**** BEGIN LOGGING AT Sun Feb 07 02:59:58 2010
Feb 07 09:06:47 <jlmoko> ~log
Feb 07 09:06:48 <apt> well, log is http://logs.nslu2-linux.org/livelogs/openmoko-cdevel/ for #openmoko-cdevel only. Maybe you meant "~logs" ?
Feb 07 09:33:37 * pieterc is listening at mickey|zzZZzz at FOSDEM
Feb 07 09:33:50 <pieterc> s/at/to/
Feb 07 09:34:32 <Weiss> :D
Feb 07 09:43:23 <radekp> larsc: hi, i think i have idea why wifi does not survive suspend/resume on 2.6.32
Feb 07 09:43:39 <radekp> i have backported the save_regs and restore_regs from http://git.openmoko.org/?p=kernel.git;a=commitdiff;h=a560d8579b4b009f847972667b7597c89030e26c
Feb 07 09:44:17 <radekp> now i have SD card working, wifi working - everything survives suspend/resume
Feb 07 09:46:11 <PaulFertser> radekp: shouldn't wifi reinitialize properly on itself, without the persist option?
Feb 07 09:49:31 <radekp> PaulFertser: good question, but i dont have answer :)
Feb 07 09:51:30 <radekp> i think the patch is related to s3cmci - because it stops working after suspend/resume
Feb 07 09:51:34 <PaulFertser> radekp: Werner said several times that ar6000 module should fully turn the device off in suspend, that it's equivalent to unbinding on unloading the module.
Feb 07 09:51:36 <pieterc> want me to ask mickey? ;-)
Feb 07 09:52:06 <larsc> mickeys talk is boring ;)
Feb 07 09:52:06 <PaulFertser> radekp: but s3cmci shouldn't stop working without some obscure module parameters anyway, should it? ;)
Feb 07 09:52:18 <pieterc> larsc: it's pretty interesting
Feb 07 09:52:22 <pieterc> although i'm irc ;-)
Feb 07 09:52:34 <pieterc> larsc: he's a good talker
Feb 07 09:52:41 <pieterc> s/talker/speaker/
Feb 07 09:52:41 <apt> pieterc meant: larsc: he's a good speaker
Feb 07 09:53:13 <pieterc> Uh I suck at English today ;)
Feb 07 09:54:19 <larsc> pieterc: tell me, when the talk is done, I'll come over then
Feb 07 09:54:22 <radekp> PaulFertser: i though that the parameter is always 1 on andytracking
Feb 07 09:54:47 <wpwrak> PaulFertser: at least it worked like this the last time i touched it (which has been a while)
Feb 07 09:55:50 <PaulFertser> radekp: it doesn't seem so, persist is 0 by default if i read the code right.
Feb 07 09:56:58 <pieterc> larsc: 2 minutes left
Feb 07 09:58:17 <lindi-> PaulFertser: have you btw ever heard that anybody would have got the "wake on wlan" support working?
Feb 07 09:58:49 <PaulFertser> lindi-: nope
Feb 07 09:58:59 <PaulFertser> lindi-: it should be perfectly possible hardware-wise
Feb 07 09:59:14 <larsc> well, if the wireless is powerdown on suspend, you won't get wol
Feb 07 09:59:41 <larsc> radekp: did you only backport the save_regs patch?
Feb 07 10:01:13 <pieterc> larsc: done
Feb 07 10:01:32 <radekp> larsc: yes
Feb 07 10:01:41 <larsc> radekp: ok, i'll take a look
Feb 07 10:01:46 <PaulFertser> larsc: there should be a module option to allow to leave wireless on in suspend.
Feb 07 10:02:29 <larsc> pieterc: ok :)
Feb 07 10:05:01 <radekp> PaulFertser: i havent found in andy-tracking where the "persist" variable is set, so question is if it is by default 0 or 1
Feb 07 10:05:22 <PaulFertser> radekp: static variables are 0 when not initialized explicitly :)
Feb 07 10:07:21 <radekp> PaulFertser: ok, thanks, so then there must be another reason why it works in andy-tracking and in 2.6.32 not
Feb 07 10:07:44 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r9a78baf0edd7 10/fsogsmd/ (4 files in 3 dirs): fsogsmd: modem_qualcomm_palm is now an optional feature (since it needs a special library)
Feb 07 10:07:45 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * rb7b0ff03493e 10/fsogsmd/ (configure.ac src/plugins/Makefile.am): 
Feb 07 10:07:45 <CIA-47> freesmartphone.org: fsogsmd: support for modem plugins requiring libgsm0710mux is now optional. Please submit --enable-libgsm0710mux to
Feb 07 10:07:45 <CIA-47> freesmartphone.org: enable those.
Feb 07 10:07:51 <pieterc> mickey|fosdem: great talk :)
Feb 07 10:08:03 <CIA-47> freesmartphone.org: 03mickey 07msmcomm * r34ee5b6c6108 10/ (7 files in 3 dirs): libmsmcomm: rename .pc file and just files; declare enums as public and fix certain things in .vapi
Feb 07 10:08:03 <CIA-47> freesmartphone.org: 03mickey 07msmcomm * rfa8c2924753f 10/openembedded/msmcommd_git.bb: msmcommd: remove OE recipe; has been committed upstream
Feb 07 10:08:03 <CIA-47> freesmartphone.org: 03mickey 07msmcomm * r15bea7a8fe3a 10/libmsmcomm/vapi/msmcomm-1.0.vapi: libmsmcommd: vapi compiles now
Feb 07 10:08:04 <CIA-47> freesmartphone.org: 03mickey 07msmcomm * r79fb807bc571 10/msmvterm/ (15 files in 3 dirs): msmvterm: add build skeleton (copied from fso-term)
Feb 07 10:08:04 <CIA-47> freesmartphone.org: 03mickey 07msmcomm * rfaaaa301e8e8 10/msmcommd/src/Makefile.am: msmcommd: remove msmc_forward, the functionality has been added to serial_forward
Feb 07 10:08:31 <CIA-47> freesmartphone.org: 03mickey 07phonesim * r529f1ef9c2e6 10/launch.sh: add launch.sh
Feb 07 10:08:58 <mickey|fosdem> pieterc: thanks a lot :)
Feb 07 10:09:16 <radekp> btw here is the patch for 2.6.32: http://activationrecord.net/radekp/pub/save_regs.patch
Feb 07 10:10:32 <pieterc> mickey|fosdem: I wonder where Rui is though
Feb 07 10:10:32 <pieterc> Met him yesterday and told me he would come
Feb 07 10:10:41 <pieterc> probably suffering from a terrible hangover ;)
Feb 07 10:10:54 <pieterc> oooh
Feb 07 10:10:55 <pieterc> there he is:D
Feb 07 10:11:27 <daniele_athome> morning..
Feb 07 10:12:42 <jlmoko> rui is near the botom port  standing up
Feb 07 10:14:19 <larsc> radekp: you can read the value of the param via sysfs, so you could check if it's either 1 or 0
Feb 07 10:14:47 <radekp> larsc: oki, i'll try
Feb 07 10:14:59 <larsc> i think it is /sys/module/s3cmci/params/persist
Feb 07 10:17:29 <mickey|fosdem> vegyraupe: hey! tried to find you yesterday, where are you hiding? :)
Feb 07 10:18:23 <vegyraupe> mickey|fosdem: hey, I am in the hacker room in the H building
Feb 07 10:18:33 <mickey|fosdem> vegyraupe: ah, 115
Feb 07 10:18:49 <vegyraupe> mickey|fosdem: no that is the one in the other buildung
Feb 07 10:18:53 <vegyraupe> building
Feb 07 10:19:05 <vegyraupe> I am in the main building, where all the distros are and usch
Feb 07 10:19:23 <mickey|fosdem> ah, not AW115, right
Feb 07 10:20:04 <mickey|fosdem> i'm pretty busy until 15 or so, perhaps we can share a coffee somewehre then
Feb 07 10:22:16 <vegyraupe> mickey|fosdem: yeah after 15 sounds good
Feb 07 10:22:19 <vegyraupe> got a talk at 5
Feb 07 10:22:31 <vegyraupe> so between 15 and 17 sounds good
Feb 07 10:28:00 <mickey|fosdem> oki
Feb 07 10:37:56 <playya__> moin
Feb 07 10:41:30 <daniele_athome> hi there, guys do you mind if I borrow your alsa source code from libphoneui?
Feb 07 10:41:49 <daniele_athome> because I see it's really complicated and I wish to focus on my application...
Feb 07 11:10:47 <playya__> daniele_athome, afaik it is GPLed ...
Feb 07 11:11:16 <daniele_athome> playya__, yes but I just wanted you to know i'm using it :)
Feb 07 11:26:37 <mrmoku> daniele_athome: just use it... and enhance it ;)
Feb 07 11:27:04 <daniele_athome> mrmoku: it's what i intend to do :)
Feb 07 12:16:09 <JaMa|Gone> ynezz: try nandwrite 37M http://build.shr-project.org/shr-unstable/images/om-gta01/shr-fso2-console-image-eglibc-ipk--20100207-om-gta01.rootfs.ubi
Feb 07 12:16:20 <JaMa|Gone> ynezz: then mount
Feb 07 12:17:45 <PaulFertser> ynezz: do not forget to flasherase_all before nandwrite
Feb 07 12:17:45 <PaulFertser> (just in case, sorry if you already know that)
Feb 07 12:18:42 <topro> PaulFertser: ping
Feb 07 12:20:10 <PaulFertser> topro: pong
Feb 07 12:20:32 <topro> PaulFertser: hi, I played a little with hstes but didn't get far
Feb 07 12:20:42 <topro> paired notebook with my car
Feb 07 12:21:25 <topro> hstest record some_filename and:the:mac:addr:of:car
Feb 07 12:21:41 <topro> --> Can't connect RFCOMM channel: Host is down 
Feb 07 12:22:42 <PaulFertser> topro: was it the same error message that car gave you right after pairing (that something was unsuccessful)?
Feb 07 12:23:06 <topro> no, pairing looked fine
Feb 07 12:23:55 <topro> the car display shows something like "Connecting..." for an indefinite ammount of time
Feb 07 12:24:23 <topro> but at that moment pairing is already completed successfully
Feb 07 12:24:46 <topro> I think it tries to get some service from the "phone"
Feb 07 12:26:05 <PaulFertser> topro: btw i'd try hstest play first
Feb 07 12:26:05 <PaulFertser> (record/play) But in fact it shouldn't affect your error.
Feb 07 12:26:05 <PaulFertser> topro: i'd gather hcidump of a full pairing attempt and then of an hstest play attempt.
Feb 07 12:26:05 <PaulFertser> topro: and then wrote a mail to the bluez folks. Most probably they'll have some ideas.
Feb 07 12:26:44 <topro> I tried play as well, same result
Feb 07 12:27:38 <topro> I'll go for hci-dumping -> bluez-folks
Feb 07 12:27:42 <topro> keep you informend
Feb 07 12:29:36 <PaulFertser> topro: thanks
Feb 07 12:30:00 <PaulFertser> topro: have you read on the ml about some advances on obexftp front to make it possible to retrieve contact list and the like?
Feb 07 12:30:16 <PaulFertser> Heiner-Fosdem: do you probably know who was involved in that?
Feb 07 12:30:52 <Heiner-Fosdem> PaulFertser: obexftp or obexd?
Feb 07 12:31:01 <topro> PaulFertser: no, I havent, sorry
Feb 07 12:32:29 <PaulFertser> Heiner-Fosdem: obexd i guess
Feb 07 12:32:49 <Heiner-Fosdem> PaulFertser: that's me who is integrating FSO support
Feb 07 12:33:50 <Heiner-Fosdem> my plugin should work, but PBAP integration in obexd itself isn't working with real devices
Feb 07 12:34:41 <Heiner-Fosdem> but work on that is going on, perhaps it will work with obexd 0.22
Feb 07 12:35:28 <PaulFertser> Heiner-Fosdem: i want to draw your attention to topro's problems, they might be related :)
Feb 07 12:36:30 <Heiner-Fosdem> topro: what are you trying to do?
Feb 07 12:36:45 <Heiner-Fosdem> just pair the phone to the car?
Feb 07 12:37:12 <topro> Heiner-Fosdem: pairing works, but using hands-free doesn't :(
Feb 07 12:37:54 <Heiner-Fosdem> topro: hmm, i never tried to use a bluetooth handsfree
Feb 07 12:38:07 <Heiner-Fosdem> i expected it to work like discribed in the wiki
Feb 07 12:38:36 <topro> as far as i know BTHFP is different from Headset
Feb 07 12:39:02 <Heiner-Fosdem> ah, then that isn't implemented
Feb 07 12:39:53 <Weiss> PaulFertser: if the battery is charging, then should current_now be negative?
