccxCZ> been using kingston and adata ones, kingston died suprisingly fast
Jan 25 14:27:57 <RST38h> ShadowJK: Why such sudden generosity? =)
Jan 25 14:28:33 <kerio> ccxCZ: sandisk
Jan 25 14:30:57 <freemangordon> wiki: "Couldn't authenticate against garage. (DB problem)"
Jan 25 14:31:01 <freemangordon> how nice :(
Jan 25 14:31:22 <ShadowJK> Skry; ya
Jan 25 14:32:04 <ShadowJK> RST38h; well I bought like half dozen card and not going to ever use this one, don't like its performance
Jan 25 14:34:06 <RST38h> Ah.. Class 6 though - should be fast
Jan 25 14:34:17 <RST38h> Ok, supposed to be fast =)
Jan 25 14:35:22 <chem|st> kerio: ShadowJK sandisk and adata were never in my range after several pendrives before, samsung class 10 32GB and kingston class 6 8BG is what I have for n900
Jan 25 14:36:07 <chem|st> both cards are pretty fast compared to old class 2-4 nabble
Jan 25 14:36:28 <freemangordon> surprisingly for me (I am SanDisk fan) it seems that this 64GB Samsung performs really well
Jan 25 14:36:44 <kerio> otoh, those class 6/class 10 cards are absolute crap at random I/O
Jan 25 14:37:04 <ShadowJK> The other sandisk I have, "Plus", "Class 10", seems decent
Jan 25 14:37:04 <freemangordon> I'll bench it again in 1-2 weeks, to see if performance degrades
Jan 25 14:37:23 <chem|st> freemangordon: I stopped to be a sandisk fan after 6 broken pendrives, I had 9, 3 are still working
Jan 25 14:37:35 <freemangordon> well, 30% not that bad :D
Jan 25 14:37:41 <chem|st> lol
Jan 25 14:38:11 <chem|st> freemangordon: garage has issues atm so bare with us for not working logins
Jan 25 14:38:16 <freemangordon> chem|st: you should try kingston ;)
Jan 25 14:38:38 <chem|st> freemangordon: I won't 'try' anything anymore
Jan 25 14:39:12 <freemangordon> lol
Jan 25 14:39:17 <kerio> for uberoptimized sequential writes it doesn't really matter, as long as it's a good brand
Jan 25 14:39:21 <ShadowJK> other samsung I meant
Jan 25 14:39:43 <chem|st> I have a verbatim pendrive working for years now and have 64gb of other pendrives/cards... and a 1GB usb3 harddrive
Jan 25 14:39:54 <chem|st> 1TB
Jan 25 14:40:34 <ShadowJK> I've been torturing a sandisk usb drive for ages now, running nilfs2 on it. more or less constant 1.5MB/s write :)
Jan 25 14:40:49 <chem|st> would kill me
Jan 25 14:40:55 <kerio> nilfs2?
Jan 25 14:41:00 <chem|st> 1.5MB/s
Jan 25 14:41:27 <ShadowJK> kerio; was best performing fs when I tested
Jan 25 14:42:00 * freemangordon is going to retest his uSD
Jan 25 14:42:11 <chem|st> it makes only sense to use external drives (no matter what kind) as long as network is slower
Jan 25 14:42:42 <chem|st> or you need something offline
Jan 25 14:42:55 <chem|st> and that is becoming a rare case nowadays
Jan 25 14:43:43 <ShadowJK> Heh, running squid on that usb flash actually
Jan 25 14:58:12 <freemangordon> seems fine to me http://pastebin.com/5nMKQsZs
Jan 25 14:58:40 <freemangordon> keep in mind this is over USB, with swap on the same device
Jan 25 15:09:25 <kerio> freemangordon: what's QD?
Jan 25 15:09:36 <freemangordon> queue depth
Jan 25 15:09:46 <freemangordon> not that I know what queue is that :D
Jan 25 15:12:01 <jonwil> pali: ping
Jan 25 15:13:23 * MrPingu wonders why is there a safelist for widgets in /etc/hildon-desktop/?
Jan 25 15:13:45 <jonwil> to whoever wanted info on libdevlock, http://pastebin.com/emnukG4E
Jan 25 15:14:21 <jonwil> damn, wiki is still refusing to work :(
Jan 25 15:16:09 <ShadowJK> 1MB of write/s is damn impressive:)
Jan 25 15:17:19 <freemangordon> ShadowJK: I wonder what will be the results if I put it in a card reader. Unfortunately I cannot find mine :D
Jan 25 15:17:49 <ShadowJK> Well, my XP box with card reader gave much higher results
Jan 25 15:18:41 <ShadowJK> btw, I just remembered seeing somewhere that while Class 2, 4 and 6 specify response time for cards, 10 and uhs dont
Jan 25 15:19:55 <freemangordon> ShadowJK: I know sequental speeds will be much higher, I am wondering about random i/o
Jan 25 15:20:41 * ShadowJK suspects uhs cards slow down on everything when run in sdhc host
Jan 25 15:21:07 <Pali> jonwil, thanks!
Jan 25 15:21:27 <jonwil> I hope that's what you need
Jan 25 15:22:01 <jonwil> My guess is that Nokia didn't publish the -dev package for that library because of security (i.e. you could use it to bypass the device lock)
Jan 25 15:22:14 <jonwil> but in our case now we do have the info :)
Jan 25 15:22:21 <Pali> you can use devlock-bin package for that
Jan 25 15:22:32 <jonwil> yeah
Jan 25 15:22:40 <Pali> it contains binary with S bit!!!
Jan 25 15:22:40 <jonwil> anyhow, its documented
Jan 25 15:23:03 <jonwil> another closed library has been given a dev package :)
Jan 25 15:23:03 <Pali> jonwil, it is hard to RE that library?
Jan 25 15:23:19 <jonwil> do you need the whole library cloned?
Jan 25 15:23:35 <Pali> I'd like to understand what doing...
Jan 25 15:23:37 <jonwil> and yes it would be harder to RE than it was to produce the dev stuff
Jan 25 15:23:50 <Pali> because I need to implement code into mce
Jan 25 15:24:18 <jonwil> you can call libdevlock from MCE no problems (its what Nokia did after all)
Jan 25 15:24:51 <Pali> yes, I know, but I must understand what that functions doing...
Jan 25 15:25:21 <jonwil> why you need to know what its doing if you just want to call it and do whatever stock MCE did?
Jan 25 15:25:58 <Pali> jonwil, I have source code of diablo mce and I'm replacing diablo code with fremantle compatible
Jan 25 15:26:14 <Pali> so I need to know which functions should be replaced with some old diablo code...
Jan 25 15:26:31 <jonwil> Having source code to diablo DSME would help with your work I suspect...
Jan 25 15:26:41 <jonwil> but yeah that library is too hard for me to RE
Jan 25 15:26:47 <Pali> ok
Jan 25 15:26:48 <jonwil> its mostly setting gconf and cal anyway
Jan 25 15:26:58 <jonwil> or calling devlock-tool to set things
Jan 25 15:28:44 <jonwil> for reference, the functions in MCE that call libdevlock are:
Jan 25 15:28:46 <jonwil> mce_devlock_init
Jan 25 15:28:54 <jonwil> mce_devlock_exit
Jan 25 15:30:36 <jonwil> and one function that has no name
Jan 25 15:34:53 <jonwil_> Have a look at what devlock.c in MCE does
Jan 25 15:35:07 <jonwil_> that should be a good starting point to figure out what to replace with calls to libdevlock
Jan 25 15:36:04 <jonwil> anything else I can help you with?
Jan 25 15:39:32 <Pali> jonwil, what is state of your vibrator mce plugin?
Jan 25 15:39:39 <Pali> working? crashing?
Jan 25 15:40:12 <jonwil> never got anywhere useful with it
Jan 25 15:40:39 <Pali> jonwil, it is implemented?
Jan 25 15:40:47 <jonwil> cant remember
Jan 25 15:40:56 <jonwil> feel free to use whatever code I posted
Jan 25 15:41:06 <jonwil> but beyond that, I have no interest in touching the mess that is MCE :)
Jan 25 16:07:03 <infobot> DocScrutinizer: infobot joined!
Jan 25 16:07:53 <Pauly1210> Hi, hows mapping on the n900. Is any of the community apps better than the old nokia maps for daily usage? 
Jan 25 16:09:33 <khm> yes.
Jan 25 16:09:36 <jacekowski> not really
Jan 25 16:09:38 <khm> I really like modrana 
Jan 25 16:09:44 <SpeedEvil> I suspect not
Jan 25 16:09:47 <jacekowski> nokia maps has no competition
Jan 25 16:09:55 <jacekowski> and it's offline
Jan 25 16:09:59 <Sicelo> yup.
Jan 25 16:10:15 <SpeedEvil> it does somewhat depend - openstreermap is good in some areas
Jan 25 16:10:19 <khm> I can't even remember the last time I bothered starting nokia's maps program on the n900
Jan 25 16:10:26 * Sicelo uses Nokia Maps + OM Voice Server
Jan 25 16:10:41 <khm> I use mappero and modrana heavily
Jan 25 16:11:47 <Pauly1210> yeah if it was like the maps on the symbian phones its just clunky
Jan 25 16:11:54 <Pauly1210> thanks guys
Jan 25 16:12:12 <Pauly1210> I'll check them out
Jan 25 16:17:49 <kerio> Pauly1210: i quite like monav
Jan 25 16:17:53 <kerio> incredibly light
Jan 25 16:18:02 <kerio> can do *everything* offline, with free datasets
Jan 25 16:18:29 <kerio> Sicelo: OM Voice?
Jan 25 16:19:15 <Pauly1210> kerio: cool its Qt
Jan 25 16:22:48 <Sicelo> kerio, yes. to give a voice to nokia maps
Jan 25 16:23:00 <kerio> Sicelo: does it add autofollow?
Jan 25 16:23:20 <Sicelo> um, what's that?
Jan 25 16:23:57 <kerio> does it do the equivalent to pushing forward in nokia maps when you pass an intersection?
Jan 25 16:24:27 <Sicelo> re-route? no.
Jan 25 16:25:17 <DocScrutinizer05> 503 ???? wtf!
Jan 25 16:25:33 <Sicelo> i have monav too, but i never downloaded the maps.
Jan 25 16:25:54 <DocScrutinizer05> who else got maemo.org  503
Jan 25 16:26:14 <Lava_Croft> yes
Jan 25 16:26:24 <DocScrutinizer05> since when
Jan 25 16:26:35 <Lava_Croft> less than 2 days
Jan 25 16:26:47 <Lava_Croft> i check about daily on your update post
Jan 25 16:26:54 <Lava_Croft> havent checked yesterday iirc
Jan 25 16:27:06 <DocScrutinizer05> yesterday everything been fine
Jan 25 16:27:26 <Lava_Croft> let's hope its nothing serious
Jan 25 16:27:53 <Lava_Croft> talk works
Jan 25 16:28:10 <Lava_Croft> but i clearly have no idea where what part of *.maemo.org is located
Jan 25 16:28:14 <DocScrutinizer05> talk is disparate
Jan 25 16:28:23 <Lava_Croft> as i figured
Jan 25 16:28:25 <DocScrutinizer05> wiki works too
Jan 25 16:28:43 <jacekowski> sounds like overload or something
Jan 25 16:28:43 <freemangordon> does it?
Jan 25 16:29:02 <freemangordon> yes, it does
Jan 25 16:29:13 <DocScrutinizer05> maybe somebody is moving repo to a new path
Jan 25 16:29:28 <DocScrutinizer05> and took down maemo.org for that
Jan 25 16:29:38 <jacekowski> 503 may mean that backend is down
Jan 25 16:29:42 <freemangordon> but why MO is 503 then?
Jan 25 16:29:52 <jacekowski> so frontend load balance/reverse proxy is running
Jan 25 16:29:52 <freemangordon> aah, I see
Jan 25 16:30:02 <jacekowski> but backend vm not so much
Jan 25 16:30:07 <jacekowski> loads of possibilities
Jan 25 16:30:22 <DocScrutinizer05> there's no load balancing anymore
Jan 25 16:30:29 <DocScrutinizer05> on new infra
Jan 25 16:30:38 <DocScrutinizer05> one VM for m.o
Jan 25 16:30:44 <DocScrutinizer05> one for wiki.m.o
Jan 25 16:30:51 <DocScrutinizer05> one for repo.m.o
Jan 25 16:30:51 <freemangordon> DocScrutinizer05: btw did you read #maemo-ssu logs from today?
Jan 25 16:31:04 <DocScrutinizer05> very partially
Jan 25 16:31:21 <DocScrutinizer05> seen the guy from Nokia dropped by
Jan 25 16:31:22 <freemangordon> the key has been found :)
Jan 25 16:31:24 <freemangordon> yep
Jan 25 16:31:40 <Lava_Croft> !
Jan 25 16:31:44 <Lava_Croft> no shit?
