**** BEGIN LOGGING AT Wed Sep 01 02:59:57 2010 Sep 01 08:00:38 morning Sep 01 09:01:34 <[Rui]> g'morning all! Sep 01 10:22:02 GNUtoo|laptop, hey! I have a friend thats wants to try out shr on htcdream. Can you pass me some info on how to install it? Some howto or similar. Sep 01 10:22:41 pespin, hi Sep 01 10:22:42 nice Sep 01 10:22:54 it's not 100% ready yet(GUI + keyboard issues) Sep 01 10:23:10 http://www.shr-project.org/trac/wiki/DreamInstall Sep 01 10:23:19 basically try it without installing it Sep 01 10:23:22 unpack on sd Sep 01 10:23:29 and then fastboot the kernel Sep 01 10:24:38 GNUtoo|laptop, uhm, no prebuilded images available? Sep 01 10:24:47 my friend doesn't have openembedded env. Sep 01 10:25:15 official SHR images availables at: Sep 01 10:25:46 http://build.shr-project.org/shr-unstable/images/htcdream/ Sep 01 10:26:05 unfortunately fastboot wasn't prebuilt Sep 01 10:26:12 just gab it from fso Sep 01 10:26:15 and build it Sep 01 10:26:18 it's x86 Sep 01 10:26:20 so... Sep 01 10:27:34 utilities/android/image-utils Sep 01 10:27:38 get the uttilities git Sep 01 10:27:56 else there are some binaries on the net Sep 01 10:28:07 but I'm not sure they give source with them Sep 01 10:29:01 GNUtoo|laptop, ok, what's needed in SD? 1 vfat part for kernel and 1 ext2 part for rootfs? Sep 01 10:29:09 no Sep 01 10:29:13 1 part ext3 Sep 01 10:29:22 great :) Sep 01 10:29:29 unpack the tarball inside Sep 01 10:29:35 ah ok, it's all explained in the web page, isn't it? :D Sep 01 10:29:38 like you do for SHR with QI Sep 01 10:29:46 yes, in generic terms Sep 01 10:29:48 I never used SHR from SD :P Sep 01 10:29:52 always NAND Sep 01 10:29:52 no command examples given Sep 01 10:29:53 jeje Sep 01 10:29:53 ok Sep 01 10:30:10 maybe you should update the tutorial Sep 01 10:30:14 as you're doing an install Sep 01 10:30:16 it's easy Sep 01 10:30:24 just note the commands you type Sep 01 10:33:46 pespin, also note the status: Sep 01 10:34:04 and the post-install stuff Sep 01 10:38:26 GNUtoo|laptop, I don't understand why fastboot is x86, as my firned is telling me he has it in android? Sep 01 10:38:52 maybe your friend meant that he got it with android Sep 01 10:39:05 but fastboot is the uttility which talks with the bootloader Sep 01 10:39:21 so it is on the computer/laptop/pc/desktop etc...not on the phone Sep 01 10:40:55 ok Sep 01 10:41:09 he confused me, he has one on the pc :) Sep 01 10:42:57 ok Sep 01 10:43:26 GNUtoo|laptop, no need for the kernel to be in SD rootfs, correct? Sep 01 10:54:57 GNUtoo|laptop, no need for the kernel to be in SD rootfs, correct? in which status has to be the htcdream before running the command? (I personally don't know anything about it :P) Sep 01 11:09:04 ok, he got it running eventually Sep 01 11:20:05 pespin, ok nice,sorry I was eating Sep 01 11:49:31 http://scap.linuxtogo.org/files/50bc9b373978469e882d3f96d477987c.png lol what about lizard and spock? xD Sep 01 11:53:14 pespin, how does dreams look? bad? Sep 01 11:53:59 GNUtoo|laptop, no idea, just talking to my colleage over jabber :P Sep 01 11:54:41 I wrote yesterday a blog entry explainging my fights against htc artemis, and i encouraged everybody to try free its phone Sep 01 11:54:48 and it seems he answered to the call :P Sep 01 11:55:08 ok nice Sep 01 11:55:09 but he told me he doesn't think he'll be using it as daily basis Sep 01 11:55:23 tell him the GUI+keyboard is highly unobtimized Sep 01 11:55:35 and that I'm working on it Sep 01 11:55:50 but as GUI depends on keyboard... Sep 01 11:55:57 I'm trying to push keyboard first Sep 01 11:56:29 great :) Sep 01 11:56:46 GNUtoo|laptop, how is