Feb 07 12:39:59 <topro> thats the reason I expected
Feb 07 12:40:10 <Heiner-Fosdem> topro: I thought that HFP is just for dialing, and recognizing the buttons and such thinge
Feb 07 12:40:31 <topro> Heiner-Fosdem: don't know for sure
Feb 07 12:41:20 <topro> Heiner-Fosdem: http://en.wikipedia.org/wiki/Bluetooth_profile#Hands-Free_Profile_.28HFP.29
Feb 07 12:41:28 <topro> maybe that gives us some clue
Feb 07 12:42:32 <Heiner-Fosdem> HFP and HSP are both providing an audio stream
Feb 07 12:42:47 <Heiner-Fosdem> so perhaps bluez4 only supports HSP till now
Feb 07 12:43:09 <Heiner-Fosdem> on bluez ML there is currently a lot about HFP
Feb 07 12:43:38 <Heiner-Fosdem> Deubeuliou: where are you now?
Feb 07 12:44:03 <PaulFertser> Weiss: "when charging the current reported should be negative"
Feb 07 12:44:28 <topro> Heiner-Fosdem: for testing I already found an interesting link to a sourceforge project to simulate the HFP kit. maybe that helps http://nohands.sourceforge.net/
Feb 07 12:44:55 <topro> didn't try in yet (as I have a real HFP-kit available ;) )
Feb 07 12:44:55 <Heiner-Fosdem> topro: i will look at that when i'm back home
Feb 07 12:44:58 <Heiner-Fosdem> thx
Feb 07 12:45:52 <topro> what can I do to help bluez-HFP
Feb 07 12:46:42 <Heiner-Fosdem> topro: i started  a plugin to support dialing, and such things with FSO
Feb 07 12:46:43 <Weiss> PaulFertser: hmm... with 2.6.32, it claims to be charging, but it's actually discharging (quite slowly though), current_now is 110812
Feb 07 12:47:15 <Heiner-Fosdem> topro: but i don't know about the rest of bluez
Feb 07 12:47:28 <PaulFertser> LOL: "this HFP
Feb 07 12:47:28 <PaulFertser> is the only BT profile installed on Maemo's N900 and actually there is no way
Feb 07 12:47:28 <PaulFertser> to have the caller name on car BT device (we try Audi/Porsche and Ferrari
Feb 07 12:47:28 <PaulFertser> California!). N900 is a 599 €uro device, not a $50 one, so for the community
Feb 07 12:47:28 <PaulFertser> this is a bug, not a feature"
Feb 07 12:47:52 <playya__> Heiner-Fosdem, did you manage to record the talks?
Feb 07 12:47:59 <Heiner-Fosdem> playya__: no
Feb 07 12:48:04 <playya__> :(
Feb 07 12:48:15 <Heiner-Fosdem> playya__: there wasen't enough time to prepare taht
Feb 07 12:48:31 <PaulFertser> topro: you can try to recompile obexd with pbap plugin to see what it gives. Probably a dummy response would suffice.
Feb 07 12:49:20 <topro> PaulFertser: I know that N900 HFP is working at least, as one of my collegues has one.
Feb 07 12:49:23 <Heiner-Fosdem> http://downloads.vdm-design.de/obexd_0.20-r2.4_armv4t.ipk this is obexd with fos-pbap plugin
Feb 07 12:49:26 <playya__> Heiner-Fosdem, do you know if mophis is there, too?
Feb 07 12:49:35 <Heiner-Fosdem> playya__: no
Feb 07 12:49:51 <Heiner-Fosdem> playya__: i'm trying to find the others :)
Feb 07 12:50:10 <playya__> hehe
Feb 07 12:50:18 <topro> PaulFertser: I think I could use Heiner-Fosdem's obexd ipk, would that be ok?
Feb 07 12:50:32 <Heiner-Fosdem> to pbap plugin: the problem is, that the methods in the plugin are never reached
Feb 07 12:50:40 <playya__> i think it's fso not fos plugin
Feb 07 12:50:48 <Heiner-Fosdem> of course
Feb 07 12:51:02 <Heiner-Fosdem> notbook keyboard is to small ;)
Feb 07 12:51:23 <PaulFertser> Weiss: gta02 driver reports charging depending on "charger status", not the current. you topped the battery, the charger is still enabled but it stops charging. I hope you read Battery Q&A page ;)
Feb 07 12:51:25 <playya__> mine is perfect
Feb 07 12:52:03 <Heiner-Fosdem> normaly i'm using the old keyboard from my old 486 :)
Feb 07 12:52:18 <topro> PaulFertser: when I got obexd, what do you expect me to test?
Feb 07 12:52:39 <playya__> gym for you fingers ;)
Feb 07 12:52:39 <ynezz> JaMa|Gone: nandwrite directly to /dev/mtd4 ?
Feb 07 12:53:04 <Heiner-Fosdem> ok, i will look for the others, bbl
Feb 07 12:53:25 <playya__> good look
Feb 07 12:53:46 <Weiss> PaulFertser: sure.. but "apm" claims only 28%, and voltage_now is 3771000 - shouldn't it have resumed charging?
Feb 07 12:53:51 <PaulFertser> topro: you need obexd with enabled pbap plugin and probably with --pbap switch on start. Probably it needs something else to actually supply the carkit with the data, i'm not sure, would need to read the code for that.
Feb 07 12:54:20 <PaulFertser> Weiss: it should have resumed, indeed, no idea how that could happen, i never saw that with andy-tracking.
Feb 07 12:57:33 <Weiss> PaulFertser: hmm.. after unplugging the charger, current_now is 572062 - does that sound too big, or OK?
Feb 07 12:57:36 <PaulFertser> topro: btw, start bluetoothd with -d switch and look carefully at the output that it produces while/after pairing.
Feb 07 12:57:54 <PaulFertser> Weiss: a running system shouldn't consume _that_ much.
Feb 07 12:58:05 <PaulFertser> Weiss: btw, do not use apm, use /sys/class/power_supply/battery
Feb 07 12:58:35 <Weiss> ah ok
Feb 07 12:59:27 <PaulFertser> Weiss: becaus apm does some magic and it can be confusing. It's deprecated anyway.
Feb 07 12:59:34 <ynezz> JaMa|Gone: http://pastebin.ca/1789009
Feb 07 12:59:40 <ynezz> JaMa|Gone: seems, it works :)
Feb 07 13:00:23 <Weiss> PaulFertser: hmm.. on the mains charger, it seems happier
Feb 07 13:00:36 <Weiss> (voltage is increasing, current_now is negative)
Feb 07 13:01:22 <PaulFertser> topro: take a look at https://bugs.maemo.org/show_bug.cgi?id=6368
Feb 07 13:01:28 <PaulFertser> (for reference purposes)
Feb 07 13:02:53 <PaulFertser> topro: anyway you won't have much success with pbap right now i'm afraid unless you're able to read the code.
Feb 07 13:03:07 <CIA-47> freesmartphone.org: 03morphis 07msmcomm * rfb62c0ee69f8 10/libmsmcomm/src/ (6 files): libmsmcomm: reduce memory footprint
Feb 07 13:03:08 <CIA-47> freesmartphone.org: 03morphis 07msmcomm * r8f7959e6b5dd 10/ (24 files in 8 dirs): Merge branch 'master' of git.freesmartphone.org:msmcomm
Feb 07 13:03:21 <topro> PaulFertser: already found that maemon-thread yesterday when I was googling about that topic
Feb 07 13:03:40 <topro> doesn't match my case as they seem to be able to use hands-free
Feb 07 13:03:52 <PaulFertser> topro: not all of them
Feb 07 13:04:51 <topro> reading the code should be ok for me, do you have an url for the most recent source including the fso-pbap plugin?
Feb 07 13:05:30 <PaulFertser> topro: i'd try without bringing fso in the mix, just pbap with some dummy data. ebook plugin seems to be also unstable, so avoid it.
Feb 07 13:06:02 <PaulFertser> topro: http://git.kernel.org/?p=bluetooth/obexd.git;a=summary
Feb 07 13:07:05 <PaulFertser> topro: meaning try to make phonebook-dummy work
Feb 07 13:07:24 <topro> I'll see what I can do
Feb 07 13:08:09 <PaulFertser> topro: of course since you can do all that with your laptop it shouldn't be hard.
Feb 07 13:10:03 <topro> PaulFertser: i assume obexd-server is the debian package to go for?
Feb 07 13:12:22 <JaMa|Gone> ynezz: good, thanks
Feb 07 13:15:34 <PaulFertser> topro: i'd compile from the sources anyway
Feb 07 13:15:49 <PaulFertser> topro: since you want the latest and the greatest code and ability to easily tweak it.
Feb 07 13:16:41 <topro> PaulFertser: do you know how to select/enable some plugins
Feb 07 13:18:22 <PaulFertser> topro: i guess configure options, but no, i'd need to read the docs/code to find out.
Feb 07 13:18:51 <mrmoku> TAsn: ping
Feb 07 13:19:21 <topro> PaulFertser: ok, thought one could select at runtime, will go for the source and recompile
Feb 07 13:24:00 <lindi-> PaulFertser: hmm, do you know if there are any command line tools for using /sys/class/power_supply/battery?
Feb 07 13:24:41 <topro> PaulFertser: I think HFP is implemented in bluez, see http://www.bluez.org/bluez-444/
Feb 07 13:24:43 <PaulFertser> lindi-: i know lshal
Feb 07 13:24:55 <lindi-> PaulFertser: but isn't hal deprecated too? ;)
Feb 07 13:24:59 <PaulFertser> lindi-: sure, it is
Feb 07 13:25:04 <PaulFertser> lindi-: i do not know others
Feb 07 13:25:36 <PaulFertser> topro: i hope you're using the latest bluez on your laptop already ;)
Feb 07 13:25:44 <lindi-> PaulFertser: libomhacks has primitive support
Feb 07 13:25:56 <PaulFertser> lindi-: but neither there're console tools to parse "apm" output, afaik.
Feb 07 13:27:23 <lindi-> PaulFertser: isn't it already a console tool?
Feb 07 13:28:13 <lindi-> PaulFertser: unfortunately the output is not very machine parseable as usual
Feb 07 13:29:05 <PaulFertser> lindi-: sysfs nodes otoh are really machine parseable, aren't they?
Feb 07 13:31:49 <lindi-> PaulFertser: they keep changing
Feb 07 13:32:31 <lindi-> PaulFertser: some sort of middleware that provides a user interface around those is needed :)
Feb 07 13:38:39 <topro> PaulFertser: bluez 4.60, obexd compiled source 0.21
Feb 07 13:38:52 <topro> s/source/from source
Feb 07 13:51:33 <morphis> heyho
Feb 07 13:51:47 <Heinervdm> morphis: why aren't you here? ;)
Feb 07 13:52:03 <morphis> busy :)
Feb 07 13:52:09 <Heinervdm> ok :)
Feb 07 13:52:15 <morphis> I would really love to be at FOSDEM
Feb 07 13:52:37 <morphis> maybe next year :)
Feb 07 13:52:38 <Heinervdm> mickey said you made some progress yesterday?
Feb 07 13:52:44 <morphis> yesterday?
Feb 07 13:52:48 <morphis> last week
Feb 07 13:52:58 <Heinervdm> he said yesterday in his talk 
Feb 07 13:53:03 <Heinervdm> don't know
Feb 07 13:53:08 <morphis> we have now pin authentication and call recognation with msmcomm :)
Feb 07 13:53:11 <morphis> jepp
Feb 07 13:53:44 <Heinervdm> all recognation i read here
Feb 07 13:53:49 <Heinervdm> call
Feb 07 13:54:01 <morphis> jepp
Feb 07 13:54:17 <morphis> make a call is currently not implemented but I will work on this next week
Feb 07 13:54:34 <Heinervdm> nice :)
Feb 07 13:54:50 <morphis> then we need only some work to do on audio routing until we have outgoing/incomming call with voice
Feb 07 13:55:37 <Heinervdm> when that is working i can try it on my phone too :)
Feb 07 13:55:43 <morphis> somebody here who can help me with understanding some arm instructions?
Feb 07 13:55:56 <morphis> Heinervdm: you even have a Palm Pre?
Feb 07 13:56:05 <Heinervdm> yes
Feb 07 13:56:16 <Heinervdm> since 1 week after release
Feb 07 13:56:18 <morphis> great
Feb 07 13:56:29 <morphis> some time to help us?
Feb 07 13:56:47 <morphis> btw. mickey published his presentation somewhere?
Feb 07 13:56:48 <Heinervdm> don't know
Feb 07 13:56:57 <Heinervdm> to both :)
Feb 07 13:57:13 <morphis> ok
Feb 07 13:57:32 <morphis> tell me if you have, we need all help we can get :)
Feb 07 13:57:52 <Heinervdm> the problem is, that the palm i my primary phone :)
Feb 07 13:58:09 <morphis> Heinervdm: no problem :) mine too
Feb 07 13:58:34 <morphis> you just need a time slot where don't be availbale to other people
Feb 07 13:58:38 <morphis> than you are find
Feb 07 13:58:43 <Heinervdm> :)
Feb 07 13:58:48 <Heinervdm> lets see
Feb 07 13:58:56 <morphis> you don't have to do any modifications to your pre
Feb 07 13:59:05 <morphis> just booting a kernel over usb with attached initramfs
Feb 07 13:59:45 <Heinervdm> do you have a howto for that at webos-internals?
