**** BEGIN LOGGING AT Thu Aug 06 02:59:57 2009 Aug 06 07:46:56 good morning openmokers Aug 06 07:57:15 morning mrmoku Aug 06 07:57:30 morning methril|work Aug 06 07:58:09 morning DocScrutinizer Aug 06 07:58:27 what are you working on? Aug 06 07:59:06 working on keeping my mood up Aug 06 07:59:52 DocScrutinizer, moin Aug 06 08:00:10 aah, and my buzzfix fell off a A7, so seems I have to rework my daily-use device Aug 06 08:00:40 s/my/DKAY's/ Aug 06 08:00:40 DocScrutinizer meant: aah, and DKAY's buzzfix fell off a A7, so seems I have to rework my daily-use device Aug 06 08:19:02 mrmoku: any good news on fsxkd-up build-process and fixing of that bugger Aug 06 08:19:03 ? Aug 06 08:28:11 mrmoku: I heard you managed to build a complete (=working?) image, after compiling abyss manually. So what's status of this enterprise now? Aug 06 08:37:24 btw.. wh's tweaing that pile of shit called opkg? TAsn? I just wondered whether it's possible (or opkg could learn) to make 'opkg install foo' learn how to keep the downloaded file in some sort of cache, so if opkg fails on first attempt - e.g. with segfault due to missing free ram - then it wouldn't start with downloading the file on second invocation, but rather notice it already did that dl a few minutes before and happily use that Aug 06 08:37:25 file from cache Aug 06 08:41:27 and yes, I'm aware I could do this manually by first downloading the files, then tell opkg to install from file, then on opkg success delete the file eventually. Alas that's not what you usually are inclined to type in on the go while using terminal-kbd and GPRS for download Aug 06 08:42:16 hehe Aug 06 08:42:24 DocScrutinizer: opkg is maintained? Aug 06 08:43:15 I didn't assume it's maintained ;-P I asked "who's tweaking it" as my conception was TAsn did some mods to opkg recently Aug 06 08:43:52 alphaone: moinmoin btw Aug 06 08:44:14 opkg would need some sort of "download" action - or has it already? Aug 06 08:44:23 moin Aug 06 08:44:33 because opkg also tries to install dependencies Aug 06 08:44:34 I seem to rememberit actually has? Aug 06 08:44:54 lemme check... Aug 06 08:44:57 but... does that one also download not-yet-installed deps? Aug 06 08:45:45 opkg install --download-only && opkg install && opkg clean Aug 06 08:45:57 that might be a way to go Aug 06 08:46:12 download Download to current directory. Aug 06 08:46:29 what i thought - only pkg and not it's dependencies Aug 06 08:46:32 under ""Informational Commands:"" ;-P Aug 06 08:46:40 O.o Aug 06 08:46:55 maybe add a --with-deps to download too# Aug 06 08:47:37 Wonka: opkg install --use-cache /tmp/opkgcache foo Aug 06 08:47:48 mh. someone needs to show me how to build .ipks from source on x86/amd64 some day... Aug 06 08:48:30 DocScrutinizer: does that exist? if yes, would be nice if one could put use-cache into the opkg config... Aug 06 08:48:38 Re flashed last night and still having issues :( running ophonekitd does not give anything obvious and it registers with the network fine etc. I then go into SHR-Settings and into phone. GSM Antenna says it is on and it shows the Operators button and Modem information button (neither seem to work) then in the Call settings part is says "Turn GSM on" :( any ideas? Aug 06 08:48:42 alas nonexistent Aug 06 08:48:51 seems Aug 06 08:52:39 duh!! "" --cache Use a package cache "" Aug 06 08:54:21 WTF? Aug 06 08:54:30 -f Use as the opkg configuration file Aug 06 08:54:41 -conf Default configuration file location Aug 06 08:54:43 is /etc/opkg/opkg.conf Aug 06 08:54:51 errr ???? Aug 06 08:56:22 nota bene, there's the ""-cache"" line between first and second option quoted above. So AIUI that's NOT like "-conf " is the long form of "-f" Aug 06 08:58:35 tough that was the only way to read this so it ctually makes sense, if only this was "--conf" instead of "-conf", and you generously assume some weirdo inserted the "--cache" at this location later on by accident Aug 06 09:00:03 well, here: Aug 06 09:00:06 -o Use as the root directory for Aug 06 09:00:08 -offline offline installation of packages. Aug 06 09:00:24 the long form also starts with single minus Aug 06 09:00:30 :-( Aug 06 09:01:16 as well here: Aug 06 09:01:20 -p Path to utilities for runing postinst Aug 06 09:01:21 -offline-path script in offline mode. Aug 06 09:03:21 so seems only way to get a clue if 2 options are meant to be synonym or are distinct from each other is to look for a "." at end of line. :-((( Aug 06 09:04:05 ~lart opkg-coders Aug 06 09:04:05 * bzzbot acting on orders from an unspecified client drags opkg-coders into court suing for $200 million Aug 06 09:07:59 tig|: version? Aug 06 09:08:11 SHR-U as of last night Aug 06 09:08:15 reflashed Aug 06 09:08:18 lol Aug 06 09:08:33 ah I take it that it is borked then Aug 06 09:09:26 [2009-08-05 20:25:55] wjbaird: I'd recommend 090703 ;ß) Aug 06 09:09:28 [2009-08-05 20:27:07] wjbaird: to me it seems it's been just regressions since then, others here might disagree Aug 06 09:10:03 hi is there a program for doing convenient openstreetmap mapping on the openmoko...specially for noting road types? Aug 06 09:10:05 tig|, where did you download that from? Aug 06 09:10:14 I bet there is but I don't remember the name Aug 06 09:10:51 shr.bearstech Aug 06 09:11:11 tig|, which one exactly Aug 06 09:11:22 (me did not sync any image yesterday...) Aug 06 09:11:38 ah now that is tricky as the actual file is on my home machine and I can't get at it from here Aug 06 09:11:47 I am better off reflashing again ? Aug 06 09:12:54 tig|, did you download from the tests/mrmoku directory or the official one? Aug 06 09:13:00 tig|: you bet, as you're using a nonexistent image it seems :-P Aug 06 09:13:02 and is frameworkd running? Aug 06 09:13:06 mrmoku: the official one Aug 06 09:13:46 at least I think it was last nights image, I was the newest one in there... Aug 06 09:13:47 * DocScrutinizer checks download-site Aug 06 09:14:04 tig|, the latest official image is from 30/07 Aug 06 09:14:43 ? what the hell did I download then :/ Aug 06 09:14:48 * tig| goes to have a look Aug 06 09:14:51 uImage-2.6.29-oe10+gitr119838+2d158aae9d8d36f575504f59884ed8e80802efe2-r3.5-om-gta02.bin 02-Aug-2009 12:27 1.7M Aug 06 09:14:53 [ ] uImage-GTA02.bin.bin 02-Aug-2009 12:27 1.7M Aug 06 09:15:03 no rootfs Aug 06 09:15:19 on http://shr.bearstech.com/shr-unstable/images/om-gta02/?C=M;O=A Aug 06 09:15:54 mrmoku: my bad then must have been an older one Aug 06 09:17:45 mrmoku: isn't it deprecated if uImage-latest doesn't match the modules in shr-image-latest? Aug 06 09:17:58 so is the consensus that go back to the 3rd July image and reflash? Aug 06 09:18:15 I'd recommend that, yes Aug 06 09:18:33 thanks will do Aug 06 09:18:57 IF you're in for a working image rather than testing cutting edge nasty Aug 06 09:19:25 MHO Aug 06 09:19:50 DocScrutinizer, I don't mind occasional breakage and like playing around with new stuff but I really need to access my SMS :) Aug 06 09:22:00 mrmoku: your notion on that? Aug 06 09:25:02 mrmoku: btw: Aug 06 09:25:07 shr-shr-lite-image-glibc-ipk--20090730-om-gta02-testlab/ 30-Jul-2009 17:03 - Aug 06 09:25:28 is the only one I find that's date of 07/30 Aug 06 09:25:42 anyone that is present do gps mapping with the openmoko? Aug 06 09:26:34 Gnutoo: afaik onen and alphaone are savvy Aug 06 09:26:46 ok thanks a lot Aug 06 09:26:57 did one of them write some programs to do some mapping? Aug 06 09:27:09 like something that note the road types Aug 06 09:27:30 because tangogps do the tracking but not more Aug 06 09:29:15 DocScrutinizer, yeah, dirty - came from removing a bad image after syncing Aug 06 09:39:33 Ping Aug 06 09:42:36 Ainulindale, ping whom? Aug 06 09:44:04 Noone :-) Aug 06 09:45:44 Pong :D Aug 06 09:50:30 dos1|away: q-master asks: "frameworkd.subsystem ERROR could not import pimd_contavts.py cannot import name BackendManager" etc. Aug 06 09:50:59 PaulFertser: o/ Aug 06 09:51:52 DocScrutinizer: morning Aug 06 09:55:55 mrmoku: it's hard for uninvolved users to figure out which SHR image to test ATM :-/ Aug 06 10:00:07 * DocScrutinizer wonders if it is a good idea maybe to move some "mature" weeks-old unstable image from U to T Aug 06 10:01:46 * mrmoku does not think so Aug 06 10:03:50 mrmoku: did you manage to understand the passwd issue? Aug 06 10:04:18 mrmoku: so let's have a recommendation at least at some prominent place. maybe edit/create new symlinks in download site of U? Aug 06 10:05:10 uImage_recommended -> uImage_20090703 or whatever Aug 06 10:05:36 s/u/_u/ Aug 06 10:05:36 DocScrutinizer meant: _uImage_recommended -> uImage_20090703 or whatever Aug 06 10:07:13 Is there a way to talk to the SIM and have a look at what SMS are on there? The info on the OM wiki seems to be out of date Aug 06 10:07:34 tig|: with FSO? Aug 06 10:07:42 Well SHR Aug 06 10:07:51 tig|: it's the same Aug 06 10:08:02 doing that with the UI is the same as doing it manually Aug 06 10:08:09 you could use mdbus though and call ogsmd directly Aug 06 10:08:20 Ainulindale: but the UI is crashing and I want to see if there is something odd in there Aug 06 10:08:40 Oh, ok :-) Aug 06 10:09:26 ah think I have found a similar bug with debug instructions Aug 06 10:10:43 ooo this is fun I think it is blowing up mickeyterm Aug 06 10:10:46 dos1: moin! ;-D Aug 06 10:10:51 unless I am not using it right Aug 06 10:10:58 tig|: use mdbus it'll be easier you'll see exactly what the problem is Aug 06 10:11:01 And if you want to be sure Aug 06 10:11:09 put fso in debug in frameworkd.conf Aug 06 10:11:47 PaulFertser: that's ok, it shouldn't make any problems. i alredy fixed that my experimental branch Aug 06 10:12:34 dos1: ok, thanks Aug 06 10:13:28 Ainulindale: is there an mdbus for utter idiots anywhere Aug 06 10:13:43 Heinervdm: morning ;) Aug 06 10:14:02 dos1: morning, reday to push git? Aug 06 10:14:11 ;) Aug 06 10:14:36 * DocScrutinizer pokes git Aug 06 10:15:14 Heinervdm: oh, i didn't finished yet :( Aug 06 10:15:40 ok :( Aug 06 10:16:36 * DocScrutinizer pushes git over the cliff Aug 06 10:19:51 Ainulindale, (passwd) well it is because OE does not run postinst for all packages Aug 06 10:20:10 if you run opkg-cl configure after first boot and reboot then everything is fine Aug 06 10:21:07 and as already said... I remember there was a initscript doing that on firstboot Aug 06 10:21:17 will have too look into what happened to it Aug 06 10:40:52 Hmm now getting : /org/freesmartphone/GSM/Device: RetrieveMessagebook failed: org.freesmartphone.Resource.NotEnabled Aug 06 10:41:10 but nothing obvious in the framework debug log Aug 06 10:45:59 tig|: nobody has GSM resource requested Aug 06 10:46:04 tig|: start any phone app. Aug 06 10:46:12 tig|: or change resource policy to enabled. Aug 06 10:46:31 PaulFertser: aha, thanks :) Aug 06 10:47:54 tig|: you have read FSO_Resources page, haven't you? Aug 06 10:49:19 PaulFertser: tried opening dialer and contacts and still get the same Aug 06 10:49:49 tig|: can you actually dial anything? Aug 06 10:49:55 PaulFertser: I am afraid not, I am still trying to get an important sms off my phone so have not done all the background reading Aug 06 10:50:01 no, dialling is dead Aug 06 10:50:44 tig|: restart ophonekitd then Aug 06 10:51:10 tig|: it's ophonekitd that requests the resource, not the gui part you started. Aug 06 10:52:04 PaulFertser: getting DEBUG: GSM not available, try again in 5s Aug 06 10:52:12 in the ophonekitd output Aug 06 10:54:16 reboot the whole thing, frameworkd is now in debug mode Aug 06 10:54:23 ? Aug 06 10:54:34 tig|: kill frameworkd Aug 06 10:54:38 tig|: kill fso-abyss Aug 06 10:54:43 tig|: kill ophonekitd Aug 06 10:54:47 tig|: start frameworkd Aug 06 10:54:50 tig|: start ophonekitd Aug 06 10:56:01 fsck, where's the alias / script for that ;-P Aug 06 10:56:05 tig|, smash phone Aug 06 10:56:11 tig|, buy new one Aug 06 10:57:49 PaulFertser: right, it has just found the GSM resource Aug 06 10:58:00 F4t: oh I have been so verrrrry close :) Aug 06 10:58:22 ךםך Aug 06 10:58:24 lol :D Aug 06 10:58:47 odd it does not seem to be registering with the network Aug 06 10:59:08 have you waited 5 mins yet? Aug 06 10:59:27 tig|: same ol'story all the time Aug 06 10:59:49 PaulFertser: this image is from july should I have killed /usr/sbin/gsm0710muxd too ? Aug 06 10:59:53 tig|: check sim with another phone! Aug 06 11:00:09 tig|: opkg install -force-reinstall fsousaged libgee0 Aug 06 11:00:33 dos1: huh?? Aug 06 11:00:52 recommended??? Aug 06 11:01:07 DocScrutinizer: it's possible he upgraded with old opkg, and that buggy opkg liked to break libgee0 package Aug 06 11:01:18 DocScrutinizer: libgee0 is necessary to run fsousaged Aug 06 11:01:24 DocScrutinizer : not got another one on that network here handy Aug 06 11:01:28 tig|: if it's there, then sure1 Aug 06 11:01:31 aaah, so it's not a mandatory bugfix Aug 06 11:01:59 dos1 : this is a reflash from the 3rd July image, both rootfs and kernel not upgraded anything yet should I still do it? Aug 06 11:02:15 hmm Aug 06 11:02:21 tig|: is fsousaged running? Aug 06 11:02:34 yes Aug 06 11:02:36 tig|: if you had gsm0710muxd running, you should have killed it too. Aug 06 11:02:38 tig|: I'd be *very* reluctant with upgrade Aug 06 11:02:55 right I will go on a killing and restarting spree then :) Aug 06 11:04:59 will I need to start /usr/sbin/gsm0710muxd manually ? Aug 06 11:05:09 and if so I assume that is before ophonekitd? Aug 06 11:07:03 hmm gsm0710 seems to be complaining : http://shr.pastebin.com/mba172d9 Aug 06 11:10:53 tig|: it will be started automatically Aug 06 11:11:13 ah that could be why it is complaining Aug 06 11:11:33 ah yes it is running Aug 06 11:11:33 ouch, my opkg-u script fails on editing "s/gsm0710muxd/fso_abyss" :-/ -- someone fixed the bug in this config file ;-P Aug 06 11:12:32 tig|: anyway I'd strongly recommend to install and use abyss rather than gsm0710muxd Aug 06 11:13:43 Right re-running mdbus Aug 06 11:18:39 Right : /org/freesmartphone/GSM/Device: RetrieveMessagebook failed: org.freedesktop.DBus.Error.NoReply Aug 06 11:18:53 * tig| bangs head on table Aug 06 11:19:19 restart dbus? Aug 06 11:19:46 if so do I need to restart frameworkd and ophonekitd Aug 06 11:19:47 ? Aug 06 11:19:58 tig|: reboot the phone :D Aug 06 11:20:17 PaulFertser: yeah good idea as it is probably in a right state Aug 06 11:20:19 I know I am Aug 06 11:25:35 i saw a comic strip once, with a guy kicking his computer and it says above "boot", and the next frame same image saying "reboot" Aug 06 11:25:57 cant find it now :( Aug 06 11:37:42 This must be a problem with the SIM here either that or my phone has a hw fault Aug 06 11:58:18 PaulFertser: no notification yet :-( o.O. But those are nice 488bytes to start with. Ask mickey about how to get PID frm dbus-name (U know dbus-noob here) Aug 06 11:58:37 oops 410bytes even Aug 06 11:58:59 DocScrutinizer: i already know that Aug 06 11:59:11 PaulFertser: I sippose fsousaged should know how to do that Aug 06 11:59:59 [2009-08-06 13:42:41] DocScrutinizer: i've no idea how to find PID by "dbus unique name". Aug 06 12:00:07 so that's obsolete :-D ? Aug 06 12:03:28 DocScrutinizer: ok, the script improved, now shows the pid and the name Aug 06 12:03:58 heh :-) Aug 06 12:04:14 My first python program ever Aug 06 12:04:59 hehe Aug 06 12:05:41 I wrote my last python proggi some 8 years ago Aug 06 12:06:16 iirc Aug 06 12:06:49 using capisuite Aug 06 12:08:35 DocScrutinizer: http://paste.debian.net/43474/ Aug 06 12:09:15 wooohooOOOOoo :-))) Aug 06 12:09:39 DocScrutinizer: not quite shell-parseable, i guess for that two separate utility should be implemented: one showing policy and state for every resource and the other akin to lsof. Aug 06 12:10:17 nah, make that options! Aug 06 12:10:49 I'm afraid that's not quite the unix-way, or do you have a well-established example? Aug 06 12:10:53 -o --open : show allocated resources only Aug 06 12:11:04 PaulFertser: how long does it take to query this over dbus? Aug 06 12:11:10 Anyway, for most purposes even the script as it is is ok. Aug 06 12:11:39 lindi-: real 0m2.680s user 0m0.995s sys 0m0.685s Aug 06 12:12:06 -r -resource : show state of named resource only Aug 06 12:12:40 -k --kill : kill all processes that allocated the resource listed Aug 06 12:14:23 -c --check : return 0 if named resource is allocated, error otherwise Aug 06 12:14:48 DocScrutinizer: -q : omit table header Aug 06 12:14:54 (implies -q --quiet) Aug 06 12:15:13 hehe Aug 06 12:15:34 DocScrutinizer: i think if you add this to the ticket probably someone someday will implement it ;) Aug 06 12:15:44 PaulFertser: I don't think you need