SHR going on yours? smooth? Sep 01 11:56:52 very smooth Sep 01 11:57:06 and great Sep 01 11:57:13 but I've lots of not-yet-commited stuff Sep 01 11:57:25 like keyboard,GUI Sep 01 11:57:28 GNUtoo|laptop, (smooth) arghh we are all jealous of you :) Sep 01 11:57:43 lol Sep 01 11:57:56 GNUtoo|laptop, how's screen compared to FR? Sep 01 11:57:58 maybe I'll do a video soon but I'd like to push keyboard Sep 01 11:57:59 smaller? Sep 01 11:58:05 480x320 but bigger Sep 01 11:58:12 and no xrandr -o 1 Sep 01 11:58:18 only xorg.conf rotation Sep 01 11:58:25 he was asking for rotation with accels Sep 01 11:58:41 and for getting big buck bunny fulscreen you need portrait mode Sep 01 11:58:47 and a not pushed patch Sep 01 11:58:53 but it was published on alsa-devel Sep 01 11:59:08 it was not pushed becuase it makes buffer underrun worse Sep 01 11:59:33 uhm what about calls? supported by default? Sep 01 11:59:38 yes Sep 01 11:59:42 works out of the box Sep 01 11:59:43 but.... Sep 01 11:59:55 ring can be problematic when you are already listenning to music Sep 01 12:00:02 3g works Sep 01 12:00:13 sms works Sep 01 12:00:30 btw tell him to do that: Sep 01 12:00:40 not to use the shr-install gui Sep 01 12:00:48 to put next->next->next Sep 01 12:00:53 otherwise keyboard popups Sep 01 12:01:30 GNUtoo|laptop, yeah he had that problem lol Sep 01 12:01:45 I really must commit keyboard Sep 01 12:01:48 but that's hard Sep 01 12:01:55 oe stuff that is getting complicated Sep 01 12:02:09 basically split packages and post-split handling Sep 01 12:02:14 which doesn't work Sep 01 12:02:25 roughly it's in gtk recipes Sep 01 12:02:28 GNUtoo|laptop, uhm isn't illume-virtual-kb another separate package from illume? Sep 01 12:02:39 it should imho Sep 01 12:02:44 I don't know...I mean the real keyboard Sep 01 12:03:05 there is a hardware keyboard on htcdream Sep 01 12:03:20 the difficulties are: Sep 01 12:03:24 GNUtoo|laptop, yeah, but popup problem i from virtual one, isn't it? Sep 01 12:03:27 add a gtk-immodule.sh Sep 01 12:03:32 yes Sep 01 12:03:45 then I need to fix some xorg bugs Sep 01 12:03:49 <[Rui]> weird... I just heardphone interference over the computer speakers, looked at the freerunner and it got dissociated from GSM network. Sep 01 12:03:51 about keyboard mapping Sep 01 12:04:03 <[Rui]> tried to ssh, failed (pings, socket is open, but doesn't fork). Sep 01 12:04:35 [Rui], got serialconsole? Sep 01 12:04:37 GNUtoo|laptop, gvery ood job with dream btw :) Sep 01 12:04:38 <[Rui]> had vala-terminal open, try to change to qwerty keyboard (doesn't react to tapping on keyboard) Sep 01 12:05:02 <[Rui]> popup top bar popsup, tried to close e, couldn't. Sep 01 12:05:04 pespin, thanks Sep 01 12:05:13 <[Rui]> I just removed the battery for a forced reboot :( Sep 01 12:05:23 <[Rui]> GNUtoo|laptop: no serial console on my fr, sadly Sep 01 12:05:26 pespin, if he wants something usable he really should follow post-install rules Sep 01 12:05:40 <[Rui]> seems to be related with receiving a text message Sep 01 12:05:55 <[Rui]> as it got connected to the network after boot, it announced a new message. Sep 01 12:06:01 <[Rui]> could also be just a coincidence... Sep 01 12:06:08 [Rui], did the test message had that iside: exec fork-bomb? Sep 01 12:06:45 <[Rui]> GNUtoo|laptop: lol, no, a friend asking if he could come up. Sep 01 12:06:48 ok Sep 01 12:06:57 <[Rui]> GNUtoo|laptop: don't tell me text message isn't safe on FSO! :) Sep 01 12:07:15 so he asked the freerunner if he could come up, and the freerunner granted him access? Sep 01 12:07:25 (I'm just joking) Sep 01 12:07:30 (because I've no clue) Sep 01 12:07:37 <[Rui]> GNUtoo|laptop: lol, np :) Sep 01 12:07:42 (if only you could copy the logs....) Sep 01 12:07:42 <[Rui]> it's *very* weird. Sep 01 12:07:57 <[Rui]> GNUtoo|laptop: but they're on ram, right? Sep 01 12:08:02 <[Rui]> so reboot == no log Sep 01 12:08:11 indeed **** BEGIN LOGGING AT Wed Sep 01 12:36:30 2010 Sep 01 12:53:48 what happend with om-gta02-2.6.32 branch on git.openmoko.org? Sep 01 12:58:50 Martix: hi, what's the problem? Sep 01 13:02:01 I remember last commit from July 2010 by JaMa, but now when I run: git checkout --track -b om-gta02-2.6.32 origin/om-gta02-2.6.32 I got last commit from Mon Apr 26 18:51:30 2010 +0200 Sep 01 13:04:48 Martix: you probably need the gitorious branch Sep 01 13:05:11 http://gitorious.org/~jama/htc-msm-2-6-32/openmoko-kernel/ Sep 01 13:06:08 ok, thanks Sep 01 13:07:35 Martix: hi Sep 01 13:07:53 Martix: about your WS Sep 01 13:08:08 gena2x: hi Sep 01 13:08:37 Martix: I have only 1 useful idea and only about report, not fix. Sep 01 13:08:59 i want to point out that register dump is incomplete Sep 01 13:09:55 gena2x: where can I get complete dump? Sep 01 13:10:33 wait a bit Sep 01 13:11:58 wow, nice larsc returned and doing things Sep 01 13:12:20 but hm. why hi commited my patch as 'author'? Sep 01 13:14:38 ok, this is not relevant to issue Sep 01 13:17:25 Martix: http://git.openmoko.org/?p=kernel.git;a=blobdiff;f=drivers/mfd/glamo-core.c;h=135856af1ee3ee663a0e19bccb499a501b5a0eff;hp=4af7c54eac1d460f1da3c8903bbf5b317abc54cc;hb=1e47708c768b5d7ff2f214ac7467a28201ba254c;hpb=d68329250120b2c92a29fc71805cd7ce043108b0 Sep 01 13:17:31 Martix: revert this one Sep 01 13:18:09 Martix: without this patch, reading regs will fail 33% of times Sep 01 13:18:27 Martix: but in other 66%, it will produce much better dump Sep 01 13:19:06 but I cant reproduce WS anymore Sep 01 13:19:20 Martix: I never pretended to fix all WSes Sep 01 13:19:45 Martix: and not really think i can help with this one Sep 01 13:20:13 but for sure better dump may help to man who will try to solve this issue later Sep 01 13:20:36 its enough for me, if WS is rare Sep 01 13:20:55 not good. several things may be tried Sep 01 13:21:15 for example raise delay i introduced Sep 01 13:21:25 and register dump may help detecting this Sep 01 13:21:39 but _full_ register dump Sep 01 13:21:48 i think your WS were on blank/unblank Sep 01 13:22:07 you may try to loop blanking with om tool Sep 01 13:22:34 btw, did it restore? Sep 01 13:22:56 i mean, were you able to restore screen with subsequent blank/unblank? Sep 01 13:24:46 no Sep 01 13:29:12 create bug report next time with register dump please Sep 01 13:29:33 btw, did you managed to do FCSE testing? Sep 01 13:29:36 gena2x: hi, the problem with power in suspend is already present in 2.6.30 kernel from openwrt, if it's any help for you Sep 01 13:30:18 gena2x: otherwise i tried some measuring of current during init, but no results Sep 01 13:30:21 radekp: hi, so you decided to bisect first :) Sep 01 13:30:48 radekp: have you tried thing i suggested - to log current_now? Sep 01 13:30:49 gena2x: we can compare 2.6.29 and 2.6.30 now which should be quite similar Sep 01 13:31:09 gena2x: yes, but battery driver is initialized very late Sep 01 13:31:19 ahhh Sep 01 13:31:24 unexpected trouble :) Sep 01 13:31:44 did you enabled PM debug ? Sep 01 13:31:51 no Sep 01 13:32:15 it may be useful, as it adds 1 line for each new/removed/suspended/resumed device Sep 01 13:32:29 to logs i mean Sep 01 13:32:30 it's also bad that the currents dont differ very much - it's only 8mA or even less Sep 01 13:32:32 gena2x: FCSE, I cant run lmbench on SHR, but on daily