Feb 07 14:00:51 <morphis> Heinervdm: currently not a really howto how to develop with this
Feb 07 14:00:56 <morphis> but I can tell you here 
Feb 07 14:01:01 <morphis> it's very simple
Feb 07 14:01:07 <morphis> you just need oe around
Feb 07 14:01:22 <Heinervdm> i have enough oe trees ;)
Feb 07 14:01:33 <Heinervdm> but not today :)
Feb 07 14:05:41 <morphis> ok :)
Feb 07 14:06:05 <morphis> I have a point "do some more documentation on all the palm pre stuff" on my todo list :)
Feb 07 14:21:21 <Weiss> PaulFertser: more about these odd 2.6.32 battery problems: "charging" from my laptop, the current is about 100mA and positive, unplugged, it's 560-odd mA and positive, and with the mains charger, it's -350mA. sounds like I have a drain of 400mA coming from somewhere?
Feb 07 14:24:04 <DocScrutinizer> what's odd with this? except the fact FR blasts on >2W
Feb 07 14:29:55 <Heinervdm> mrmoku|away: there is an error in this commit http://git.shr-project.org/git/?p=pyphonelog.git;a=commitdiff;h=2d73b801fa1a76bd9fbd802cf493ac20d6e8698b it should be phonelog.mo instead of messages.mo
Feb 07 14:31:55 <DocScrutinizer> Weiss: power down modem, GPS, WLAN, do a top and sigstop all tasks. Then check again
Feb 07 14:32:46 <DocScrutinizer> Weiss: switch off all LED, make sure vib is off. Dim screen
Feb 07 14:33:49 <DocScrutinizer> Weiss: that's definitely *not* a _battery_ problem
Feb 07 14:34:30 <Weiss> yeah, I agree
Feb 07 14:36:01 <DocScrutinizer> GPRS with massive data thruput easily can cause such behaviour - to name just one
Feb 07 14:36:55 <DocScrutinizer> or running aircrack, or mplayer ;-P
Feb 07 14:37:29 <DocScrutinizer> to name just one other cause, by 2 examples
Feb 07 14:37:40 <Weiss> it's sitting idle on my desk with no SIM in..
Feb 07 14:38:06 <DocScrutinizer> so now it starts to get starnge
Feb 07 14:38:40 <DocScrutinizer> cpufreq?
Feb 07 14:39:42 <lindi-> trying to charge n900 can also generate that :)
Feb 07 14:40:48 <DocScrutinizer> if it's kernel-rev dep (lol, yeah lindi- right) I'd guess it might be kernel spinning idle busy. This should show up in top system percentage though I'd guess
Feb 07 14:41:49 <lindi-> Weiss: how large is the consumption in suspend?
Feb 07 14:42:25 <DocScrutinizer> lindi-: you have a talent to ask questions that are hard to answer
Feb 07 14:42:32 <Weiss> hehe
Feb 07 14:42:38 <Weiss> I don't have the means to measure that right now
Feb 07 14:44:05 <DocScrutinizer> Weiss: actually you have
Feb 07 14:44:11 <DocScrutinizer> bq27000
Feb 07 14:44:40 <Weiss> hmm?
Feb 07 14:45:09 <DocScrutinizer> read out the capacity, suspend for 15min, read again. multiply diff * 4
Feb 07 14:45:52 <Weiss> capacity as in /sys/class/power_supply/battery/capacity?
Feb 07 14:45:58 <DocScrutinizer> yup
Feb 07 14:46:01 <DocScrutinizer> e.g.
Feb 07 14:46:18 <DocScrutinizer> or use bq27k-detail script, for convenience
Feb 07 14:47:54 <DocScrutinizer> or just check SHR-settings-power (don't miss the "update" button)
Feb 07 14:48:12 <Weiss> in progress..
Feb 07 14:48:23 <lindi-> DocScrutinizer: hard to answer?
Feb 07 14:48:44 <lindi-> Weiss: just read current_now  on resume
Feb 07 14:49:01 <DocScrutinizer> lindi-: that's extremely unreliable
Feb 07 14:49:48 <lindi-> DocScrutinizer: why?
Feb 07 14:50:19 * DocScrutinizer suggests to ponder about the meaning of "now"
Feb 07 14:50:35 <lindi-> DocScrutinizer: yes it lags several seconds
Feb 07 14:50:43 <DocScrutinizer> how many?
Feb 07 14:50:58 <lindi-> i do not know
Feb 07 14:51:06 <DocScrutinizer> :-/&
Feb 07 14:51:18 <lindi-> but usually over 10
Feb 07 14:51:19 <DocScrutinizer> *I* do
Feb 07 14:51:24 <DocScrutinizer> nope
Feb 07 14:52:57 <lindi-> DocScrutinizer: hmm?
Feb 07 14:53:40 <DocScrutinizer> lindi-: read bq27k datasheet and you'll get to know
Feb 07 14:54:58 <DocScrutinizer> lindi-: basically you're relying on a hidden clock scheme inside bq27k which you never know the "phaseshift" to your 'real world' probing
Feb 07 14:55:37 <DocScrutinizer> the clock sheme is even different for the various readings from bq27k
Feb 07 14:55:58 <DocScrutinizer> for current it's some 5 sec iirc
Feb 07 14:56:17 <lindi-> i just measured 14
Feb 07 14:56:25 <DocScrutinizer> so you lag 5<x<10 sec
Feb 07 14:57:15 <lindi-> Weiss: any results?
Feb 07 14:57:57 <Weiss> ok.. capacity went from 28 to 19 in about 10 mins of suspend
Feb 07 14:58:08 <DocScrutinizer> and then it still might give you an average over a timespan that's not representative
Feb 07 14:58:09 <Weiss> (current_now was 566250 just after resume)
Feb 07 14:58:15 <lindi-> Weiss: ok
Feb 07 14:58:25 <lindi-> Weiss: that is quite high current
Feb 07 14:58:29 <DocScrutinizer> forget it
Feb 07 14:58:29 <Weiss> yeah
Feb 07 14:58:40 <Weiss> the left-hand side of the touchscreen seems quite warm, as well
Feb 07 14:58:45 <lindi-> Weiss: did the current rise after resume?
Feb 07 14:59:06 <Weiss> the current_now?
Feb 07 14:59:13 <lindi-> Weiss: yep
Feb 07 14:59:38 <DocScrutinizer> lindi-: you seem to develop a habit to spoil my mood and throw stick tween my legs
Feb 07 14:59:45 <mrmoku> Heinervdm: ohh :/
Feb 07 14:59:50 <lindi-> Weiss: but now might be a good idea to remove the battery and figure out where the energy is going :)
Feb 07 15:00:01 <mrmoku> Heinervdm: btw... how was your presentation?
Feb 07 15:00:31 <lindi-> DocScrutinizer: sorry, im trying to be as polite as i can
Feb 07 15:00:41 <DocScrutinizer> yeah and don't forget to set those loops on fire, so Weiss will look even smarter
Feb 07 15:02:18 <DocScrutinizer> FSCK I did that hundereds of times. I tell you it's not reliable. You go on with that shit
Feb 07 15:02:36 * DocScrutinizer wanders off searching for the red nose
Feb 07 15:02:55 <DocScrutinizer> suckers
Feb 07 15:03:20 <Weiss> :S
Feb 07 15:03:23 <wpwrak> DocScrutinizer: is lindi trying to pick up the current while suspended ?
Feb 07 15:03:37 <DocScrutinizer> yeah
Feb 07 15:03:41 <Weiss> FWIW.. current_now actually went DOWN slightly a few seconds after resuming
Feb 07 15:03:56 <DocScrutinizer> overriding my suggestion
Feb 07 15:04:07 <lindi-> Weiss: are you charging from usb?
Feb 07 15:04:13 <wpwrak> DocScrutinizer: i think you could make this work. i actually did some of that ...
Feb 07 15:04:45 <Weiss> lindi-: yep
Feb 07 15:04:51 <DocScrutinizer> you can, but it's not as simple as lindi- suggests. and weiss hoping the loops
Feb 07 15:05:15 <DocScrutinizer> hopping
Feb 07 15:05:32 <Weiss> the capacity method works too, and also gave a number that was too high?  Weiss> ok.. capacity went from 28 to 19 in about 10 mins of suspend
Feb 07 15:06:40 <DocScrutinizer> 28 to 19 WHAT?
Feb 07 15:06:46 <DocScrutinizer> mAh???
Feb 07 15:06:58 <lindi-> percent probably
Feb 07 15:07:12 <Weiss> whatever the units of the "capacity" sysfs node are :S
Feb 07 15:07:33 <DocScrutinizer> fuck %
Feb 07 15:07:43 <lindi-> Weiss: when did this start to happen?
Feb 07 15:08:04 <Weiss> just after I flashed a 2.6.32 kernel
Feb 07 15:08:28 <lindi-> Weiss: does it stop if you switch to older one?
Feb 07 15:08:31 * DocScrutinizer afk until lindi- finished confusing people
Feb 07 15:20:57 <Weiss> flashed from 2.6.32+KMS to 2.6.29+KMS, now it seems happy
Feb 07 15:21:09 <Weiss> will try 2.6.32 (no KMS) in a minute to check it's not my fault
Feb 07 15:36:57 <lindi-> Weiss: will be interesting if it really is a software issue
Feb 07 16:44:10 <TAsn> mrmoku, pong
Feb 07 16:49:05 <mrmoku> TAsn: disturbing your studies?
Feb 07 16:50:10 <TAsn> I wish you were
Feb 07 16:50:13 <TAsn> I just came back from work :(
Feb 07 16:50:36 <mrmoku> ohh :/
Feb 07 16:50:59 <mrmoku> TAsn: I'm having a problem with the hashtable phonelog passes in for new contacts...
Feb 07 16:51:38 <TAsn> what did I do wrong ? :P
Feb 07 16:51:45 <TAsn> oh, I probably didn't use gvalues
Feb 07 16:51:47 <TAsn> have i?
Feb 07 16:51:53 <Serdar> hi there
Feb 07 16:52:05 <Serdar> Deubeuliou, back again?
Feb 07 16:52:16 <mrmoku> TAsn: don't know how that should look like in python... but I would say no :P
Feb 07 16:52:29 <TAsn> mrmoku, no idea either (how it should look like in python)
Feb 07 16:52:35 <TAsn> I'll give it a look later on
Feb 07 16:52:36 <TAsn> :P
Feb 07 16:52:38 <mrmoku> ok
Feb 07 16:52:39 <TAsn> though actually
Feb 07 16:52:44 <TAsn> I think it worked in the past
Feb 07 16:52:49 <TAsn> so I don't think that is it.
Feb 07 16:52:52 <mrmoku> well it works partly
Feb 07 16:53:35 <mrmoku> it shows the field fine... but on g_hash_table_insert to change/add a value it segfaults somewhere in glib
Feb 07 16:54:04 <Weiss> does anyone know where the sysfs node for powering off GSM went in 2.6.32?
Feb 07 16:54:09 <mrmoku> TAsn: btw. what is the difference between g_hash_table_insert/replace ?
Feb 07 16:54:20 <mrmoku> Weiss: moment
Feb 07 16:54:50 <TAsn> mrmoku, look at the docs, I don't remember exactly, but I think one calls the free of the previous object, and one does not.
Feb 07 16:54:54 <mrmoku> Weiss:         sed_file $MODEMPY "s#/neo1973-pm-gsm.0/#/gta02-pm-gsm.0/#g"
Feb 07 16:55:02 <mrmoku> so it is gta02-pm-gsm.0
Feb 07 16:55:15 <mrmoku> TAsn: well looking at the docs is what confuses me ;)
Feb 07 16:55:27 <TAsn> sec, I'll look
Feb 07 16:55:29 <mrmoku> and the docs describe the same
Feb 07 16:56:02 <Weiss> mrmoku: ah, thanks a lot. should've spotted it, but didn't
Feb 07 16:56:18 <mrmoku> np
Feb 07 16:56:25 <TAsn> mrmoku, replace frees the old key
Feb 07 16:56:31 <TAsn> and value
Feb 07 16:56:34 <TAsn> while insert
Feb 07 16:56:56 <mrmoku> does the same :P
Feb 07 16:57:11 <TAsn> no
Feb 07 16:57:23 <TAsn> insert only deletes the value
Feb 07 16:57:28 <TAsn> I think I got the docs right
Feb 07 16:57:45 <TAsn> (the second part of the docs talks about remove, I think, or maybe hope :P)
Feb 07 16:58:15 <mrmoku> ahhh... the yeah, see the difference now :)
Feb 07 16:58:48 <mrmoku> TAsn: got confused by insert talking about deleting key/value too... but not the old ones
Feb 07 16:58:52 <mrmoku> did not see that part
Feb 07 16:59:00 <TAsn> mrmoku, :)
Feb 07 16:59:05 <TAsn> It's terribly confusing
Feb 07 16:59:15 <TAsn> maybe I didn't get it right as well, though I think Ihave.
Feb 07 17:00:05 <mrmoku> ok, then replace is what I want :)
Feb 07 17:01:09 <mrmoku> hmm
Feb 07 17:04:52 <TAsn> mrmoku, to be honest
Feb 07 17:05:05 <TAsn> I prefer you'll use remove and an insert instead of a replace
Feb 07 17:05:16 <TAsn> It's better to use something we know for sure
Feb 07 17:05:19 <TAsn> how it works
Feb 07 17:05:26 <TAsn> and can assume everyone will understand
Feb 07 17:05:33 <TAsn> then using a weirdly named function
Feb 07 17:05:47 <TAsn> that has a secret behavior :)
Feb 07 17:05:50 <TAsn> dont you think?