Jan 25 16:31:52 <DocScrutinizer05> hope you all were kind and reasonable to him
Jan 25 16:32:22 <freemangordon> I think so. actually the guy was very polite and cooperative
Jan 25 16:33:00 <DocScrutinizer05> no shit what?
Jan 25 16:33:00 <DocScrutinizer05> sure, why not?
Jan 25 16:33:13 <freemangordon> just mentioning
Jan 25 16:33:16 <DocScrutinizer05> not everybody is like djszapi
Jan 25 16:33:18 <DocScrutinizer05> ;-)
Jan 25 16:33:29 <Lava_Croft> no shit key found
Jan 25 16:33:34 <freemangordon> (whoever he is :D )
Jan 25 16:33:43 <freemangordon> Lava_Croft: read the logs
Jan 25 16:33:57 <DocScrutinizer05> ooh, you never been in #harmattan
Jan 25 16:34:20 <freemangordon> well, been there a couple of times, for a couple of minutes :D
Jan 25 16:34:36 <freemangordon> but usually I pester arcean when I join there :D
Jan 25 16:34:47 <DocScrutinizer05> then it probably been djszapi who made you quit
Jan 25 16:35:36 <DocScrutinizer05> I *think* you can find him here in this chan's logs too, long before HARM appeared
Jan 25 16:35:56 <Lava_Croft> the name doesnt ring a negatively sounding bell
Jan 25 16:35:59 <DocScrutinizer05> asking insane questions and bitching at us when he didn't get right answers
Jan 25 16:36:15 <freemangordon> never seen him
Jan 25 16:36:15 <Lava_Croft> oh that does ring a bell
Jan 25 16:36:34 <freemangordon> i've missed the party it seems :D
Jan 25 16:36:52 <DocScrutinizer05> like "how do I install xterm?" "it's already installed!" "not here, sucker!"
Jan 25 16:37:18 <freemangordon> really?
Jan 25 16:37:20 <freemangordon> omg
Jan 25 16:37:23 <Lava_Croft> maybe he is culturally inclined to talk like that
Jan 25 16:37:53 <Sicelo> where did pupnik go?
Jan 25 16:39:21 <Sicelo> ~seen pupnik
Jan 25 16:39:27 <infobot> pupnik <PugVader@p54B29D69.dip.t-dialin.net> was last seen on IRC in channel #meego, 483d 16h 48m 40s ago, saying: 'tegra chipset?'.
Jan 25 16:45:58 <kerio> DocScrutinizer05: the guy seemed supportive
Jan 25 16:46:12 <kerio> but the key he found is for some internal thing nokia has, so he didn't know if he'll be able to use it
Jan 25 16:46:14 <DocScrutinizer05> sure
Jan 25 16:55:18 <tommis> i today broke my n900's stand thingy?
Jan 25 16:55:33 <tommis> were could i get new back plate for cheap?
Jan 25 16:56:32 <kerio> tommis: try ebay
Jan 25 16:58:29 <kerio> for instance, on the italian ebay there's an original one taken from a broken n900 for 6,43€
Jan 25 17:04:41 <kgu> is repos still behind the flaky firewall? 
Jan 25 17:07:28 <DocScrutinizer05> right now it seems down
Jan 25 17:08:03 <DocScrutinizer05> and it will stay down until Eero/Nemein moved it to exposed-IP
Jan 25 17:08:18 <DocScrutinizer05> probably
Jan 25 17:08:37 <kgu> they are the 'grid-guys'?
Jan 25 17:08:48 <DocScrutinizer05> Seems Eero gave up on the Endian FW and decided to expose the repo IP to internet
Jan 25 17:09:07 <DocScrutinizer05> well, basically one of them is
Jan 25 17:09:31 <FIQ> the repository worked yesterday, how come?
Jan 25 17:09:42 <FIQ> it was slow, but it worked :p
Jan 25 17:09:58 <kgu> ok, lets hope it will come alive again
Jan 25 17:14:51 <DocScrutinizer05> FIQ: we had a short time where repo 'worked', immediately after FW reboot. yes
Jan 25 17:15:13 <FIQ> ah
Jan 25 17:15:30 <FIQ> then 50000 N900s decided to DDoS the server, including mine I guess?
Jan 25 17:15:58 <DocScrutinizer05> FIQ: see update12
Jan 25 17:17:18 <Pali> DocScrutinizer05, please look at mail from @nokia. Council or board should answer
Jan 25 17:29:31 <ShadowJK> the fw sounds like linksys stuff. tcp/udp expire times so high all conntrack tables get full
Jan 25 17:30:44 <kerio> hey, linksys is good
Jan 25 17:30:55 <kerio> perhaps dlink
Jan 25 17:31:00 <kgu> I cant see how the N900s can bring rmo down. They dont try to update at the same time
Jan 25 17:59:09 <DocScrutinizer05> Pali: in 2 hours
Jan 25 17:59:18 <Pali> ok
Jan 25 17:59:19 <DocScrutinizer05> Pali: had not even time to check mail yet
Jan 25 17:59:28 <DocScrutinizer05> meeting  now
Jan 25 18:00:27 <Pali> DocScrutinizer05, see discussion with nokia guy: http://mg.pov.lt/maemo-ssu-irclog/latest.log.html#t2013-01-25T13:34:56
Jan 25 18:26:17 <RST38h> Meanwhile: Unlocking Your Phone Is Illegal Starting Tomorrow
Jan 25 18:26:24 <RST38h> (in the US of course, where else?)
Jan 25 18:27:20 <freemangordon> RST38h: whaat?
Jan 25 18:27:54 <tommis> wat
Jan 25 18:27:57 <tommis> >us
Jan 25 19:40:53 <kerio> warfare: install cssu-thumb!
Jan 25 19:41:52 <warfare> kerio: yesterday HAM complained about some repos not working. 
Jan 25 19:42:04 <kerio> hm
Jan 25 19:42:42 <kerio> well, you can enable mirrors
Jan 25 19:42:56 <kerio> and disable the "ignore packages from wrong domain" in the hidden settings
Jan 25 19:44:16 <warfare> kerio: that could do the trick, I'll try later. 
Jan 25 20:23:17 <GeneralAntilles> The difference between Maemo and Harmattan:
Jan 25 20:24:58 <GeneralAntilles> Both make simple tasks (decompressing an archive, searching for a URL in the browser history), but with Maemo those things can be worked around
Jan 25 20:25:06 <GeneralAntilles> Harmattan they're just impossible.
Jan 25 20:25:20 <kerio> they make tasks?
Jan 25 20:25:21 <GeneralAntilles> Want to decompress a rar file? Good luck. (Dumbass)
Jan 25 20:26:05 <Estel_> Pali,  where is the "usual place" for latest kernel-power, and why it's not linked in first post in any kp thread?:P
Jan 25 20:26:54 <GeneralAntilles> kerio, er, make them difficult
Jan 25 20:26:56 <Estel_>  kernel-power development thread proudly states - in opening post - that currently, latest version is kp50. Kernel-power "main" thread talks about kp 51. I'm using kp51r1, and I'm supposed to find kp52 "somewhere"
Jan 25 20:27:41 <Estel_> sorry for pestering, just wanted to highlight, that from perspective of someone who haven't read every single post of kp thread - unlike the old days - it's nearly impossible to find it
Jan 25 20:28:01 <Pali> here is link: http://talk.maemo.org/showpost.php?p=1305042&postcount=288
Jan 25 20:28:08 <Estel_> opening posts of kp threads just beg to have static link to place, where it's to be found
Jan 25 20:28:12 <Estel_> thanks a lot
Jan 25 20:28:15 <Pali> I wanted to push it to extras-devel...
Jan 25 20:28:21 <Estel_> yea, I know
Jan 25 20:28:24 <Pali> ...but autobuilder not working
Jan 25 20:28:33 <Estel_> I think you can forget it for next ~month :P
Jan 25 20:28:36 <Pali> after will be inextras-devel I will create new thread
Jan 25 20:28:40 <Estel_> autobuilder? devel doesnt work, too
Jan 25 20:28:45 <Estel_> and extras, at all
Jan 25 20:28:53 <Estel_> and few other things, too :P
Jan 25 20:29:18 <Estel_> and anyone supposing that it will be fixed in less than few weeks is overly optimistic.
Jan 25 20:29:46 <Pali> Ferenc today fixed garage git repos
Jan 25 20:30:06 <Pali> I got mail and tried to push kp52 to garage and it worked
Jan 25 20:30:13 <Estel_> nice to hear
Jan 25 20:30:29 <Estel_> btw, in post you've submitted, same link is repeated 4 times in row, or is it just me?
Jan 25 20:31:09 <Estel_> well, when it comes to maemo infrastructure - unlike in real life - I'm pesimistic, so I can be only positively surprised
Jan 25 20:32:18 <Estel_> I start in position of not expecting anything on Maemo infra to work, at all, including this irc channel - or expect it all to break anytime soon - so every line of irc here or every time wiki doesn't fail to load (like it does 50% of times) I'm positively surprised. It's good for health, me thinks.
Jan 25 20:33:08 <Sicelo> hmm
Jan 25 20:33:14 <SpeedEvil> :-) 
Jan 25 20:33:56 <Sicelo> after GeneralAntilles' talk about harmattan, this infrastructure move has affected N9 users or not?
Jan 25 20:35:32 <Estel_> thanks Pali, going to rock on new kp and updated usbmode, will report about results, sometime in future
Jan 25 20:36:05 <Estel_> question to anyone accustomed to it - current state of bme replacement is usable for every-day, if properly handled?
Jan 25 20:36:13 <Estel_> i.e. no showstoppers?
Jan 25 20:36:24 <kerio> not that i've experienced
Jan 25 20:39:52 <Estel_> lame question again, where are .item things for uboot or multiboot or?
Jan 25 20:40:06 <Estel_> it always appeared automagically after kernel-power-bootimg install
Jan 25 20:40:15 <Estel_> not that I can't create it myself, just curious
Jan 25 20:40:39 <Pali> /etc/bootmenu.d/
Jan 25 20:41:03 <Estel_> thanks, noticed this myself at the same time, could check twice before asking
Jan 25 20:41:30 <kerio> Estel_: by the way, peripheral without charging will kinda drain the n900's battery
Jan 25 20:41:43 <Estel_> 80 mA isnt much
Jan 25 20:41:48 <kerio> 80mA is a lot
Jan 25 20:42:04 <Estel_> ~3000/80
Jan 25 20:42:04 <infobot> 37.5
Jan 25 20:42:21 <Estel_> 37.5 hours is still more than average notebook battery :P
Jan 25 20:42:33 <Estel_> also, no one is forced to use it
Jan 25 20:43:07 <Estel_> just trigger, when you want to connect N900 as client through USB to something battery powered, without vampiric drain on said device power
Jan 25 20:43:23 <kerio> Estel_: just disable the charging and connect
Jan 25 20:43:27 <Estel_> 500 mA after negotiation is overkill for many netbooks/notebooks
Jan 25 20:43:42 <Estel_> wut? charging will auto-re enable
Jan 25 20:43:45 <kerio> also, i wonder if it's possible to charge very little
Jan 25 20:43:52 <kerio> Estel_: why is that?
Jan 25 20:44:06 <Estel_> well, N900, as client, need boost to estabilish connection
Jan 25 20:44:10 <Estel_> i.e. 5V
Jan 25 20:44:13 <kerio> echo none > /sys/class/power_supply/bq24150x_charger/mode
Jan 25 20:44:24 <Estel_> = enabling auto mode
Jan 25 20:45:08 <Estel_> see bq24150 documentation, if I haven't messed somethink
Jan 25 20:45:36 <kerio> then how would this "peripheral without charging" mode be implemented?
Jan 25 20:45:46 <Estel_> BTW, trick with charging very low is possible, using steps provided by chip (50 or 150 is lowest)
Jan 25 20:45:57 <kerio> oh, wouldn't that be even better?
Jan 25 20:46:06 <kerio> you won't drain the n900's battery
Jan 25 20:46:11 <Estel_> but if you're sharing internet connection to notebook or something in train, you don't want N900 to draw *any* power
Jan 25 20:46:16 <kerio> usually a netbook's battery is much bigger
Jan 25 20:46:20 <kerio> hm
Jan 25 20:46:28 <Estel_> as N900 battery will still last much longer than notebook one, by factors of few
Jan 25 20:46:43 <Estel_> yea, it's bigger, and netbook uses much more :P
Jan 25 20:46:55 <kerio> use the netbook to charge the n900, use the n900
Jan 25 20:47:01 <kerio> :D
Jan 25 20:47:26 <Estel_> well, no problem for me, as I don't have netbook, and my notebook is acting like desktop, since I got N900 :P
Jan 25 20:47:39 <kerio> seriously though, how would you implement this?
Jan 25 20:47:47 <Estel_> N900 + small usb keyboard + usb/bluetooth mouse is my travel kit
Jan 25 20:47:53 <kerio> disable the charging once the connection is established?