any lessons from me ;-P Aug 06 12:17:23 DocScrutinizer: tbh i was quite puzzled by the bq27k parsing script. I learned bash by the manpage, missed quite some important features obviously. Aug 06 12:18:00 So i do Aug 06 12:18:07 (need some lessons) Aug 06 12:19:13 hte real hacker knows how to write FORTRAN no matter which language he is urged to use ;-) Aug 06 12:19:18 TAsn: hmm Aug 06 12:19:42 TAsn: i searched though synaptic to install full kde4.3 instalation Aug 06 12:19:51 TAsn: and i saw libphone-utils0 ;) Aug 06 12:20:38 * DocScrutinizer is now known as DocScrutinizer-breakfast Aug 06 13:05:24 PaulFertser: ouch, " using **pending_return in dbus... strongly discouraged" is messing up whole lsfsor output here :-((( Aug 06 13:06:18 DocScrutinizer: hehe, probably you can do it like ./lsfsor.py 2>/dev/null Aug 06 13:06:24 k, can fix it with 2>dev/null, but anyway Aug 06 13:06:52 hehe Aug 06 13:06:52 DocScrutinizer: anyway that's SHR bindings that should be fixed. Aug 06 13:07:11 20090703 here Aug 06 13:07:15 dunno Aug 06 13:07:52 very nice anyway Aug 06 13:08:14 could we do the 2> *inside* script? Aug 06 13:09:11 Could we fix the bindings ;) Aug 06 13:15:02 PaulFertser: bindings? WHich bindings? Aug 06 13:15:10 That's a dbus problem not a SHR one Aug 06 13:15:21 (IIRC) Aug 06 13:16:34 Ainulindale: to me it looks like SHR is using some deprecated dbus stuff. Aug 06 13:16:48 Ainulindale: probably outdated python dbus bindings. Aug 06 13:17:10 IIRC we had that with python/Fso as well Aug 06 13:17:28 Ainulindale: sure, that's why i suspect python bindings. Aug 06 13:17:59 Not SHR's fault then, but FSO's :-p Aug 06 13:18:01 as ever =) Aug 06 13:18:33 i have written a recipe for the latest fbreader (the one in epenembedded is more than two years old) - what is the best way to publish it? Aug 06 13:19:24 all dbus bindings... not just python Aug 06 13:19:33 (as it hits connman too) Aug 06 13:20:12 mrmoku: so whose fault is that? Not FSO as they don't do distro. Aug 06 13:20:30 OE's fault! :-p Aug 06 13:20:37 Ainulindale: ah, right. Aug 06 13:20:44 PaulFertser, well... distributed fault Aug 06 13:20:44 But anyways Aug 06 13:20:54 FSO was doing a distro until now :-) Aug 06 13:21:19 our fault because we include that verbose patch that fixes the weird dbus error that has been observed around FSOSHRUDCON Aug 06 13:21:39 dbus's fault because it does not include that patch (maybe someone should check the state on that) Aug 06 13:21:50 and the bindings fault as they don't do it right Aug 06 13:22:23 PaulFertser, we could just drop the verbosity of that patch... but then it will never get a proper fix :P Aug 06 13:22:54 mrmoku: thanks for explanations :) Aug 06 13:23:26 PaulFertser, check https://bugs.freedesktop.org/show_bug.cgi?id=19796 Aug 06 13:23:45 mickey|zzZZzz, reopened the discussions on that topic :) Aug 06 13:30:02 mickeyl, time for a small question? Aug 06 13:30:50 mrmoku, Ainulindale: Problems if I commit that one into shr/import? It disables the printk's during for suspend resume. Saves you 3 seconds of output. :) Aug 06 13:31:21 stefan_schmidt, no problem for me Aug 06 13:31:34 (me likes saving time :) Aug 06 13:31:37 mrmoku: ok Aug 06 13:31:40 * stefan_schmidt also Aug 06 13:31:45 will push it now Aug 06 13:32:29 stefan_schmidt: printks? Aug 06 13:32:36 stefan_schmidt: kernel messages? Aug 06 13:32:47 mrmoku: sure Aug 06 13:32:55 PaulFertser: yes Aug 06 13:32:58 cd ???? && wget http://trac.freesmartphone.org/raw-attachment/ticket/461/lsfsor.py && mv lsfsor.py lsfsor && chmod +x lsfsor Aug 06 13:33:05 mickeyl, we get lots of unnecessary -dev packages pulled into the shr image Aug 06 13:33:06 stefan_schmidt: what messages are you talking about? Aug 06 13:33:13 mickeyl, and I don't understand why Aug 06 13:33:18 stefan_schmidt: those going to the klogd or those going to console? Aug 06 13:33:24 PaulFertser: yes Aug 06 13:33:27 mickeyl, somehow I suspect vala stuff for pulling it into Aug 06 13:33:30 mrmoku: often this comes because of a wrong DEPENDS or RDEPENDS Aug 06 13:33:32 stefan_schmidt: i've nothing on console here. Aug 06 13:33:41 you have to take care not to use _${PN} Aug 06 13:33:52 err Aug 06 13:33:53 wrong Aug 06 13:33:58 to _use_ ${PN} Aug 06 13:34:00 PaulFertser: beacuse you disbaled klogd yourself? Aug 06 13:34:17 mickeyl, any hint on how to find out the culprit package? Aug 06 13:34:18 stefan_schmidt: so you mean the slowness comes from logging kernel messages to the log or what? Aug 06 13:34:50 PaulFertser: Yes, they do. We did that for fso before Aug 06 13:35:04 mrmoku: PaulFertser told there's a dependencies logging feature in bb Aug 06 13:35:07 PaulFertser: The major time for suspend resume is indeed printing the messages Aug 06 13:35:17 stefan_schmidt: printing the messages to the file? Aug 06 13:35:30 DocScrutinizer, yes, but it does not show me the -dev packages Aug 06 13:35:39 :-/// Aug 06 13:35:49 PaulFertser: on the fb tty Aug 06 13:36:09 stefan_schmidt: what kernel loglevel do you use? Aug 06 13:36:27 mrmoku: mv *dev *dev-away; see what complains? Aug 06 13:36:30 stefan_schmidt: Qi has loglevel=4 Aug 06 13:36:41 PaulFertser: The image default. That's the whole point for cho "echo 0 0 0 0 >/proc/sys/kernel/printk" Aug 06 13:36:55 PaulFertser: Not everybody uses qi.... Aug 06 13:36:59 stefan_schmidt: and i thought i've already moved all the messages to the appropriate loglevel so you shouldn't see anything on FB by default. Aug 06 13:37:12 stefan_schmidt: those using u-boot can tweak loglevel= parameter to their liking. Aug 06 13:37:18 PaulFertser: On the default shr image with u-boot I see them all the time Aug 06 13:37:29 stefan_schmidt: that means you want to see them. Aug 06 13:37:40 stefan_schmidt: because if you didn't you'd tweaked loglevel= kernel parameter. Aug 06 13:37:52 PaulFertser: Well I don't expect users to tweak the u-boot env Aug 06 13:38:06 PaulFertser: and then there's still 'klogd -c 1' or whatever level you like Aug 06 13:38:20 stefan_schmidt: if some messages are shown in an appropriate loglevel that should be noted and fixed, not hidden! Aug 06 13:38:36 s/appropriate/inappropriate/ Aug 06 13:38:37 PaulFertser meant: stefan_schmidt: if some messages are shown in an inappropriate loglevel that should be noted and fixed, not hidden! Aug 06 13:38:43 ack Aug 06 13:38:47 :) Aug 06 13:39:26 stefan_schmidt: if somebody is using u-boot it means by definition he wants flexibility that Qi can't offer. So let him finetune loglevel and whatever he wants himself. Aug 06 13:39:28 Then have fun with it guys. Just revert my commit and work it out... Aug 06 13:39:56 PaulFertser: It does not by definition. People using u-boot because a lot devices were shipped with it. Aug 06 13:40:08 Not everyone likes to flash his bootloader. Aug 06 13:40:14 mrmoku: i blame you for the wrong commit hiding what should be exposed Aug 06 13:40:24 PaulFertser, huh? Aug 06 13:40:25 stefan_schmidt: i can understand concerns of gta01 users. Aug 06 13:40:30 first of all scrolling seems fubar to me. It does incredibly nasty stupid things obviously, like re-renderinf the *whole* screen Aug 06 13:40:45 stefan_schmidt: can't understand why gta02 users can have any problem flashing the bootloader. Aug 06 13:40:53 PaulFertser, git blame will tell otherwise ;) Aug 06 13:41:20 mrmoku: i know the truth and i have irc logs Aug 06 13:41:44 PaulFertser: I'm not on the party for forcing users to mess with u-boot just because it is possible Aug 06 13:42:14 s/whole/whole non-blank/ Aug 06 13:42:14 DocScrutinizer meant: first of all scrolling seems fubar to me. It does incredibly nasty stupid things obviously, like re-renderinf the *whole non-blank* screen Aug 06 13:42:21 stefan_schmidt: what if your users don't use klogd? Will your dirty hack work then? Aug 06 13:42:49 mickeyl, _${PN} for PROVIDES and RPROVIDES too? Aug 06 13:42:51 echo 0 0 0 0 >/proc/sys/kernel/printk Aug 06 13:43:03 stefan_schmidt: i see, this will work regardless. Aug 06 13:43:04 That tells the _kernel_ to stop printing Aug 06 13:43:07 stefan_schmidt: sure Aug 06 13:43:15 I have not pushe dit yet Aug 06 13:43:15 Still i think that's fucking wrong. Aug 06 13:43:36 stefan_schmidt: i also very much dislike the way FSO treats charger/battery status. Aug 06 13:43:39 And I don't waste time on