Feb 07 17:06:28 <DocScrutinizer> ouch
Feb 07 17:06:55 <TAsn> DocScrutinizer, ?
Feb 07 17:10:10 <Deubeuliou> Serdar: I'm here
Feb 07 17:10:15 <Deubeuliou> (in the hacker room)
Feb 07 17:10:28 <Serdar> ah.
Feb 07 17:11:13 <Deubeuliou> aw ... we are being thrown out
Feb 07 17:11:42 <Serdar> hehe same for yesterday
Feb 07 17:11:52 <Serdar> I am at Bar 2
Feb 07 17:12:10 <Deubeuliou> ok, incoming
Feb 07 17:17:42 <mrmoku> TAsn: yeah, agree, ok
Feb 07 17:32:19 <alabd> hello , can we use this on shr-t http://www.mozilla.com/en-US/mobile/features/ ?
Feb 07 17:49:19 <alabd> any opinion ?
Feb 07 18:20:08 <mrmoku> alabd: IIUC no
Feb 07 18:21:18 <alabd> why not correctly ?
Feb 07 18:22:20 <mrmoku> because it is using some ARM instructions for newer arm cpus...
Feb 07 18:22:34 <mrmoku> at least that's what I read in here some time ago
Feb 07 18:23:11 <lindi-> mrmoku: what's the name of the program?
Feb 07 18:23:29 <mrmoku> lindi-: fennec?
Feb 07 18:23:40 <lindi-> that url does not mention fennec
Feb 07 18:23:45 <lindi-> so i was bit unsure
Feb 07 18:23:49 <mrmoku> hmm
Feb 07 18:23:53 * mrmoku is unsure too :)
Feb 07 18:24:33 <topro> didn't fennec get renamed to firefox mobile?
Feb 07 18:24:51 <topro> recently, for release of 1.0
Feb 07 18:26:08 <lindi-> wonder if we'll have iceweasel mobile soon :)
Feb 07 18:26:21 <mrmoku> lindi-: http://www.mozilla.com/en-US/mobile/faq/
Feb 07 18:26:24 <mrmoku> mentions fennec
Feb 07 18:26:34 <mrmoku> Fennec is the codename of Mozilla’s mobile project
Feb 07 18:26:46 <mrmoku> , however, the official product name is Firefox.
Feb 07 18:27:40 <alabd> http://ftp.mozilla.org/pub/mozilla.org/mobile/1.0rc3/fennec-1.0rc3.en-US.linux-i686.tar.bz2
Feb 07 18:27:45 <alabd> that is for linux 
Feb 07 18:27:55 <alabd> android version is on the way seems
Feb 07 18:37:47 <Weiss> so... my battery drain doesn't seem to be from GPS, GSM, BT or WLAN - all are switched off or rfkilled. is there a better way to ensure the plug is properly pulled on them, or are there any other candidates which can be turned off to test?
Feb 07 18:38:45 <PaulFertser> Weiss: pmu register dump
Feb 07 18:39:05 <PaulFertser> Weiss: plus soc gpio list.
Feb 07 18:39:32 <PaulFertser> Weiss: first you get with sysfs node, the second is by s3c24xx-gpio utility.
Feb 07 18:39:43 <PaulFertser> And someone'd need to make sense of the output...
Feb 07 18:40:56 <lindi-> Weiss: turn them on and see if you get even more consumption?
Feb 07 18:41:12 <lindi-> ;)
Feb 07 18:41:24 <lindi-> Weiss: however, 500 mA is already very high
Feb 07 18:41:38 <UberNeo> can anybody suggest me a good ebook reader on Neo
Feb 07 18:41:48 <PaulFertser> UberNeo: fbreader
Feb 07 18:42:05 <UberNeo> but that doesnt support .pdf
Feb 07 18:42:47 <PaulFertser> UberNeo: good ebooks do not come in .pdf usually ;)
Feb 07 18:42:58 <PaulFertser> UberNeo: xpdf works btw, i used it to read some datasheets.
Feb 07 18:43:22 <UberNeo> and anything to read .doc or .xls
Feb 07 18:43:30 <UberNeo> does openoffice is the answer
Feb 07 18:45:32 <pespin> UberNeo, you have abiword in SHR feeds
Feb 07 18:45:45 <UberNeo> PaulFertser, any .ipk available for xpdf
Feb 07 18:46:34 <PaulFertser> UberNeo: no idea, i use Debian. As to the doc, i'd recommend to convert to text or html with wv-utils (or something like that, i can look up the name) and then use fbreader.
Feb 07 18:46:50 <PaulFertser> UberNeo: abiword is a ~lightweight alternative to oo.o writer.
Feb 07 18:46:55 <PaulFertser> (fyi)
Feb 07 18:47:28 <UberNeo> which all formats does fbreader supports
Feb 07 18:48:09 <pespin> for pdfs you have epdfview in shr feeds too :P
Feb 07 18:49:24 <PaulFertser> UberNeo: formats: http://www.fbreader.org/about.php
Feb 07 18:49:49 <lindi-> PaulFertser: mupdf is faster and lighter than xpdf
Feb 07 18:50:34 <PaulFertser> lindi-: thanks!
Feb 07 18:55:27 <Weiss> PaulFertser: like these: http://pastebin.de/3796
Feb 07 18:55:30 <Weiss> :)
Feb 07 18:55:59 <PaulFertser> Weiss: yeah, now i'd like somebody to interpret the values :)
Feb 07 18:56:21 <Weiss> sadly I don't know very much about PMUs and GPIOs :(
Feb 07 18:56:46 <lindi-> Weiss: compare against working situtation?
Feb 07 18:56:52 <lindi-> s/situtation/situation/
Feb 07 18:56:52 <apt> lindi- meant: Weiss: compare against working situation?
Feb 07 18:57:11 <PaulFertser> Weiss: what you really must know about the GPIOs is that no GPIO should output 1 to a device that has no power supplied to it.
Feb 07 18:57:14 <Weiss> ah, good idea
Feb 07 18:59:06 <wpwrak> PaulFertser: ... unless you want the latter condition to be false ;-)
Feb 07 18:59:44 <PaulFertser> wpwrak: supplying power through a gpio seems like a bad idea :)
Feb 07 19:00:58 <wpwrak> PaulFertser: heh, sometimes it actually makes sense. if you need very very little power :)
Feb 07 19:01:42 <PaulFertser> wpwrak: some uCs are capable of supplying some 20mA i think, not that little ;)
Feb 07 19:02:34 <wpwrak> yup, it could be less :)
Feb 07 19:11:48 <Weiss> hmm... I see quite a lot of things that were zero with 2.6.29 becoming something else with 2.6.32..
Feb 07 19:14:50 <PaulFertser> Weiss: i'll try to make sense out of your data a bit later, but today. Too bad nobody else is interested right atm.
Feb 07 19:14:56 <PaulFertser> (because i'm a lazy ass)
Feb 07 19:15:32 <Weiss> I'm formatting it nicely at the moment to show the differences
Feb 07 19:16:18 <PaulFertser> There's not that much devices that can be powered on/off, so direct interpretation of results, without comparing with "good" case should be ok too.
Feb 07 19:16:50 <Weiss> yeah
Feb 07 19:16:57 <Weiss> although there seem to be quite a few differences..
Feb 07 19:28:08 <Weiss> PaulFertser, DocScrutinizer: http://pastebin.de/3799
Feb 07 19:28:16 <Weiss> (definitely going to need your expertise in interpreting this)
Feb 07 19:29:12 <DocScrutinizer> duh, what???
Feb 07 19:31:51 <DocScrutinizer> OMG, you want me to check each and every GPIO of SoC for you???
Feb 07 19:32:23 <PaulFertser> DocScrutinizer: only those that can affect power managment, given that all the optional devices should be turned off by the moment.
Feb 07 19:32:25 <DocScrutinizer> damn, that's an expensive task
Feb 07 19:33:16 <Weiss> yeah :(
Feb 07 19:34:09 <Weiss> I'll go through it register by register of course, if that's the only way, but if anything immediately jumps out as being wrong to either of you..
Feb 07 19:34:42 <DocScrutinizer> how should it. that's a friggin number desert
Feb 07 19:36:30 <DocScrutinizer> messed up by linebreaks, completgely illegible
Feb 07 19:39:55 <Weiss> ok, never mind
Feb 07 19:40:58 * Weiss gets comfy with a few datasheets...  doesn't seem to be much other option (?)
Feb 07 19:44:32 <DocScrutinizer> Weiss: the way would be to 1) list the changes in a nice list like "B5:F0;  D11:>1;  " for the pins that changed (new state). 2) either a) check with schematics (pins are named there, like "B5") ans see what it does, or b) go to the sources and look up the #define that gives a meaningful name to GPIO-B5
Feb 07 19:45:36 <DocScrutinizer> no need to read datasheets first instance
Feb 07 19:46:02 <Weiss> ah, ok - that seems more sane
Feb 07 19:46:15 <Weiss> what do the ">"s and "R"s mean?
Feb 07 19:46:53 <DocScrutinizer> ask wpwrak 
Feb 07 19:47:02 <lindi-> Weiss: manual page would be definitely in order :)
Feb 07 19:47:29 <DocScrutinizer> I think it's input, output, high-Z, and special-function (like UART-TX)
Feb 07 19:48:14 <DocScrutinizer> try gpio --help
Feb 07 19:48:47 <Weiss> aha, ok
Feb 07 19:49:22 <DocScrutinizer> If you prepare the list as mentioned above, I'd be willing to do the schematics lookup
Feb 07 19:50:07 <DocScrutinizer> (preferrably one pin / line)
Feb 07 19:50:27 <DocScrutinizer> so s/;/\n/
Feb 07 19:51:25 * DocScrutinizer ->diner
Feb 07 19:51:44 <DocScrutinizer> s/n/nn/
Feb 07 19:51:45 <apt> DocScrutinizer meant: so s/;/\nn/
Feb 07 19:51:57 <DocScrutinizer> pfft
Feb 07 19:56:24 <wpwrak> Weiss: should be output and pull-up/down
Feb 07 20:15:23 <topro> how to disable sim card contacts in shr-u 'contacts'-app?
Feb 07 20:15:55 <DocScrutinizer51> disable backend
Feb 07 20:16:04 <TAsn> topro, what doc said, from shr-settings
Feb 07 20:16:15 <topro> looked in /etc/frameworkd.conf, but its not self-explaining to me
Feb 07 20:16:34 <topro> ah, theres gui support, didn't think about
Feb 07 20:20:48 <topro> DocScrutinizer51: TAsn: thx
Feb 07 20:20:57 <TAsn> np.
Feb 07 20:21:03 <TAsn> topro, btw, it requires a reboot
Feb 07 20:21:04 <TAsn> :|
Feb 07 20:21:22 <topro> thats what I just realized
Feb 07 20:21:36 <TAsn> I know it's very windowsih, we'll attend to that sometime in the future :P
Feb 07 20:21:42 <TAsn> (Actually, it's dos1's job)
Feb 07 20:21:57 <PaulFertser> TAsn: not full reboot, just frameworkd reboot
Feb 07 20:22:07 <PaulFertser> So it's not too windowsish
Feb 07 20:22:08 <PaulFertser> :)
Feb 07 20:22:10 <topro> PaulFertser: too late, power off
Feb 07 20:22:18 <TAsn> PaulFertser, but restarting frameworkd is a hassle
Feb 07 20:22:35 <TAsn> you gotta know what you are doing
Feb 07 20:22:45 <PaulFertser> TAsn: if only you knew how often i do that. I've a handy script to restart the muxer+frameworkd+ophonekitd
Feb 07 20:22:58 <TAsn> So I too just prefer a full restart
Feb 07 20:23:05 <TAsn> PaulFertser, ophonekitd = phonefsod? :P
Feb 07 20:23:43 <PaulFertser> TAsn: i'm too fucking cowardish to upgrade though i do not like what i have now a lot :(
Feb 07 20:24:02 <TAsn> PaulFertser, you really do have ophonekitd running?
Feb 07 20:24:10 <PaulFertser> TAsn: yes
Feb 07 20:24:15 <TAsn> omg
Feb 07 20:24:33 <TAsn> that's ~4 months old
Feb 07 20:24:35 <TAsn> even more maybe.
Feb 07 20:24:52 <TAsn> PaulFertser, that's very windowsish of you :P
Feb 07 20:25:00 <TAsn> (using old and buggy sw)
Feb 07 20:26:38 <PaulFertser> TAsn: that's very windowsish of you to still use uber-buggy insanely designed opkg :|
Feb 07 20:28:00 <TAsn> at least I'm using latest buggy version :)
Feb 07 20:28:37 <TAsn> PaulFertser, actually, a moment before you said that, I realized how much I like pacman
Feb 07 20:28:41 <TAsn> it's just so awesome
Feb 07 20:28:49 <TAsn> I can't even tell why, it's just perfect.