Jan 25 20:48:00 <Estel_> kerio, exactly
Jan 25 20:48:13 <kerio> well, that's kinda easy
Jan 25 20:48:15 <ShadowJK> N900 charging chip lowest charge current setting 550mA, but also input curren limits 100, 500 and 800mA available
Jan 25 20:48:20 <Estel_> if it wouldn't work, last resort is for N900 to provide boost for itself
Jan 25 20:49:14 <Estel_> well, when I got hit by bug of "peripherial mode with boost", my device immediately though it got connected to something
Jan 25 20:49:26 <ShadowJK> if you kill bme after connecting to laptop, charging will drop to 100mA input limit in 30s or so?
Jan 25 20:49:34 <Estel_> and presented me with screen inviting to choose mode (mass storage, pc suite)
Jan 25 20:49:51 <Estel_> ShadowJK,  yes, but after 30 minutes, charging chip or omething will kick reboot
Jan 25 20:49:57 <Estel_> or 32 minutes
Jan 25 20:50:10 <kerio> ShadowJK: we're using pali's bq24k module anyway
Jan 25 20:50:16 <Estel_> I remember vaguely, that something doesn't like charging in emergency mode for too long
Jan 25 20:51:33 <Estel_> kerio,  if "just disabling charging" wouldn't work, then, peripherial + boost should work - after all, nothing would drain this 200mA, so only part of it would get wasted
Jan 25 20:51:55 <ShadowJK> nah dunno, I've had various charge script mods for faster charging, and had failsafes that triggered and made my script abort, leaving device without bme or anything.. and device was still alive in the morning :P
Jan 25 20:52:36 <Estel_> without bme but connected, charging in emergency, with bright orange led?
Jan 25 20:52:43 <ShadowJK> yes
Jan 25 20:52:46 <kerio> Estel_: you probably also want pali's battery applet
Jan 25 20:52:53 <kerio> it is good
Jan 25 20:52:57 <Estel_> interesting, worth trying
Jan 25 20:53:01 <Estel_> kerio, sure
Jan 25 20:53:15 <kerio> except that if your bq27k is properly calibrated, you'll want to modprobe -r rx51_battery
Jan 25 20:53:18 <Estel_> it's probably in bme replacement section of his site? where is his site, again?
Jan 25 20:53:21 <kerio> because it's less accurate
Jan 25 20:53:22 <ShadowJK> bright orange led is the "JK's charge algorithm failed and failsafes triggered" indicator :P
Jan 25 20:53:22 <kerio> ~pali
Jan 25 20:53:23 <infobot> from memory, pali is http://atrey.karlin.mff.cuni.cz/~pali/
Jan 25 20:53:53 <kerio> there's no section for that, though :c
Jan 25 20:54:00 <Estel_> kerio,  rx-51 is totally inacurrate, as it use "design capacity", which is *never* right
Jan 25 20:54:12 <kerio> Estel_: tell him, maybe you'll be more successful than me
Jan 25 20:54:16 <Estel_> no idea why Pali use it at all, maybe I don't know something
Jan 25 20:54:38 <Estel_> well, early mugens had thermpresistor there, so "design capacity" changed with temperature...
Jan 25 20:54:57 <Estel_> Pali, I'm telling you, that "design capacity" is depreciated and shouldn't be used for anything
Jan 25 20:55:09 <kerio> <Pali> rx51_battery reports value which is not wrong as value from bq27x
Jan 25 20:55:15 <Estel_> even nokians screwed it, when upgraded bl-5j to 1420 mAh
Jan 25 20:55:39 <kerio> (should read "not as wrong" i think)
Jan 25 20:55:48 <Estel_> they use same resistor, that was in 1320 mAh bl-5j
Jan 25 20:55:54 <Estel_> define not wrong or not as wrong
Jan 25 20:56:08 <Estel_> calibrated bq27x00 is accurate, rx_51 never is
Jan 25 20:56:14 <kerio> indeed
Jan 25 20:56:59 <Estel_> design capacity is read from value of resistor between BMI and battery -, so whatever is put there, it's design capacity. Some manufacturers even put thermoresistor there, which confuses hell out of rx_51
Jan 25 20:57:11 <ShadowJK> hm, I should check on my other device where I have new bl-5j, but IIRC they upped the design capacity resistor in it.
Jan 25 20:57:19 <Estel_> Pali, what madness made you to use it and give it priority over bq27x00?
Jan 25 20:57:41 <Estel_> ShadowJK,  I got 100% original 1420 BL-5J from brand new nokia asha
Jan 25 20:57:46 <Estel_> and it had same resistor
Jan 25 20:57:53 <Estel_> maybe it differs from batch to batch
Jan 25 20:58:25 <Estel_> I measured battery to determine if they really upped the capacity, then, I disassembled it
Jan 25 20:58:26 <Pali> Estel_, bq27k chip not reporting correct max values if battery is not calibrated
Jan 25 20:58:43 <Pali> also calibration can change max value only for some %
Jan 25 20:58:45 <Estel_> Pali,  sure - but rx_51 *never* reports correct values
Jan 25 20:58:52 <Estel_> no matter if calibrated or not
Jan 25 20:58:59 <ShadowJK> Estel_; iirc 1320mAh batt had 1200mAh design capacity :)
Jan 25 20:59:03 <Pali> so you need to do more cycles for better values
Jan 25 20:59:04 <warfare> kerio: cssu-thumb complains about missing packages (libmameosec0, upstart, etc.). I have community, community-testing, community-thumb and extras enabled. 
Jan 25 20:59:18 <warfare> any hints?
Jan 25 20:59:20 <Pali> ShadowJK, do you have rx51_battery kernel module?
Jan 25 20:59:39 <kerio> warfare: enable redpill mode, and disable "ignore packages from wrong domains"
Jan 25 20:59:39 <Estel_> Pali, but rx_51 take "design capacity" from some random resistor, that no one, even manufactures, care of
Jan 25 20:59:54 <Estel_> They put whatever there
Jan 25 21:00:01 <ShadowJK> Pali; no, I mean as reported by bme
Jan 25 21:00:03 <Estel_> whatever around 100kOhm
Jan 25 21:00:10 <kerio> warfare: also, how did you install community-ssu-enabler correctly? a mirror for community-testing?
Jan 25 21:00:23 <Estel_> but it reads "design capacity" from same source - resistor
Jan 25 21:00:33 <kerio> warfare: either have community or community-testing enabled
Jan 25 21:00:34 <Estel_> it is only thing used to "code" design capacity
Jan 25 21:00:53 <kerio> both are simply useless, -testing will trump stable anyway
Jan 25 21:00:53 <Estel_> except that no one cares about it and no manufacturer, even Nokia, uses it properly
Jan 25 21:01:27 <Estel_> Whatever rx_51 shows, is not related to battery at all, it's just "something around 1320"
Jan 25 21:01:29 <kerio> warfare: also, /j #maemo-ssu
Jan 25 21:01:31 <warfare> kerio: Somehow I've installed the cssu-enabler by hand (cssu-testing) and did the upgrade with HAM. 
Jan 25 21:01:48 <Estel_> less or more, depending on higher or lower resistor put there
Jan 25 21:01:53 <Estel_> Pali ^
Jan 25 21:01:56 <Pali> if you have some idea how to fix this problem.... you have calibrated 12xx mAh battery, then you change it with 14xx or 1yyy battery, how should design capacity be read?
Jan 25 21:02:34 <Estel_> Pali, no physical way to read it, just re-calibration. But, rx_51 doesn't help here
Jan 25 21:02:51 <Estel_> as 1yyy battery manufacturer doesn't give a shit about coding design capacity into it
Jan 25 21:02:55 <Pali> bq27k chip will not increase max capacity to normal new by one cycle and calibration
Jan 25 21:02:58 <Estel_> just threat it as non-existent
Jan 25 21:03:09 <Estel_> Pali, normally, it will
Jan 25 21:03:25 <Pali> so what to show? old capacity?
Jan 25 21:03:27 <Estel_> it won't only, if you change battery by very high values, like 1200 -> 2700 mAh
Jan 25 21:03:31 <Pali> for new better battery??
Jan 25 21:03:48 <Estel_> Pali, sure, until it's calibrated. And use *need calibration* flag
Jan 25 21:03:54 <Estel_> it is exported by driver
Jan 25 21:03:56 <Estel_> afaik
Jan 25 21:04:01 <Pali> and what to show if you totally discharge battery and bq chip will be reseted?
Jan 25 21:04:06 <Estel_> chip have flag that shows "need calibration"
Jan 25 21:04:28 <Pali> after that max capacity wil be read from design register which is 2048 (or something) mAh
Jan 25 21:04:32 <Estel_> vdq is 1 during proper discharge cycle
Jan 25 21:04:43 <Estel_> so, when chip acknowledge new capacity, vdq changes to 0
Jan 25 21:04:48 <Pali> Estel_ calibration flag will be deleted after vdq
Jan 25 21:04:52 <ShadowJK> I've never managed to reset bq on purpose by power starvation :)
Jan 25 21:05:16 <Estel_> Pali, it is 2048, but only if you put battery out ans bupbat is dead
Jan 25 21:05:16 <Pali> but bq chip will not set correct max capacity with one cycle
Jan 25 21:05:24 <Estel_> so, alongside clock, chip will reset
Jan 25 21:05:38 <Pali> ShadowJK, I was able to do that
Jan 25 21:05:42 <Estel_> for regular batteries, it will, during 2 cycles at worst
Jan 25 21:06:01 <Estel_> heck, even from 1300 to 3100 I need two cycles only
Jan 25 21:06:14 <kerio> Estel_: so this battery is only 1420 mAh? :(
Jan 25 21:06:20 <kerio> goddammit, vi____ said it was 1600
Jan 25 21:06:40 * ShadowJK has a resetbq.sh for when changing battery
Jan 25 21:06:42 <Estel_> maybe his relative capacity was 1600 on device, constant current test show ~1400
Jan 25 21:07:04 <Estel_> Pali, there is no perfect solution to this, but rx_51 is making things even worse
Jan 25 21:07:25 <Estel_> not onlynot helping at all, but just outright lying about capacity, taking it from moon
Jan 25 21:08:15 <Estel_> and consider confusion, when someone use battery with thermoresistor here, i.e. "design capacity" changing every few minutes or hours
Jan 25 21:08:39 <Estel_> or bug reports, that it shows 1320 for their new, oryginal, Nokia's 1400 batteries :P
Jan 25 21:08:56 <Estel_> or 1200, as Shadowjk said about design capacity of 1320 one
Jan 25 21:08:59 <Estel_> it's a mess
Jan 25 21:09:17 <Estel_> never ever use design capacity for anything, except scarry'ing children
Jan 25 21:09:17 <kerio> Estel_: Pali's battery applet - now including a thermometer
Jan 25 21:09:37 <Pali> so propose solution how to get design capacity
Jan 25 21:09:45 <Estel_> Pali, there is no such thing
Jan 25 21:09:53 <Pali> which wil work in every situation
Jan 25 21:09:54 <kerio> why do we even want the design capacity anyway
Jan 25 21:09:58 <Estel_> no thing in chip or battery itself reports capacity (properly). period
Jan 25 21:10:06 <Estel_> consider it non-existent
Jan 25 21:10:06 <kerio> we want the current capacity
Jan 25 21:10:25 <Pali> current capacity is calculated from design
Jan 25 21:10:32 <Estel_> w-aaaat!
Jan 25 21:10:49 <Pali> bq chip doing it
Jan 25 21:11:01 <Estel_> so, my 3000mAh battery, which reports design of 1200, will show 600 mAh left on 50%?!
Jan 25 21:11:14 <kerio> bq27k *counts* the mAh that pass through, aiui
Jan 25 21:11:25 <Estel_> no no no, current capacity is and always was full capacity - used capacity
Jan 25 21:11:42 <Estel_> yea, bq counts capacity used, variable one
Jan 25 21:12:04 <Pali> percentage is calculated from current and design
Jan 25 21:12:04 <Estel_> ShadowJK can confirm ;)
Jan 25 21:12:12 <Estel_> absolutely no
Jan 25 21:12:20 <ShadowJK> kerio; if it was repeatable (the adc and resistor thermal drift seems to say otherwise), and accurate, it'd be nice gauge of battery health to compare current capacity vs design... On N9, the bq chip has "Impedance Track" thingy which does a more useful thing of comparing battery impedance. SpeedEvil's ohm.sh research path is similar, in determining the useful usable capacity of a battery and getting useful "battery low" thresholds so you dont get surpris
Jan 25 21:12:21 <kerio> and that's another thing that's dead wrong
Jan 25 21:12:26 <ShadowJK> e shutdowns or sudden unexpected battery low warnings :)
Jan 25 21:12:29 <Estel_> or my 3000 mAh battery would get 280%
Jan 25 21:12:32 <kerio> as my "110%" charge can report
Jan 25 21:12:36 <kerio> Estel_: yep :D
Jan 25 21:13:24 <Estel_> Pali,  using bq27x00 or later i2c get raw data from chip, I never, ever, saw using design capacity to calculate anything
Jan 25 21:13:27 <Estel_> over years
Jan 25 21:13:44 <ShadowJK> Design Capacity = battery capacity when brand new. Capacity = current battery capacity. Charge = current amount of charge stored in the battery (where maximum = capacity, maximum when brand new = design capacity)
Jan 25 21:14:02 <Estel_> percentage is derived from full capacity/current capacity
Jan 25 21:14:17 <Estel_> ShadowJK,  but lets stick to practice, not theory
Jan 25 21:14:23 <kerio> Estel_: just rmmod rx51_battery :)
Jan 25 21:14:29 <ShadowJK> In general, when Capacity divided by  Design Capacity is less than 0.8, it's time to replace batt
Jan 25 21:14:34 <Estel_> how bq27200.sh calculates % of battery?
