**** BEGIN LOGGING AT Thu Jun 30 02:59:57 2011 Jun 30 06:10:26 heyho Jun 30 06:14:16 yo morphis Jun 30 06:46:49 mickey|office: ping Jun 30 06:52:43 * mrmoku off to a client Jun 30 08:20:54 freesmartphone.org: 03morphis 07cornucopia * re7ed20ba2a10 10/fsousaged/src/plugins/dbus_service/resource.vala: fsousaged: dbus_service: add some more debug statements in resource class Jun 30 08:20:55 freesmartphone.org: 03morphis 07cornucopia * ra5b5213e33a2 10/fsousaged/src/plugins/dbus_service/plugin.vala: Jun 30 08:20:55 freesmartphone.org: fsousaged: dbus_service: even respect shadow resources when doing the suspend Jun 30 08:20:55 freesmartphone.org: Not enabled shadowed resources have the state UNKNOWN. When doing a suspend this always Jun 30 08:20:55 freesmartphone.org: fails as the code checks only for a state not equal to SUSPENDED or ENABLED. This adds the Jun 30 08:20:55 freesmartphone.org: check for the UNKNOWN state to be able to suspend the device when shadowed resources are Jun 30 08:20:56 freesmartphone.org: not enabled. Jun 30 08:20:57 freesmartphone.org: 03morphis 07cornucopia * r53fe2b2aa38a 10/fsodeviced/src/plugins/palmpre_quirks/powersupply.vala: fsodeviced: palmpre_quirks: respect connected charger when updating power status Jun 30 08:35:11 freesmartphone.org: 03morphis 07aurora * rf8e0237848b7 10/aurora-daemon/aurora/extensions/agents.py: aurora-daemon: when we have charged completely we don't charge anymore Jun 30 11:28:49 hi, DocScrutinizer Jun 30 11:28:53 hi Jun 30 11:29:19 nslu2-logs are a PITA Jun 30 11:29:44 xchat-gnome is a PITA Jun 30 11:29:53 mgedmin to the rescue :-D Jun 30 11:30:25 oh, when I click that link, chromium *downloads* the .txt file instead of opening it in a window Jun 30 11:30:28 I can see how that's a pain Jun 30 11:30:38 indeed Jun 30 11:31:12 who owns logs.nslu2-linux.org? I could help him/her set up irclog2html for a pretty HTMLized version Jun 30 11:31:16 ~logs Jun 30 11:31:16 logs are on http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged. logs are updated daily --- livelogs are http://logs.nslu2-linux.org/livelogs/openmoko-cdevel.txt and http://hentges.net/tmp/logs/irc/livelogs/#openmoko-cdevel.livelog Jun 30 11:32:02 AAAH, that's been hentges-log which had mgedmin nice formatting Jun 30 11:32:09 alas it's down since ages Jun 30 11:32:14 morphis: you are still misusing "even" :-) Jun 30 11:32:20 http://ibot.rikers.org/%23openmoko-cdevel/20110629.html.gz uses irclog2html.pl Jun 30 11:32:43 which was written by Jeff Waugh a long time ago Jun 30 11:32:56 all I did was ported it to Python and added some small features/output formats Jun 30 11:33:02 is laggy like dunnowhat as well Jun 30 11:33:03 iirc Jun 30 11:33:03 antrik: I know, but I on the way doing it right Jun 30 11:33:38 for some odd reason I never liked ibot logs Jun 30 11:34:33 morphis: I'm still puzzled though why you have trouble with this in the first place... it's not really any different than in German Jun 30 11:35:26 bad habit .. Jun 30 11:35:45 and it's not that easy to kick it Jun 30 11:35:47 mgedmin: I'd appreciate much to get nice chanlogs as defined by your awesomeness :-) Jun 30 11:37:37 morphis: just try to remember that "even" means "sogar"... then it should be obvious when it doesn't fit Jun 30 11:39:31 DocScrutinizer: try http://irclog.netripper.com/openmoko-cdevel/ Jun 30 11:40:10 jluis: I'll try that to find out which bot to kick Jun 30 11:40:19 that looks nice Jun 30 11:40:41 * mgedmin afk for a little bit Jun 30 11:41:36 jluis: