**** BEGIN LOGGING AT Mon Jul 06 02:59:56 2009 Jul 06 06:49:50 morning Jul 06 06:59:32 morning Jul 06 11:23:33 hello Jul 06 11:24:10 I just installed SHR (daily from 20090703) Jul 06 11:24:28 I see icons (the ones I selected as being available) Jul 06 11:24:44 but I don't see any icons for mokomaze and other programs I installed via opkg later on :( Jul 06 11:25:06 even though their desktop file is in /usr/share/applications everything looks ok. Jul 06 11:25:32 also - some have Categories=Office - some Games etc. Jul 06 11:25:42 but those that show up - all just show up on the desktop. Jul 06 11:25:54 FiXion: i mean you must change the .desktop files Jul 06 11:25:58 is there a "menu" - like the kde (or gnome) menu where I can find these categories? Jul 06 11:26:02 s/mean/think/ Jul 06 11:26:02 tobii_ meant: FiXion: i think you must change the .desktop files Jul 06 11:26:10 tobii_: I found an entry in tweaks about this - but they seem to be fine Jul 06 11:26:34 then i don't know Jul 06 11:26:57 also - when I install a lot of programs - I don't want icons for all of them in the "quicklaunch" area.. it's too small. Jul 06 11:27:10 is there a menu of some sorts that I can setup? Jul 06 11:27:22 do you have icons for ALL your programs in that small desktop ? Jul 06 11:27:30 or do you start them via terminal? Jul 06 11:27:49 i had all on my desktop :) Jul 06 11:28:08 tobii_: had? Jul 06 11:28:30 have Jul 06 11:28:30 ^ Jul 06 11:28:58 anyone know of a menu - which would order the icons according to their given categories? Jul 06 11:29:09 (like on normal linux desktop) - so I can browse games category Jul 06 11:29:11 f.ex. :) Jul 06 11:30:38 and anyone know of a replacement for mofi that actually works with WEP? Jul 06 11:36:08 do you all run shr -and then just do opkg update&& opkg upgrade? Jul 06 11:36:18 * FiXion_ just got his gta02 :) Jul 06 11:40:28 FiXion_: yes Jul 06 11:49:22 I'd like to code for it myself - but I'm a sysadmin - so I usually code shell/perl.. and don't mess much with gui. Jul 06 11:49:39 any pointers to docs for illume Jul 06 11:58:28 tobii_: btw. where do I report bugs like opkg upgrade failing? http://pastebin.com/m25194de4 Jul 06 11:58:47 cannot satisfy dependencies.. Jul 06 11:59:24 it's the frameworkd-config-shr-dev package which has a dependency on a version of frameworkd-config-shr that isn't available Jul 06 12:10:59 FiXion_: you will find the menu determining the content of the desktop in /etc/xdg/menus Jul 06 12:11:38 theoretically you are supposed to even sort them by categories... some way. Though no one ever managed to acomplish that :P Jul 06 12:11:48 if you have success in that... tell us :D Jul 06 12:28:48 mrmoku: I will. I will be looking to add an icon for a popup menu. Jul 06 12:46:18 what controls the layout in the "illume" top-bar ? Jul 06 12:47:30 shr-messages was interrupted by a segfault :( Jul 06 12:47:47 FiXion_: right after booting? Jul 06 12:48:44 mrmoku: I never opened the "messages" before. it says this when I try to start it, using the icon Jul 06 12:49:16 mrmoku: running it from cli says more: http://pastebin.com/m7767e293 Jul 06 12:49:35 resource devices is not enabled Jul 06 12:49:42 status is 'disabled' Jul 06 12:50:00 not very helpful :( Jul 06 12:50:04 to me atleast Jul 06 12:55:43 I miss bash.. is it too heavy for the freerunner? Jul 06 12:56:00 and what makes opkg a much better choice than apt* from Debian? Jul 06 12:56:18 it worked like a charm with neovento (but neovento isn't a very stable "phone") Jul 06 13:03:39 any hints as to where I add keyboard layouts to the keyboard app? (it only offers english as is) Jul 06 13:15:21 FiXion_, /usr/lib/enlightenment/modules/illume/keyboards/ Jul 06 13:15:26 contains the layouts Jul 06 13:15:45 searching the openmoko wiki for illume keyboard might help too Jul 06 13:16:38 mrmoku: I tried googling lists.shr-project.org and google in general Jul 06 13:16:55 * FiXion_ is working on finetuning his google-fu for the freerunner and shr :) Jul 06 13:17:48 http://wiki.openmoko.org/wiki/Illume_keyboard Jul 06 13:18:03 mrmoku: yup - found it. Jul 06 13:18:10 mrmoku: thank you for your help Jul 06 13:18:50 DocScrutinizer: ping Jul 06 13:22:30 mrmoku: if I made a danish keyboard layout (would be for the Terminal edition) - should I send it to devel@shr-project.org? Jul 06 13:22:50 I can't find | (pipe) on the Terminal keyboard :) Jul 06 13:22:56 and I'd like danish chars :) Jul 06 13:23:45 FiXion_: | is on the right side with shift on Jul 06 13:24:11 and yes, if you have a danish keyboard layout feel free to send it to devel@shr-project.org :) Jul 06 13:24:52 mrmoku: I found | - but I think I'll find some time and get my "feet wet" with a small project Jul 06 13:25:12 adding a "menu applet" to the top illume bar would be nice Jul 06 13:25:37 it seems the different illume themes can change the topbar contents.. Jul 06 13:25:47 so I was thinking of having a look at them Jul 06 13:26:20 FiXion_: those thingies in the topbar are called gadgets Jul 06 13:26:32 you can enable/disable them via the illume config Jul 06 13:26:49 which is accessible if you open the topbar and tap on the wrench Jul 06 13:26:55 so basicly I "just" need to add a "menu gadget" Jul 06 13:27:13 yep Jul 06 14:09:30 hi, I've a problem with the list...some I linked to some mailing list post on my blog...and now what's at the address(the message content) have changed Jul 06 14:10:55 s/some/some posts Jul 06 14:11:16 oops that's very confused...I'll retry Jul 06 14:11:48 :) and tell *which* list Jul 06 14:12:20 hi, I've some problem with the mailing list archives at http://lists.openmoko.org/pipermail/community/2008-November/ I had link that pointed to the thread on