Jan 25 21:14:59 <Estel_> you wrote it, and you never used design capacity there, as reported by resistor
Jan 25 21:15:22 <ShadowJK> Estel_; as reported by bq. When calibrated, it's (current charge / current capacity) * 100
Jan 25 21:15:40 <Estel_> kerio, rc51_battery get autoloaded after appled installation? need to blacklist it?
Jan 25 21:15:50 <Estel_> Pali, see? bq doesnt use design capacity for it
Jan 25 21:15:51 <kerio> Estel_: it's autoloaded by the bme replacement
Jan 25 21:15:57 <kerio> it's not "auto"loaded
Jan 25 21:16:02 <kerio> it's modprobed by /etc/event.d/bme
Jan 25 21:16:12 <kerio> but the battery applet is quite flexible
Jan 25 21:16:17 <ShadowJK> "Design Capacity" is ILMD in bq, but the preprogrammed value is silly :)
Jan 25 21:16:21 <kerio> you can do whatever you want and it'll switch after a handful of seconds
Jan 25 21:16:22 <Estel_> so just need to edit this event.d script, ok
Jan 25 21:17:09 <Estel_> I respect and admire Pali's work - it's just that using so-called design capacity for anything is wrong, because, as ShadowJK said, it's just silly value not related to reality
Jan 25 21:17:20 <Pali> you need rx51_battery which reporting battery temperature to dsme and pulseaudio
Jan 25 21:17:46 <Estel_> what? I feel terrorized :P why pulseaudio want my battery temperature?!
Jan 25 21:18:08 <Estel_> also, couldn't bq27x00 report it, like in good old times?
Jan 25 21:18:14 <Pali> I do not know
Jan 25 21:18:39 <Pali> temperature from bq chip is not battery temperature (this wrote nokia bme devs)
Jan 25 21:18:47 <Estel_> ok, so couldn't rx51_battery just fulfil those needs, without touching design capacity with a long stick, and messing with battery applet?
Jan 25 21:18:52 <Pali> and I implemeted battery temperature reading to rx51_battery
Jan 25 21:18:58 <ShadowJK> Estel_; if the bq chip had been used as intended, and as OpenMoko did it, integrated into the battery, it would've been properly programmed with ILMD which could be used as a proper source of "design cpacity" :)
Jan 25 21:19:10 <Estel_> hell! there is no other temp chip in device
Jan 25 21:19:10 <kerio> Estel_: it's just the applet that prioritizes rx51_battery data over bq27x00_battery data
Jan 25 21:19:27 <Estel_> Pali, only one source of temp indication is inside bq
Jan 25 21:19:28 <kerio> ShadowJK: no it wouldn't've
Jan 25 21:19:33 <kerio> they would've put a random value in, as usual
Jan 25 21:19:40 <kerio> especially the crap batteries
Jan 25 21:19:45 <Estel_> so whatever rx51_battery reads, it's the same as bq27x00 reads
Jan 25 21:19:49 <ShadowJK> There's a temperature sensor connected to madc, according to the schematics
Jan 25 21:19:50 <Pali> Estel_, not true
Jan 25 21:20:01 <Estel_> w-h-at?
Jan 25 21:20:11 <Pali> rx51_battery and bme using temperature sensor from madc
Jan 25 21:20:15 <Estel_> so we got *two* of them?
Jan 25 21:20:30 <Estel_> ok. Sorry. My mistake. wasn't aware of that, and lived in lie about it
Jan 25 21:20:32 <Estel_> anywa
Jan 25 21:20:44 <Estel_> <Estel_> ok, so couldn't rx51_battery just fulfil those needs, without touching design capacity with a long stick, and messing with battery applet?
Jan 25 21:20:48 <Estel_> still applies
Jan 25 21:20:59 <kerio> rx51_battery doesn't "touch" anything
Jan 25 21:21:02 <kerio> it just reports data
Jan 25 21:21:08 <kerio> it's the battery applet that's in denial
Jan 25 21:21:09 <Estel_> kerio, you know what I mean
Jan 25 21:21:12 <Pali> all: I'm going to hacking kernel. If you have better solution how to determinate "design capacity" ping me back
Jan 25 21:21:20 <Pali> solution should work in any state
Jan 25 21:21:25 <Estel_> there is no design capacity
Jan 25 21:21:30 <kerio> THE BATTERY APPLET SHOULDN'T CARE ABOUT THE DESIGN CAPACITY FOR FUCK'S SAKE
Jan 25 21:21:32 <Estel_> why the hell applet prioritize anything?
Jan 25 21:21:51 <Estel_> just let users set what it should care about
Jan 25 21:22:00 <kerio> Estel_: it goes rx51 module -> bq27k module -> bme -> bq27k over i2c, aiui
Jan 25 21:22:04 <Estel_> if someone want to use rx51_battery data for battery applet, let it be
Jan 25 21:22:11 <Estel_> hardcoded?
Jan 25 21:22:20 <kerio> probably
Jan 25 21:22:25 <kerio> patches welcome :)
Jan 25 21:22:47 <Estel_> Pali, better solution, as you asked - either doesn't care for design capacity in battery applet, or at least let user change priorities
Jan 25 21:23:03 <ShadowJK> I agree, UI shouldn't care about "Design Capacity". On N9, the UI does care, but only as far as to display "Battery Health: OK" or something like that.
Jan 25 21:23:05 <Estel_> so one could set bq27x00 as higher priority over rx51_battery for applet
Jan 25 21:23:11 <Pali> Estel_ I need something as "design" capacity for calculating percentages
Jan 25 21:23:28 <kerio> you shouldn't use "design", design becomes useless in two months
Jan 25 21:23:46 <Estel_> Pali...
Jan 25 21:23:47 <Estel_> <ShadowJK> Estel_; as reported by bq. When calibrated, it's (current charge / current capacity) * 100
Jan 25 21:24:01 <Pali> and then is not calibrated?
Jan 25 21:24:03 <kerio> yeah, i'd use the bq27k fuel gauge
Jan 25 21:24:05 <Estel_> here are your percents
Jan 25 21:24:15 <Pali> note that CI flag is not enought for calibration
Jan 25 21:24:23 <Estel_> shit. If it's not calibrated, "design capacity" doesn't help!
Jan 25 21:24:29 <ShadowJK> Pali; you don't really, percentage should be calculated against "last measured capacity".  Even lapotops do that :)
Jan 25 21:24:38 <Estel_> why healing cold with cancer?
Jan 25 21:24:51 <Estel_> yes...
Jan 25 21:25:36 <Pali> what about max(bq_max, rx_desing) ?
Jan 25 21:25:36 <kerio> the issue is the usual dumb user who doesn't know what a battery is
Jan 25 21:25:43 <Pali> as design capacity?
Jan 25 21:25:58 <kerio> but those will use bme data, anyway
Jan 25 21:26:39 <freemangordon> Pali: you'd better use bq if calibrated, otherwise design reported by rx51
Jan 25 21:27:10 <Estel_> freemangordon,  design reported by rx51 is even worse than not calibrated bq
Jan 25 21:27:36 <ShadowJK> and "battery capacity" as LMD != ILMD ? BSI : LMD
Jan 25 21:27:39 <Estel_> or, calculate it in any way you want, just allow user to overwrite priority settings in battery applet, even if by simple text file
Jan 25 21:28:03 <freemangordon> Estel_: no, it is not, for genuine batteries
Jan 25 21:28:05 <kerio> if bq27k isn't calibrated, don't show the charge max and use the percentage from bq27k
Jan 25 21:28:14 <ShadowJK> (use LMD if LMD is not same as ILMD, if LMD same as ILMD use BSI reported capa)
Jan 25 21:28:16 <Estel_> to allow users set bq27x00 as higher priority over rx51 *always*, if desired. Problem solved
Jan 25 21:28:30 <Estel_> freemangordon,  nope, as said, genuine batteries have it screwed too
Jan 25 21:28:33 <kerio> ShadowJK: what's ILMD?
Jan 25 21:28:37 <freemangordon> Estel_: if bq not calibrated, values reported are bogus
Jan 25 21:28:38 <Pali> if bq27k is not calibrated kernel driver is not reporting any capacity
Jan 25 21:28:45 <Pali> only voltage and current
Jan 25 21:28:57 <Estel_> 1320 ones got 1200 in design capacity, 1400 ones use same, but some of them, reportedly, say 1320
Jan 25 21:28:59 <freemangordon> this ^^^ too
Jan 25 21:29:06 <kerio> Pali: it's still reporting the percentage, right?
Jan 25 21:29:14 <Pali> no
Jan 25 21:29:18 <kerio> wat
Jan 25 21:29:23 <Pali> -ENODATA
Jan 25 21:29:28 <kerio> hm
Jan 25 21:29:28 <freemangordon> Estel_: all nokia batteries I have here report its design capacity through rx51 correctly
Jan 25 21:29:30 <Estel_> hm, strange, using i2cget raw readings, I have capacity no matter what
Jan 25 21:29:36 <Estel_> even if not calibrated
Jan 25 21:29:42 <freemangordon> ve've verified that on a couple of other devices
Jan 25 21:29:50 <Estel_> ShadowJK,  how do you deal with it in bq27200.sh?
Jan 25 21:29:51 <freemangordon> *we
Jan 25 21:30:15 <freemangordon> Estel_: but it is bogus
Jan 25 21:30:26 <Estel_> well, zillions of devices using bq27200.sh over years can confirm, that capacity should be reported even on uncalibrated (but wrong capacity)
Jan 25 21:30:53 <Estel_> nothing more bogus than using some hardcoded resistor put inside battery to determine it's capacity and percentage
Jan 25 21:31:07 <ShadowJK> Estel_; i don't. bq27200.sh is specifically reporting raw bq27200 stuff. The only "modified truth" I do is that the chip reports a flag status for charging/discharging, and I apply that to the reported current
Jan 25 21:31:08 <Estel_> you could as good screw it and hardcode 1200 as this value into applet
Jan 25 21:31:18 <freemangordon> Estel_: wrong
Jan 25 21:31:27 <freemangordon> mi 1360 reports as 1360
Jan 25 21:31:30 <freemangordon> *my
Jan 25 21:31:37 <freemangordon> my 1430 reports as 1430
Jan 25 21:31:44 <freemangordon> genuine nokia
Jan 25 21:31:52 <Estel_> so wtf is with kernel driver, that it doesn't get same things as i2c can get from bq chip?
Jan 25 21:32:13 <kerio> Estel_: you can still read the "registers" file, it's a text file
Jan 25 21:32:14 <Estel_> freemangordon,  nice. My 100% genuine 1420 bl-5j from Nokia asha, brand new, from carrier, reports 1200
Jan 25 21:32:15 <Estel_> so what?
Jan 25 21:32:19 <Estel_> it's guessworking
Jan 25 21:32:30 <kerio> and it has the raw data
Jan 25 21:32:42 <Estel_> mugens use thermoresistor, and design capacity can change every minute during device runtime
Jan 25 21:32:45 <freemangordon> Estel_: there is no 1420 mAh in ashas ;)
Jan 25 21:32:50 <freemangordon> it is 1430
Jan 25 21:32:59 <freemangordon> so, check once again if it is genuine
Jan 25 21:33:00 <Estel_> ok
Jan 25 21:33:02 <Estel_> 1430
Jan 25 21:33:09 <kerio> trolls trolling trolls trolling Pali
Jan 25 21:33:10 <Pali> everybody with kp52 please paste your outputs of: $ cat /sys/class/power_supply/rx51-battery/uevent
Jan 25 21:33:21 <Estel_> doest it change anything re design capacity being silly value, that shouldn't be used for anything in UI?