clearly an unadvertised unsolicited chanlog - violates #freenode rules Jun 30 11:43:14 logbots have to be advertised and accepted by chanops, and need to be noticed to chan users in topic Jun 30 11:43:15 I just discovered it yesterday searching in google Jun 30 11:44:39 actually ALL bots need clearance by chanops Jun 30 11:44:54 and I'm taking this seriously Jun 30 11:45:46 I think you should update the topic to mention that the channel is logged. see e.g. http://purl.rikers.org/%23openmoko-cdevel/ Jun 30 11:46:05 the fact nrirclog wasn't listed in topic clearly tells me none of my chanop colleagues invited it Jun 30 11:46:14 so that's an immediate ban Jun 30 11:47:10 lindi-: wut? Jun 30 11:47:15 antrik: ok Jun 30 11:47:20 see e.g. /topic? Jun 30 11:47:44 Chan Logs: ~logs & http://logs.nslu2-linux.org/livelogs/openmoko-cdevel.txt | Jun 30 11:47:51 is that good enough? Jun 30 11:48:48 ah well, it's too long to fit to my screen :) Jun 30 11:56:59 DocScrutinizer: I'm pretty sure nothing in the policy says each bot has to be listed separately. the users only care whether the channel is logged at all, which is clearly the case -- how many bots log it doesn't matter Jun 30 11:57:36 I care, as I want to benefit from a log ;-D Jun 30 11:57:39 (TBH, I think the policy is quite ridiculous anyways -- it should be obvious to anyone that a public channel might be logged...) Jun 30 11:58:16 and generally bots are not allowed unless accepted by chanops or owner Jun 30 11:58:17 you want to benefit, and that's why you killed it? totally makes sense :-P Jun 30 11:58:56 adding a bot to a channel without informing chanops is an offense Jun 30 11:59:45 and clearly violates #freenode rules Jun 30 12:00:19 in case such bot misbehaves, you need to know whom to ping about it Jun 30 12:01:01 unsolicited bot may even get klined Jun 30 12:02:07 TBH, I never have seen trouble with unsolicited bots... it's the solicited ones that tend to be a pain in the arse, as some people (usually the owners) feel an irresistable urge to play with them constantly Jun 30 12:02:22 I hope bouncers are not seen as bots :) Jun 30 12:02:29 ~botsnack Jun 30 12:02:29 DocScrutinizer: :) Jun 30 12:02:31 hehe Jun 30 12:02:52 this quassel even logs to an SQL dabase :) Jun 30 12:02:57 database Jun 30 12:02:59 lindi-: private bouncers are perfectly legal Jun 30 12:03:20 that's why bots are not klined automatically Jun 30 13:45:22 sigh sigh Jun 30 13:45:26 I've no icons anymroe Jun 30 13:45:34 on fresh htc dream image Jun 30 13:46:24 fixed Jun 30 13:46:27 by switching theme Jun 30 13:46:34 and putting GRY back Jun 30 14:53:55 I think SHR needs aegis aka TC to become a true mature OS ;-P Jun 30 14:54:27 (just kidding, pissed by Nokia and got pissed by some nokian) Jun 30 14:54:41 nevermind Jun 30 14:55:50 DocScrutinizer: not running the GUI as root would be a nice start, SELinux and friends can come later Jun 30 14:56:12 SElinux is a *good* thing Jun 30 14:56:27 trusted platform is a *bad* thing in my book Jun 30 14:56:43 well, depends on who holds the keys Jun 30 14:56:55 and the difference between TP/TC and SElinux is exactly the bad parts Jun 30 14:57:22 indeed, TC means the manufacturer of the hw holds the keys Jun 30 14:57:37 there's no better/other way really to implement TC Jun 30 14:58:01 I prefer the term traitor-ware Jun 30 14:58:22 you can not give root cert to user, as that would mean a rogue program could snatch it away and compromise a whole hw series Jun 30 15:01:38 you *could* ship hw with individual