sisvel/mp3 and they now point to something else Jul 06 14:12:50 <[Rui]> damn, even hidden objects occupy space in windows :( Jul 06 14:13:00 (http://lists.openmoko.org/pipermail/community/2008-November/035618.html and http://lists.openmoko.org/pipermail/community/2008-November/035635.html) now points to some others posts Jul 06 14:17:06 I guess they were all regenerated after Sean's recent announcement post got truncated and they ran some regenerate script or something Jul 06 14:17:17 (unfortunately) Jul 06 14:25:07 FiXion_: ("menu gadget") add "quickStart" Jul 06 14:25:27 it's already there Jul 06 14:27:58 shr-u (20090624), I've expericenced serval screen dithering in recent week. Jul 06 14:29:24 The condition is: audo dimming on, auto suspending off, hot day (>=30 centigrade), phone was put in a portable bag Jul 06 14:30:14 Gnutoo: http://admin-trac.openmoko.org/trac/ticket/1740 Jul 06 14:30:37 DocScrutinizer, ok lol thanks Jul 06 14:31:51 Gnutoo: http://www.google.de/search?q=site%3Ahttp%3A//lists.openmoko.org/pipermail/+"" Jul 06 14:32:47 DocScrutinizer, ok thanks Jul 06 14:35:38 I agree broken pipermail-URLs are a real PITA Jul 06 14:36:09 and an absolute NoGo Jul 06 14:44:36 ~seen morphis Jul 06 14:44:36 morphis was last seen on IRC in channel #openmoko-cdevel, 1h 25m 45s ago, saying: 'DocScrutinizer: ping'. Jul 06 14:45:08 another drive-by shooter ;D Jul 06 14:51:37 DocScrutinizer: he wanted to ask you about a method of "repairing" a testpoint suggested on m800 ML. Jul 06 14:51:45 (probably) Jul 06 14:51:53 yup Jul 06 14:52:29 he could have posted a pointer Jul 06 14:53:29 DocScrutinizer: he did it 2 days ago Jul 06 14:53:51 ooops Jul 06 14:54:20 DocScrutinizer: i'll try to find it Jul 06 14:54:22 well, low traffic on #, might backscroll Jul 06 14:54:52 DocScrutinizer: http://lists.gnufiish.org/pipermail/gnufiish-devel/2009-July/000158.html Jul 06 14:55:09 PaulFertser: thanks :-)) Jul 06 15:04:33 PaulFertser: here in germany we have a prank to call some stupid people, tell them we are telco service and now they urgently need to ground their phone by placing it into a bucket of water ;-) Jul 06 15:05:53 the suggested repair procedure is somewhat similarly usefull for the particular case Jul 06 15:08:11 probably the OP even meant exactly what he wrote: soldering-oil (an extremely corrosive flux used for welding copper tubbing) Jul 06 15:11:19 and with sandpaper or any similar mechanical means you very likely will remove isolation from some other layer and *not* contact the trace which the blind via is contacting to Jul 06 15:12:21 how about not soldering anything to TPs but build some spring-loaded connector for this TP array? ;) Jul 06 15:13:32 Wonka: that's what these TP are built for in the first instance ;-D. Alas it's too late for Morphis' M800 now :-/ Jul 06 15:13:38 ack Jul 06 15:15:08 still he can use it in a standard way (as a "phone"), until he starts trying to fix it with sandpaper and welding-oil Jul 06 15:17:27 I'd bet the OP of the sandpaper advice broke the GND-pad of the mentioned transistor due to plane dissipating much heat when he tried to remove the component. That's why he succeeded with his crude fix Jul 06 15:18:39 he had a whole outer-layer GNDplane to simply remove varnish and soler a new pad to it next to removed one Jul 06 15:23:58 roh: are you still handling git.openmoko.org these days? (if not, who do i need to talk to?) Jul 06 15:25:27 Weiss: it's rather unclear who's in duty to handle git. Roh, jc, and I have root access Jul 06 15:26:03 ah, thanks Jul 06 15:26:23 I'm really a noob in administrating git, so I'm rather reluctant to do anything Jul 06 15:26:37 i'm in need of a key update, since (to my shame) i lost my old private key Jul 06 15:27:16 uh, fail :) Jul 06 15:27:20 indeed Jul 06 15:27:45 the shame burns me inside Jul 06 15:28:20 aaah, ok. Probably I could do that. Let's wait though if roh is willing to handle that (if it's not that urgent for you) Jul 06 15:29:20 yeah, any time is fine - it just means the DRM stuff on git.openmoko is out of date wrt the current state of the art (currently available on git.bitwiz) Jul 06 15:29:45 Weiss: anyway I suggest you send your new pubkey to roh, jc, and me Jul 06 15:32:42 ok, will do Jul 06 15:34:09 Weiss: also a ticket on admin-trac might help. It's the most reliable way to handle those issues (well at least has been in the past) Jul 06 15:34:45 .oO( any form of authentication? ) Jul 06 15:35:20 .oO( maybe send a challenge to a perviously known mail address of the owner of the key to be replaced? ) Jul 06 15:36:04 hmm.. admin-trac would be better (authenticated by my Trac account) Jul 06 15:36:08 that's the way it has been handled Jul 06 15:54:10 ~seen jc Jul 06 15:54:10 jc was last seen on IRC in channel #openmoko-cdevel, 4d 2h 21m 14s ago, saying: 'man, don't again.'. Jul 06 15:56:08 Wonka: btw we luckily don't need to replace the key (as Weiss doesn't have to revoke it), it's enough to simply add the new one Jul 06 17:00:25 Weiss: did you open a admintrac-ticket? I don't find any new ticket component account Jul 06 17:01:39 not yet.. sorry, i'm juggling a few things at the moment Jul 06 17:02:09 np, just curious Jul 06 17:07:47 Weiss: for your convenience: http://admin-trac.openmoko.org/trac/query?status=accepted&status=assigned&status=new&status=reopened&group=component&component=accounts&order=changetime&col=id&col=summary&col=status&col=owner&col=type&col=priority&col=reporter&col=changetime&report=10&desc=1 Jul 06 17:10:12 <[Rui]> I don't read python well, does anyone know what widget is used in the paroli-msgs2 message list which allows that horizontal button reveal? Jul 06 17:11:52 tracfeed: Ticket #540 (default backlight percent 