Feb 07 20:28:50 <topro> PaulFertser: what are you using if not opkg? no shr?
Feb 07 20:28:54 <TAsn> does everything terribly well.
Feb 07 20:28:56 <PaulFertser> topro: debian here
Feb 07 20:29:09 <PaulFertser> topro: with its own unique mix of problems :(
Feb 07 20:29:29 <TAsn> PaulFertser, that's soo windowsish of you to use slow apps (apt on the moko)
Feb 07 20:29:42 <topro> tried debian on freerunner but at a moment when debian unstable didn't install correctly due to xor changes and tslib input module
Feb 07 20:30:15 <topro> cause I'd prefer apt-get, too
Feb 07 20:30:15 <PaulFertser> TAsn: all my fr is so slow that rarely used apt is not an issue.
Feb 07 20:30:33 <pespin> Heinervdm, will you please ping me when evopedia is in OE? ;)
Feb 07 20:30:55 <Heinervdm> pespin: yes
Feb 07 20:31:01 <topro> what keeps me from using debian is that shr is way more up-to-date concerning fso stuff
Feb 07 20:31:07 <dos1> TAsn: enabling and disabling backends doesn't require frameworkd restart
Feb 07 20:31:49 <TAsn> PaulFertser, ;)
Feb 07 20:31:58 <TAsn> dos1, sure it does
Feb 07 20:32:05 <PaulFertser> topro: i used fso from git on my Debian but now that's not too easy because of all that vala deps.
Feb 07 20:32:31 <dos1> TAsn: changing default backend from or to SIM-Messages-FSO does, but enabling/disabling backends doesn't
Feb 07 20:33:49 <dos1> TAsn: well, maybe it's just my mind being foolish now, but i think i remember when i implemented disabling backends on fly :P
Feb 07 20:33:58 <dos1> TAsn: enabling backends always worked on fly
Feb 07 20:36:40 <TAsn> dos1, but it's a config, isn't it?
Feb 07 20:36:44 <TAsn> do you read configs on the fly?
Feb 07 20:36:49 <dos1> TAsn: no
Feb 07 20:37:04 <dos1> TAsn: but shr-settings uses opimd dbus method to enable/disable backends
Feb 07 20:37:13 <TAsn> i c.
Feb 07 20:37:14 <TAsn> so maybe.
Feb 07 20:37:16 <dos1> TAsn: so opimd enables/disables, and then saves config
Feb 07 20:37:21 <TAsn> but I don't think it removes already cached contacts
Feb 07 20:40:06 <TAsn> dos1, anyhow, what about more type related (and not type related) opimd functions?
Feb 07 20:40:13 <TAsn> we still miss probing by type
Feb 07 20:40:25 <TAsn> i.e where type = "05423213234"
Feb 07 20:40:32 <TAsn> which will search all the fields of that type
Feb 07 20:41:02 <TAsn> and searching for multiple stuff at the same time
Feb 07 20:41:15 <TAsn> i.e where Phone = 2113 or 234234 or 4354354 or 43543534
Feb 07 20:41:20 <TAsn> and get back a table
Feb 07 20:42:53 <dos1> well, probing by type can be implemented client-side :P
Feb 07 20:43:10 <dos1> (but i agree, i'll implement that)
Feb 07 20:43:20 <TAsn> dos1, it's slow on client side
Feb 07 20:43:27 <TAsn> I started implementing it here
Feb 07 20:43:30 <dos1> hmm?
Feb 07 20:43:31 <dos1> why?
Feb 07 20:43:35 <TAsn> but having multi calls
Feb 07 20:43:35 <dos1> it shouldn't be much slower
Feb 07 20:43:39 <TAsn> ^
Feb 07 20:43:40 <dos1> only one call more
Feb 07 20:43:49 <TAsn> how come?
Feb 07 20:43:54 <TAsn> you don't support "or"
Feb 07 20:43:59 <TAsn> so I have to probe each field
Feb 07 20:44:11 <TAsn> one call to get the types
Feb 07 20:44:12 <dos1> but i support _at_least_one (or something like that, don't remember exact name :P)
Feb 07 20:44:16 <TAsn> and one for each field
Feb 07 20:44:27 <dos1> so it works like "or"
Feb 07 20:44:32 <TAsn> dos1, no idea what you are talking about :)
Feb 07 20:44:51 <dos1> TAsn: IIRC i implemented it over your request ;P
Feb 07 20:44:55 <TAsn> though anyhow, please wrap it, will make our life easier :)
Feb 07 20:45:02 <dos1> on the same time when implementing sorting
Feb 07 20:45:12 <TAsn> dos1, I do tend to have good idea ;p but I don't remember what you are talking about atm.
Feb 07 20:45:23 <Heinervdm> pespin: today at my talk someone said that a user doesn't care about mac address in emtooth. So perhaps you can make a button that switches the view between mac and name and showing name is default
Feb 07 20:45:45 * DocScrutinizer51 points to his recommendation to implement regex based free text search, plus a cleartext type attribute for each field
Feb 07 20:46:03 <dos1> DocScrutinizer51: it's already implemented
Feb 07 20:46:14 <dos1> regexp for values
Feb 07 20:46:23 <dos1> and each field has type attribute
Feb 07 20:46:47 <dos1> you just can't regexp in field names :P
Feb 07 20:47:02 <TAsn> and you don't want to, either
Feb 07 20:47:06 <dos1> yup
Feb 07 20:47:10 <TAsn> assuming fields will be Phone and such is bad.
Feb 07 20:47:30 <TAsn> btw, where's the pisi project located at?
Feb 07 20:47:41 <TAsn> I want to suggest an idea
Feb 07 20:47:49 <dos1> TAsn: yup, _at_least_one=True
Feb 07 20:47:50 <TAsn> nvm
Feb 07 20:47:52 <TAsn> found it.
Feb 07 20:47:56 <TAsn> dos1, which does what?
Feb 07 20:48:07 <dos1> TAsn: {'Phone':'666', 'Name':'dos', '_at_least_one':True}
Feb 07 20:48:11 <TAsn> ok
Feb 07 20:48:12 <TAsn> thanks.
Feb 07 20:48:21 <dos1> TAsn: you'll get results with Phone=666 or Name=dos or both :P
Feb 07 20:48:22 <TAsn> so it is reasonable
Feb 07 20:48:31 <TAsn> dos1, sweet :)
Feb 07 20:48:44 <TAsn> (I'd just prefer having a search by type method, but this is good enough atm)
Feb 07 20:49:09 <DocScrutinizer51> nota bene though: name and type of a field is distinct things
Feb 07 20:49:12 <dos1> TAsn: and thanks, i just noticed name resolving in opimd-utils doesn't use _at_least_one - it probes every field in loop... ;/
Feb 07 20:49:29 <TAsn> dos1, :P
Feb 07 20:49:54 <TAsn> I wake up early tomorrow, so I'll probably be tired
Feb 07 20:50:01 <TAsn> but if not, I guess I'll start using this.
Feb 07 20:50:27 <DocScrutinizer51> the name is user definable for new fields. for the type you have a list of predefined to select from
Feb 07 20:50:49 <DocScrutinizer51> and only apps can install new types and register ffor them
Feb 07 20:51:56 <dos1> DocScrutinizer51: looks like it's already implemented like that ;)
Feb 07 20:52:04 <dos1> (just apps can't install new types yet)
Feb 07 20:52:17 <DocScrutinizer51> cool
Feb 07 20:54:03 <TAsn> DocScrutinizer51, yeah, it's starting to get awesome :)
Feb 07 20:54:22 <TAsn> this way is very sane, we finally feel like opimd "delivers"
Feb 07 20:54:45 <Heinervdm> talked to mickey today about opimd
Feb 07 20:54:56 <TAsn> and?
Feb 07 20:55:11 <Heinervdm> and we thought about removing all the backends :)
Feb 07 20:55:28 <Heinervdm> And only use sqlite
Feb 07 20:55:49 <Heinervdm> and then use a predefined set of fields
Feb 07 20:56:20 <Heinervdm> for those fields there is an colume in the table
Feb 07 20:56:35 <dos1> Heinervdm: what about SIM then? :P
Feb 07 20:56:50 <Heinervdm> that should then be integrated in the app
Feb 07 20:57:23 <Heinervdm> when you insert a new SIM you can choose to import those Contacts
Feb 07 20:57:29 <dos1> btw, we're already using predefinied set of fields in sqlite :P
Feb 07 20:57:42 <Heinervdm> and it's posible to export several contacts to sim
Feb 07 20:58:18 <Heinervdm> dos1: yes, but i've choosen those fields according to the deprected documentation ;)
Feb 07 20:58:42 <Heinervdm> So most of that fields aren't used, because you've choosen other names
Feb 07 20:59:05 <dos1> but atm that doesn't change much...
Feb 07 20:59:50 <Heinervdm> mickey said that he will write an mail about those ideas to the ml
Feb 07 21:01:43 <DocScrutinizer51> (removing all backends) another of my recommendations
Feb 07 21:02:23 <Heinervdm> i know :)
Feb 07 21:04:15 <DocScrutinizer51> fixed record structure will kill opimd though
Feb 07 21:04:38 <DocScrutinizer51> unless each record allows for 512 fields of arbitrary type, name, and content
Feb 07 21:06:50 <DocScrutinizer51> which maybe isn't as bad an idea as it sounds
Feb 07 21:08:17 <DocScrutinizer51> allow user to config fields_max *nd even supply a sqlite 3liner script to modify existing record structure)
Feb 07 21:09:22 <DocScrutinizer51> then set fieldtype of new record's fields to the template - most of the fields type=unused
Feb 07 21:10:50 <Weiss> DocScrutinizer51: I traced the GPIOs which had changed: (raw values in your preferred format are also there): http://pastebin.de/3808
Feb 07 21:11:07 <Weiss> looks like sysfs might be blatently lying about power being on or off to certain things...
Feb 07 21:11:45 <DocScrutinizer51> oh shit
Feb 07 21:12:49 <Weiss> is there a significant difference between ">0" and "F0"?
Feb 07 21:13:13 <TAsn> DocScrutinizer51, (romeving all backends) another one of my recommendations as well.
Feb 07 21:14:03 <TAsn> Heinervdm, I agree, though there is one big issue, they should not appear in one table, that's just a terrible waste of hd space :)
Feb 07 21:14:32 <Heinervdm> TAsn: but the query would be much faster
Feb 07 21:14:51 <TAsn> Heinervdm, I'm against the current behavior as well
Feb 07 21:14:56 <Heinervdm> any other way is missuse of a database
Feb 07 21:15:05 <TAsn> I think we should just split them to different columns
Feb 07 21:15:11 <TAsn> *tables
Feb 07 21:15:14 <TAsn> and join them
Feb 07 21:15:15 <TAsn> when needed
Feb 07 21:15:18 <TAsn> with one core table
Feb 07 21:15:32 <TAsn> Heinervdm, the current way is a misuse of a db, my suggestion is valid
Feb 07 21:16:01 <mrmoku> DocScrutinizer51: what do you mean by 'will kill opimd though' ?
Feb 07 21:16:20 <TAsn> mrmoku, because it won't let us use many cool custom columns
Feb 07 21:16:33 <TAsn> but then he suggested a solution
Feb 07 21:16:59 <DocScrutinizer51> Weiss: no much diff tween F and >   afaik
Feb 07 21:17:08 <TAsn> DocScrutinizer51, Heinervdm: this is exactly what gave me the motivation to "push" the types changes to opimd
Feb 07 21:17:19 <TAsn> this will allow us fast probing
Feb 07 21:17:30 <TAsn> and very defined fields
Feb 07 21:18:03 <mrmoku> TAsn: btw. I'm implementing change notification for contacts in libphone-ui now... to update the list
Feb 07 21:18:04 <Heinervdm> perhaps we can turn that whole thing around
Feb 07 21:18:13 <TAsn> mrmoku, supersweet :)
Feb 07 21:18:14 <Heinervdm> one colume for every type
Feb 07 21:18:21 <mrmoku> TAsn: need a name though :P
Feb 07 21:18:24 <DocScrutinizer51> Weiss: I'll check this pretty list thoroughly as soon as I'm back home
Feb 07 21:18:26 <TAsn> Heinervdm, and?
Feb 07 21:18:36 <TAsn> suggest one :
Feb 07 21:18:38 <TAsn> :P
Feb 07 21:18:56 <mrmoku> _phoneui_contacts_refresh is what is there (unused)
Feb 07 21:18:57 <TAsn> I don't remember our convention for signals
Feb 07 21:19:06 <TAsn> with a prefixed _
Feb 07 21:19:07 <TAsn> ?
Feb 07 21:19:15 <TAsn> odd.
Feb 07 21:19:16 <mrmoku> the prototype in phoneui.c
Feb 07 21:19:19 <DocScrutinizer51> mrmoku: kill in a sense of 'mess  up functionality fubar'
Feb 07 21:19:26 <mrmoku> DocScrutinizer51: ahh, ok :)
Feb 07 21:19:26 <TAsn> mrmoku, phoneui-utils.c ?