rootcert, and ship the key for that printed on last page of the manual, so user could use it to sign his stuff on a windows PC or sth. Won't ever happen though Jun 30 15:02:18 I seem to remember OLPC did something like that, or maybe it was generated on first boot Jun 30 15:03:28 yeah, a possible use of a chain-of-trust cert based security framework. Not aegis though Jun 30 15:04:28 and x-fade bullying me out of #maemo at large cause I call aegis a thing from hell is just ridiculous Jun 30 15:04:50 cya Jun 30 15:06:12 seems like an N950 bootloader unlock is coming/probable: according to n900 community edition irc meeting on tuesday, things are still in progress to get required tools & software released. should be done soon Jun 30 15:33:25 pabs3: where from is that? Jun 30 15:33:40 #meego-handset Jun 30 15:33:54 never looked there Jun 30 15:35:10 quim pointed me at it for asking about the bootloader thing Jun 30 15:35:25 http://irclogs.meego.com/meetbot/meego-meeting/2011/ 28.6. 10:56 Jun 30 15:35:33 seems they got no chanlog Jun 30 15:35:39 aah Jun 30 15:39:35 can't spot it Jun 30 15:39:39 have to run Jun 30 15:39:44 o/ Jun 30 15:42:05 DocScrutinizer: http://irclogs.meego.com/meetbot/meego-meeting/2011/meego-meeting.2011-06-28-10.56.log.html#l-136 Jun 30 16:04:50 GNUtoo: hmm... I hear something now... but I dislike what I hear :-P Jun 30 16:04:56 ok Jun 30 16:05:11 personally I revisited the dream and .mmaped the alsa driver Jun 30 16:05:14 but.... Jun 30 16:05:28 mplayer uses like 50% CPU just by decoding ogg vorbis Jun 30 16:05:34 *now uses Jun 30 16:05:39 before it was lower Jun 30 16:06:21 I've got lot's of very loud clicking noise Jun 30 16:06:28 ah ok Jun 30 16:07:38 I can hear the operator of my voice box in the background Jun 30 16:07:45 ok Jun 30 16:14:16 GNUtoo: the play_silence thing did work? Jun 30 16:16:04 yes Jun 30 16:16:22 git commit and git push Jun 30 16:17:28 ok Jun 30 16:19:45 bbl Jun 30 16:21:01 freesmartphone.org: 03mok 07gsmvoice_alsa_cmtspeechdata+buffer_threads * r366638eb6689 10cornucopia/fsoaudiod/vapi/libcmtspeechdata.vapi: Jun 30 16:21:01 freesmartphone.org: fsoaudiod: libcmtspeechdata.vapi: extend EventData Jun 30 16:21:01 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jun 30 16:21:05 freesmartphone.org: 03mok 07gsmvoice_alsa_cmtspeechdata+buffer_threads * rfbecc5534748 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_cmtspeechdata/cmthandler.vala: Jun 30 16:21:05 freesmartphone.org: fsoaudiod: gsmvoice_alsa_cmtspeech: use a RingBuffer Jun 30 16:21:05 freesmartphone.org: This adds a ringbuffer class and uses it to buffer audio from Jun 30 16:21:05 freesmartphone.org: and to the modem. The part to the modem is still disabled to Jun 30 16:21:05 freesmartphone.org: get playback right first. Jun 30 16:21:06 freesmartphone.org: Gives horrible sound right now - lot's of alsa clicking noise. Jun 30 16:21:06 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jun 30 16:21:14 * mrmoku bbl too Jun 30 16:49:01 yeah I have qt4-embedded-gles running on the Palm Pre! Jun 30 16:52:02 congrats Jun 30 17:01:59 congrats??? the gles driver are proprietary! Jun 30 17:02:20 s/freesmartphones.org/proprietarysmartphones.org/ Jun 30 17:03:31 bbl Jun 30 17:07:56 GNUtoo: yes, I am using the properietary drivers from webOS Jun 30 17:09:11 don't get me wrong, dealing with properitary shit is nothing I like Jun 30 17:09:22 but it's the only way getting the powervr running atm Jun 30 17:09:28 pabs3: still not even any >kimju< on that webpage Jun 30 