100%: waste power) created Jul 06 17:27:53 anybody able to get microphone working on freerunner with debian squeeze? My speakers are working fine, when I run 'arecord -d 5 foobar.wav' it doesn't complain, but only records silence. Jul 06 17:28:42 troy|work: voip*state Jul 06 17:29:21 I have tried one from the Manual Debian Wiki page... is there another somewhere that I might try? Jul 06 17:29:32 troy|work: one? Jul 06 17:29:51 another voip-handset.state file? Jul 06 17:30:13 54807d1f5dbcd0642f41af9da6bbff0a works Jul 06 17:30:56 sorry... I'm not up on the lingo. What does that refer to? Jul 06 17:31:08 and arecord -q -c 1 -f S16_LE -c1 -r44100 a.wav Jul 06 17:32:56 lindi-_: what does 54807d1f5dbcd0642f41af9da6bbff0a refer to? Jul 06 17:35:42 troy|work: md5sum Jul 06 17:36:24 lindi-_: sorry to pester... md5 sum of what exactly? Jul 06 17:49:46 Weiss japp.. just send me your new one and i'll add it later (wrong computer here) Jul 06 17:54:40 roh: thanks. a Trac ticket is coming up... Jul 06 18:20:12 tracfeed: Ticket #540 (default backlight percent 100%: waste power) updated Jul 06 18:28:15 tracfeed: Ticket #541 (MoFi disables usb0 permanently) created Jul 06 18:40:12 tracfeed: Ticket #542 (MoFi is deprecated and should die) created +> Jul 06 18:40:22 :{ Jul 06 18:40:53 Mofi never worked for me... Jul 06 18:41:14 why not include wifiman.py... Jul 06 18:41:15 Heinervdm, Mofi never worked, period. it only sometimes connected to networks :) Jul 06 18:41:27 Heinervdm, it's not perfect either Jul 06 18:41:29 though I use it. Jul 06 18:41:41 Heinervdm, it doesn't set up routing tables/dhcp Jul 06 18:41:50 (although I may be using an old version) Jul 06 18:42:24 But if it's in the image we are forced to work on it ;) Jul 06 18:42:32 haha true. Jul 06 18:42:32 :) Jul 06 18:42:48 Heinervdm, please create a ticket/send a mail to the community list :) Jul 06 18:43:28 * TAsn meant shr-user Jul 06 18:43:41 shr-devel Jul 06 18:43:41 ffs Jul 06 18:43:42 ;) Jul 06 18:43:50 to join in with app-bashing: filemanager is stupid, completely unusable and useless, and segfaults on cd'ing to a 4GB-vfat uSD Jul 06 18:44:40 DocScrutinizer, aye. Jul 06 18:44:44 both should die. Jul 06 18:44:49 though I hope you agree mofi should die first :) Jul 06 18:45:39 Ok, and next: I want to have a dates application in the image Jul 06 18:46:45 propably an app using an opimd-dates Jul 06 18:46:51 :) Jul 06 18:48:27 same here :) Jul 06 18:49:10 actually, I have many issues I want to address, though I have a crazy calculus test next week (though after that, I'm free, weepee!) Jul 06 18:50:49 I've looked at a dates app and a opimd-dates domain but dates is a lot different to messages and contacts Jul 06 18:51:12 So i have given up Jul 06 18:51:38 and now i'm looking for a webbrowser Jul 06 18:51:56 midori is nice :) Jul 06 18:52:01 though buggy Jul 06 18:52:44 speaking about webbrowser choise for the neo? Jul 06 18:52:48 the font size in midori is much to big for me Jul 06 18:53:36 (mofi) honestly I don't get it what might be so damn complicated whith simply porting any reasonably well performing wifi-manager app out there to OM Jul 06 18:53:55 my version of woosh fits my needs the best currently Jul 06 18:54:04 but it has no fingerscrolling Jul 06 18:54:33 DocScrutinizer, actually Jul 06 18:54:36 for some reason Jul 06 18:54:40 shr uses connman Jul 06 18:54:49 and not a working connection manager ;( Jul 06 18:54:56 we can use networkmanager Jul 06 18:55:02 which works just fine Jul 06 18:55:07 ack Jul 06 18:58:51 * TAsn votes for network manager Jul 06 18:58:52 :) Jul 06 18:59:24 we should start focusing on features + usability and not always trying to accept beta software :) Jul 06 19:10:16 hi Jul 06 19:12:01 DocScrutinizer: (remove varnish from GND layer) not the best idea to get TP back, full ack :) Jul 06 19:14:00 PaulFertser: I strongly recommend glueing the original TP to it's original position correctly oriented, then use silver varnish to contact to the via stub that's probably still there Jul 06 19:14:00 DocScrutinizer: on morphis's photo it looks like there's a microvia in the left part of the place where TP was. Do you have an idea of its diameter? Probably he can find a very thin wire to stuff in it? Jul 06 19:14:28 it's a laser-microvia Jul 06 19:14:44 DocScrutinizer: not metalised? I'm unfamiliar with this technology :( Jul 06 19:15:03 and probably filled with gold coating Jul 06 19:15:29 inside is filled, outside is glue Jul 06 19:15:54 DocScrutinizer: how can he use silver varnish _after_ glueing the TP back? Jul 06 19:16:07 so a 0.05mm top flat surface is all that's probably there to contatct Jul 06 19:16:54 why not? the via should fit into the hole it left inside the TP Jul 06 19:18:08 I mean i don't see how to ensure contact between the via and the TP Jul 06 19:18:36 if there's no hole inside TP (i.e. via torn ahalf "below ground"), then morphis needs to resect a very small part of TP to enaple varnish flow to via and TP surface Jul 06 19:19:47 btw the via on my M800 is located on the radius of TP at 16:00 position Jul 06 19:19:48 DocScrutinizer: what if he applies that varnish first and then puts the TP back, securing it there by other means? Jul 06 19:20:09 other means won't work Jul 06 19:20:30 Hi folk Jul 06 19:20:32 s Jul 06 19:20:56 I tried upgrading my SHR unstable, and got: ERROR: Cannot satisfy the following dependencies for frameworkd-config-shr-dev: frameworkd-config-shr (= 0.8.5.1+gitr1473+130fb4c50f81b803d42b32c69f20e0713930f741-82+060495f1ddcf6db79c1cd9d12e51b47adac8e2ed-r6) Jul 06 19:21:05 PaulFertser: I have a silver varnish here that states to be even solderable at low temp Jul 06 19:21:07 how do i activate the sixteen theme? Jul 