discussing this to death. If nobody wants it, I keep it for myself. Aug 06 13:43:57 PaulFertser: what has battery todo with this? Aug 06 13:44:00 I got this error message while trying to send my PIN with libfso-glib's free_smartphone_gsm_sim_send_auth_code: Aug 06 13:44:02 You are jumping... Aug 06 13:44:04 Error 19 while authenticating: Method "SendAuthCode" with signature "s" on interface "org.freesmartphone.GSM.SIM" doesn't exist Aug 06 13:44:14 stefan_schmidt: is it the new FSO trend to hide things from the users? Aug 06 13:44:14 this is weird, because that's exactly what is in the specs Aug 06 13:44:20 http://git.freesmartphone.org/?p=specs.git;a=blob_plain;f=html/org.freesmartphone.GSM.SIM.html;hb=HEAD#SendAuthCode Aug 06 13:44:21 stefan_schmidt, it will stop kprinting to log too? Aug 06 13:44:23 Well, forget it. I'll not push it. Work it out in a sane way if you prefer. Aug 06 13:44:38 stefan_schmidt: i think i already did. Aug 06 13:44:40 we need a decent set of utilities to edit bootloader envs from userspace! Aug 06 13:45:06 stefan_schmidt: specifically tested numerous times to find all debugging messages with inappropriate levels and altered them accordingly. Several months ago. Aug 06 13:45:12 PaulFertser: it's only that just a few users can use this. Anyway, not my problem Aug 06 13:45:33 mrmoku: RPROVIDES, yeah Aug 06 13:45:44 mickeyl, ok, found one then :) Aug 06 13:45:45 mrmoku: the main problem is the RDEPENDS though Aug 06 13:45:47 stefan_schmidt: i'm sorry to sound harsh/rude/whatever Aug 06 13:46:12 stefan_schmidt: i just don't know how to properly argue probably. Aug 06 13:46:16 dos1, in ubuntu?! Aug 06 13:46:16 :) Aug 06 13:46:17 mrmoku: you can always grep through the Packages file to see where a non-dev package depends on a -dev package Aug 06 13:46:31 mickeyl, so RDEPENDS should always be RDEPENDS_${PN} ? Aug 06 13:46:44 ok Aug 06 13:47:00 that's annoying Aug 06 13:47:15 well Aug 06 13:47:16 not always Aug 06 13:47:24 you just have to think about what you want to say Aug 06 13:47:38 But now I understand why we had to set a PREFERRED_PROVIDER for a package to that package Aug 06 13:47:52 That's because we didn't put RPROVIDES_${PN} Aug 06 13:47:53 since without ${PN} an R statement will be used for all subpackages Aug 06 13:47:56 I didn't think about that Aug 06 13:47:58 stefan_schmidt: my opinion wrt u-boot is: if you're a u-boot user, that means you're on your own as using u-boot makes support much harder. Also i don't know any active system-level dev who cares about u-boot. So, as i see it u-boot is for special purposes and advanced users anyway -> no problem to tweak env. Aug 06 13:48:05 and that's almost always not what you want Aug 06 13:48:45 mickeyl: Ainulindale an idea about this error ? Aug 06 13:48:45 Method "SendAuthCode" with signature "s" on interface "org.freesmartphone.GSM.SIM" doesn't exist Aug 06 13:48:46 mickeyl, ahh, very enlightening - thanks :) Aug 06 13:49:12 PaulFertser: I see it completely different. People that are having u-boot have it because they want to avoid flashing a new bootloader. They want to have nothing to do with such things. Aug 06 13:49:16 Deubeuliou: what's telling you that? Aug 06 13:49:38 What will happen is that the majority of user can not profit from your solution Aug 06 13:49:39 stefan_schmidt: agree Aug 06 13:49:42 Deubeuliou: are you sure you're talking to the right object? Aug 06 13:49:54 stefan_schmidt: then they're screwed anyway, using old u-boot is especially deprecated anyway. Aug 06 13:49:54 But another idea is coming into my mind Aug 06 13:50:09 PaulFertser: says who? Aug 06 13:50:31 mickeyl: yes, I can use GetAntennaPower on GSM.Device on the same object. but I'll check for typo errors Aug 06 13:50:35 PaulFertser: that is not the way to handle users Aug 06 13:50:37 DocScrutinizer: there were quite some problems fixed wrt initialisation of pcf50633, some gpios etc. Aug 06 13:50:52 actually the "new" uboot has issues not seen on "old" one Aug 06 13:50:52 stefan_schmidt: you might also claim they shouldn't upgrade GSM firmware? Aug 06 13:51:07 mrmoku: btw., i have started with shr/import -> oe.dev migration Aug 06 13:51:09 Ah, well, we have the developer-only-gta02 again :) Aug 06 13:51:18 mrmoku: some python diffs you have are strange Aug 06 13:51:24 - SECTION = "devel/python" Aug 06 13:51:27 + SECTION = "python-devel" Aug 06 13:51:28 ?? Aug 06 13:51:35 i ignored all these Aug 06 13:51:36 mickeyl, where? Aug 06 13:51:39 PaulFertser: It's just about making some stuff easier to use Aug 06 13:51:45 mrmoku: dunno, some python-foo.bb Aug 06 13:51:55 mickeyl, heh, ok Aug 06 13:52:01 alphaone: Qi is about making stuff easier to use. U-boot is about sooner or later messing one's environment. Aug 06 13:52:01 mrmoku: what about the patch to python-ecore that adds some virtual keyboard thingy? should we propose that upstream rather? Aug 06 13:52:32 I'm still using qi Aug 06 13:52:33 i have upstream access, but i'm not sure whether this function belongs in there Aug 06 13:52:38 s/qi/u-boot/ Aug 06 13:52:38 alphaone meant: I'm still using u-boot Aug 06 13:52:42 mickeyl: solved orally Aug 06 13:52:48 Ainulindale: oki Aug 06 13:52:50 * stefan_schmidt lost interest. bye Aug 06 13:52:54 Ainulindale: sicko :-) Aug 06 13:53:03 mickeyl, the more upstream the better... on the other hand that means we have to bump EFL... which is always a pain Aug 06 13:53:12 and by the way mickeyl Aug 06 13:53:15 He's using your article :-) Aug 06 13:53:22 Ainulindale: oops ;) Aug 06 13:53:32 DocScrutinizer: http://git.openmoko.org/?p=u-boot.git;a=shortlog;h=refs/heads/stable Aug 06 13:53:39 mrmoku: true enough, i have some motivation to patch out the SOnames anyways *cough* Aug 06 13:53:58 DocScrutinizer: and i remember for sure how i read the ML during 2008 and the first advice was to update u-boot. Aug 06 13:54:00 alphaone: when will you fix #1024 on my phones? :-p Aug 06 13:54:04 mickeyl, don't tell raster ;) Aug 06 13:54:19 Ainulindale: As soon as you invite me to paris ;-.) Aug 06 13:54:21 mrmoku: i won't :) Aug 06 13:54:28 mickeyl, with current EFL libeflvala will build again? Aug 06 13:54:31 he'll see it sooner or later Aug 06 13:54:43 alphaone: well when you want Aug 06 13:54:50 mrmoku: yes, I added the necessary stuff upstream Aug 06 13:54:54 I'm fully available around the 19th of august Aug 06 13:54:55 some weeks ago Aug 06 13:55:00 My girlfriend is going back to Russia for a month Aug 06 13:55:04 So I'll have extra room Aug 06 13:55:35 WTF am i wrong and causing conflicts again when in fact i'm after doing the right thing? Shit. Aug 06 13:55:35 mickeyl, good :) Aug 06 13:56:45 we need a decent set of utilities to edit bootloader envs from userspace! Aug 06 13:56:57 PaulFertser: Sometimes there's more than one right thing... Aug 06 13:57:02 Or different perspectives Aug 06 13:57:05 DocScrutinizer: ACK Aug 06 13:57:19 DocScrutinizer: That would help as well I guess Aug 06 13:57:35 alphaone: yes but i'm now feeling like i did the wrong thing expressing my opinion. Aug 06 13:57:36 PaulFertser: hi. I still don't understand why you ask to use loglevel (if I read correctly the discussion). Why can't you make sure everyone is using *_debug macros and use dynamic printk() ? Aug 06 13:58:10 rtp: debug messages are appropriately using debug macroses. Info messages are using info loglevel. Aug 06 13:58:26 rtp: the problem is that Glamo FB is so slow that most folks don't want to see even INFO messages on it. Aug 06 13:58:37 rtp: as it considerably slows down suspend and resume. Aug 06 13:59:06 PaulFertser: I think it was a misunderstanding. I think Stefan felt that you were presenting it as the truth. Aug 06 13:59:18 alphaone: i know there's no truth at all. Aug 06 13:59:24 mrmoku: (env edit util) lindi knew about some tools that do exactly this. Alas not available in OM repo. Could you check? Aug 06 13:59:26 alphaone: everything is subjective for me. Aug 06 13:59:31 PaulFertser: There are many truths :-) Aug 06 13:59:44 PaulFertser: so the real problem is rather than glamofb is using info when it should not. no matter glamo or not, fb is slow so printing a lot of things on resume will make it slow. Aug 06 13:59:48 git question, can i diff a file living in two arbitrary branches or do i need to checkout one or