Feb 07 21:19:30 <mrmoku> yup
Feb 07 21:19:47 <Heinervdm> TAsn: yes it's not that easy, because we need the value and the name of the field
Feb 07 21:20:06 <TAsn> Heinervdm, and that'll too cause redundancy
Feb 07 21:20:12 <Weiss> DocScrutinizer51: cool - thanks a lot
Feb 07 21:20:12 <TAsn> we need to split to multiple databases
Feb 07 21:20:35 <TAsn> probably to
Feb 07 21:20:39 <Heinervdm> but that will result in really big joins
Feb 07 21:20:45 <TAsn> meant table
Feb 07 21:20:45 <mrmoku> phoneui_backend_contacts_contact_changed(int changetype, char *path)
Feb 07 21:20:53 <TAsn> Heinervdm, nah, because the primary ids are indexed
Feb 07 21:20:55 <Heinervdm> then we have to optimize the querys
Feb 07 21:21:08 <TAsn> and with views it's practically cached (I hope)
Feb 07 21:21:24 <TAsn> Heinervdm, people don't optimize queries, computers do :P (or at least in some sense)
Feb 07 21:21:50 <TAsn> anyhow, I think we should have a "core", "common" and "exotic" tables
Feb 07 21:21:55 <TAsn> or something like that :)
Feb 07 21:22:03 <Heinervdm> i that case _we_ should optimize the query
Feb 07 21:22:43 <TAsn> core will include name, surname, Phone and everything else we'll define as core
Feb 07 21:22:57 <TAsn> common will have stuff that many contacts have but not all
Feb 07 21:23:01 <TAsn> and exotic will have the rest
Feb 07 21:23:12 <TAsn> this will hopefully mean we won't waste space for nothing.
Feb 07 21:23:24 <TAsn> Heinervdm, when I'm talking about space, I'm also talking about cache space wasted
Feb 07 21:23:42 <TAsn> and reading parts from sd
Feb 07 21:23:44 <TAsn> which both take time.
Feb 07 21:23:49 <Heinervdm> but we should also use the right datatype for a field
Feb 07 21:23:55 <TAsn> Heinervdm, of course
Feb 07 21:24:06 <TAsn> that's part of our reason to have types in opimd
Feb 07 21:24:14 <ynezz> sqlite don't have datatypes 
Feb 07 21:24:17 <TAsn> they are strictly defined.
Feb 07 21:24:22 <TAsn> ynezz, ?
Feb 07 21:24:45 <TAsn> Heinervdm, i.e a phone number is a X long string
Feb 07 21:24:54 <TAsn> a birthday is a date
Feb 07 21:25:08 <TAsn> name is X long string
Feb 07 21:25:22 <TAsn> and anyother field you can think of is strictly defined.
Feb 07 21:25:41 <Heinervdm> ynezz: http://www.sqlite.org/datatype3.html
Feb 07 21:25:55 <DocScrutinizer51> all datatypes of opimd are text   - except primarykey id maybe
Feb 07 21:26:25 <Heinervdm> currently yes
Feb 07 21:26:35 <Heinervdm> but that has to be changed
Feb 07 21:26:57 <ynezz> well I'm wrong, but it's possible to insert string into number filed
Feb 07 21:26:58 <DocScrutinizer51> it's not opimd's task to check types and values. app does that
Feb 07 21:27:01 <ynezz> s/filed/field
Feb 07 21:28:09 <DocScrutinizer51> nope. but you can insert number to a string
Feb 07 21:28:19 <Heinervdm> DocScrutinizer51: but if you predefine some fields, then you can predefine the type of those fields too. And then you can things like <, > in the query too for numeric types
Feb 07 21:28:24 <TAsn> DocScrutinizer51, but if opimd can assume the value, it's great.
Feb 07 21:28:33 <TAsn> so that's why enforcing a date to be legit is a great idea.
Feb 07 21:28:48 <TAsn> (in opimd)
Feb 07 21:29:27 <DocScrutinizer51> well, ok. date actually might be usefull
Feb 07 21:29:34 <DocScrutinizer51> also filename
Feb 07 21:29:52 <Heinervdm> we also have timestamps
Feb 07 21:30:09 <TAsn> DocScrutinizer51, aye.
Feb 07 21:30:11 <DocScrutinizer51> that's a date
Feb 07 21:30:12 <TAsn> Heinervdm, date = timestamps
Feb 07 21:30:15 <TAsn> yeah.
Feb 07 21:30:16 <TAsn> :)
Feb 07 21:30:26 <Heinervdm> no, timestamps are int ;)
Feb 07 21:30:37 <TAsn> Heinervdm, int = date
Feb 07 21:30:39 <ynezz> DocScrutinizer51: http://pastebin.ca/1789459
Feb 07 21:30:48 <TAsn> ynezz, I know you are right
Feb 07 21:30:50 <Heinervdm> TAsn: ok :P
Feb 07 21:30:54 <PaulFertser> Weiss: ping
Feb 07 21:30:57 <TAsn> (about being able to insert strings to int fields)
Feb 07 21:31:10 <PaulFertser> Weiss: pmu register 0x34 specifies regulator for bluetooth. 
Feb 07 21:31:11 <TAsn> sqlite somewhat support field types
Feb 07 21:31:17 <TAsn> *column types
Feb 07 21:31:19 <PaulFertser> Weiss: in your dump it seems BT is enabled.
Feb 07 21:31:20 <DocScrutinizer51> so what's date?
Feb 07 21:31:23 <TAsn> although every cell has a type
Feb 07 21:31:28 <DocScrutinizer51> ynezz: won't fly. edge here atm
Feb 07 21:31:29 <TAsn> in reality
Feb 07 21:32:00 <ynezz> yep, but you've to be careful, it's not mysql
Feb 07 21:32:07 <TAsn> ynezz, yep.
Feb 07 21:32:15 <Heinervdm> on dbus unix timestamps are used for dates
Feb 07 21:32:29 <Heinervdm> there is no explecite date type
Feb 07 21:32:31 <TAsn> Heinervdm, we also use that internally
Feb 07 21:32:42 <TAsn> Heinervdm, also in sqlite
Feb 07 21:32:43 <Heinervdm> yes, that was our convention
Feb 07 21:32:51 <TAsn> Heinervdm, that's the best convention
Feb 07 21:33:07 <Weiss> DocScrutinizer51: yep... BT_EN GPIO is enabled and that register is 1 instead of 0... let me double-check that it's switched off according to sysfs..
Feb 07 21:33:42 <TAsn> (except for the 2038 bug)
Feb 07 21:34:15 <Heinervdm> has sqlite a limit?
Feb 07 21:34:35 <Heinervdm> yes 8 bytes :)
Feb 07 21:35:28 <TAsn> huh?
Feb 07 21:35:49 <Heinervdm> INTEGER. The value is a signed integer, stored in 1, 2, 3, 4, 6, or 8 bytes depending on the magnitude of the value.
Feb 07 21:36:09 <Heinervdm> according to the sqlite docu
Feb 07 21:36:34 <pespin> Heinervdm, (mac address) ok, I'll think about it :) but before I want to solve some problems I'm having when adding an scroller to a vbox in elementary, maybe you can help me.. I've asked at #e for some days but got no answer :S
Feb 07 21:36:42 <TAsn> Heinervdm, oh, ok.
Feb 07 21:36:59 <Heinervdm> TAsn: 8byte = 64bit
Feb 07 21:37:20 <DocScrutinizer51> Weiss: doesn't makee up for 500mA. nasty enough nevrtheless
Feb 07 21:37:29 <DocScrutinizer51> limit?
Feb 07 21:37:30 <DocScrutinizer51> for int maybe
Feb 07 21:37:57 <TAsn> Heinervdm, thanks for the math lesson ;P
Feb 07 21:38:32 <Heinervdm> so with 2^63 we shouldn't be affected by the year 2038 bug
Feb 07 21:38:47 <DocScrutinizer51> btw I don't see why you couldn't sort or relatively compare on strings - thus on sane text dateformats as well
Feb 07 21:39:16 <TAsn> Heinervdm, yeah :)
Feb 07 21:39:35 <TAsn> Heinervdm, we won't last to see the next y2k38 bug
Feb 07 21:39:37 <Heinervdm> DocScrutinizer51: we can do that but then we have to cache all the stuff and then compare
Feb 07 21:40:00 <Heinervdm> with correct types sqlite can compare for us
Feb 07 21:40:07 <DocScrutinizer51> huh?
Feb 07 21:40:31 <Heinervdm> SELECT * FROM table WHERE timestamp < 1234567890
Feb 07 21:40:42 <DocScrutinizer51> string *is* a correct type
Feb 07 21:41:20 <DocScrutinizer51> SELECT * FROM table WHERE timestamp < "1234567890"
Feb 07 21:41:41 <DocScrutinizer51> or better
Feb 07 21:42:22 <DocScrutinizer51> SELECT * FROM table WHERE timestamp < "2010-01-01-19:30:55,126864"
Feb 07 21:42:28 <PaulFertser> Weiss: with your .29 BT is off while with .32 it's on.
Feb 07 21:42:37 <PaulFertser> Weiss: (0x34 register)
Feb 07 21:43:51 <Heinervdm> pespin: there is an bug in the setup.py of evopedia. It should you only relative path's
Feb 07 21:45:06 <pespin> Heinervdm, don't understand second sentence. are you saying it should use only relative paths?
Feb 07 21:45:27 <Heinervdm> yes, forgot a word :)
Feb 07 21:45:31 <pespin> ok ;)
Feb 07 21:45:36 <pespin> I'll send a mail to author
Feb 07 21:46:33 <Heinervdm> pespin: applications instead of /usr/share/applications and ../lib/static instead of /usr/lib/static
Feb 07 21:47:09 <PaulFertser> Weiss: GSM is on in both cases if i understand it correctly
Feb 07 21:47:22 <Heinervdm> DocScrutinizer51: will that work? never tried < on strings...
Feb 07 21:47:51 <TAsn> Heinervdm, until the year 10000 it will
Feb 07 21:48:16 <DocScrutinizer51> 23::@85ms used to sort contacts on name alpha collation all the time
Feb 07 21:49:06 <DocScrutinizer51> grr
Feb 07 21:49:07 <DocScrutinizer51> Heinervdm: it's used to sort contacts on name alpha collation all the time
Feb 07 21:50:07 <Heinervdm> ok, but then 2345678 will be bigger then 12345678
Feb 07 21:50:56 <DocScrutinizer51> of course you have to use sane formats
Feb 07 21:51:56 <DocScrutinizer51> also you usally have funtions:
Feb 07 21:52:38 <DocScrutinizer51> SELECT * FROM table WHERE str2int(timestamp) < 1234567890
Feb 07 21:53:11 <DocScrutinizer51> if you really really want to do that botch
Feb 07 21:54:59 <Weiss> PaulFertser: you're right... testing now, if I make absolutely sure I've turned BT off first, GPB6 and 0x34 both go to zero
Feb 07 21:55:07 <DocScrutinizer51> even:
Feb 07 21:56:15 <DocScrutinizer51> SELECT name, firstname, unixtimestamp2date(timestamp) as birthday  FROM table WHERE str2int(timestamp) < 1234567890
Feb 07 21:56:21 <PaulFertser> Weiss: SD card is powered in both cases too
Feb 07 21:56:54 <TAsn> DocScrutinizer, Heinervdm: I'm not sure what you are talking about
Feb 07 21:56:57 <TAsn> but just know this:
Feb 07 21:57:02 <Weiss> PaulFertser: GSM is definitely off though, according to "cat /sys/bus/platform/devices/gta02-pm-gsm.0/power_on"
Feb 07 21:57:17 <TAsn> sqlite naturally supports saving dates as strings, ints and real
Feb 07 21:57:34 <TAsn> so all your discussion is probably unneeded :)
Feb 07 21:57:39 <DocScrutinizer51> PaulFertser: sdcard on sounds sane?
Feb 07 21:57:42 <Weiss> PaulFertser: SD card, or WLAN? (there's no SD card present..)
Feb 07 21:58:05 <PaulFertser> Weiss: ok, let me check.
Feb 07 21:58:12 <PaulFertser> Weiss: SD card == wlan
Feb 07 21:58:37 <DocScrutinizer51> TAsn: I'm arguing against using prior knowledge about fieldtypes in opimd
Feb 07 21:59:15 <TAsn> oh, I disagree with you.
Feb 07 21:59:21 <TAsn> but I'm too busy to take part of this discussion
Feb 07 21:59:29 <DocScrutinizer51> on user defined fields you open a can of worms this way
Feb 07 21:59:30 <TAsn> so please revert my last comment (disagreement)
Feb 07 21:59:56 <Weiss> PaulFertser: WLAN should be powered off according to rfkill? (/sys/bus/platform/devices/gta02-pm-wlan.0/rfkill/rfkill1/state = 0)
Feb 07 22:00:28 <PaulFertser> Weiss: (gsm on/off): http://paste.debian.net/59069/
Feb 07 22:02:33 <PaulFertser> Weiss: (sd card on) stupid me, sorry, it's indeed the SD card, not wifi.
Feb 07 22:02:38 <PaulFertser> DocScrutinizer51: ^
Feb 07 22:03:11 <DocScrutinizer51> so?
Feb 07 22:04:13 <PaulFertser> Wifi is always powered on it seems.