06 19:21:33 settings -> look -> theme and ? Jul 06 19:22:16 oh, and "Package frameworkd-config-shr-dev is depended upon by packages: frameworkd-config-shr-dev" Jul 06 19:22:17 DocScrutinizer: what if he places a wire on the board, glues it to the pcb, bends the leaded end downward and fills the place where TP was with the varnish? Jul 06 19:22:27 sybren: ignore it. Jul 06 19:22:38 sybren: it works well regardless. Jul 06 19:22:52 but still I strongly suggest to glue part of TP to PCB, and just leave out the area of via so you can contact TP and via with silver Jul 06 19:23:15 PaulFertser: also feasible Jul 06 19:23:40 will spoil the other TP in vicinity though Jul 06 19:23:45 PaulFertser: thanks, will do. I'm uninstalling the frameworkd-config-shr-dev package with force-depends, as it doesn't contain any files anyway ;-) Jul 06 19:24:56 DocScrutinizer: would you mind if i send him this converstation via email right now until he damaged everything with sand paper? Jul 06 19:25:16 go ahead. appreciated Jul 06 19:26:03 maybe you like to CC me Jul 06 19:29:06 Mail sent Jul 06 19:30:25 just 30sec ago got a mail, but was spam of my mail-rovider :-P Jul 06 19:31:19 Hm, one russian guy goes to Barcelona 11-18 Jul, now if only i could find somebody there who wants to sell me his gta01... Jul 06 19:34:51 * rtp doesn't want to sell his gta01bv[34] :) Jul 06 19:36:27 rtp: :D Jul 06 19:36:37 rtp: but you don't want to really hack on it too ;) Jul 06 19:37:45 rtp: btw, i know know about problems and how they should be solved with battery meter. Do you have anything else in mind? Does shr unstable work for you? Jul 06 19:38:59 reinstalled a fresh unstable on mmc (on mmc because shr guy doesn't seem to produce images fitting in the mtd) and it doesn't seem to work Jul 06 19:39:33 but it's not a matter of gta01 only. from time to time I'm using my gta01 battery on my gta02 and get same behaviour Jul 06 19:39:49 dos1: hey! do you know rtp is not the first to complain that gta01 image doesn't fit? Can you take care of notifying those who can fix it (mrmoku?)? Jul 06 19:40:06 rtp: gta01 battery can't be monitored on gta02. Jul 06 19:40:45 DocScrutinizer: btw, SpeedEvil says that battery current measurement is extremely and hopelessly noisy on gta01. The logs guys showing me seem to confirm that :-/ Jul 06 19:40:53 PaulFertser: ouch. annoying. I'm using my gta01 battery as backup so it would be nice if I can more or less rely on it Jul 06 19:41:27 PaulFertser: also using gta02 battery on gta01 doesn't work too :) Jul 06 19:41:30 DocScrutinizer: it's 0.22 shunt and the current reading jumps like mad +- 100mA even (that'd be 22mV difference). Jul 06 19:42:28 rtp: i think you can approximately tell the remaining charge with gta01 bat in gta02 by reading "voltage_now" sysfs node. Jul 06 19:42:30 mompl Jul 06 19:42:57 PaulFertser: well, same on gta01. I'm looking at voltage/current_now Jul 06 19:43:32 rtp: i'll provide percentage estimate soon. You know the one present in .24 is completely bogus (fully linear). Jul 06 19:44:11 PaulFertser: btw, it looks like the gta01 is charging the battery forever. the status always contains "charging" :( Jul 06 19:45:24 PaulFertser: too bad. linear is not good. You're looking at the current or voltage to make your curves ? Jul 06 19:45:55 rtp: voltage. poly-segment approximation. Jul 06 19:46:38 why not current ? iirc it's better than voltage Jul 06 19:46:58 rtp: http://www.ibt-power.com/Battery_packs/Li_Ion/Lithium_ion_tech.html Jul 06 19:47:49 anyone know if Paval Mechack?? is on irc? Jul 06 19:47:58 Pavel Jul 06 19:48:03 tmzt: Pavel Machek? Jul 06 19:48:07 rtp: (forever) same on 02. That's bogus info from a silly broken kernel-hack, that should have been removed by PF long ago ;-) Jul 06 19:48:22 DocScrutinizer: not quite so :( Jul 06 19:48:23 yeah, sorry Jul 06 19:49:13 DocScrutinizer: i fixed Qi only according to your (and Werner's?) suggestion to change minimal charging current threshold. Jul 06 19:49:42 PaulFertser: if it's a broken hack, how am I supposed to know it's fully charged ? :) Jul 06 19:50:01 rtp: current doesn't help anything for capacity estimation, as you'll lose track to integrate it anyway during suspend at very least Jul 06 19:50:06 rtp: is it with GSM on? Jul 06 19:50:37 tracfeed: Ticket #542 (elmdentica Package request) created Jul 06 19:50:45 (noisy current) well probably current actually _is_ that noisy Jul 06 19:52:30 DocScrutinizer: i can't imagine how you can get 350, a second later 450, second later 500 and then 350 again :S Jul 06 19:52:35 PaulFertser: (fix) we need to remove that nasty silly kernel-charger-restart that makes state display "charging" exactly 15min after PMU stopped charger due to chrg-current-limit-end Jul 06 19:52:40 DocScrutinizer: without gsm or gps used. Jul 06 19:52:47 DocScrutinizer: i know, i know... Jul 06 19:53:25 DocScrutinizer: gotta do it. Still keeping the relevant email in my inbox and Werner's detailed code-level analisys marked in gnus. Jul 06 19:53:32 PaulFertser: yes, it's on as I'm using my gta01 as phone. my gta02 will be buzz-fixed soon (should have done it sooner, I know :P) Jul 06 19:53:34 PaulFertser: Werner mentioned some compile-time define REENABLE-CHARGE=0 or sth like that Jul 06 19:55:05 (noisy) hmm, without GSM it's a little hard to figure how those jumps might happen Jul 06 19:55:50 rtp: quite possibly your wonderful gta01 sucks so much current from the bat directly that PMU never stops charging. Or probably just one register tweak is needed as on gta02... It's too hard to debug low-level stuff remotely, you know. Jul 06 19:55:52 tracfeed: Ticket #540 (default backlight percent 100%: waste power) updated Jul 06 19:56:10 might