Jan 25 21:33:35 <kerio> Pali: it's a bunch of lines
Jan 25 21:33:41 <freemangordon> Estel_: what I am saying is to ise it as a last resort
Jan 25 21:33:47 <freemangordon> *use
Jan 25 21:33:48 <Pali> Estel_, freemangordon, kerio ^^^^
Jan 25 21:33:56 <freemangordon> Pali: no kp52 here :(
Jan 25 21:34:04 <Pali> ok
Jan 25 21:34:17 <Estel_> freemangordon,  could agree (although don't like forcing it as last resort, would like to keep bq readings or no readings at all, until calibrated)
Jan 25 21:34:33 <kerio> http://pastebin.mozilla.org/2086151 , and the battery is *not* 1267mAh, it's around 1400
Jan 25 21:34:46 <Estel_> as design capacity is just bullshit, not related to reality at all. Or related, if you got lucky, but not physically
Jan 25 21:34:58 <freemangordon> kerio: I guess it is that chinese one?
Jan 25 21:35:00 <kerio> yep
Jan 25 21:35:19 <kerio> it's a pretty nice battery, and it costs 10USD
Jan 25 21:35:24 <kerio> (shipped)
Jan 25 21:35:35 <kerio> vi____ was luckier than me i suppose
Jan 25 21:35:40 <freemangordon> not here :D (shipped)
Jan 25 21:35:43 <kerio> (calibrated 1600mAh)
Jan 25 21:35:44 <Milhouse_01> hi
Jan 25 21:35:48 <kerio> vi____: i hate you
Jan 25 21:35:51 <kerio> Milhouse_01: hi
Jan 25 21:35:54 <Estel_> Pali, second, I'll reboot device to get kp52 modules loaded etc
Jan 25 21:36:01 <ShadowJK> Estel_; amusingly even before bq27200.sh I didn't use statusbar batterymeter, I had a monstrous script that read data from "hal-device bme" every minute and applied heurestics to make it make more sense and be more useful :)
Jan 25 21:36:16 <Milhouse_01> sorry, i just recognized that the maemo server is down
Jan 25 21:36:32 * ShadowJK unfortunately lost that with first dead n900 :(
Jan 25 21:36:41 <freemangordon> ShadowJK: I guess once calibrated bq reports correct values.
Jan 25 21:36:48 <Milhouse_01> is there any chance the get a backup of pr1.3 firmware for n900
Jan 25 21:36:48 <ShadowJK> freemangordon; yes
Jan 25 21:36:49 <kerio> Pali: it would be neat to have a "flags" dir for easy access to flags and a "raw" dir for easy access to the content of the registers
Jan 25 21:36:55 <Milhouse_01> :-(
Jan 25 21:37:04 <kerio> Milhouse_01: tablets-dev.nokia.com being down is recent
Jan 25 21:37:19 <kerio> there's plenty of backups, i'm not sure if any of them are public, considering that it's not legal to redistribute the images
Jan 25 21:37:51 <freemangordon> so, what I am trying to say, is to use the lower reading from rx51, until bq gets calibrated, not those bogus > 200mAh reported by bq when not calibrated
Jan 25 21:37:54 <Pali> kerio, I sent bq27k patch with file "registers" to upstream but was rejected
Jan 25 21:38:02 <freemangordon> *2000mAh
Jan 25 21:38:12 <kerio> the registers file is in here, i wanted *more* :3
Jan 25 21:38:29 <kerio> freemangordon: if bq27k isn't calibrated, don't show the full capacity
Jan 25 21:38:35 <kerio> put a question mark
Jan 25 21:38:41 <freemangordon> kerio: this is the driver
Jan 25 21:38:44 <kerio> oh
Jan 25 21:38:45 <Milhouse_01> hmm... i'll never buy anything from Nokia again ... waiting for Jolla
Jan 25 21:38:49 <freemangordon> not the bq itself afaik
Jan 25 21:39:03 <kerio> Milhouse_01: it's not like you can buy more n900s from nokia
Jan 25 21:39:20 <ShadowJK> freemangordon; specifically, if reported LMD (Last measured discharge) isn't a value exactly equal to that of LMD=ILMD; and/or loop: LMD=LMD-LMD/6; (iirc)
Jan 25 21:40:27 <freemangordon> ShadowJK: you lost me :). I am not that much involved in bq stuff(like you, Pali and doc) 
Jan 25 21:41:11 <freemangordon> but what i have seen to be reported by an uncalibrated bq is something like 2057 mAh
Jan 25 21:41:15 <kerio> Milhouse_01: we're in the middle of a bigass transition from nokia-owned stuff to community-owned stuff
Jan 25 21:41:20 <kerio> confusion is to be expected
Jan 25 21:41:39 <Pali> I got 20xx max capatity too
Jan 25 21:41:52 <kerio> Pali: well perhaps you should calibrate your bq27k :P
Jan 25 21:41:53 <freemangordon> which is insane to be shown to an enduser
Jan 25 21:42:04 <Pali> and because BME turning device off before EDV1, it wss never correct
Jan 25 21:42:14 <freemangordon> :nod:
Jan 25 21:42:37 <kerio> freemangordon: so just don't show it
Jan 25 21:42:40 <Pali> so we should not use uncalibrated max capacity from bq27k
Jan 25 21:42:54 <Pali> and then we should not use % from bq27k
Jan 25 21:43:07 <Pali> (because % is calculated from max capacity)
Jan 25 21:43:16 <kerio> the percentage from bq27k is taken from LMD aiui
Jan 25 21:43:42 <freemangordon> that's why I think it is better to either show rx51 readings (which will be somehow relevant for new genuine batteries) or as kerio says, show "capacity unknown"
Jan 25 21:43:43 <Milhouse_01> hmm ok. thanks i will see
Jan 25 21:43:44 <Pali> percentage in bq27k is current/max*100
Jan 25 21:44:09 <freemangordon> until calibrated that is
Jan 25 21:44:31 <kerio> ShadowJK: what's the percentage calculated by bq27k when it's not calibrated but has done a couple of cycles of charge-discharge?
Jan 25 21:44:48 <Milhouse_01> bye! CU
Jan 25 21:44:53 <kerio> Milhouse_01: hold on
Jan 25 21:45:03 <Pali> kerio, % is calculated by ^^^^ always
Jan 25 21:45:10 <kerio> merlin1991: do you have the fiasco images for pr1.3 somewhere?
Jan 25 21:45:41 <ShadowJK> freemangordon; so, "last measured discharge" (LMD) is the capacity bq thinks battery has. ILMD (I=initial) is a preprogrammed default value of 2000-something mAh. On each valid full to empty charge, LMD gets adjusted based on how much juice the battery had before reachign empty. There's, however, a safeguard preventing "large" capacity adjustments (because chip was designed to be integrated into battery, and would expect small gradual degradation of capac
Jan 25 21:45:47 <ShadowJK> ity, not big steps).. IIRC that limit is 1/6th of LMD.. so, if LMD=ILMD, or if LMD is a value which can be derived by starting with LMD=ILMD and sbtracting 1/6th of LMD from LMD, it would indicate the proper capacityy hasnt been learned
Jan 25 21:46:44 <ShadowJK> Pali; bq % is not calculated from max capacity. it's calculated from last measured capacity
Jan 25 21:47:07 <Pali> ah, yes I mean that (in kernel there is max and design)
Jan 25 21:47:19 <Pali> max = bq_last and design = bq_max
Jan 25 21:47:28 <kerio> ShadowJK: but it requires a vdq-edv1 cycle, right?
Jan 25 21:47:50 <kerio> Estel_: what's the value reported by rx51 on your mugen?
Jan 25 21:47:57 <freemangordon> ShadowJK: thanks
Jan 25 21:48:14 <Estel_> ok, back, I'm on kp52 now
Jan 25 21:48:31 <Estel_> kerio, i don't use mugen at the moment - dual scud
Jan 25 21:48:37 <Estel_> it reports 1200, obviously
Jan 25 21:48:45 <kerio> hm
Jan 25 21:48:50 <Pali> I have there another problem: now hald-addon-bme turn off device when EDVF is set. But I got problem with "enter pin code" when battery was between EDV1 and EDVF
Jan 25 21:48:51 <Estel_> mugens got thermoresistors there, so design capacity changes with temperature
Jan 25 21:48:55 <Pali> and GSM modem not worked
Jan 25 21:48:59 <kerio> so bq27k's value of 2000 is more accurate than rx51's value of 1200
Jan 25 21:49:03 <Pali> after I charged battery
Jan 25 21:49:07 <kerio> :D
Jan 25 21:49:12 <Pali> s/before/
Jan 25 21:49:25 <Estel_> Pali, what was this thing to cat
Jan 25 21:49:25 <kerio> Pali: not even after charging a bit?
Jan 25 21:49:26 <Estel_> ?
Jan 25 21:49:32 <freemangordon> Estel_: hmm, why 1200?
Jan 25 21:49:39 <kerio> cat /sys/class/power_supply/rx51-battery/uevent
Jan 25 21:49:44 <ShadowJK> Pali; too low battery
Jan 25 21:49:46 <freemangordon> I guess you got those in parallel
Jan 25 21:49:53 <kerio> freemangordon: it has the value of a single scud
Jan 25 21:49:54 <Pali> so I will change code to turn off device when EDV1 will be set
Jan 25 21:49:58 <Estel_> freemangordon, yea, but resistor is only one
Jan 25 21:50:00 <Pali> what do you think? ^^^
Jan 25 21:50:03 <freemangordon> aah, ok
Jan 25 21:50:04 <Estel_> I don't use two connectors
Jan 25 21:50:07 <Estel_> just two cells
Jan 25 21:50:10 <freemangordon> ok, got it
Jan 25 21:50:11 <kerio> Pali: in the applet? in hald-addon-bme?
Jan 25 21:50:14 <kerio> i'm not sure
Jan 25 21:50:20 <Pali> hald-addon-bme
Jan 25 21:50:35 <Estel_> single scud is 1500 in reality, btw
Jan 25 21:50:36 <kerio> Pali: it would screw up by calibration script
Jan 25 21:50:43 <ShadowJK> Pali; EDV1 (3248 mV) is already pretty marginal for modem. EDVF at 2800 or something is just too low. the modem is getting undervolted and crashing, then getting rebooted. UI presents enter pin code.
Jan 25 21:50:49 <Estel_> 1400 by design, and 1200 by design reported via resistor :P
Jan 25 21:51:13 <Estel_> crashing modem due to undervolting may be not healthy for it
Jan 25 21:51:13 <kerio> Pali: perhaps wait a tiny bit after edv1 is set?
Jan 25 21:51:17 <Estel_> just wild guess
Jan 25 21:51:20 <Pali> Estel_: cat /sys/class/power_supply/rx51-battery/uevent
Jan 25 21:51:24 <Estel_> kk
Jan 25 21:51:25 <freemangordon> Estel_: well, curse mugen for not putting the correct resistor then :P
Jan 25 21:51:28 <kerio> yeah, we shouldn't get to the brownout point
Jan 25 21:51:45 <kerio> freemangordon: no, mugen has a thermistor there
Jan 25 21:51:48 <freemangordon> or scud
Jan 25 21:51:49 <kerio> even more fun
Jan 25 21:51:59 <Estel_> Pali, no reason why device couldn't shutdown at edv1
Jan 25 21:52:10 <Estel_> bme killed it at higher voltage, most of the times
Jan 25 21:52:18 <kerio> Estel_: to calibrate bq27k
Jan 25 21:52:25 <freemangordon> I wonder why the hell they put a thermistor there
Jan 25 21:52:26 <Estel_> or, even better, lets shutdown device at sysfs value set by user :)
Jan 25 21:52:28 <kerio> Pali: can you make it wait like 30 seconds?
Jan 25 21:52:35 <kerio> Estel_: hald-addon-bme is userland
Jan 25 21:52:47 <ShadowJK> With japod batt and recent nokia batt, there's seconds of useful life available after 3248mV. With mugen and crap ancient bl-5j, device can last hour or two after 3248 (ignoring modem unreliability)
Jan 25 21:52:51 <freemangordon> Pali: can;t we put the device offline?
Jan 25 21:52:52 <Estel_> freemangordon,  cause they had a batch of those lying around? :P
Jan 25 21:53:04 <kerio> Pali: my (and most) calibration scripts work by charging until vdq, then disabling the charging, then polling every 5 or 10 seconds until edv1, and then restarting the charging
Jan 25 21:53:14 <kerio> all of this with the charger attached
Jan 25 21:53:15 <ShadowJK> freemangordon; historically Nokia used thermistor too.
Jan 25 21:53:26 <ShadowJK> freemangordon; back in nimh days :)
Jan 25 21:53:50 <freemangordon> ShadowJK: I see. but in n900 it has a different purpose
Jan 25 21:53:53 <kerio> do you have a reasonable way to make it wait like a minute before shutting down?
Jan 25 21:54:14 <Estel_> Pali
Jan 25 21:54:17 <Estel_> POWER_SUPPLY_NAME=rx51-battery
Jan 25 21:54:31 <Estel_> POWER_SUPPLY_CHARGE_FULL_DESIGN=1339436
Jan 25 21:54:40 <Estel_> other lines interesting too?
Jan 25 21:55:10 <freemangordon> Estel_: 1200?