17:11:49 morphis: ack Jun 30 17:12:52 DocScrutinizer: great to see you are with me :) Jun 30 17:13:17 it's really nice to see all these graphical effects we get with accelerated graphics Jun 30 17:13:44 mickey|daddy: ping Jun 30 17:13:48 morphis: I'm pragmatic on those things, better rip out some closed blob to get sth working than not getting it working at all Jun 30 17:13:59 yes Jun 30 17:14:02 thats the way Jun 30 17:14:14 maybe there is a solution at the end of this year Jun 30 17:14:27 as someone promises open source drivers for the powervr Jun 30 17:14:37 I can feel with gnutoo's sentiment, nut in the end... Jun 30 17:14:40 but* Jun 30 17:15:23 nobody will give you a dime for sitting in the corner with a frown Jun 30 17:16:00 yes Jun 30 17:16:11 I's not like to see FSO/SHR depending on it though Jun 30 17:16:17 I'd Jun 30 17:16:30 thats something I don't want too Jun 30 17:16:42 but in the accelerated graphics area there is currently not other option Jun 30 17:16:53 modem stuff is possible to reverse engineer Jun 30 17:16:58 but graphics are quite more complex Jun 30 17:17:33 I know Jun 30 17:18:04 just make sure your desktop e.g. has a non-accelerated mode alternative that really works as well Jun 30 17:19:43 the meego way to make GFX accel mandatory for any system running meego is quite a bit ... *odd* Jun 30 17:21:15 on a broader view, I can live quite comfortably without compiz on my laptop with SiS graca Jun 30 17:22:07 I wouldn't get compiz to work, and I probably had to shoot somebody if ever I find KDE not running without compiz Jun 30 17:22:32 hehe :) Jun 30 17:23:02 making accelerated graphics the only way is the ugliest thing you could do Jun 30 17:23:21 thats even I am critize on todays desktop like unity and gnome-shell Jun 30 17:23:29 apply this to SHR and I'm fine with whatever rip-out you get working for GFX accel Jun 30 17:23:30 (btw. I am using unity and it's great) Jun 30 17:23:53 there is not real need for those to depend on accelerated graphics Jun 30 17:23:57 look at E17 Jun 30 17:24:09 it has nice effects without requiring accelerated graphics Jun 30 18:07:04 GNUtoo: about the proprietary bits Jun 30 18:07:11 GNUtoo: it's nothing I want Jun 30 18:07:22 GNUtoo: and it's not the way to succeed Jun 30 18:08:31 GNUtoo: to cite DocScrutinizer: "I'm pragmatic on those things, better rip out some closed blob to get sth working than not getting it working at all" Jun 30 18:10:40 I'm not going to destroy myself to free some stuff, just to get some proprietary stuff on top of it Jun 30 18:11:02 ok Jun 30 18:11:17 I respect your view and I am also a fan of it Jun 30 18:11:24 but what should I do? Jun 30 18:11:35 buy those features and never use them? Jun 30 18:11:44 make it not mandatory Jun 30 18:11:50 I don't Jun 30 18:11:50 would be the best option Jun 30 18:12:03 for both of us Jun 30 18:12:03 it isn't as far as I understood Jun 30 18:12:12 on meego it is Jun 30 18:12:13 it's just working Jun 30 18:13:09 first it isn't on my hardware, which means that I'm less concerned, but still I won't touch that hardware if it requires something proprietary Jun 30 18:13:15 DocScrutinizer: the ring buffer seems to work now... or not... at least it does not segfault anymore :P Jun 30 18:13:30 I mean what's the use of non-free 3d Jun 30 18:13:30 GNUtoo: it's not about that it is required Jun 30 18:13:34 you can use it when you want Jun 30 18:13:39 ok Jun 30 18:13:44 and you don't have to if you don't want it Jun 30 18:13:51 how will you handle the