be some form of aliasing with a regular (sawtooth, squarewave) current pattern caused by usual sheduler and UART activity Jul 06 19:56:25 even converters maybe Jul 06 19:57:03 PaulFertser: sure iirc it's sucking a lot of current but I think it was managing to stop in early days Jul 06 19:57:54 rtp: i think i can do the same tweak as for gta02 for gta01 qi, i don't want to touch u-boot though. And somebody gotta test it. Jul 06 19:58:20 after some hours a ultimate ratio emergency shutoff timer for batcharge should expire Jul 06 19:59:00 it's default for 50606 as well I think Jul 06 19:59:05 PaulFertser: why do you need to tweak something in uboot or qi and not in the kernel ? Jul 06 19:59:06 it's hw Jul 06 19:59:29 rtp: it's a long story... Jul 06 19:59:38 hehehe Jul 06 19:59:51 PaulFertser: like the batget stuff ? /o\ Jul 06 19:59:53 rtp: there's a custom in OM devices to setup some core pmu parameters in the bootloader. Jul 06 20:00:18 rtp: it seems stupid but there were some arguments for i don't remember. Jul 06 20:00:19 tracfeed: Ticket #540 (default backlight percent 100%: waste power) updated Jul 06 20:00:30 <|Marco|> this is strange, after activating gprs over dbus andthen deactivating successfully, I can't activate it again or poll directly with mdbus, it just hangs :/ Jul 06 20:00:34 as we start charging on BL (as early as possible) Jul 06 20:01:05 <|Marco|> this is with shr-shr-lite-image-glibc-ipk--20090703-om-gta02.rootfs.tar.gz Jul 06 20:01:06 |Marco|: not here Jul 06 20:01:11 <|Marco|> sorry Jul 06 20:01:25 works for me Jul 06 20:01:37 (via shr-settings) Jul 06 20:01:41 PaulFertser: I would have seen that in the machine file in the kernel but I think some reasons for doing it in uboot/qi can be found. bios on x86 are hiding a lot of stuff for instance Jul 06 20:02:00 <|Marco|> DocScrutinizer: tryed using mdbus on it ? Jul 06 20:02:01 rtp: and we all hate that, don't we ;) Jul 06 20:02:17 nope, no need to do that here ;-) Jul 06 20:02:24 PaulFertser: indeed :) Jul 06 20:03:13 mrmoku|away: just FYI: the shr-messages app worked after a reboot Jul 06 20:03:17 good evening :) Jul 06 20:04:03 how could add a small feature to the different apps in shr ? Jul 06 20:04:08 bye Jul 06 20:04:23 shr-messages f.ex. - should really look in the shr-contacts for a "hit" on the phonenumber in the sms Jul 06 20:04:38 <|Marco|> DocScrutinizer: got a nice message from mdbus now while it's hanging: http://pastebin.com/m3a5eddc4 Jul 06 20:04:54 FiXion_: afaik it does Jul 06 20:05:17 DocScrutinizer: not in my shr - daily from 0703 with opkg update && opkg upgrade just run. Jul 06 20:05:20 rtp: i think we can ask wpwrak why it is the case here. Jul 06 20:05:46 I have some weirdness in the shr-contacts though - before it just showed no entries - consistently. But now - after several reboots - it suddenly shows them again Jul 06 20:06:05 and it crashed - with "no message" - do you want to save? Jul 06 20:06:06 PaulFertser: I already mentioned it's for early start of bat-chg Jul 06 20:06:06 :) Jul 06 20:06:20 DocScrutinizer: only that? Jul 06 20:06:32 DocScrutinizer: that's no good reason for not duplicating pmu init in kernel. Jul 06 20:06:39 DocScrutinizer: and "unfortunately" it's a compiled app - so I need to get the source and setup a build environment Jul 06 20:06:41 well, at least hat's one of the points Jul 06 20:06:46 I suppose you're not building on your phones :) Jul 06 20:06:59 PaulFertser: we need to set up some parameters so that we can boot at all. i think the kernel should reset the pmu as well, though. and boot loader and kernel should share as much of the same data (at compile time) as possible. Jul 06 20:07:42 muuaahahaa, same annoying story. Yes we've all been there Jul 06 20:07:43 wpwrak: why share data? If we got bootloader that can boot us, we can concentrate on getting the right values in kernel and not care about bootloaders anymore, can't we? Jul 06 20:08:23 well, bye fellas, now for real Jul 06 20:08:58 ohh - and mofi does absolute not work at all with WEP networks Jul 06 20:09:21 and for some reason wpa_supplicant wouldn't work either - even though the same worked fine on om2009t5 and neovento5.6 (and works on my laptop) Jul 06 20:09:32 but manually conneting with iwconfig works perfectly Jul 06 20:09:45 just ignore my rants if they are misplaced :) Jul 06 20:09:47 [2009-07-06 20:40:12] tracfeed: Ticket #542 (MoFi is deprecated and should die) created DocScrutinizer: see you Jul 06 20:10:00 I'm not quite sure where I should report what - if anything of this kind Jul 06 20:10:08 PaulFertser: just to make sure the two don't diverge needlessly. also, if we're in the 100 mA envelope, any divergence is a warning sign Jul 06 20:10:21 I'd just like to see if I could start fixing the things I can figure out.. :) Jul 06 20:10:30 DocScrutinizer, you snitch :) Jul 06 20:10:59 wpwrak: because we are too afraid to diverge we now don't fix kernel at all and count on bootloader that is too old and doesn't include our fixes anyway for most users. Jul 06 20:11:43 FiXion_, feel more than welcomed to report a bug concerning wpa_spplicant, though concerning mofi, at least in my opinion, don't :) (as DocScrutinizer stated by pasting my quote) Jul 06 20:13:12 wpwrak: i remember Andy tried to boot in 100mA and the kernel started but died on enabling BL. Do you know if anybody tried to not enable BL, was it enough to boot? Jul 06 20:13:13 ahh - I just see you have a build your own distro guide on the trac page as well Jul 06 20:13:44 TAsn: thanks. Jul 06 20:13:45 PaulFertser: afaik people are not informed when there are needed updates of qi/uboot. that doesn't help :) Jul 06 20:14:00 TAsn: you know if there's any network-manager