Jan 25 21:55:15 <Estel_> hey, hey, wait a second. When I last checked there was sysfs entry for bq45... charger
Jan 25 21:55:21 <Estel_> freemangordon,  seems it deteriorated
Jan 25 21:55:28 <Estel_> resistor changed resistance over years of usage
Jan 25 21:55:34 <Estel_> = this strange value
Jan 25 21:55:41 <freemangordon> yeah, sure
Jan 25 21:55:47 <Estel_> ?
Jan 25 21:56:01 <freemangordon> rx51 driver is out for 2-3 months or something
Jan 25 21:56:06 <Estel_> I can check it again after two weeks and you will see it changed slightly
Jan 25 21:56:12 <Estel_> no, no
Jan 25 21:56:17 <Estel_> it was readable for ages
Jan 25 21:56:30 <freemangordon> it uses a different formula compared to nokia's binary (whatever the name was)
Jan 25 21:56:34 <Estel_> via haldon-bme or something, don't remember
Jan 25 21:56:35 <Estel_> ah
Jan 25 21:56:46 <Estel_> so, I though about that nokia's value
Jan 25 21:56:58 <freemangordon> don't :P
Jan 25 21:56:58 <Estel_> as for rx51 it's first time I check it now
Jan 25 21:57:09 <Estel_> anyway, it's not related to reality, at all
Jan 25 21:57:14 <freemangordon> it is
Jan 25 21:57:14 <Estel_> nor for dual or single scud
Jan 25 21:57:24 <Estel_> how come?
Jan 25 21:57:31 <kerio> Pali: what about something even lower, like 3000mV? or would that be too low for the modem, still?
Jan 25 21:57:33 <freemangordon> well, those "duals" are some kind of exotic
Jan 25 21:57:34 <FIQ> if anyone asked about fiasco images, I have one...
Jan 25 21:57:38 <Estel_> single scud have 1500 as new
Jan 25 21:57:46 <kerio> FIQ: the guy threw a tantrum and left
Jan 25 21:57:51 <FIQ> ah
Jan 25 21:58:00 <FIQ> RX-51_2009SE_10.2010.13-2.VANILLA_PR_EMMC_MR0_ARM.bin  RX-51_2009SE_20.2010.36-2_PR_COMBINED_MR0_ARM.bin
Jan 25 21:58:04 <FIQ> is what I have
Jan 25 21:58:05 <Estel_> freemangordon,  even without exotic things, scuds are 1500
Jan 25 21:58:07 <FIQ> though not public
Jan 25 21:58:10 <freemangordon> Estel_: nokia genuine batteries report almost correct values
Jan 25 21:58:33 <freemangordon> and that value will stay only for 2-3 recharge cycles
Jan 25 21:58:34 <Estel_> sure, we can expect non-whitstanding support of genuine batteries, for next years :|
Jan 25 21:58:48 <freemangordon> until bq gets calibrated
Jan 25 21:58:54 <Estel_> I haven't said that your idea is bad
Jan 25 21:59:09 <Estel_> it's that I see no reason why not allow user to change priority settings if desired
Jan 25 21:59:16 <Estel_> so, let it be as you said by default
Jan 25 21:59:26 <freemangordon> for sure it is better than reporting > 2000mAh :)
Jan 25 21:59:35 <Pali> kerio, I need logs why it report 110%
Jan 25 21:59:41 <Estel_> and let ones like me, or kerio, or dr_frost_dk, or hal.f maemo, change it to never use rx51 values (battery applet)
Jan 25 21:59:43 <freemangordon> on a cheapu 800mAh chinese
Jan 25 21:59:48 <Pali> without it I cannot fix problem
Jan 25 21:59:51 <freemangordon> *cheapo
Jan 25 22:00:13 <kerio> Pali: the problem is that the design reported for my battery is 1200mAh
Jan 25 22:00:17 <Estel_> well, as ShadowJK said, even laptops use last measured capacityp now design one
Jan 25 22:00:20 <ShadowJK> freemangordon; Nokia battery management is plagued by legacy. I suspect N9 and lumia are only sane devices. N900 is a hybrid :)
Jan 25 22:00:20 <kerio> when it's ~1400 in reality
Jan 25 22:00:28 <Estel_> s/now/not/
Jan 25 22:00:29 <infobot> Estel_ meant: well, as ShadowJK said, even laptops use last measured capacityp not design one
Jan 25 22:00:36 <Estel_> when not calibrated
Jan 25 22:00:40 <Pali> kerio, negative and hight value as 100 should not be reported
Jan 25 22:00:42 <Pali> never
Jan 25 22:01:16 <freemangordon> ShadowJK: sure BME in n900 is crap. But we have the chance to fix that
Jan 25 22:01:22 <Estel_> well, my battery is 3000 mAh and reported as 1300, does it count? :P
Jan 25 22:01:35 <ShadowJK> freemangordon; well it's not just sw, it's hw too
Jan 25 22:02:02 <Estel_> ok, so don't fix it by enforcing other crap on users, even for few cycles. Use default for dummies, and allow others to overwrite priorities, even by hidden config file or sysfsentrey or whatever
Jan 25 22:02:27 <Estel_> hello sixwheeledbeast :)
Jan 25 22:02:35 <FIQ> I'm still very curious how my situation is handled by bme
Jan 25 22:02:44 <FIQ> as I have two different batteries and charge both externally
Jan 25 22:03:03 <freemangordon> guys, imagine, you enter a shop and bye a battery. you put it in the device. what capacity you'd expect to be reported?
Jan 25 22:03:08 <FIQ> (both are genuine nokia batteries)
Jan 25 22:03:21 <Estel_> freemangordon,  so i said, let it be as you proposed for dummies
Jan 25 22:03:21 <sixwheeledbeast> Estel_: Hi you will have to excuse me (poor inet connection due to snow)
Jan 25 22:03:32 <Estel_> then, allow others to overwrite priority settings for applet
Jan 25 22:03:38 <Estel_> instead of hardcoding it
Jan 25 22:03:42 <kerio> freemangordon: who said that it should report a capacity? :)
Jan 25 22:03:46 <freemangordon> Estel_: ask pali for that :P
Jan 25 22:04:00 <Pali> we need %
Jan 25 22:04:14 <Pali> and percentage is caluclated as current/design
Jan 25 22:04:15 <freemangordon> kerio: because it is better to know if you've been sold an 800 mAh battery with changed sticker
Jan 25 22:04:16 <kerio> Pali: % comes from bq27k, and is as accurate as you can get, considering that the design capacity is FUBAR
Jan 25 22:04:21 <FIQ> My experience is that my charge very rarely show "full", though that can be my external charger's problem
Jan 25 22:04:28 <FIQ> though I don't think so
Jan 25 22:04:28 <Estel_> I just want to be able to set it to use bq always, and ignore rx51. SET IT, not demanding it being like that by default
Jan 25 22:04:29 <kerio> freemangordon: you can't know that until you do a cycle or two
Jan 25 22:04:35 <kerio> FIQ: update to the very latest kp52
Jan 25 22:04:39 <kerio> (january 2013)
Jan 25 22:04:49 <freemangordon> kerio: you can if only the sticker is changed
Jan 25 22:04:51 <kerio> Estel_: +1
Jan 25 22:04:54 <FIQ> kerio: hmmmmmm why?
Jan 25 22:05:10 <kerio> freemangordon: this battery reports 1200, but it's 1400 in reality
Jan 25 22:05:24 <kerio> FIQ: the charge termination sense is set at 100 instead of 150, maybe it's that
Jan 25 22:05:28 <Estel_> if we're getting rid of damn bme, lets not do enforcing hardcoded things again
Jan 25 22:05:41 <freemangordon> kerio: so? did you think it is genuine nokia 1430 mAh?
Jan 25 22:05:44 <Estel_> safe default for dummies that doesn't want to see 2048 after chip reset...
Jan 25 22:05:59 <Estel_> and configurability, even if by text file or whatever, for others
Jan 25 22:06:02 <freemangordon> kerio: I guess no
Jan 25 22:06:05 <FIQ> kerio: charge termination.. as in the time for when the charging is stopped?
Jan 25 22:06:14 <FIQ> i.e. when it's considered "full"
Jan 25 22:06:14 <kerio> FIQ: yeah
Jan 25 22:06:15 <Estel_> freemangordon,  we're using non-genuine code
Jan 25 22:06:20 <FIQ> uhh but how is that relevant
Jan 25 22:06:21 <Estel_> instead of genuine bme :P
Jan 25 22:06:27 <FIQ> I use an external charger
Jan 25 22:06:30 <Estel_> so who said we need to use genuine hw battery?
Jan 25 22:06:33 <kerio> FIQ: i have no idea, it might be a hald-addon-bme discrepancy between stock and replacement
Jan 25 22:06:34 <FIQ> <FIQ> My experience is that my charge very rarely show "full", though that can be my external charger's problem
Jan 25 22:06:47 <FIQ> "external charger"
Jan 25 22:06:56 <freemangordon> Estel_: I am not against configuratio ;)
Jan 25 22:07:04 <Pali> now hald-addon-bme doing this: when battery is not calibrated it calculate bars (0-8) from battery voltage from hardcoded values
Jan 25 22:07:04 <FIQ> anyway, I guess I can update to kp52 anyway
Jan 25 22:07:06 <Estel_> good to hear ;)
Jan 25 22:07:09 <FIQ> There's just one problem...
Jan 25 22:07:11 <kerio> FIQ: mind you, "full" is a lie, except for like the very minute the charge ends
Jan 25 22:07:16 <FIQ> extras-devel is done
Jan 25 22:07:18 <kerio> once the charge ends, the battery starts to discharge
Jan 25 22:07:20 <Pali> and I think this can be ok
Jan 25 22:07:24 <FIQ> *down
Jan 25 22:07:31 <kerio> FIQ: hold on, is your issue with the replacement bme?
Jan 25 22:07:32 <Estel_> your idea for default behavior - rx51 when not calibrated, then bq27x00 when calibrated - is sane
Jan 25 22:07:36 <FIQ> kerio: yeah
Jan 25 22:07:40 <kerio> k
Jan 25 22:07:42 <freemangordon> Pali: no other option aiui
Jan 25 22:07:42 <FIQ> kerio: Uh I think I use the stock one
Jan 25 22:07:44 <Estel_> just let us overwrite it with whatever priorities we feel fancy ;)
Jan 25 22:07:54 <FIQ> "yeah" was @ once the charge ends, the battery starts to discharge
Jan 25 22:08:01 <kerio> FIQ: then "full" is even less of a reality for you
Jan 25 22:08:02 <Estel_> Pali,  it looks good
Jan 25 22:08:06 <freemangordon> Estel_: toldya, ask Pali for that :)
Jan 25 22:08:07 <FIQ> I see
Jan 25 22:08:09 <kerio> it takes power to boot the n900 :)
Jan 25 22:08:17 <FIQ> right
Jan 25 22:08:18 <Pali> yes, so "current" mAh is calulated from "bars"
Jan 25 22:08:24 <Pali> in non caibrated state
Jan 25 22:08:25 <kerio> and bme doesn't really know wtf to do with a newly-inserted battery
Jan 25 22:08:39 <freemangordon> kerio: neither is bq :)
Jan 25 22:08:43 <FIQ> Oh well, just wondering what it tries to report
Jan 25 22:08:43 <freemangordon> aiui
Jan 25 22:08:49 <Estel_> Pali,  good idea, but why make yourself trouble by coding that, instead of just exporting priorities to be configurable by user?
Jan 25 22:09:06 <FIQ> as my batteries, while theoretically being the same, have different life lengths and probably different charges in practice
Jan 25 22:09:25 <Estel_> Pali,  even by single hidden file, we won't tell anyone :P
Jan 25 22:09:33 <freemangordon> Estel_: wh should have some sane values reporteb to not-that-smart users :)
Jan 25 22:09:39 <freemangordon> *we
Jan 25 22:09:54 <Estel_> freemangordon,  so we have "sane" rx51 now reported to less smart users
Jan 25 22:10:05 <FIQ> and I happily swap them every charge, they essentially never stay in the N900 for more than one charge as I never charge on the N900
Jan 25 22:10:16 <freemangordon> Pali was talking about bars and charge
Jan 25 22:10:23 <Estel_> smart ones can do config change. Or code it, if you want, no problem with it - just don't sacrifice configurability
Jan 25 22:10:31 <Pali> in hald-addon-bme is this code:   if (capacity < 0)    capacity = 0;  else if (capacity > 100)    capacity = 100;
Jan 25 22:10:39 <Pali> so 110 % is impossible ^^^
Jan 25 22:10:39 <Estel_> bars are nice, as long as we can disable them :P
Jan 25 22:11:14 <freemangordon> Estel_: well, i propose that for some future version
Jan 25 22:11:15 <FIQ> having an accurate % showing would probably be impossible for me, but it would be nice anyway :p
Jan 25 22:11:28 <Estel_> whatever you will do will be good, as long as one can set up - even by config file - to use bq27x00 and screw rest
Jan 25 22:11:31 <freemangordon> lets have everything in place (as logic) i guess it can be tweaked later on
Jan 25 22:11:38 <Estel_> freemangordon,  why not
Jan 25 22:11:46 <Pali> kerio ^^^^^
Jan 25 22:11:53 <kerio> i agree
Jan 25 22:12:05 <Estel_> :)
Jan 25 22:12:12 <kerio> although i still think that the default behaviour when facing a calibrated bq27200 should be to blindly report what it reports
Jan 25 22:12:13 <Estel_> now tell me you all missed me :P
Jan 25 22:12:28 <Estel_> kerio, enough :P
Jan 25 22:12:53 <Estel_> even if I agree :P
Jan 25 22:12:55 <kerio> because a calibrated bq27200 IS THE LAW
Jan 25 22:13:04 <kerio> like judge dredd
Jan 25 22:14:22 <freemangordon> kerio: I guess Pali is doing that not only for you and Estel_ :P. And if I got it right, CSSU is the long-term goal. Thus we MUST think for guys and gals who have NFC what bq or rx51 is
Jan 25 22:14:38 <Estel_> until you calibrated it with low usage, and then, do constant 1A power draw - battery will die, while reporting 600 mAh still left :
Jan 25 22:15:02 <kerio> Estel_: i usually calibrate with ~200mA of constant draw
Jan 25 22:15:03 <freemangordon> kerio: BTW did you manage to go over 470 mA?