the other? Aug 06 14:00:00 alphaone: i feel like i forced to not apply the patch by some inappropriate means. Aug 06 14:00:54 PaulFertser: FOSS is always a struggle between opinions. Aug 06 14:01:41 But I agree that the kernel messages on suspend/resume really suck from a user POV. Aug 06 14:01:43 opinions and options Aug 06 14:01:54 rtp: with loglevel=4 supplied by bootloader i still see INFO messages in dmesg but only higher levels on FB. I think that's exactly what should be the default. Aug 06 14:01:56 where can I find all existing object paths ? Aug 06 14:01:56 mickeyl, git diff branch1..branch2 -- file ? Aug 06 14:02:07 It takes >5 seconds for resume Aug 06 14:02:12 PaulFertser: it's like the "*Masking IRQ*" of the pcf50606. I need to check what's used here but it looks wrong to have them in syslog all the time Aug 06 14:02:53 PaulFertser: one should not have to play with loglevel. You're not messing with it on x86, don't you ? Aug 06 14:02:58 rtp: exactly. If the user won't see any messages anywhere we won't find and fix inappropriate loglevels ever. Aug 06 14:03:04 mickey@andromeda:/local/pkg/oe/ms5$ git diff fso/milestone5.5..shr -- recipes/python/python-ecore_svn.bb Aug 06 14:03:04 fatal: bad revision 'fso/milestone5.5..shr' Aug 06 14:03:13 rtp: because my console is fast enough. Aug 06 14:03:19 fwiw, it's wrong to change VT in the first place Aug 06 14:03:23 Nils did patch that out for the 01 Aug 06 14:03:30 but it crashed when he tried to patch it out for the 02 Aug 06 14:03:48 mickeyl, shr? Aug 06 14:03:51 s/crahsed/didn't reinitialize Aug 06 14:04:11 PaulFertser: that or because the right printk levels are using in all your drivers :) Aug 06 14:04:15 mickeyl: we'll soon have proper X driver so it won't be an issue. Aug 06 14:04:26 mrmoku: ah, got it now thanks. git diff origin/fso/milestone5.5..origin/shr/import -- recipes/python/python-ecore_svn.bb Aug 06 14:04:47 PaulFertser: hopefully. changing VTs during suspend/resume is just wrong on mobile consumer devices IMO . Aug 06 14:04:58 rtp: no. I see lots of INFO messages on my console and it doesn't slow down the boot so i'm ok with it. If i see all the same messages on FR fb then yes, it slows down. Aug 06 14:05:15 mickeyl: many laptops require that or else s2ram doesn't work :( Aug 06 14:05:31 Deubeuliou: you can browse through the objects a busname exports with mdbus Aug 06 14:05:31 mickeyl: but yes, it's wrong Aug 06 14:05:42 Deubeuliou: mdbus Aug 06 14:06:24 nah, I like it (switch vt) Aug 06 14:06:37 PaulFertser: changing vt on suspend on x86 is also to allow progress bar during suspend :) Aug 06 14:07:16 people want a seamless user experience Aug 06 14:07:21 rtp: never seen progress bar on suspend, i don't think there's any. IIRC s2ram doesn't have anything like that. Aug 06 14:07:26 feel free to patch it out on your devel images Aug 06 14:07:45 mickeyl: k, :-) Aug 06 14:08:50 mickeyl: i can't see why should you care about "end-users" before the software is actually end-user ready. Do you think i enjoy using my FR as a phone. You bet i don't. Aug 06 14:08:51 anyway, fb scrolling is fubar. I'd even dare to call it buggy Aug 06 14:09:04 PaulFertser: i think we need to start somewhere Aug 06 14:09:07 PaulFertser: s2ram can use splashy or bootsplash (or plymouth I think) if it's there Aug 06 14:09:13 i enjoy a seamless boot process Aug 06 14:09:14 myself Aug 06 14:09:21 so for me it's important Aug 06 14:09:26 same goes for suspend/resume Aug 06 14:10:16 mickeyl: i only see 4 messages on boot (about gether since it's compiled as a module) and no messages on suspend/resume using an appropriate loglevel. And yet i see all proper info messages in dmesg. To me it seems seamless enough. Aug 06 14:10:42 i see a blank screen Aug 06 14:10:47 PaulFertser: btw, did you compare with quiet ? :) Aug 06 14:10:48 then screen goes dark Aug 06 14:10:53 comes to life again on a blank screen Aug 06 14:10:55 then switches back Aug 06 14:10:59 that sucks Aug 06 14:11:06 (again, for me personally) Aug 06 14:11:36 mickeyl: ok, but how is that important? I still have to restart frameworkd from time to time because of some TIMEOUT or other misfortune... Aug 06 14:12:15 it becomes important for me if it annyos me enough Aug 06 14:12:24 mickeyl: and i'm still using Zhone because SHR is using opkg which i don't want to have anything with and no decent telephony apps packaged for Debian yet. Aug 06 14:13:07 there are multiple ways to approach a goal. I think there are ways to do polishing and developing in parallel. Aug 06 14:13:09 mickeyl: and anyway i want incremental search in contacts which shr lacks. Aug 06 14:13:16 etc. Aug 06 14:13:26 I can't see how a user can enjoy this device, honestly. Aug 06 14:13:37 http://free-electrons.com/blog/elc-2009-videos/ Aug 06 14:13:47 I do only because i'm a free software geek/tinkerer. Aug 06 14:13:49 i can enjoy the freedom, but still be annoyed by blank screens Aug 06 14:14:53 I'm annoyed by the end-users who don't even want to read docs. I can't see why anybody should care about them. Aug 06 14:15:34 Too many of those lusers in OM "community" unfortunately. Aug 06 14:16:07 I can understand and sometimes help a novice. I don't want to help those who doesn't want to learn. Aug 06 14:16:11 DocScrutinizer: there are two utilities to modify u-boot env. one is established and one is new Aug 06 14:20:25 PaulFertser: The vast majority of cell phone users never read a manual... Should the free-runner be any different? Aug 06 14:20:50 When some python/whatever program doesn't work for me i first try to understand it learning python/whatever. When my car needs repair i first learn how to do that even if i'm going to have technicians do that for me. Etc. That's the life attitude thing. Either you're consumer or hacker. Aug 06 14:20:55 wjbaird: sure Aug 06 14:21:02 wjbaird: cell phone manuals are awful crap. Aug 06 14:21:08 wjbaird: useless. Aug 06 14:21:28 wjbaird: FR is small portable GSM-equipped GNU/Linux-running platform. Aug 06 14:22:26 PaulFertser: I guess the question is - do you want the FR to be usable by the average cell phone user? Personally, I'd really like it to be... However, if your goal is only appeal to the geeks/hackers like us, then I guess that makes sense... Aug 06 14:25:25 wjbaird: my experience shows that average users can't really enjoy powerful tools. Aug 06 14:25:51 wjbaird: also i'm for "art for the sake of art" attitude wrt developing. Aug 06 14:26:13 wjbaird: if the result happens to be usable by an "average" end-user, good. If not, well not a big deal. Aug 06 14:27:35 See you guys, i hope our talks will be more pleasant next time. Aug 06 14:27:47 PaulFertser: I disagree completely. My wife uses and loves a macbook pro 17" - an amazingly powerful tool with unix under the hood. She's quite non-technical, and has no need to be. However, if I need to get 'under the hood' and edit some config files in vi, I can... I agree that it is far harder to build a powerful tool that *can* be enjoyed by a non-technical user, but in my opinion it's important... Aug 06 14:28:23 wjbaird: all the cool guys i know had to install Gentoo on their macbooks anyway. And gentoo is quite geeky. Aug 06 14:29:13 They were forced to install Gentoo because of Mac OS X limitations, not because they love gentoo too much. Aug 06 14:30:06 PaulFertser: and it's true that my co-husband installed fedora on his macbook - but he's definitely a geek. I use mac os x and don't find it limiting at all - as do the vast majority of mac users... Aug 06 14:30:11 many of them very geeky... Aug 06 14:30:56 wjbaird, there are many flavors of geeks Aug 06 14:31:22 I don't think many tolkien geeks find mac os x very limiting Aug 06 14:32:56 TAsn: true. and I'm not necessarily saying it's an invalid goal to only aim the FR at people who can hack Python... But in my opinion, it's aiming way too low... I personally would love to see a distro on the FR that can be easily used by the average person... Aug 06 14:34:28 and, quite frankly, I'm self-motivated in that... I *can* hack python... but frankly I don't really want to in order to use my phone... Aug 06 14:34:49 rtp, PaulFertser: s2disk has progress bar, i don't think s2ram has it Aug 06 14:34:55 TAsn: I'm a tolkien geek Aug 06 14:35:16 First thing I did