package for arm? Jul 06 20:14:20 ohh - and anyone know why almost everybody is using opkg instead of apt* Jul 06 20:14:23 FiXion_, thank you for reporting Jul 06 20:14:27 rtp: that's my point. I think we can just reintroduce all pmu initialization in kernel and don't worry if it diverges. Jul 06 20:14:33 FiXion_, it has been said that apt is heavy Jul 06 20:14:38 haven't checked that myself Jul 06 20:14:40 I tried neovento 5.6 - which is a debian - and it ran perfectly Jul 06 20:14:51 no idea then. :) Jul 06 20:14:51 and honestly opkg is a bit short.. Jul 06 20:15:00 FiXion_, that's not being honest Jul 06 20:15:05 opkg is amazingly short Jul 06 20:15:10 and debian/neovento had a rather large arm repository Jul 06 20:15:13 it sucks in every possible aspect :) Jul 06 20:15:20 PaulFertser: (diverge) naw, just change the build process such that kernel and qi can share the pmu init. done :) Jul 06 20:15:23 FiXion_, ipk = deb 2 packages Jul 06 20:15:28 ofcourse not all "optimized" for freerunner though Jul 06 20:15:30 i.e, I think you can install them with apt-get Jul 06 20:15:36 PaulFertser: why not. would be easier to handle Jul 06 20:15:47 wpwrak: it's easier said than done. I can do that ad-hoc on my host but i'm yet to find a "proper" way. Jul 06 20:16:01 PaulFertser: (100 mA) if the kernel got as far as switching on the backlight, it's probably okay. i don't know if anyone tried. probably not. Jul 06 20:16:13 TAsn: ipk? is that a tool? Jul 06 20:16:40 FiXion_, ipk is the package file extension, .ipk :) Jul 06 20:16:43 afaik opkg can handle .deb packages as well - and ipk's are "almost" debs Jul 06 20:17:02 wpwrak: strange, seems to be so easy and fun... Jul 06 20:17:07 FiXion_, I think ipks are exactly debs or maybe extended debs Jul 06 20:17:14 PaulFertser: +1 for your approach Jul 06 20:17:27 and I'd also like to test out bash Jul 06 20:17:33 really miss history and ctrl+r Jul 06 20:18:03 then just test out bash :) Jul 06 20:18:09 you are the only one stopping yourself Jul 06 20:18:11 btw Jul 06 20:18:13 it doesn't seem to be in the opkg repository Jul 06 20:18:15 opkg install bash Jul 06 20:18:20 FiXion_, first of all Jul 06 20:18:26 if you don't see a package you want Jul 06 20:18:30 just open a bug in the shr trac Jul 06 20:18:36 and we'll (someone'll) add it Jul 06 20:18:38 dooh Jul 06 20:18:42 wpwrak: so, if we do some initial settings in uboot/qi and change them inside the kernel, it can prevent it to boot ? is that what you meant ? Jul 06 20:18:45 I was sure I tried opkg installing it Jul 06 20:18:54 TAsn: I'd rather just build it myself Jul 06 20:19:00 second of all, you can just search the angstrom repositories for nice packages :) Jul 06 20:19:07 FiXion_, don't you want it to be in the repos? Jul 06 20:19:09 I've built tonnes of deb and rpms for different distro's Jul 06 20:19:20 TAsn: then I'll just give you a link so you can pull it. Jul 06 20:19:30 FiXion_, that's not how we work Jul 06 20:19:35 we need bitbake recipes :) Jul 06 20:19:47 if you just want a package, angstrom repository is the way to go ;) Jul 06 20:19:47 TAsn: you get the debian folder package as well Jul 06 20:19:50 if opkg does that Jul 06 20:20:03 that's why I wanted to setup my own build environment Jul 06 20:20:06 I don't think I quite get you. Jul 06 20:20:13 TAsn: what do you mean? Jul 06 20:20:20 TAsn: you get the debian folder package as well Jul 06 20:20:25 not sure I got what you meant :) Jul 06 20:20:41 anyhow, setting up a build environment is easy Jul 06 20:20:42 FiXion_: http://people.openmoko.org/joerg/unsorted/opkg-u Jul 06 20:20:45 TAsn: a debian package is build - by creating a debian folder (within the source folder) - containing the files for the deb tool to build a package Jul 06 20:20:57 rpm's are easier - they just have a .spec file Jul 06 20:21:03 * FiXion_ goes helping his wife Jul 06 20:21:15 actually, deb packs are easy Jul 06 20:21:27 I don't remember the exact compression format atm Jul 06 20:21:41 ar Jul 06 20:21:46 though it's just a compressed archive that consists of files and dirs Jul 06 20:21:48 Heinervdm, thanks Jul 06 20:21:50 TAsn: debs are ar archives and .tar.gz Jul 06 20:21:53 rtp: yeah, you basically have to make sure you don't consume more than 100 mA before you know you have a good (i.e., > 100 mA) power source Jul 06 20:22:06 and a CONTROL dir Jul 06 20:22:23 for postinst/postrm/info and such scripts/configs Jul 06 20:22:23 :) Jul 06 20:22:27 wpwrak: i still can't see why "diverging" is considered bad. Jul 06 20:22:28 rtp: u-boot implements a whole USB stack, so it can find out. Qi mercifully doesn't to this. that's why the kernel has to stay in the 100 mA envelope Jul 06 20:22:49 TAsn: the control stuff is in the control.tar.gz file iirc :) Jul 06 20:22:59 rtp, oh right! Jul 06 20:23:01 it's a package Jul 06 20:23:06 *an archive Jul 06 20:23:12 that consists of two archives Jul 06 20:23:14 right? Jul 06 20:23:16 one is dat Jul 06 20:23:17 a Jul 06 20:23:21 the other is control Jul 06 20:23:31 (haven't handled that for quite some time) Jul 06 20:23:32 PaulFertser: just to keep things simple. keep things at one place, change them only once :) Jul 06 20:24:16 wpwrak: but if the kernel is booting to the step one can change some settings, can't we assume that the power source is ok ? Jul 06 20:24:24 wpwrak: do you have any example of such a setting? Jul 06 20:25:53 wpwrak: and yes, i join rtp's question. Jul 06 20:26:23 TAsn: debs are easy to build - there's all those nice tools so it's no problem :) Jul 06 20:27:06 same goes for ipks Jul 06 20:27:07 btw Jul 06 20:29:02 SHR: 03mikael.berthe 07shr * r2cedd20ce8e7 10/libframeworkd-phonegui-efl/data/ussd.edc: libframeworkd-phonegui-efl: handle