Jan 25 22:15:20 <kerio> freemangordon: camera, torch, gps, wifi, cpu
Jan 25 22:15:25 <kerio> i got to 950mA
Jan 25 22:15:25 <kerio> :D
Jan 25 22:15:30 <freemangordon> nice :D
Jan 25 22:15:35 <kerio> for a bit
Jan 25 22:15:51 <ShadowJK> FIQ; bme has voltage curve thresholds. Primarily it tracks by counting current in/out, within specific voltage profile of a specific profiled battery. If battery is fresher or older than the profile (fixed profile, not adjustable), the reported charge level either freezes or makes a big jump. In border cases you get funny behaviour like hundred percent forever until empty :)
Jan 25 22:16:06 <xes> kerio: ..with torch application?
Jan 25 22:16:13 <Estel_> freemangordon,  sure. That's why I said configurability as an option, not enforcing anyone
Jan 25 22:16:24 <FIQ> ShadowJK: I see
Jan 25 22:16:32 <Estel_> xes, torch is quite power-savyy
Jan 25 22:16:57 <kerio> xes: with cssu's nicocam
Jan 25 22:17:01 <FIQ> the light draws much power?
Jan 25 22:17:01 <kerio> push t
Jan 25 22:17:02 <Estel_> one need recording video in hd (camera + cpu + dsp usage), gps, bluetooth, wifi without power saving, screen on at max brightness
Jan 25 22:17:04 <Estel_> etc
Jan 25 22:17:06 <Estel_> + torch
Jan 25 22:17:12 <Estel_> to get above 1A
Jan 25 22:17:26 <Estel_> light is around 100mA IIRC
Jan 25 22:17:39 <freemangordon> ooh, camera focus is about 100mA too, iirc
Jan 25 22:17:43 <SpeedEvil> and download a file over 3g
Jan 25 22:17:46 <Estel_> which makes N900 a 30h flashlight with dual-scud
Jan 25 22:17:54 <SpeedEvil> with the vibrator on
Jan 25 22:17:55 <kerio> heh, i didn't really want to stress the system *too* much
Jan 25 22:18:05 <kerio> vibrator and camera focus engine all firing up at the same time
Jan 25 22:18:06 <Estel_> SpeedEvil,  or send zeroes or urandom through 3G :P
Jan 25 22:18:11 <freemangordon> kerio: why not, don;t you have a second device? :P
Jan 25 22:18:23 <kerio> freemangordon: i should really do something with that one
Jan 25 22:18:28 <Estel_> lets write app "battery killer"
Jan 25 22:18:34 <Estel_> that enables everything at once
Jan 25 22:18:39 <Estel_> and keeps it that way
Jan 25 22:18:48 <xes> there is way to measure the usb host mode current?
Jan 25 22:18:55 <Estel_> sure
Jan 25 22:18:58 <kerio> xes: 200mA
Jan 25 22:18:59 <kerio> flat
Jan 25 22:19:03 <kerio> ...kinda
Jan 25 22:19:03 <Estel_> check what bq27x00 reports
Jan 25 22:19:04 <freemangordon> Estel_: I think it is way easier to use a hammer to kill the battery :D
Jan 25 22:19:18 <Estel_> freemangordon,  as a discharge tool
Jan 25 22:19:24 <Estel_> + activate it via ssh
Jan 25 22:19:28 <Estel_> to scarry people
Jan 25 22:19:28 <DocScrutinizer05> xes: no, you can't measure VBUS-BOOST
Jan 25 22:19:45 <Estel_> DocScrutinizer05,  increased usage is shown as difference in current-now
Jan 25 22:19:48 <DocScrutinizer05> you can detect overcurrent cutout condition though, which will happen at 200mA
Jan 25 22:19:54 <Estel_> mA draw of devixe changes
Jan 25 22:20:13 <Estel_> with known values without device attached, one can measure how much device attached uses
Jan 25 22:20:28 <DocScrutinizer05> sure, but that will usually dissolve in noise of normal powerconsumption of device
Jan 25 22:20:30 <xes> i'm playing with an rtl-sdr receiver dongle but sometimes hangs ..i was supposing it was a problem with the usb current..
Jan 25 22:20:38 <Estel_> s/devixe/device
Jan 25 22:21:10 <DocScrutinizer05> xes: use a powerd hub
Jan 25 22:21:13 <kerio> xes: put a powered usb hub between the receiver and the n900
Jan 25 22:21:18 <Estel_> or y-cable
Jan 25 22:21:19 <DocScrutinizer05> :-P
Jan 25 22:21:22 <kerio> or a y-cable, yes
Jan 25 22:21:28 <kerio> goddammit guys stop typing so fast
Jan 25 22:21:33 <Estel_> my y-cable got name, even
Jan 25 22:21:38 <Estel_> "Hammer of Thor"
Jan 25 22:21:47 <Estel_> need to see it to understand
Jan 25 22:22:03 * kerio sometimes uses two y-cables to draw 1A from a computer
Jan 25 22:22:19 <DocScrutinizer05> my Y-cable got a anme as well: Joerg's Y-cable (invented in 2008 iirc, for Openmoko)
Jan 25 22:22:20 <Estel_> ...and you end up drawing 200 mA
Jan 25 22:22:21 <xes> i'll try..but the fun would be a complete portable sdr receiver ;)
Jan 25 22:22:27 <Estel_> due to lack of negotiation
Jan 25 22:22:56 <Estel_> whatever sdr is
Jan 25 22:22:56 <kerio> Estel_: nah, my laptop and/or os x aren't picky
Jan 25 22:23:05 <kerio> xes: battery pack?
Jan 25 22:23:14 <Estel_> doesn't matter, usb specification won't let it
Jan 25 22:23:14 <kerio> there's very little you can do, if 200mA aren't enough
Jan 25 22:23:24 <kerio> Estel_: apple doesn't care about your puny specifications
Jan 25 22:23:26 <Estel_> battery pack is cheap
Jan 25 22:23:34 <kerio> or something
Jan 25 22:23:36 <xes> SDR? http://superkuh.com/rtlsdr.html
Jan 25 22:23:37 <Estel_> and 5V stabilizer is free from maxim
Jan 25 22:24:57 <kerio> xes: neat!
Jan 25 22:25:17 <Estel_> xes, looks neat, but what does it do?
Jan 25 22:25:22 <Estel_> what samples over what?
Jan 25 22:25:33 <Estel_> radio amateur?
Jan 25 22:25:38 <Pali> Estel_, kerio, freemangordon: create wiki package and possible solutions what to show in battery widget and how to calculate it
Jan 25 22:25:51 <Pali> also what should hald-addon-bme report
Jan 25 22:25:55 <Estel_> wiki works @ 7b/s for me
Jan 25 22:26:03 <Pali> and what should be "0" as unknown value
Jan 25 22:26:04 <Estel_> so expect it in dawn of 2015
Jan 25 22:26:12 <kerio> Pali: would you be ok with not displaying a capacity in some situations?
Jan 25 22:26:32 <freemangordon> Pali: would you mind to put what's curren in cssu-devel?
Jan 25 22:26:37 <freemangordon> *current
Jan 25 22:26:44 <Pali> kerio, percentage must be always reported
Jan 25 22:26:55 <kerio> i meant capacity as in max capacity
Jan 25 22:26:55 <Pali> unknown design capacity is "0"
Jan 25 22:26:57 <kerio> the number after /
Jan 25 22:27:05 * DocScrutinizer05 announcing service downtime of... DocScrutinizer05
Jan 25 22:27:05 <Pali> yes, it can be 0
Jan 25 22:27:08 <xes> Estel: yes.... imagine you can receive any kind of signal....and decode with a software block of code..
Jan 25 22:27:11 * kerio waves
Jan 25 22:27:14 <kerio> take care doc
Jan 25 22:27:18 <kerio> and sleep!
Jan 25 22:27:23 <Hurrian> Pali, for user experience sake, could you set Design Capacity to 1300?
Jan 25 22:27:26 <DocScrutinizer05> live!
Jan 25 22:27:28 <Estel_> xes, what, cryptography?
Jan 25 22:27:30 <Hurrian> (if unset)
Jan 25 22:27:30 <DocScrutinizer05> sleep is overrated
Jan 25 22:27:37 <freemangordon> Hurrian: why?
Jan 25 22:27:42 <kerio> sleep is overrated until you're tired ;)
Jan 25 22:27:46 <freemangordon> that is what BME does
Jan 25 22:27:51 <Hurrian> freemangordon, ah
Jan 25 22:27:56 <freemangordon> and it drives me nuts
Jan 25 22:27:58 <freemangordon> :D
Jan 25 22:27:59 <Estel_> xes, cryptography over analog (or digital?) radio signal?
Jan 25 22:28:08 <Pali> for now I wil not touch desing/max capacity code in status applet and hald-addon-bme
Jan 25 22:28:13 <xes> Estel: the limit is fantasy and... cpu power
Jan 25 22:28:15 <freemangordon> Hurrian: well, 1260, but it is the same
Jan 25 22:28:25 <DocScrutinizer05> kerio: I had lots of it since yesterday, none since monday before that
Jan 25 22:28:28 <Pali> create wiki page with solutions and we can discuss about variants
Jan 25 22:28:38 <Estel_> xes, whoa, interesting, so it's kind of long-range wifi, just low on speed?
Jan 25 22:28:49 <DocScrutinizer05> kerio: but my mind need a defragmentation
Jan 25 22:29:02 <DocScrutinizer05> so: o/
Jan 25 22:29:08 <Estel_> cpu power of computer, or of dongle chip?
Jan 25 22:29:14 <freemangordon> Pali: push the current version in cssu-devel please, so we can work on changes/improvements based on some real-life scenarios
Jan 25 22:29:20 <Hurrian> I was simply thinking, if replacement BME plus open battery module gets pushed into mainline CSSU, there'll be an entire storm of people bitching about their battery always being at 0%
Jan 25 22:29:32 <Pali> freemangordon, last version is in cssu-devel
Jan 25 22:29:43 <Pali> of status applet
Jan 25 22:29:44 <DocScrutinizer05> meh, for the friggin bot:
Jan 25 22:29:45 <kerio> Hurrian: nah, we want to avoid that
Jan 25 22:29:46 <kerio> :)
Jan 25 22:29:49 <DocScrutinizer05> kerio: but my mind need a defragmentation, so o/
Jan 25 22:30:04 <kerio> ~seen DocScrutinizer05 
Jan 25 22:30:08 <Pali> it should be version 1.0-2
Jan 25 22:30:08 <freemangordon> Pali: aah, the one I deinstalled because of unusable bars with BME? :)
Jan 25 22:30:10 <infobot> docscrutinizer05 is currently on #maemo #n9 #harmattan #meego #openmoko #maemo-ssu #infobot #openmoko-cdevel. Has said a total of 49 messages. Is idling for 20s, last said: 'kerio: but my mind need a defragmentation, so o/'.
Jan 25 22:30:42 <xes> Estel: with that dongles you can receive from about 60MHz to 2000MHz... and decode any kind of modulation or signal
Jan 25 22:30:50 <xes> bye DocScrutinizer05!
Jan 25 22:31:23 <Estel_> xes, whoa
Jan 25 22:31:36 <Estel_> need to read more about it, looks like my kind of heaven for modder
Jan 25 22:31:40 * freemangordon waves
Jan 25 22:31:45 <kerio> xes: how can you send stuff, though?
Jan 25 22:31:51 <Estel_> ~60 mhz should be quite long-range capable
Jan 25 22:32:01 <Estel_> kerio, those dongles send it
Jan 25 22:32:07 <Estel_> instead of receiving
Jan 25 22:32:07 <kerio> oh
Jan 25 22:32:10 <kerio> ...oh
Jan 25 22:32:12 <Estel_> when properly treated
Jan 25 22:32:14 <kerio> OH!