Feb 07 22:04:17 <PaulFertser> :-O
Feb 07 22:04:50 <DocScrutinizer51> it's indeed the wifi not sdcard??
Feb 07 22:04:51 <DocScrutinizer51> sorry lost track. http takes ages as I'm on edge atm
Feb 07 22:05:23 <PaulFertser> DocScrutinizer51: SD card voltage is on, that's normal. What's strange is that wifi vdd is connected to io_3v3, so it's always on too.
Feb 07 22:06:02 <Weiss> GSM is behaving oddly.. /sys/bus/platform/devices/gta02-pm-gsm.0/power_on=0, but pmu 15 says 0x07 (I killed ophonekitd)
Feb 07 22:06:13 <DocScrutinizer51> can't comment as no schematics here as well
Feb 07 22:07:18 <DocScrutinizer51> to me it seems this kernel is pretty much completely fucked up 
Feb 07 22:07:49 <PaulFertser> Weiss: that's on .32? ask larsc ;)
Feb 07 22:07:58 <Weiss> yep, .32
Feb 07 22:08:16 <PaulFertser> Weiss: same wit BT probably
Feb 07 22:08:39 <Weiss> BT seems to behave now... the GPIO switched off when I told it to
Feb 07 22:08:57 <Weiss> I think I forgot to disable it when doing the dumps
Feb 07 22:09:08 <PaulFertser> Weiss: (bluetooth): http://paste.debian.net/59070/
Feb 07 22:09:40 <DocScrutinizer51> that would be even the end of nastiness if the gpio didn't switch when you tell it to do
Feb 07 22:11:41 <PaulFertser> Weiss: what else should we check?
Feb 07 22:12:06 <Weiss> accelerometers..
Feb 07 22:12:25 <Weiss> CS and INT lines changed for both accels 
Feb 07 22:12:38 <PaulFertser> Weiss: what current difference do you get btw?
Feb 07 22:12:52 <Weiss> between .29 and .32?
Feb 07 22:13:31 <Weiss> a bit under half an amp
Feb 07 22:15:00 <PaulFertser> Weiss: half an amp, incredible...
Feb 07 22:15:17 <PaulFertser> Weiss: accellerometers seem to be powered all the time, both on my device and for you in both cases.
Feb 07 22:15:29 <PaulFertser> (meaning pmu supplies power here)
Feb 07 22:16:00 <Weiss> well... enough of a current difference that it can't charge from USB (only from mains)
Feb 07 22:18:07 <Weiss> (in other news, 2.6.32+DRM makes the lock screen slider super-smooth)
Feb 07 22:18:53 <PaulFertser> DocScrutinizer51: do you probably have an idea how such a huge current difference might occur?
Feb 07 22:19:36 <DocScrutinizer> CPU, GSM
Feb 07 22:19:55 <PaulFertser> Weiss: try turning off GSM by pmu 15=0
Feb 07 22:19:56 <DocScrutinizer> (GLAMO?)
Feb 07 22:21:05 <DocScrutinizer> actually 2W you schould be able to spot by your fingertip, even without removing the case
Feb 07 22:21:05 <mwester> :p  blasted kernel devs again
Feb 07 22:21:10 <Weiss> (in other news, 2.6.32+DRM makes the lock screen slider super-smooth)
Feb 07 22:21:19 <Weiss> whoops, mis-paste
Feb 07 22:21:30 <Weiss> DocScrutinizer: yep, it's noticeably warm.. around the left hand side of the screen
Feb 07 22:21:42 <Weiss> very roughly... it's not too concentrated
Feb 07 22:22:05 <PaulFertser> Weiss: that's where pmu is
Feb 07 22:22:31 <Weiss> PaulFertser: pmu 15=0 might have given a small current change, but not much. current_now=563062
Feb 07 22:22:39 * DocScrutinizer wonders if by any odds the device could have entered selfcharging mode
Feb 07 22:22:53 <PaulFertser> Weiss: and gpio H1=0R; gpio H2=0
Feb 07 22:23:19 <DocScrutinizer> nah that's exclusive or by hw, iirc
Feb 07 22:23:20 <PaulFertser> DocScrutinizer51: i checked EN_USBHOST, seem to be correct in both cases.
Feb 07 22:23:34 <DocScrutinizer> ~praise PaulFertser 
Feb 07 22:23:35 <apt> All hail PaulFertser!
Feb 07 22:24:14 <PaulFertser> Too early, we didn't found the magic current sink yet
Feb 07 22:24:17 <PaulFertser> :|
Feb 07 22:24:22 <PaulFertser> find
Feb 07 22:24:52 <Weiss> what does "secondary function" mean for a GPIO?
Feb 07 22:24:59 <DocScrutinizer> cpu is next to pmu, no?
Feb 07 22:25:15 <DocScrutinizer> Weiss: like UART
Feb 07 22:25:22 <DocScrutinizer> or LCD-data
Feb 07 22:25:35 <PaulFertser> Yeah, the same pin can play either role.
Feb 07 22:26:13 <PaulFertser> Weiss: have you tried the manual way to turn off gsm? 15=0, gpio H1=0R; gpio H2=0 should do that.
Feb 07 22:26:34 <Weiss> PaulFertser: I've done 15=0, but not the others. trying now...
Feb 07 22:27:42 <DocScrutinizer> I still don't see any quote on 'top' cpu loads
Feb 07 22:28:20 <Weiss> ah.. yep, nothing leaping out in top
Feb 07 22:28:21 <DocScrutinizer> and I really wonder why :-/
Feb 07 22:29:01 <PaulFertser> Weiss: what about the idea it's glamo consuming that much?
Feb 07 22:29:31 <DocScrutinizer> if cpu is running cyrcles somewhere in kernel, there's few chances to spot that via s2c2442-gpio
Feb 07 22:29:39 <Weiss> H1H2=0, still a 0.4A missing..
Feb 07 22:30:33 <Weiss> r.e. Glamo: it does it or without DRM/KMS (I tried both), which are completely different paths as far as graphics is concerned.. could be something in glamo-core though
Feb 07 22:31:36 <DocScrutinizer> I recommend opening the case, opening the cans if really needed, and spot the chip that heats up to 90°C
Feb 07 22:31:59 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r0165c943a4fe 10/tools/mdbus2/src/main.vala: mdbus2: upstream readline.vapi is broken; work around it for now
Feb 07 22:32:00 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r887ef8c04d6f 10/tools/mdbus2/vapi/readline.vapi: mdbus2: add fixed readline.vapi; scheduled for upstream inclusion
Feb 07 22:32:01 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * reccce6689f50 10/tools/mdbus2/ (src/main.vala vapi/readline.vapi): mdbus2: tab completion in interactive mode almost works
Feb 07 22:32:02 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r0002bc3f731b 10/tools/mdbus2/src/main.vala: mdbus2: fix completion for interfaces
Feb 07 22:32:03 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r532f36f522b6 10/tools/mdbus2/src/main.vala: mdbus2: calling methods from interactive shell now works
Feb 07 22:32:06 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * rac8b50b4bbeb 10/ (19 files in 14 dirs): complete AUTHORS and MAINTAINERS file; please add yourself
Feb 07 22:32:06 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r16dcb1ea80ee 10/tools/mdbus2/src/main.vala: mdbus2: unify error messages and fix parameter delivery
Feb 07 22:32:14 <DocScrutinizer> aaah, forgot:
Feb 07 22:32:23 <DocScrutinizer> ~spank kernel-devels
Feb 07 22:32:24 * apt bends kernel-devels over his knee and tatoos 'ibot' on kernel-devels's pasty white buttocks.
Feb 07 22:33:03 <Weiss> hehe
Feb 07 22:33:29 <max_posedon> git.openmoko.org still unavaible? (for git?)
Feb 07 22:34:20 <PaulFertser> Weiss: is something changing when you completely blank the screen?
Feb 07 22:35:36 <Weiss> PaulFertser: difference of about 7000 in current_now..
Feb 07 22:36:10 <max_posedon> mickey|sofa, hi! have you seen http://trac.freesmartphone.org/ticket/516 ? Should I add smth more info about this issue? 
Feb 07 22:36:46 <DocScrutinizer> probably each kernel devel that doesn't ship his module/whatever with a set of publically validated unit test scripts should be banned from git commit
Feb 07 22:37:16 <mickey|sofa> max_posedon: no, thanks, that's enough info for now. i need to take a look on my own with an updated distro
Feb 07 22:38:27 <PaulFertser> Weiss: btw, PMU is still using wrong values for LDOxENA, they work by chance, not by purpose :|
Feb 07 22:38:28 <DocScrutinizer> Weiss: could you please use bq27k-detail.sh to give use units to the values you quote
Feb 07 22:41:24 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * rdea0ee29b552 10/tools/mdbus2/src/main.vala: mdbus2: add some simple checks for bus name and object path validity
Feb 07 22:43:36 <Deubeuliou> hi !
Feb 07 22:43:42 <mickey|sofa> heya Deubeuliou 
Feb 07 22:43:43 * Deubeuliou is back from the fosdem
Feb 07 22:43:51 <mickey|sofa> welcome back
Feb 07 22:43:53 * mickey|sofa as well
Feb 07 22:44:02 <Deubeuliou> dat wuz great :)
Feb 07 22:44:08 <mickey|sofa> thanks for your nice presentation
Feb 07 22:44:17 <mickey|sofa> i think we had a good program given the short time
Feb 07 22:45:01 <Deubeuliou> we were pushed out very quickly, though
Feb 07 22:45:04 <mickey|sofa> yeah
Feb 07 22:45:20 <Deubeuliou> I even forgot Q&A :/
Feb 07 22:45:24 <mickey|sofa> who knows, perhaps next year we get a full room
Feb 07 22:45:32 <Deubeuliou> yeah !\o/
Feb 07 22:45:55 <DocScrutinizer> ([2010-02-07 23:35:36] <Weiss> PaulFertser: difference of about 7000 in current_now..)  this makes no sense neither for uA nor for mA
Feb 07 22:45:56 <Deubeuliou> maybe a "telephony" room
Feb 07 22:46:02 <mickey|sofa> yeah
Feb 07 22:46:08 <PaulFertser> Weiss: i briefly checked all changed pmu registeres but couldn't find anything really important.
Feb 07 22:46:12 <mickey|sofa> with more platforms
Feb 07 22:47:15 <Deubeuliou> that surely would help
Feb 07 22:49:04 <Weiss> DocScrutinizer: "Average current" is the useful one here?
Feb 07 22:49:26 <DocScrutinizer> I'd guess yes. lemme check
Feb 07 22:49:32 <pespin> H
Feb 07 22:50:15 <pespin> ops, typo ^ looking for Heinervdm ;)
Feb 07 22:50:44 <Deubeuliou> 23:36 -!- Heinervdm [~thomas@pD9E13742.dip.t-dialin.net] has quit []
Feb 07 22:51:16 <DocScrutinizer> AI yes
Feb 07 22:51:47 <PaulFertser> DocScrutinizer: probably 7mA difference was because gsm wasn't really enabled and was sleeping (as it does automatically).
Feb 07 22:52:16 <DocScrutinizer> errr, 7mA diff for completely dimmed screen ?
Feb 07 22:52:49 * DocScrutinizer feels like watching 'lost in translation"
Feb 07 22:53:05 <PaulFertser> DocScrutinizer: oh, it's me stupid, we're talking about the screen now :)
Feb 07 22:53:21 <PaulFertser> Yes, this doesn't make sense.
Feb 07 22:54:48 <PaulFertser> DocScrutinizer: in fact watching "lost in translation" was somewhat pleasant iirc. Interesting music, ~good plot and actors, great dialogues, excellent film quality overall.
Feb 07 22:58:26 <PaulFertser> Weiss: (dimmed screen diff) that doesn't sound right
Feb 07 22:59:07 * DocScrutinizer had a daydream nightmare of misconfiged SDRAM eating power in selfrefresh like mad, and a dozen other even more nasty ideas showing up on mental horizon
Feb 07 22:59:24 <Weiss> :S
Feb 07 22:59:37 * Weiss needs to go to bed, but will continue this tomorrow (evening)
Feb 07 22:59:57 <Weiss> including taking the PCB out fully to see what's not
Feb 07 23:00:06 <DocScrutinizer> good, make that a veritable nightmare for tonight then ;-P
Feb 07 23:00:16 <Weiss> :o
Feb 07 23:00:36 <PaulFertser> Weiss: so do you really mean turning backlight on and off made 7mA diff?
Feb 07 23:00:57 <Weiss> PaulFertser: I mean that was roughly how much current_now changed by.. 
Feb 07 23:01:31 <DocScrutinizer> this kernel sysfs is FUBAAAAAAAR
Feb 07 23:01:36 <PaulFertser> Weiss: backlight should be on an order of 70mA and if it doesn't turn on or off it's hardly unnoticeable.
Feb 07 23:02:05 <Weiss> but.. the units of sysfs are "abitrary"?
Feb 07 23:02:13 <DocScrutinizer> nope
Feb 07 23:02:15 <Weiss> or are they supposedly uA?