with the powerbook I own is to wipe MacOS X from it Aug 06 14:35:29 And install Gentoo Aug 06 14:35:29 But because I like it =) Aug 06 14:36:26 wjbaird, we are going this way Aug 06 14:36:26 don't worry ;) Aug 06 14:36:43 Ainulindale, you are a tolkien geek Aug 06 14:36:46 but not only! Aug 06 14:36:53 Well I'm mainly a geek anyay Aug 06 14:36:54 you are also a computer geek! :) Aug 06 14:36:55 +w Aug 06 14:37:09 As well as a coffee, music, history geek Aug 06 14:37:13 I have too many hats Aug 06 14:37:19 Ainulindale, aye. Aug 06 14:37:23 On the other hand, I specifically bought a macbook pro because I can get Aperture on it - a pro photography tool... I spent some time trying to get f-spot working well enough to use for my photography and decided I'd rather spend the time working on my photography than working on f-spot... Aug 06 14:40:00 mickeyl: blank screen on boot? it's fixed in SHR ;> Aug 06 14:41:38 we only have to add symlink to xsplash-vga.ppm to have one splash from beginning of boot to Illume Aug 06 14:44:58 good for boot, suspend/resume still ugly though Aug 06 14:44:59 perhaps go home and try get that patch submitted Aug 06 14:46:01 * ceda_wrk would prefer to be able to control volume during a phone call than have the suspend/resume be pretty Aug 06 14:46:14 mickeyl, I have another problem/question Aug 06 14:46:20 mickeyl, is opkg-cl configure on firstboot needed? Aug 06 14:46:33 or read concatenated SMS. These are the only two things keeping from using my openmoko instead of my SonyEricsson Aug 06 14:46:40 mrmoku: pretty much, yes Aug 06 14:47:01 mrmoku: just take a look at all the postinsts we can't do at image generation time Aug 06 14:47:26 mickeyl, yep, I'm wondering where that initscript disappeared to Aug 06 14:47:27 of course, we could think about doing more w/ qemu Aug 06 14:48:56 mickeyl: suspend/resume is ~2 seconds here Aug 06 14:49:14 mickeyl: err, resume ~seconds, suspend < 1 second Aug 06 14:49:25 no VT switching? Aug 06 14:49:35 with vt switching Aug 06 14:49:36 yep Aug 06 14:49:41 then it could be ever faster and more beautiful... Aug 06 14:49:45 s/ever/even/ Aug 06 14:49:47 mickeyl meant: then it could be even faster and more beautiful... Aug 06 14:50:17 i think we can disable vt switching now, as we have quite stable suspend Aug 06 14:51:22 if glamo is cooperative... Aug 06 14:51:26 but vt switching on suspend is the last thing which could annoy me :P Aug 06 14:52:09 and for me it's good to see that it suspended instead of just blanking Aug 06 14:52:15 glamo is cooperative if you're using the right xserver Aug 06 14:52:55 uhh Aug 06 14:53:02 when the hell we'll switch to xorg ;/ Aug 06 14:53:42 dos1, Aug 06 14:53:47 where did you find a libphone pack? Aug 06 14:53:57 dos1: (and for me) same here Aug 06 14:54:01 TAsn: in debian ;D Aug 06 14:54:09 sweet :) Aug 06 14:54:16 debian arm? Aug 06 14:54:17 or both? Aug 06 14:54:51 TAsn: no, on FR I have only SHR ATM. it was on PC ;) Aug 06 14:55:07 dos1, are you sure that was libphone and not linphone? Aug 06 14:55:52 TAsn: linphone couldn't have SHR mentioned in description... :P Aug 06 14:56:00 oh ;) Aug 06 14:56:44 dos1, supersweet! :) Aug 06 14:57:17 I think it's time to write documentation Aug 06 14:57:27 libphone-utils0 - library for basic phone operations Aug 06 14:57:37 and maybe add more useful stuff... :) Aug 06 14:58:20 Maintainer: Sebastian Reichel Aug 06 14:58:24 lindi-: (2 tools) could you tell mrmoku which of both is the one we want to get on OM please Aug 06 14:58:26 cool. Aug 06 14:58:29 Version: 0.1+git20090801-1 Aug 06 14:59:54 dos1, it's probably there because it's a libframeworkd-phnogeui dep Aug 06 15:04:54 TAsn: there is no libframeworkd-phonegui0 ;D Aug 06 15:05:04 even sweeter! :) Aug 06 15:05:17 ppl actually liked this software! :) Aug 06 15:10:15 mickeyl: i think i'll need new dependency in opimd in near future Aug 06 15:10:29 on what? Aug 06 15:10:30 libphone-utils0, for normalizing phone numbers Aug 06 15:11:26 do you have a python binding yetß Aug 06 15:11:39 mickeyl: yep, i wrote it some time ago Aug 06 15:11:48 sounds good, go for it Aug 06 15:11:53 it's already commited Aug 06 15:12:19 dos1: where can I review the code? Aug 06 15:12:54 DocScrutinizer: that's TAsn's code ;) Aug 06 15:13:03 mompl Aug 06 15:13:10 yup, I seem to remember Aug 06 15:13:58 DocScrutinizer: http://git.shr-project.org/git/?p=libphone-utils.git;a=summary Aug 06 15:14:08 dos1: thnx :) Aug 06 15:14:24 hmm Aug 06 15:14:51 i'd like to see a formatting function here Aug 06 15:15:19 and something that allows us to set the country prefix on demand, since we get this info from fso Aug 06 15:15:26 TAsn: ^^^ ;) Aug 06 15:15:36 mickeyl: setting country is possible AFAIK Aug 06 15:16:15 DocScrutinizer, the internal names and external documentation still don't really exist, but soon enough, we'll prevail ! :) Aug 06 15:16:15 mickeyl, you can easily do that Aug 06 15:16:15 ya, seen it Aug 06 15:16:20 first of all you have the configuration file Aug 06 15:16:33 but even without it, you can use the API to change and even dump to config file Aug 06 15:16:56 37 static char *trailing_delimiters = "wWpP;,"; Aug 06 15:16:57 38 static char *possible_chars = "0123456789+"; Aug 06 15:16:57 39 static char *filler_chars = " -()"; Aug 06 15:17:06 these will probably reachable from the API soon. Aug 06 15:17:22 though atm, I thought this list will suffice Aug 06 15:17:39 mickeyl: There shall be different CCs, the one you obtain from FSO usually isn't needed as all numbers derived from FSO usually won't need any normalization Aug 06 15:18:09 i like how some other phones format the number during entry, so we should add an area code database to it Aug 06 15:18:11 then there's also a CC that's assumed associated to your addressbook Aug 06 15:18:13 oops Aug 06 15:18:21 this one won't change arbitrarily Aug 06 15:18:22 possible memory leak Aug 06 15:18:24 :( Aug 06 15:18:30 fixing, sec. Aug 06 15:18:40 DocScrutinizer, Aug 06 15:18:44 you can get the CC from sim Aug 06 15:19:11 I think that's what he meant. Aug 06 15:19:11 (I mean: the SIM's CC, not the networks) Aug 06 15:19:41 dhh, for *what*? Obviously for your network, but NOT for your own habits how to deal with numbers in phonebook Aug 06 15:20:08 TAsn: AIUI there's no such thing like a SIM's CC Aug 06 15:20:17 for setting the users country code automatically... Aug 06 15:20:35 (i.e the config setting) Aug 06 15:20:49 how TF should SIM know *MY* CC?? o.O Aug 06 15:21:35 it doesn't, but the issuer number contains a provider field, which you can derive a CC from Aug 06 15:22:01 and what would I be any whiser than? Aug 06 15:22:15 dunno Aug 06 15:22:26 you know where your home is, then :D Aug 06 15:22:28 it's the CC of the provider's home-net at the best Aug 06 15:22:41 not *my* ome Aug 06 15:22:55 DocScrutinizer, maybe I confused country code Aug 06 15:22:57 your home is where your SIM is from Aug 06 15:22:59 :D Aug 06 15:23:07 is israel's CC 972? Aug 06 15:23:08 don't move! Aug 06 15:23:11 if so, it's available from sim Aug 06 15:23:14 I honestly dare to disagree Aug 06 15:23:29 as the providers stamp it in there :) Aug 06 15:24:05 WTF has my (sql-based) phonebook to deal with (TPE prepaid) SIM's CC???? Aug 06 15:25:39 I hope you don't seriously suggest the CC of all my short-form contacts change on every sim swap! :-( Aug 06 15:25:40 DocScrutinizer, it'll just set a default to your CC Aug 06 15:26:00 DocScrutinizer, I'm just saying Aug 06 15:26:05 it's better than guessing 49 Aug 06 15:26:06 :) Aug 06 15:26:08 FSCK defaults. I *hate* that automatism even now Aug 06 15:26:10 DocScrutinizer: uboot-envtools Aug 06 15:26:55 fix first-boot settings wizard to query user about his CC. Most easy and clean thing in the world Aug 06 15:27:10 DocScrutinizer: did you saw it in any other phone? Aug 06 15:27:19 asking for country? Aug 06 15:27:29 dos1, ++ Aug 06 15:27:42 dos1, though I think Aug 06 15:27:48 it's not because of the sim Aug 06 15:27:50 did you ever get a phone that's not been preconfigured for the area you're using it in? Aug 06 15:27:55 but because Aug 06 15:27:57 exact/y Aug 06 15:27:59 exactly Aug 06 15:28:10 DocScrutinizer: yes Aug 06 15:28:15 they have specially crafted firmwares Aug 06 15:28:19 dos1, ++ I agree that for the majority of people, using