newlines in USSD window Jul 06 20:29:37 woooohoooo, finally :-) Jul 06 20:29:55 what's ussd? Jul 06 20:30:15 * TAsn is puzzled as he heard this name before but never bothered asking Jul 06 20:31:08 McKael: are you now into edc hacking? ;) Jul 06 20:31:27 PaulFertser: haha Jul 06 20:31:27 btw. I added the "start" gadget to the top - but when it opens the apps menu - it is shown "off screen" :( Jul 06 20:31:40 instead of switched to the other side Jul 06 20:31:49 is it just borken or is there a setting? Jul 06 20:31:55 dontbestupid=1 Jul 06 20:31:56 :) Jul 06 20:31:59 borked Jul 06 20:31:59 dos1: thanks for applying Jul 06 20:32:01 McKael: when i tried to change something in the Zhone edc i had a hard time understanding anything about it ;) Jul 06 20:32:15 McKael: no problem, thanks for patch :) Jul 06 20:32:20 DocScrutinizer: would be even better if navigating through ussd "menus" were more intuitive :) Jul 06 20:32:36 yup Jul 06 20:32:45 PaulFertser: I just looked at other files -- I still know nothing about edc ;) Jul 06 20:32:48 I think I opened a ticket for that? Jul 06 20:32:56 McKael, haha, I know this feeling :) Jul 06 20:33:03 (from edc hacking that is) Jul 06 20:33:05 DocScrutinizer: don't know Jul 06 20:33:15 couldn't find any decent documentation Jul 06 20:33:33 errrks, docs for E* Jul 06 20:33:40 :-( Jul 06 20:33:51 DocScrutinizer, never heard about these kind of things. Jul 06 20:33:57 only rumors Jul 06 20:34:01 never seen ones myself Jul 06 20:34:18 simply because there isn't? Jul 06 20:34:38 McKael: next step is porting mcabber to elementary, i hope? ;) Jul 06 20:34:39 aren't the sources considered as docs ? :) Jul 06 20:34:56 DocScrutinizer, probably :) Jul 06 20:34:59 muhaha, same as on ALSA Jul 06 20:35:01 rtp, probably Jul 06 20:35:57 oldest coders' joke ever: "the source is the doc" Jul 06 20:36:38 DocScrutinizer: don't know for userspace but when I wrote a asoc driver, I found some docs in the kernel Documentation dir :) Jul 06 20:36:40 usually it's simpler to code it new than to understand *that* type of docs Jul 06 20:37:07 rtp, the kernel ppl are generally nicer :) Jul 06 20:37:29 alsa != kernel, for that instance Jul 06 20:37:33 PaulFertser: not sure about that ;) Jul 06 20:38:02 TAsn: I'm not really sure of that :) Jul 06 20:38:16 BTW It took me 2 days or 3 to get the openembedded environment ready Jul 06 20:38:17 btw FSO docs aren't much better on that behalf Jul 06 20:38:20 rtp: oh, you're into asoc stuff. Wanna fix one funny bug on gta02? Jul 06 20:38:21 btw. why does free -m say I only have 118MB ram? Jul 06 20:38:28 DocScrutinizer, at least they somehow doc the api Jul 06 20:38:30 got lots of failures Jul 06 20:38:31 :) Jul 06 20:38:32 I thought the free runner came with more? Jul 06 20:38:35 and there's always mdbus Jul 06 20:38:44 118 under the total column Jul 06 20:39:02 rtp, It's something I find good to believe :) Jul 06 20:39:17 PaulFertser: which bug ? Jul 06 20:39:50 rtp: it may be okay, but perhaps it could be better. every once in a while, someone finds some optimization, etc. Jul 06 20:40:03 TAsn: though neither the API doc nor mdbus tell you about semantics, nor teach any detail about e.g. xaml-rules Jul 06 20:40:17 how do I find out which package on opkg contains libgps.so.16 ? Jul 06 20:40:36 DocScrutinizer, aye, the config files are not even slightly documented (at least from what I could see) Jul 06 20:40:44 I tested the navit 0.1.0-svn2309 - and it's borken.. it says it doesn't support "binfile" :) Jul 06 20:40:45 rtp: but what I mean is more a way for keeping the code sane than to force boot loader and kernel into agreement Jul 06 20:41:03 rtp: http://www.mail-archive.com/openmoko-kernel@lists.openmoko.org/msg08658.html Jul 06 20:41:06 they've just released 0.1.1 - I installed that from them - and it is build with libgps.so.16 Jul 06 20:41:18 rtp: it's a lot of parameter and you already have the whole version matrix to worry about, so why make it even more complex ? Jul 06 20:44:00 Does the ring volume setting work for someone, with shr-unstable? Jul 06 20:45:17 wpwrak: the problem is to find the best solution possible for everyone. I'm not convinced playing with uboot/qi and asking to people to update them is a bonus point. updating the kernel is far easier Jul 06 20:45:48 McKael: don't worry about ring volume, soon will be obsolete anyway. But no, at least sms-notify is completely borked for me Jul 06 20:45:58 DocScrutinizer: ok Jul 06 20:46:50 PaulFertser: I'm reading the thread but seeing that Broonie answered to this thread, I believe he has given better answer than me. Jul 06 20:47:21 rtp: he has given an answer ("aliasing is tricky stuff and should be done right") but not the patch, unfortunately Jul 06 20:48:19 PaulFertser: too bad. broonie knows a lot about asoc and wm codecs for sure Jul 06 20:48:31 Sure :D Jul 06 20:49:01 Guys, Pavel Machek writes he now can boot his HTC Dream with 2.6.31 and Zaurus userland. Jul 06 20:51:32 PaulFertser: the palm pre would be more interesting than the htc. iirc, it's a omap3 and the kernel doesn't has a lot a "funky" special kernel patches Jul 06 20:52:12 rtp: agree. But nothing's better than gta02 anyway ;) Jul 06 20:52:13 rtp: i didn't say you have to ask them to upgrade :) but eventually, they might. and you can't necessarily predict which boot loader and kernel combinations they'll run. Jul 06 20:52:20 PaulFertser, haha. ;) Jul 06 20:52:48 PaulFertser: for a one year old, it already has quite a story ;-) Jul 06 20:54:27 Wow, good words about SHR and gta01 from Pavel Machek: "shr unstable from may 9 works quite nicely. It needs gllin 1.1, but with that GSM&GPS just works... Python is also pre-installed, along with pygtk. Good." Jul 