Feb 07 23:02:19 <DocScrutinizer> yep
Feb 07 23:02:35 <DocScrutinizer> at least afaik iirc etc pp
Feb 07 23:02:55 <PaulFertser> Weiss: they must be uA, yes
Feb 07 23:04:21 <DocScrutinizer> gentlemen, check if kernel devels went havoc another time and simply messed up sysfs pragmas to make life more funny
Feb 07 23:04:47 <PaulFertser> DocScrutinizer: we're working pmu and gpio level now, so that shouldn't be an issue...
Feb 07 23:05:25 <DocScrutinizer> sure, if they decided current_now should report 10s of uA now, rather than uA
Feb 07 23:06:31 <DocScrutinizer> that's a further reason I recommended using bq27k-detail, btw
Feb 07 23:08:38 <PaulFertser> Nah, they couldn't have changed bq27k driver in such a strange way.
Feb 07 23:08:40 <DocScrutinizer> I suspect a "iron in spinach" type of error on Weiss' side though :-P
Feb 07 23:11:50 <DocScrutinizer> anyway I don't see any chance for LCD bl to burn 500mA
Feb 07 23:12:13 <DocScrutinizer> otoh when the 500mA figure is from same sysfs node.. :-S
Feb 07 23:14:00 <DocScrutinizer> aaah, and capacity going down from a supposed % 28 to 19 in 10 minutes also doesn't match the picture
Feb 07 23:15:10 <DocScrutinizer> well, maybe yes
Feb 07 23:15:28 <DocScrutinizer> would make for some 600mA drain from bat
Feb 07 23:15:52 <DocScrutinizer> ?
Feb 07 23:16:04 <DocScrutinizer> can't calculate anymore
Feb 07 23:16:33 <PaulFertser> That's 108mAh
Feb 07 23:16:40 <DocScrutinizer> *6
Feb 07 23:16:46 <PaulFertser> 1.8 A?
Feb 07 23:16:54 <PaulFertser> Unrealistic :)
Feb 07 23:17:23 <PaulFertser> 9% of 1200mAh bat in 10 min. Unlikely.
Feb 07 23:17:44 <DocScrutinizer> 108mAh/10minutes*60min
Feb 07 23:18:36 <PaulFertser> Oh yeah, /6, i thought it /10, assuming an hour has 100 min apparently.
Feb 07 23:18:46 <DocScrutinizer> or let me put it that way: 120mAh/2h
Feb 07 23:18:50 * PaulFertser should sleep now, talking nonsence
Feb 07 23:19:04 <DocScrutinizer> 1200mAh even
Feb 07 23:20:15 <PaulFertser> DocScrutinizer: so yes, it looks like 648mA consumption indeed.
Feb 07 23:21:46 * DocScrutinizer wonders if the modem could see a constant nasty #1024
Feb 07 23:22:39 <PaulFertser> DocScrutinizer: we made sure the modem is disabled by cutting power to it. pmu 15=0 and appropriate gpio tricks.
Feb 07 23:22:56 <DocScrutinizer> either that, or GPRS (or spyware to keep a call all the time ;-P ), or CPU burning cycles like aircracking for its life
Feb 07 23:23:21 <DocScrutinizer> or one of my nightmares
Feb 07 23:23:52 <PaulFertser> Yes, the load might be 0 but if "idle task" doesn't execute the "halt instruction" it would likely result in a really high consumption :)
Feb 07 23:23:59 <DocScrutinizer> aka glamo havoc, RAM eats refresh
Feb 07 23:24:01 <DocScrutinizer> etc
Feb 07 23:24:17 <DocScrutinizer> PaulFertser: exactly
Feb 07 23:24:49 <DocScrutinizer> NOPREEMPT? :-P
Feb 07 23:25:05 <DocScrutinizer> or somesuch?
Feb 07 23:25:24 <DocScrutinizer> cpufreq
Feb 07 23:25:42 <DocScrutinizer> tickless
Feb 07 23:26:35 <DocScrutinizer> I bet a bottle of good spanish brandy it's the CPU
Feb 07 23:26:49 <DocScrutinizer> or GLAMO
Feb 07 23:27:36 <DocScrutinizer> (all those magic machines in there. /me moans)
Feb 07 23:32:20 * DocScrutinizer points to R1813/R1814 and starts to moan even more
Feb 07 23:33:17 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * rb85302942670 10/tools/serial_forward/ (AUTHORS MAINTAINERS README configure.ac): serial_forward: bump version, authors, README
Feb 07 23:33:32 <PaulFertser> 80k6???
Feb 07 23:33:40 <DocScrutinizer> yo
Feb 07 23:33:44 <DocScrutinizer> :-(
Feb 07 23:34:44 <DocScrutinizer> picks up spikes and noise like hell
Feb 07 23:36:00 <PaulFertser> Is it some weird requirement or another fuckup?
Feb 07 23:38:58 <CIA-47> freesmartphone.org: 03mickey 07cornucopia * r2dc73d97c64b 10/tools/serial_forward/src/Makefile.am: serial-forward: make distcheck happy
Feb 07 23:39:33 <DocScrutinizer> PaulFertser: level
Feb 07 23:39:48 <DocScrutinizer> 3V3 -> 1V8
Feb 07 23:39:51 <DocScrutinizer> or sth
Feb 07 23:40:09 <PaulFertser> DocScrutinizer: ah, makes some sense then.
Feb 07 23:40:18 <DocScrutinizer> nah, fuckup
Feb 07 23:41:17 <DocScrutinizer> one of the things that *ought* have been fixed in A7 the latest
Feb 07 23:41:26 <PaulFertser> DocScrutinizer: what is the correct method?
Feb 07 23:41:35 <DocScrutinizer> levelshifter
Feb 07 23:42:18 <PaulFertser> I thought there's something "simpler"
Feb 07 23:42:37 <DocScrutinizer> prolly there is
Feb 07 23:43:16 <DocScrutinizer> like use one of the dozens unused GPIO in one of the 1V8 chips
Feb 07 23:43:39 <DocScrutinizer> or scatch it all together
Feb 07 23:43:44 <DocScrutinizer> and tie high
Feb 07 23:44:35 <DocScrutinizer> find a better way anyway, than doing a tradeof between S/N and quiescent current
Feb 07 23:45:57 <DocScrutinizer> now we got the quiescent current and *stikk* we did a reset on every resume as the glamo catches glitches during suspend - at least that's what my cloudy memory reveals
Feb 07 23:46:09 <DocScrutinizer> the worst of both worlds
Feb 07 23:46:23 <DocScrutinizer> *still*
Feb 07 23:47:22 <DocScrutinizer> the story goes like "we dunno if it caught a reset by a spike, so we do a reset on our own to know where we are"
Feb 07 23:47:50 <DocScrutinizer> iirc
Feb 07 23:48:47 <DocScrutinizer> mompl, will scan my 89k mails for R1813
Feb 07 23:48:56 <DocScrutinizer> takes 30min :-(
Feb 07 23:54:21 <PaulFertser> It's still should be quite a powerful spike/noise to override the soc signal (/me confused)
Feb 07 23:55:30 <PaulFertser> Anyway, sleeptime :/
Feb 07 23:55:46 <PaulFertser> DocScrutinizer: enjoy your 'evening', see you tomorrow :)
Feb 07 23:56:20 <DocScrutinizer> PaulFertser: google: 
Feb 07 23:56:22 <DocScrutinizer> It seems that our choice about divider network (R1813 / R1814) to
Feb 07 23:56:23 <DocScrutinizer> level-translate 3.3V CPU GPIO to 1.8V Glamo RST# might not have been the
Feb 07 23:56:25 <DocScrutinizer> brightest thing we ever did.  I can reset the Glamo (provoking sticky
Feb 07 23:57:02 <DocScrutinizer> andy on [om-kernel]
Feb 07 23:57:14 <DocScrutinizer> 2008-07-29
Feb 07 23:58:02 <DocScrutinizer> subject: Wacky Racers on Glamo
Feb 07 23:59:09 <DocScrutinizer> I should have pasted a lil more :-(((
Feb 07 23:59:14 <DocScrutinizer> brightest thing we ever did.  I can reset the Glamo (provoking sticky
Feb 07 23:59:15 <PaulFertser> DocScrutinizer: np
Feb 07 23:59:16 <DocScrutinizer> WSOD) by touching Glamo RST# with my scope probe.  Glamo data says you
Feb 07 23:59:17 <DocScrutinizer> need to assert RST# for 1us to get a reset, but it appears to reset the
Feb 07 23:59:19 <PaulFertser> Googled it already
Feb 07 23:59:19 <DocScrutinizer> thing on any old glitch and is just recommending to hold it 1us to be
Feb 07 23:59:20 <DocScrutinizer> sure everything got reset.
Feb 08 00:00:16 <DocScrutinizer> WSOD --- WAAAAAAAH
Feb 08 00:00:42 <DocScrutinizer> seems I completely forgot about that bugger
Feb 08 00:01:16 <DocScrutinizer> PaulFertser: anyway, enjoy your nap
Feb 08 00:01:30 <PaulFertser> Thanks
Feb 08 00:02:24 <PaulFertser> DocScrutinizer: though i can't really understand how that divider network is any worse than 100k pullups used in other places.
Feb 08 00:03:06 <DocScrutinizer> I guess the problem is they missed on the input load of the RST#
Feb 08 00:03:24 <DocScrutinizer> which decalibrates the divider
Feb 08 00:03:36 <PaulFertser> Andy'd have noticed that with his scope
Feb 08 00:03:41 <DocScrutinizer> (a shot from the hip)
Feb 08 00:14:31 <DocScrutinizer> > > R1813, R1814: We see problems with sensitivity to spikes and EMI here, 
Feb 08 00:14:32 <DocScrutinizer> causing 
Feb 08 00:14:34 <DocScrutinizer> > > inadvertent reset of Glamo. Maybe we can find better idea, like FET for 
Feb 08 00:14:36 <DocScrutinizer> > > levelshifter. Should ask Andy. Impact: BOM and/or Layout.
Feb 08 00:14:37 <DocScrutinizer> result of meeting: won't get done any change on this for A8 today. Keep topic 
Feb 08 00:14:38 <DocScrutinizer> for A9 ;
Feb 08 00:15:17 <DocScrutinizer>  [me] on a mail about a EE meeting, 2008-11-19
Feb 08 00:15:56 <DocScrutinizer> retransmit for PaulFertser :-)
Feb 08 00:16:01 <DocScrutinizer> > > R1813, R1814: We see problems with sensitivity to spikes and EMI here, 
Feb 08 00:16:03 <DocScrutinizer> causing 
Feb 08 00:16:04 <DocScrutinizer> > > inadvertent reset of Glamo. Maybe we can find better idea, like FET for 
Feb 08 00:16:06 <DocScrutinizer> > > levelshifter. Should ask Andy. Impact: BOM and/or Layout.
Feb 08 00:16:07 <DocScrutinizer> result of meeting: won't get done any change on this for A8 today. Keep topic 
Feb 08 00:16:09 <DocScrutinizer> for A9 ;
Feb 08 00:16:10 <DocScrutinizer> [2010-02-08 01:15:17] <DocScrutinizer>  [me] on a mail about a EE meeting, 2008-11-19
Feb 08 00:21:13 <DocScrutinizer> hooray, kmail/Kontaact segfaulted another time. Seems >80k mails isn't exactly the usecase it's designed for
Feb 08 02:37:10 <wpwrak> DocScrutinizer: you need mutt :)
Feb 08 02:37:24 <DocScrutinizer> hmm
Feb 08 02:37:48 <wpwrak> (admittely, I'm at only 44 kmails so far)
Feb 08 02:38:30 <DocScrutinizer> you purged [community] ? :-D
Feb 08 02:38:48 <wpwrak> naw, that goes to a different folder
Feb 08 02:39:21 <wpwrak> only 11 k there, though. 1.5 k in "noise" (buglog and commitlog)
Feb 08 02:41:07 <DocScrutinizer> yup, actually 80k was exaggerated
Feb 08 02:42:02 <DocScrutinizer> [admin] was a nightmare :-)
Feb 08 02:42:03 <wpwrak> i have some folders with 67 k, though (the quarterly catch from my unfiltered mailbox)
Feb 08 02:47:13 <DocScrutinizer> nah, not exaggerated. [community] has 11844 
Feb 08 02:47:18 <DocScrutinizer> *unread*
Feb 08 02:47:43 <DocScrutinizer> plus 57559 read mails
Feb 08 02:50:18 <DocScrutinizer> 57559 _read_ messages...
Feb 08 02:50:23 * DocScrutinizer shudders
Feb 08 02:50:56 <wpwrak> what a way to waste one's life, eh ? ;-)
Feb 08 02:51:09 <DocScrutinizer> yup
Feb 08 02:52:17 * DocScrutinizer wonders if that's enough of a reasoning to get official retiring status
Feb 08 02:53:35 <wpwrak> taking the invalidity route ? :)
Feb 08 02:53:48 <DocScrutinizer> just pondering
Feb 08 02:55:36 <DocScrutinizer> you can not read 280MB of emails in 2 years and stay sane
Feb 08 02:57:02 <wpwrak> sanity is overrated anyway
Feb 08 02:57:12 <DocScrutinizer> now that I've realized and accepted that, I feel better instantly ;-P
**** ENDING LOGGING AT Mon Feb 08 02:59:57 2010