the CC from the sim will be the right behaviour... as long as it's easy to tweak... Aug 06 15:28:22 DocScrutinizer: my gf has phone from britain Aug 06 15:28:37 though they might just probe for network CC. Aug 06 15:28:40 wjbaird, aye. Aug 06 15:29:09 oops Aug 06 15:29:16 so you suppose we should mimic whatever crap anybody at nokia or whereever is doing wrt to countrycode (which btw is poorly handled on most phones anyway)? Aug 06 15:29:18 a merge will happen in libphone Aug 06 15:29:20 SHR: 03tom 07libphone-utils * rc316cb06a8bb 10/src/phone-utils.c: Fixed a rare-never-would-have-happenned memory leak. fixed a little inconsistency concering the coding style, and removed useless glib casts (no idea how they got there...) Aug 06 15:29:22 nvm ;) Aug 06 15:29:30 SHR: 03tom 07libphone-utils * r4d17cc3721db 10/ (7 files in 5 dirs): Merge branch 'master' of git+ssh://shr.bearstech.com/libphone-utils Aug 06 15:30:13 DocScrutinizer: i can't see crap in that behaviour Aug 06 15:30:33 I'm starting to feel upset, so tme for my breakfast finally. Seems a day with a default mood to argue Aug 06 15:31:10 so I'll follo PaulFertser's example and quit channel for some hours Aug 06 15:31:43 bye Aug 06 15:31:49 DocScrutinizer: probing sim for cc will work in 90% of people, others can very easly tweak it for their needs Aug 06 15:32:24 so build in the means to tweak **FIRST**!!!!!! >:-( Aug 06 15:32:50 DocScrutinizer, I have a suggestion Aug 06 15:32:51 DocScrutinizer: tweaking it is python's two-liner Aug 06 15:32:54 what if we Aug 06 15:32:55 then I give a shit on whatever damn defaults you extract from anywhere Aug 06 15:32:59 first set the sim CC Aug 06 15:33:06 and then let the users change them in shr-settings? Aug 06 15:33:21 and on the first boot we suggest them to check out shr-settings for phone specific stuff? Aug 06 15:33:22 :) Aug 06 15:33:24 yep, SETTINGS is the magic word Aug 06 15:33:31 :) Aug 06 15:34:25 agreed - definitely want it easily tweakable by a user... i.e. in settings, not some config file somewhere... Aug 06 15:34:41 my words Aug 06 15:34:45 bye Aug 06 15:34:48 i could writ shr-settings module even now Aug 06 15:35:11 if only i could not have to hold my eye with one hand Aug 06 15:35:28 so i can write ony with one hand ;P Aug 06 15:36:55 please replace CC etc by any defaults *only* if current setting is NAN (or maybe 0). So once I select my CC, no darn stupid wizard is interfering no way never ever! Aug 06 15:37:43 DocScrutinizer: of course, that's what i had in my mind from begining Aug 06 15:37:56 dos1: good guy Aug 06 15:37:59 :) Aug 06 15:39:56 * DocScrutinizer remembers the insult when swapping Sim with frien's reset the TZ to GMT and no means on earth (except fsck or reflash) was able to get it right again, *sigh* and bye Aug 06 15:43:53 dos1: so if you like to please me, you might include a setting option to change that TZ thing as well (both disable the automatic settings policy option as well as actual TZ setting) Aug 06 15:45:26 DocScrutinizer: dunno if it's reachable by fso dbus interface, i'll check Aug 06 15:45:33 hmm. nice, someone send to me [PATCH] add switch to enable or disable SIM card authentification on startup Aug 06 15:45:38 in shr-settings Aug 06 15:45:40 :) Aug 06 15:48:02 dos1, :) Aug 06 15:48:32 dos1, I'm waiting for someone to send you a patch to make opimd load in parallel to framework Aug 06 15:48:36 and not jam the whole process ;) Aug 06 15:48:39 that's the beauty of open source Aug 06 15:49:21 Ainulindale, waiting for someone to write patches for stuff you are too lazy to do yourself? :) Aug 06 15:51:42 Yep Aug 06 15:51:51 Hey you lazy student! Aug 06 15:51:54 I've got a job Aug 06 15:53:52 I also got a job! Aug 06 15:54:02 dos1: well I guess you'll edit /etc/frameworkd-phonegui.conf directly, no? Aug 06 15:54:16 DocScrutinizer, phone-utils.conf Aug 06 15:54:30 and furthermore, you got API to change it Aug 06 15:54:32 ... Aug 06 15:55:05 it's called save-settings Aug 06 15:55:06 or something :) Aug 06 15:55:15 dos1: if only I could manage to even locate where to disable that annoyance of automatic TZ setting (prolly via otimed) I'd suggest you edit this exactly he same way Aug 06 15:55:49 DocScrutinizer: it's in frameworkd.conf (timesources, or something like that) Aug 06 15:56:25 I grepped for "time" and scrutinized frameworkd.conf via less. Maybe I'm blind :-( Aug 06 15:56:38 DocScrutinizer: and CC i'll do with phone-utils api, as i already checked it to work with python bindings ;) Aug 06 15:56:53 dos1, then do it! :) Aug 06 15:57:01 DocScrutinizer: i think it isn't in SHR config at all, look at example in fso git repo Aug 06 15:57:04 dos1: yes, do it! cool :-) Aug 06 15:57:42 WHERE'S THE MANPAGE??? Aug 06 15:58:44 in former days *nothing* went upsteream without a ancompanying manpage Aug 06 15:59:17 DocScrutinizer, shr-unstable isn't really upstream Aug 06 15:59:23 especially since it's in the shr git Aug 06 15:59:32 (if you were talking about libphone-utils) Aug 06 15:59:33 FSO is though, no? Aug 06 15:59:36 if you were talking abotu fso Aug 06 15:59:45 yeah, no idea where's the man at :) Aug 06 16:00:50 DocScrutinizer: http://git.freesmartphone.org/?p=framework.git;a=blob;f=conf/example/frameworkd.conf;h=b16d26492cee6068a1cfe056a2e419b70241e5bd;hb=HEAD Aug 06 16:01:16 zonesources = NONE in [otimed] Aug 06 16:02:01 dos1: PLEASE pretty please with sugar on top include a way to edit this to settings Aug 06 16:02:32 who is otimed maintainer? Aug 06 16:16:52 * Weiss just made a phone call in his KMS/DRI2 FreeRunner :D Aug 06 16:16:58 on* Aug 06 16:18:25 i'd got used to working without *any* acceleration in my daily use, and it feels SOOOO fast now that EXA works for me again Aug 06 16:42:15 Ainulindale, it is frameworkd-config-shr pulling in all those -dev packages Aug 06 16:42:20 just FYI Aug 06 16:42:43 have to go out with family now... will try to solve that then Aug 06 16:42:58 * mrmoku is optimistic to get some new image until WE :) Aug 06 16:49:39 Weiss: hehe :) how usable is kms/dri2 now? Aug 06 16:50:11 mrmoku|away: how did you check that? Aug 06 16:52:45 dos1: as of half an hour ago, it "works", but the KMS-aware DDX needs a lot of stabilisation (crashes on resume, crashes if TangoGPS is started, doesn't control the LCD properly i think) Aug 06 16:53:04 Weiss: ok Aug 06 16:53:10 Weiss: keep doing great work :) Aug 06 16:53:18 but the KMS/DRI2 utopia is in sight now :) Aug 06 16:53:40 s/LCD/backlight* Aug 06 16:53:58 dos1, I'm so happy I got my phonelog back! :) Aug 06 17:19:47 Weiss: wow nice Aug 06 17:55:37 is anyone interested that enlightenment just SEGV'd when adding bluetooth to the shelf gadgets? Aug 06 17:55:57 shr-u, not updated for a few days Aug 06 17:56:21 boot, bluetooth on, add shelf gadget, crash Aug 06 17:56:27 happened twice in a row Aug 06 17:59:56 and again, wasn't doing anything that time, but had managed to add the shelf gadget Aug 06 18:00:04 something's not right Aug 06 18:09:31 [OT] someone here with inkscape experience? i am trying to get inkscape's perspective effect to work Aug 06 18:49:34 any known problems with Wifi on SHR? I can successfully connect with OM2009, but when I try mokonnect on SHR-U (build from July 3rd) I appear to connect correctly (get an IP, resolve.conf looks alright, etc., route looks ok), but I don't get any packets back when I try to ping the outside world... Aug 06 19:17:01 tim_abell, adding BT widget to the shelf has crashed shr-u/e for at least 7 months Aug 06 19:47:05 hmm submitted the first shr draft Aug 06 19:51:55 i cant get a fix here with shr Aug 06 19:53:55 hello all Aug 06 19:54:35 please help me to make Neo as Bluetooth GPS relay by SHR ? Aug 06 20:01:13 fruct: what exactly do you mean by 'bluetooth gps relay'? Passing gps data to another system via bluetooth? Aug 06 20:05:46 wjbaird: yes! As GlobalSat BT-338 - it possible? Aug 06 20:07:09 I test do as wroted here http://wiki.openmoko.org/wiki/Neo_FreeRunner_GPS#Bluetooth_GPS_relay Aug 06 20:07:31 but SHR used another daemon Aug 06 20:08:03 fruct: afraid I can't help you there... Not sure what the equiv on SHR is... Aug 06 20:08:53 so =( Aug 06 20:31:37 mrmoku|away: apparently the RPROVIDES thing worked out nicely Aug 06 20:33:21 Has anyone looked at getting woosh added to the shr