06 20:54:31 wpwrak: well, if upgrading allows to get battery related stuff in better state, would be stupid to not upgrade :) Jul 06 20:55:17 PaulFertser, I don't know the dude, who is he? Jul 06 20:55:25 (except for an htc hacker) Jul 06 20:55:53 TAsn: quite famous Linux kernel dev. Jul 06 20:56:22 oh, I'm not into celebs :) Jul 06 20:56:33 I barely know who's esr Jul 06 20:56:39 * rtp is wondering if someone should tell him to _not_ upgrade its shr image Jul 06 20:56:51 Ahaha Jul 06 20:57:44 I really missed the oabi compat in current kernel. Jul 06 20:58:34 rtp: is it still missing? Jul 06 20:58:44 rtp: i thought mrmoku|away fixed the config Jul 06 20:59:17 Hm... could somebody explain how how to connect debug board to neo, for very stupid person? I want kernel serial console... Jul 06 20:59:25 PaulFertser: I though too until I flashed this week current kernel. I've replied today to his mail on shr-user Jul 06 21:00:08 max_posedon: very carefully is the key here. Try to not bend the flexi cable at all. Jul 06 21:00:41 PaulFertser: I fail to understand how this has been changed back to disabled :( Jul 06 21:00:44 my main problem, I don't understand what usb cable where I should put Jul 06 21:01:40 max_posedon: you take a regular usb - miniusb cable and connect it to the debug board Jul 06 21:02:30 max_posedon: http://wiki.openmoko.org/wiki/Connecting_Neo_1973_with_Debug_Board_v2 Jul 06 21:07:01 weird...if i try to start frameworkd with multiple subsystems("frameworkd -s ousaged,ogsmd") it throws an DBusException: http://www.pastebin.ca/1486105 Jul 06 21:07:03 why? Jul 06 21:09:03 doxy2: /usr/bin/python /usr/bin/frameworkd -s ousaged,ogs ? Jul 06 21:09:06 doxy2: ogs? Jul 06 21:09:21 PaulFertser: but there's no typo Jul 06 21:09:29 oops, sorry, so I'm not sure how to connect usb cables, and how to get serial console Jul 06 21:10:02 PaulFertser: maybe it only prints a fixed length :) ? Jul 06 21:10:03 doxy2: ah, ok it's just truncated. Anyway looks like the problem with dbus permissions, look at your /etc/dbus/... configs. Jul 06 21:11:00 max_posedon: take a cable like you use to connect your FR to your PC. Connect flexicable between FR and dboard. Plug the usb cable in the board and in your PC. That's all, the console is at /dev/ttyUSB0. Jul 06 21:11:26 PaulFertser: frameworkd.conf? policy_user is set to root for all. and i'm running frameworkd as root Jul 06 21:12:30 doxy2: probably you don't have an entry for org.freesmartphone.Resource interface at all Jul 06 21:13:04 PaulFertser: well, dang. i dont Jul 06 21:13:45 doxy2: i'm not a dbus expert, just a guess based on your paste. It seems like you can fix it after some experimenting with that policy file. Jul 06 21:16:00 PaulFertser: hmm wonder why it was missing from the file. Resource is listed under org.freesmartphone.Usage API in the FSO DBus specification. does it use ousaged object? Jul 06 21:18:53 doxy2: no idea tbh. Did adding it help? Jul 06 21:21:18 PaulFerser: nope Jul 06 21:24:34 PaulFertser: still, i can start frameworkd solely with ousaged. it's fails only when there's ousaged and others Jul 06 21:25:40 doxy2: strange thing is that it fails because of the policy. I'd say that the policy is possibly wrong. Jul 06 21:26:00 PaulFertser: for example : "frameworkd -s ogsmd,odeviced" would work Jul 06 21:29:48 PaulFertser: yeah. could be. this is what i have there now: http://www.pastebin.ca/1486132 Jul 06 21:31:37 doxy2: i'd also try to add allow_own for org.freesmartphone.Resource but then i'd give up ;) Jul 06 21:32:04 PaulFertser: but then again..."dbus-send --system --print-reply --dest=org.freesmartphone.ousage org/freesmartphone/Usage org.freedesktop.DBus.Introspectable.Introspect" only returns Usage's methods. and replacing /org/freesmartphone/Usage with /org/freesmartphone/Resource returns nada Jul 06 21:33:28 PaulFertser: didn't help Jul 06 21:33:32 damn Jul 06 21:44:07 hmm what should i make of this then. nobody evert starts frameworkd with only some of the subsystem or never with ousaged anyway :P Jul 06 21:48:00 but this is interesting. if i don't specify the subsystem to be started frameword fails outputting this error: http://www.pastebin.ca/1486188 Jul 06 21:48:37 heh Jul 06 21:48:39 hello Jul 06 21:50:39 don't you find the mobile day of the desktop summit a little light ? Jul 06 21:51:06 does anyone here goes there ? Jul 06 21:51:30 pfff...it was the same error, stupid,stupid,stupid :P Jul 06 21:51:57 doxy2: the same as the previous? Jul 06 21:52:11 doxy2: so it means frameworkd didn't start for you at all Jul 06 21:53:03 PaulFertser: no, i meant the last error i posted is the same as the previous(when i specified ousaged explicitly) Jul 06 21:53:24 PaulFertser: it starts as long as i don't start ousage or start only ousage Jul 06 21:53:35 doxy2: got it... Jul 06 21:59:57 PaulFertser: btw, i checked a frameworkd.confg from a fso image and didn't have an entry Resource either Jul 06 22:05:19 PaulFertser: do you know if we have an dbus expert here who knows about frameworkd :P ? Jul 06 22:06:05 doxy2: sorry, i don't Jul 06 22:06:41 PaulFertser: np. thanks for the help man Jul 06 22:07:17 doxy2: this time it was more of a mis-help, but i'll try to be more useful next time :) Jul 06 22:08:11 suppose mickey could know. he has his name inserted in way too many frameworkd files :P Jul 06 22:12:07 Is there an elementary widget that can be used to create tabs, or I have to create it myself Jul 07 00:12:52 roh: thanks :D Jul 07 01:19:49 PaulFertser: I wonder if gta01 could use NTCSW output and BATTMP input of PMU to talk to GTA02 bq27k. Might be possible to exploit these I/O by a rather tricky driver **** ENDING LOGGING AT Tue Jul 07 03:00:06 2009