optionality? Jun 30 18:13:51 thats all Jun 30 18:13:52 mrmoku: fine Jun 30 18:14:01 well... sound is horrible :P Jun 30 18:14:02 like in oe for instnace Jun 30 18:14:07 GNUtoo: provide two versions of aurora Jun 30 18:14:16 DocScrutinizer: I get a lot of alsa clicks Jun 30 18:14:16 one with qt4-embedded and one with qt4-embedded-gles Jun 30 18:14:21 ok Jun 30 18:14:30 and the user won't accidentaly get the non-free one Jun 30 18:14:31 ok Jun 30 18:14:32 mrmoku: that's for sure not the fifio's fault, if it is bugfree Jun 30 18:14:35 yes Jun 30 18:14:47 than it's less an issue for me Jun 30 18:15:08 mrmoku: plaster it with printf's Jun 30 18:15:23 :) Jun 30 18:15:25 GNUtoo: the kernel driver btw. is GPL Jun 30 18:15:30 morphis, I know Jun 30 18:15:32 make the buffer length sth odd. like 1000 e.g. Jun 30 18:15:40 morphis, I'll PM you to give you more details Jun 30 18:15:46 GNUtoo: ah great Jun 30 18:16:13 mrmoku: maybe even temporarily do a simple checksum over every packet in and out Jun 30 18:16:46 hmm... that's an interesting idea :) Jun 30 18:16:49 or simply print the raw 160 long as pairs of char Jun 30 18:17:40 replacing control chars by their transposed representation Jun 30 18:17:56 like ^M->M Jun 30 18:19:06 or - even more simple - open two files, and write all the stuff that goes into fifo to file1 and al the stuff as read from fifio to file2 Jun 30 18:19:44 ok, that is the best way Jun 30 18:19:51 I's start with verbose debug-printouts for both pointers Jun 30 18:19:54 and makes it easily verifyable by a diff Jun 30 18:23:09 [debug cmtspeech] fifo_write: readptr:200 writeptr:600 calculated_free_space_in_fifo:1300 memcpy(0x0045600, 0x232323, 160) Jun 30 18:24:16 [debug cmtspeech] fifo_read: readptr:200 writeptr:600 calculated_data_avail_fifo:1300 memcpy(0x232323, 0x0045600, 160) Jun 30 19:03:35 morphis: how do I printf the address of something in vala? Jun 30 19:04:08 | cmthandler.vala:121.59-121.83: error: Argument 2: Cannot convert from `uint8*' to `uint' Jun 30 19:04:11 | stdout.printf( "---| memcpy( %X, %X, %d )\n", &x[ring_size - ring_tail], ring, count - (ring_size - ring_tail) ); Jun 30 19:04:26 just cast to uint8 ? Jun 30 19:46:08 morphis: anyone working on reverse engineering those? Jun 30 19:46:40 morphis: open source drivers for the powervr???? got a link? Jun 30 19:50:06 morphis: nevermind, found something at phoronix Jun 30 20:03:48 freesmartphone.org: 03mok 07gsmvoice_alsa_cmtspeechdata+buffer_threads * rcc47a8f202d2 10cornucopia/fsoaudiod/src/plugins/gsmvoice_alsa_cmtspeechdata/cmthandler.vala: Jun 30 20:03:48 freesmartphone.org: fsoaudiod: gsmvoice_alsa_cmtspeech: give writei the correct length when playing back modem data Jun 30 20:03:48 freesmartphone.org: Signed-off-by: Klaus Kurzmann Jun 30 20:03:56 DocScrutinizer, GNUtoo: how stupid one can be :P; Jun 30 20:04:08 playback works fine now Jun 30 20:04:10 HAH found a bug? Jun 30 20:04:21 ah you messed up with bytes and frames Jun 30 20:04:22 well... a brain bug prolly Jun 30 20:04:45 I bet a frams vs bytes oopsie Jun 30 20:04:57 GNUtoo: if you have time and want to test I would be glad to hear what you think Jun 30 20:05:11 yeah... fed writei with bytes... instead of frames Jun 30 20:05:18 ok Jun 30 20:05:42 mrmoku: I'm pleased to hear it works fine, as that moves my fifo concept out of firing line Jun 30 20:05:46 I finish debugging why the htcdream is so slow and then I look Jun 30 20:05:56 ok Jun 30 20:06:17 will activate and play with the toModem part meanwhile Jun 30 20:30:09 mrmoku: any fifo xruns? Jun 30 20:30:26 mrmoku: any ALSA xruns? Jun 30 20:30:55 mrmoku: I'd actually be worried to have no fifo xruns at all Jun 30 20:34:13 let me check the log Jun 30 20:34:41 1970-01-01T08:17:52.630340Z [WARN] CmtHandler : RingBuffer error: Buffer has only 0 bytes available (320 requested) Jun 30 20:34:49 DocScrutinizer: yes, there are some underruns Jun 30 20:47:08 pabs3: jepp, the news on phoronix is what I meant Jun 30 20:47:29 mrmoku: you got it already? Jun 30 20:47:34 hi all Jun 30 20:48:00 angelox|n900: heyho Jun 30 20:49:08 morphis: hi,btw recompiling qt didn't work with aurora :) Jun 30 20:49:19 angelox|n900: why? Jun 30 20:49:52 morphis: got same error.... Jun 30 20:50:24 angelox: can you show me the dump again? Jun 30 20:51:17 morphis: sure,let me find it :) powered up laptop right now :) Jun 30 20:52:19 morphis: http://pastie.org/2141359 Jun 30 20:55:20 angelox: hm, I recompiled pyside today and I didn't got that error Jun 30 20:55:49 yes,that is strange i guess... Jun 30 20:55:59 can show me the complete tmp/work/armv7a-oe-linux-gnueabi/python-pyside-embedded-1.0.2-r2.4/temp/log.do_configure Jun 30 20:56:14 sure,let me pastebin it Jun 30 20:57:18 ok Jun 30 20:57:23 morphis: http://pastebin.com/pEWi6WLg Jun 30 20:57:27 angelox: otherwise you can try to revert the relevant patch Jun 30 20:57:34 with "git revert f56183151b355f6ab87e4f044c918d2f91380478" Jun 30 20:57:41 and build qt and pyside again Jun 30 20:57:44 but clean them before! Jun 30 20:57:49 mrmoku: cool Jun 30 20:58:01 morphis: but i'll lose some feature ? Jun 30 20:58:32 morphis: else if, i can do it! :) Jun 30 20:58:46 no, you won't lose a feature Jun 30 20:59:05 I just disabled webkit as we don't need it currently I don't want to recompile it everytime Jun 30 20:59:12 but maybe it's better to enable it Jun 30 20:59:15 again Jun 30 20:59:25 will revert that commit Jun 30 20:59:54 morphis: ok,thank you! Jun 30 21:01:38 pushed Jun 30 21:01:58 pulling Jun 30 21:06:22 so I am off for today Jun 30 21:06:23 gn8 Jun 30 21:08:27 good night Jun 30 21:08:29 and thank you Jun 30 21:24:27 mrmoku: found out about semantics of cmt timing messages? Jun 30 21:25:36 aah, and how's cpu load for now? Jun 30 23:04:39 DocScrutinizer: actually, I think handing out the TPM key to the machine owner is totally pointless. AIUI the whole purpose of TPM is remote attestation, which is used to give control of the machine to someone else than the person having physical possession of it... Jun 30 23:05:24 exactly Jul 01 00:15:00 antrik: tell me one thing - why the heck are FOSS coders contributing and supporting TPM? Like e.g. on meego Jul 01 00:15:38 are they so extremely stupid so they think this is a good thing they can keep control over it? Jul 01 00:16:32 are they ignorant and just don't see all the implications you can learn about in wikipedia with just 2 clicks? Jul 01 00:16:59 or are they evil and think this is the right thing to do? Jul 01 00:17:12 I'm lost on this topic Jul 01 01:25:33 * angelox|n900 tries aurora Jul 01 01:36:49 DocScrutinizer: well, for Nokia it's quite clear: they want to sell DRM-protected media files, without resorting to secret code. they said so from the beginning Jul 01 01:38:57 yo Jul 01 01:39:02 I hate it Jul 01 01:42:33 I guess that's just another negative effect from their idiotic Ovi ambitions... Jul 01 01:50:17 i think that there's a lot of errors in aurora's makefile install atm Jul 01 01:53:57 or it isn't finished **** ENDING LOGGING AT Fri Jul 01 02:59:57 2011