repo? From what I've seen, it's the best browser for the FR out there... not that this says much... ;-/ Aug 06 20:33:44 Well Aug 06 20:33:46 I packaged it Aug 06 20:33:54 A month or so ago Aug 06 20:34:02 IIRC he didn't build properly, maybe I'm mistaken, I should retry Aug 06 20:34:49 it's a lot lighter-weight than midori, and the UI works much better on a small screen... Aug 06 20:35:01 it's based on webkit too IIRC Aug 06 20:35:08 So it shouldn't be that more lightweight Aug 06 20:36:41 Hmm. could be mistaken about the lightweight bit - but the UI is definitely nicer - gives you more real-estate for the web-page... Aug 06 20:36:59 :-) Aug 06 20:37:12 meanwhile I'm rebuilding an image Aug 06 20:37:15 and I'll test it Aug 06 20:37:19 Ok - thanks!!! Aug 06 20:44:09 What about background images / wallpapers? Some of the SHR screenshots I've seen have that - but I haven't figured out how to enable it yet... Aug 06 20:45:56 it's in the menus :-) Aug 06 20:46:53 can you give me a hint? :-) Aug 06 20:48:38 I dug through the settings app and couldn't find anything... am I just blind? Aug 06 20:50:55 it's in illume settings Aug 06 20:50:59 unfold the top menu bar Aug 06 20:51:07 click on the topmost left icon, the wrench Aug 06 20:51:14 then it's hidden somewhere in there Aug 06 20:52:02 Ah! thank you... I forgot Illume had it's own set of settings... Aug 06 20:52:41 sigh. Pity illumes setting panels don't work with the FR's resolution... Aug 06 20:54:20 yep Aug 06 20:55:23 and that enlightenment segfaults when you change your wallpaper... :-( Aug 06 20:58:31 wjbaird: no, it segfaults only when using x11-16 engine Aug 06 20:59:23 dos1: is that on by default? I haven't tweaked my SHR very much yet, and I'm sure I haven't changed that... But it did just segfault when I tried to change the wallpaper... Aug 06 20:59:47 wjbaird: hmm... no, it isn't default Aug 06 20:59:52 wjbaird: and here it works with x11 Aug 06 21:01:01 <[Rui]> any hints on why libelm looks like crap? Aug 06 21:01:12 <[Rui]> not generally.... Aug 06 21:02:00 dos1: well - second time is the charm --- this time I was able to set the wall paper, and it worked fine. *shrug* Aug 06 21:02:41 [Rui]: try to elm_finger_size_set(80) at begining of your app Aug 06 21:03:17 <[Rui]> dos1: but why don't other apps looks so bad? Aug 06 21:03:31 <[Rui]> dos1: I shouldn't need to do that Aug 06 21:03:51 [Rui]: dunno, but shr-settings uses finger_size_set AFAIR Aug 06 21:03:55 <[Rui]> hms... wasn't there some env variables? let's check on the wiki Aug 06 21:04:05 [Rui]: and SHR apps don't need to cause they uses edj layout Aug 06 21:04:06 <[Rui]> dos1: it's supposed to be defined on your environment Aug 06 21:04:23 <[Rui]> dos1: that function exists so you can go away from the "standard" defined by your environment Aug 06 21:04:30 no, it shouldn't Aug 06 21:04:50 env variable should be only when you want to override default ;p Aug 06 21:11:42 any interest in bundling up some wallpaper/background images for inclusion with SHR? I've created some FR-resolution images based on my art-work and released them under CC-SA. There's an opkg on opkg.org - you can see a sample at http://www.opkg.org/images/screenshots/tmp32_om.jpg Aug 06 21:14:07 <[Rui]> dos1: the default should be defined by the environment Aug 06 21:14:30 <[Rui]> dos1: then you can reset the environment for what you want by redefining the variable :) Aug 06 21:14:41 <[Rui]> dos1: if the variables are not set, then its an SHR bug :) Aug 06 21:16:38 Image is almost good here... Aug 06 21:17:12 <[Rui]> dos1: ok, verified. I will file a bug Aug 06 21:17:29 [Rui]: if variable is not set then elementary should fall back to default Aug 06 21:17:30 <[Rui]> it fixed the button sizes Aug 06 21:17:42 hmm Aug 06 21:17:45 or maybe you are tight Aug 06 21:17:47 right* Aug 06 21:17:58 [Rui]: but then your app shouldn't rely on finger size :P Aug 06 21:18:00 <[Rui]> dos1: it quite probably does, but libelm may have a lower default than what the openmoko screen would require Aug 06 21:18:04 <[Rui]> dos1: it doesn't Aug 06 21:18:33 <[Rui]> dos1: or better yet, under a certain point of view it does Aug 06 21:18:36 <[Rui]> dos1: but it should Aug 06 21:19:05 <[Rui]> dos1: as the same application code, without changing a preference, can be made to look good on one system and and another just due to different scales Aug 06 21:20:09 <[Rui]> so now the buttons are fixed, but the bubbles still don't expand to the full wideness of the screen Aug 06 21:20:42 <[Rui]> «ELM_FINGER_SIZE - This is sets the size of a "finger" in pixels. This should be tuned for the intended target and use. If Elementary is being used on a desktop, a finger size of 0 would be right. The same would apply for using on a touchscreen with a stylus. If you use it with an actual finger, then the size would be recommended to be bigger» Aug 06 21:21:18 <[Rui]> but maybe both can be fixed with ELM_SCALE Aug 06 21:21:23 <[Rui]> http://trac.enlightenment.org/e/wiki/Elementary Aug 06 21:26:28 <[Rui]> Where should I send the current elmdentica recipe for inclusion in shr? Aug 06 21:26:44 <[Rui]> or who? Aug 06 21:29:16 on the shr-devel ML Aug 06 21:29:18 as a patch for git Aug 06 21:29:41 It'll be applied in your name :-) Aug 06 21:34:10 Hi all, I just build shr-lite-image at home and discovered that neither 'opkg' or 'ipkg' is installed in the rootfs-image ! The /etc/opkg/*.conf files are there and so is /usr/lib/libopkg.so.0.0.0. Any knows why the 'opkg' isn't there ? Aug 06 21:35:39 We're currently investigating that issue :-) Aug 06 21:35:54 Ah, ok :-) Aug 06 21:36:19 I rebuilt an image today here and it worked properly Aug 06 21:36:28 Reducing the amount of -dev packages sadly included in the image Aug 06 21:36:39 Hopefully to 0 on the new one Aug 06 21:36:45 I'm currently flashing an intermediate version Aug 06 21:36:55 Then I already assume that you know about the missing /etc/frameworkd.conf but /usr/bin/frameworkd-binary exists... :-) Aug 06 21:37:42 Do you use the 'bb -c rebuild shr-xxx-image' command to rebuild an image ? Aug 06 21:42:20 nope Aug 06 21:43:18 ingi: for the time being I suggest the following Aug 06 21:43:23 make update (twice to be sure) Aug 06 21:43:28 then cd shr-unstable Aug 06 21:43:36 rm -rf tmp/stamps/*/task-shr* Aug 06 21:43:41 rm -rf tmp/stamps/*/shr-splash* Aug 06 21:43:55 rm -rf tmp/deploy/glibc/ipk/*/shr-splash* Aug 06 21:44:00 rm -rf tmp/deploy/glibc/ipk/*/task-shr* Aug 06 21:44:06 and bitbake shr-lite-image Aug 06 21:44:11 should work properly Aug 06 21:44:21 (I didn't bump PR yet hence the necessity to remove stamps and packages) Aug 06 21:44:48 Yeeha :-) thanks, I'll be doing that :-) Aug 06 21:45:03 I'm flashing kernel here Aug 06 21:45:09 ingi: to explain the situation Aug 06 21:45:22 frameworkd-config-shr and frameworkd-config-shr-dev were both providing frameworkd-config Aug 06 21:45:30 The preferred provider for this specific package was not set Aug 06 21:45:36 Hence frameworkd-config-shr-dev was included Aug 06 21:45:40 whereas it didn't contain anything Aug 06 21:46:06 ok image booting Aug 06 21:46:14 I'll keep you in touch in the following minutes Aug 06 21:46:22 ok good luck. Aug 06 21:49:33 X won't boot Aug 06 21:49:43 Ok, wrong passwd :-) Aug 06 21:49:57 That I know mrmoku|away had Aug 06 21:50:45 Didn't know you needed one ! Aug 06 21:51:06 well for dbus particularly Aug 06 21:51:13 which may come in handy for FSO :-) Aug 06 21:52:57 So the password was hidden somewhere down in some *-dev.bb recipe or what ? Aug 06 21:53:31 No idea yet Aug 06 21:54:15 mrmoku|away: issue with -dev packages is fixed Aug 06 21:54:30 passwd issue/postinst isn't Aug 06 21:55:24 Hmmm no evas-engine again Aug 06 21:55:40 Funny I fixed that yesterday Aug 06 21:55:51 <[Rui]> Ainulindale: ok, thanks :) Aug 06 21:56:01 I guess that's task-shr-minimal failing again Aug 06 21:57:36 and yep it is Aug 06 21:57:39 Collected errors: Aug 06 21:57:39 * ERROR: Cannot satisfy the following dependencies for task-shr-minimal-apps: Aug 06 21:57:42 * frameworkd-config-shr (= 0.9.5.9+gitr1553+aec532c0ba245da9771b56d028d3d51808ecbda6-93+0c164b58a270c689e7381e1c63e0a94ba1947af3-r7) * Aug 06 22:21:47 Ainulindale: good luck with the hunt, I appreciate your effort. **** ENDING LOGGING AT Fri Aug 07 02:59:57 2009