Jan 25 22:32:31 <Estel_> with simple mod (very easy) and kernel modules and userland bits, approriate
Jan 25 22:32:42 <Estel_> using this on the go with N900
Jan 25 22:32:49 <Estel_> = more than mesh network
Jan 25 22:32:54 <kerio> Estel_: maybe you can
Jan 25 22:33:02 <kerio> with your bigass battery
Jan 25 22:33:07 <xes> no kernel modules... only libs reading directly from usb
Jan 25 22:33:12 <Estel_> just make yourself liIon belt
Jan 25 22:33:19 <Estel_> xes, lol
Jan 25 22:33:23 <Estel_> even better
Jan 25 22:33:36 <kerio> Estel_: played minecraft ftb lately? :D
Jan 25 22:33:43 <xes> until i'm receiving fm radio...and fmtx :)
Jan 25 22:33:47 <Estel_> kerio, buy batteries best in price/capacity ration, and prepare bracelet/belt
Jan 25 22:33:55 <Estel_> 20 000mAh in your belt :P
Jan 25 22:34:23 <Estel_> kerio, I ot only vague idea of what minecraft is
Jan 25 22:34:41 <kerio> a game where you go around and build things
Jan 25 22:34:49 <kerio> there's a popular modpack that adds a bunch of tecnical things
Jan 25 22:34:52 <Estel_> xes, receiving and transmitting fm via those dongles is fun, but sending digital radio signal is more fun
Jan 25 22:35:09 <kerio> including a bigass battery pack that you can wear instead of your chestplate, to automatically charge the electrical tools you use
Jan 25 22:35:13 <Estel_> kerio, make it work on N900, the open source version :P
Jan 25 22:35:20 <Estel_> in multiplayer
Jan 25 22:35:24 * SpeedEvil has a 528000mAh 'USB battery'
Jan 25 22:35:28 <kerio> i know some of the minetest dudes
Jan 25 22:35:30 <Estel_> we have too few online multiplayer games for N900
Jan 25 22:35:33 <kerio> (minecraft clone)
Jan 25 22:35:34 <SpeedEvil> it is not very portable.
Jan 25 22:35:47 <Estel_> SpeedEvil,  @ what voltage?
Jan 25 22:35:53 <SpeedEvil> 5v
Jan 25 22:36:03 <kerio> Estel_: anyway, scuds are still 20€/piece, or something, right?
Jan 25 22:36:10 <Estel_> 528A? and it weight ten kilos?
Jan 25 22:36:13 <SpeedEvil> it's really 2*12v*110ah, with a SMPS
Jan 25 22:36:21 <SpeedEvil> Estel_: no, not 10kg
Jan 25 22:36:23 <Estel_> kerio, yes, but buying new, I recommend polarcell
Jan 25 22:36:28 <kerio> right
Jan 25 22:36:31 <Estel_> some claims new ones are 1700 mAh
Jan 25 22:36:41 <kerio> but there's a 10€ battery on aliexpress that's still decent :D
Jan 25 22:36:51 <kerio> vi____ got a 1600mAh one
Jan 25 22:37:02 <Estel_> kerio, price/quality battery for belt should cost around 3 euro
Jan 25 22:37:07 <Estel_> if bl-5j
Jan 25 22:37:13 <Estel_> but for belt you can use other types
Jan 25 22:37:18 <Estel_> with better density
Jan 25 22:37:25 <Estel_> SpeedEvil,  seriously, how much it weight?
Jan 25 22:37:39 <Estel_> and 528A is after taking into account conversion losses?
Jan 25 22:37:41 <kerio> Estel_: i'm still waiting for the case redesign for the two 18650s, btw
Jan 25 22:38:07 <SpeedEvil> Estel_: knock off 10*ish
Jan 25 22:38:10 <Estel_> kerio, wait hapilly, it's proving so irritating, that I'm thinking about dropping it in favor of 3 flat batteries
Jan 25 22:38:12 <SpeedEvil> 10%
Jan 25 22:38:28 <SpeedEvil> Estel_: maybe 40kg
Jan 25 22:38:35 <Estel_> SpeedEvil,  I see. How much it weight, approxi...?
Jan 25 22:38:40 <Estel_> lol
Jan 25 22:38:56 <Estel_> I though "no, no 10kg" was about that it's less heavy :P
Jan 25 22:39:00 <SpeedEvil> it's got a raspberry pi plugged into it
Jan 25 22:39:14 <SpeedEvil> and a 750w ups
Jan 25 22:39:22 <Estel_> your pi is going to get quite overweight fedding into it
Jan 25 22:39:39 <Estel_> kerio, btw
Jan 25 22:39:49 <Estel_> do you have any random 18650of whatever state
Jan 25 22:39:54 <Estel_> living or not
Jan 25 22:40:02 <Estel_> to send me for some tests?
Jan 25 22:40:11 <kerio> nope
Jan 25 22:40:22 <kerio> i'm not a hardware guy at all
Jan 25 22:40:24 <Estel_> damn, will need to scavenge some
Jan 25 22:40:33 <Estel_> from old battery or whatever
Jan 25 22:40:43 <Estel_> thickness it's not problem
Jan 25 22:40:48 <Estel_> they're just little too long
Jan 25 22:41:12 <Estel_> hard to put them there, considering secondary board with camera, even in V configuration
Jan 25 22:41:29 <kerio> Estel_: what about shorter batteries?
Jan 25 22:41:29 <Estel_> well, will give it some more love
Jan 25 22:41:52 <Estel_> shorted would do it, but 14... ones doesn't have benefits of 18
Jan 25 22:42:02 <Estel_> more expensive, less popular etc
Jan 25 22:42:13 <Estel_> so they stop being better than flat ones, at all
Jan 25 22:42:34 <Estel_> choosing between two 14... and 3 flatties, I take 3 flatties
Jan 25 22:42:46 <Estel_> althoug 18650 x 2 own even 3 flatties
Jan 25 22:42:57 <kerio> Estel_: what can you do with the space of the stock cover?
Jan 25 22:43:10 <kerio> say, something that fits inside an otterbox
Jan 25 22:43:23 <Estel_> stock covera? nothing, if it comes to battery
Jan 25 22:43:30 <kerio> :(
Jan 25 22:43:34 <Estel_> jst better kickstand
Jan 25 22:43:36 <kerio> is it *that* tightly packed?
Jan 25 22:43:39 <Estel_> etc
Jan 25 22:44:04 <Estel_> well, look at your n900 with stock cover, where you want to put another battery
Jan 25 22:44:07 <xes> Estel_: http://www.g0hww.net/2012/06/rtl-sdr-port-of-gr-air-modes-feeding.html
Jan 25 22:44:22 <Estel_> if you mean body replacement, it will have all goddies in space of stock cover...
Jan 25 22:44:27 <FIQ> i've considered buying higher capacity batteries, but now that I have an external charger with not so optimal layout, there's no point
Jan 25 22:44:27 <Estel_> except battery
Jan 25 22:44:41 <Estel_> xes, thanks
Jan 25 22:45:07 <xes> ;) feel the fantasy growing
Jan 25 22:46:23 <Estel_> hole shit, 137 miles
Jan 25 22:46:41 <Estel_> holy*
Jan 25 22:47:08 <Estel_> I just hope that enough talented coders work on userland coding/decoding bits for it
Jan 25 22:47:54 <Estel_> I wonder, btw, how high thoroughtput (bytes per second) for wireless communication would be achievable through it
Jan 25 22:48:15 <Estel_> even if it would work for  10 miles reliably, it would be nice revolution for mesh networks
Jan 25 22:48:34 <xes> it is only a receiver..
Jan 25 22:48:46 <xes> no transmission
Jan 25 22:48:49 <Estel_> and "we" would be ready to recreate free internet, when "they" take control over it ;)
Jan 25 22:48:51 <Estel_> I know
Jan 25 22:48:53 <Estel_> but
Jan 25 22:49:11 <Estel_> transmission is also possible, or not? that is how I understood it from first link
Jan 25 22:49:25 <Estel_> "sending 8bit samples at frequencies..." etc
Jan 25 22:49:40 <xes> not with a 6 euro device :)
Jan 25 22:49:48 <futpib_> so powerfull transmitter on usb-power, no way
Jan 25 22:50:14 <Estel_> still, receiving capabilities are quite amazing
Jan 25 22:51:03 <xes> in a few days i'll post all the packages compiled fot the n900
Jan 25 22:51:11 <xes> *for
Jan 25 22:51:16 <Estel_> whoa, nice
Jan 25 22:51:41 <Estel_> just ping me with link if possible (or PM on TMO, if it will still work and I won't get banned in my sleep)
Jan 25 22:52:11 <Estel_> next step - receiving signal from ISS ;)
Jan 25 22:53:24 <xes> someone has already received the iss with this...
Jan 25 22:53:54 <ShadowJK> Ugh, i've got massive backlog here. I'll reply in order, and apologies for redundant replies...
Jan 25 22:54:33 <ShadowJK> FIQ: Boot-time bme current charge level (reporting.current) is pure guess. It's +-20% in best case.
Jan 25 22:56:05 <ShadowJK> Pali, not calibrated, battery bars from voltage, sounds like one half of bme's current algo (which is half current based and half voltage based :))
Jan 25 22:56:19 <ShadowJK> (good approach overall tho for uncalibrated sensors)
Jan 25 22:58:49 <ShadowJK> FIQ, in theory, accurate metering for your batteries (and constant swapping) would be possible, but it would be alot of work, and prone to error :)
Jan 25 22:59:01 <ShadowJK> (tl:dr; not worth the effort)
Jan 25 23:02:45 <FIQ> ok
Jan 25 23:04:34 <ShadowJK> On diablo I do this: Measure instantaneous current every 1s, integrate that. If accumulated value < 0, set value to 0. If voltage < 3400, set value to 0. If value > (user determined maximum capacity), set value to maximum user determined capacity.
Jan 25 23:04:48 <ShadowJK> It tends to drift upwards :(
Jan 25 23:05:02 <ShadowJK> Which is why I have max capa thing
Jan 25 23:05:20 <SpeedEvil> 5 seconds
Jan 25 23:05:27 <ShadowJK> Diablo
Jan 25 23:05:31 <SpeedEvil> the ADC polls every 5.12s IIRC
Jan 25 23:05:33 <SpeedEvil> oh
Jan 25 23:05:33 <ShadowJK> does not run on N900
Jan 25 23:05:34 <SpeedEvil> nvm
Jan 25 23:05:38 <SpeedEvil> sortty
Jan 25 23:05:48 <Estel_> xes, just add strong, directional transmitter, and build radar :)
Jan 25 23:06:10 <Estel_> thing with planes is half of it, already
Jan 25 23:06:17 <Estel_> based on receiving
Jan 25 23:06:40 <ShadowJK> On N800 max capa is 800 (originally 1200), N810 1500 (should be 1300 or so, bought new batt last year)
Jan 25 23:07:21 <ShadowJK> I don't have have "full charge detection" in my script on Diablo... something that bq runs in hw
Jan 25 23:08:28 <ShadowJK> Ran my diablo script (with 5s period instead of 1s) on fremantle too (with mods), and "full charge" could vary by +-100mAh depending on system load and stuff
Jan 25 23:23:59 <Estel_> ~Pali
Jan 25 23:24:00 <infobot> pali is, like, http://atrey.karlin.mff.cuni.cz/~pali/
Jan 26 00:36:16 <pauly1210> Can i use Maemo 5 SDK on x64 linux?
Jan 26 00:37:04 <pauly1210> the maemo wiki is down but the nokia developer page says a 32 bit ubnutu would work>
Jan 26 00:37:09 <pauly1210> ?
Jan 26 00:38:38 <ShadowJK> I used the VM image on linux
Jan 26 00:40:51 <pauly1210> ShadowJK: ya i dont think scratchbox would work, i forgot about the vm thanks
Jan 26 00:42:09 <ShadowJK> All the infra is in a state of flux right now, so I dunno what's up and what's down or whether you'll be able to download the vm or not :/
Jan 26 00:44:45 <pauly1210> ShadowJK: how about just getting Qt Creator ?
Jan 26 00:46:05 <ShadowJK> heh
Jan 26 00:46:12 <ShadowJK> If it works for you, why not
Jan 26 00:55:32 <bef0rd> get qtcreator from nokia
Jan 26 00:55:45 <bef0rd> Nokia Qt SDK its called I believe
Jan 26 00:56:16 <bef0rd> http://www.developer.nokia.com/info/sw.nokia.com/id/da8df288-e615-443d-be5c-00c8a72435f8/Qt_SDK.html
Jan 26 01:01:49 <pauly1210> bef0rd: thanks downloading now
**** ENDING LOGGING AT Sat Jan 26 02:59:59 2013