user testing is stable Sep 01 13:33:32 radekp: hm. i had 8mA with .29 and 20mA with .34 Sep 01 13:33:42 radekp: so it is 10 or so Sep 01 13:33:58 Martix: if it 'feels' more responsive? Sep 01 13:34:04 s/if/is/ Sep 01 13:34:04 gena2x meant: Martix: is it 'feels' more responsive? Sep 01 13:34:31 radekp: ok, i'll try to do it today Sep 01 13:34:35 radekp: now in fact Sep 01 13:34:43 gena2x: imho yes Sep 01 13:35:27 Martix: ok, i'll do lmbench here, and publish results Sep 01 13:35:55 Martix: no patch for my .34 :( may be .32 will work, or i'll contact author. Sep 01 13:36:22 gena2x: for example, when I am poking in SHR-settings, it seems like last visited categories are cached Sep 01 13:37:17 gena2x: I'll send a .32 patch to devel and try to apply it on .34 Sep 01 13:38:24 I should send mail to author and report my experience Sep 01 13:38:26 Martix: i can do lmbench really fast Sep 01 13:38:40 Martix: and send copy to maillist please :) Sep 01 13:39:00 Martix: so, all i need is kernel patch Sep 01 13:39:48 ok Sep 01 14:44:43 ~seen JaMa Sep 01 14:44:49 http://wiki.freesmartphone.org/index.php/SHR/Meeting Sep 01 14:44:50 jama <~martin@161-24.13.24.78.awnet.cz> was last seen on IRC in channel #openmoko-cdevel, 8d 17h 59m 39s ago, saying: 'tomorrow I have to arrive to work before 7 :/'. Sep 01 14:44:58 leviathan: ping Sep 01 14:45:08 dcordes: pong Sep 01 14:45:19 I'm recently doing a IPC presentation Sep 01 14:45:22 dcordes: do you know Sep 01 14:45:28 if there is a recent block diagram Sep 01 14:45:34 leviathan: what's up with our chit chat at 2000CET ? Sep 01 14:45:36 about the architecture of FSO2? Sep 01 14:45:42 no, sorry Sep 01 14:45:49 hmm Sep 01 14:45:54 mickey|zzZZzz: might though ? Sep 01 14:45:54 today?? Sep 01 14:46:05 Wednesday 09/01/10 20:00 CET ? OK OK ? NAK Sep 01 14:46:21 OK Sep 01 14:46:27 I don't know :) Sep 01 14:46:29 monday, tuesday, wednesday Sep 01 14:46:30 hmm Sep 01 14:46:45 english day count... ^_^" Sep 01 14:46:58 is now 20.00 CET? Sep 01 14:47:19 my gnome panel app tells me it's 16.44 CET now Sep 01 14:47:36 sometimes I dont even know what day it is in german Sep 01 14:47:46 how am I supposed to know it in english ;-P Sep 01 14:47:53 its a much deeper problem Sep 01 14:48:01 something with time feeling behind your pc Sep 01 14:48:03 ^^ Sep 01 14:50:04 leviathan, I've a video for you, it tallks a little bit about buisness practices: Sep 01 14:50:13 mhmm Sep 01 14:50:20 ahh, btw Sep 01 14:50:21 :-) Sep 01 14:50:31 new patch in codeaurora Sep 01 14:50:35 some quallcomm guy Sep 01 14:50:42 wrote usb gadget driver Sep 01 14:50:52 now I've got functional usb ethernet in upstream Sep 01 14:50:53 :-D Sep 01 14:51:10 http://balance.fsf.org/video/lp2010-sat-symbian.ogv Sep 01 14:51:14 ok Sep 01 14:51:17 wow Sep 01 14:51:19 !!! Sep 01 14:51:20 <[Rui]> GNUtoo|laptop: that weird problem I had earlier today? happened again. Sep 01 14:51:36 ah ok Sep 01 14:51:40 so put the logs on sd Sep 01 14:51:47 and put that in background: Sep 01 14:51:58 while true;do dmesg -c;done >> dmesg Sep 01 14:52:16 <[Rui]> GNUtoo|laptop: perhaps with a sleep it would be wiser ;) Sep 01 14:52:25 no Sep 01 14:52:28 dmesg -c Sep 01 14:52:38 note the -c Sep 01 14:52:46 you want it as soon as it crashes right Sep 01 14:52:51 so no sleep Sep 01 14:53:39 <[Rui]> GNUtoo|laptop: I don't know *when* it'll happen Sep 01 14:53:52 that's not an issue Sep 01 14:54:04 if you know it will happen Sep 01 14:54:04 <[Rui]> GNUtoo|laptop: what's the gist on that video? Sep 01 14:54:11 gist? Sep 01 14:54:14 what does it means? Sep 01 14:54:36 they talk a bit about buisness Sep 01 14:54:37 <[Rui]> GNUtoo|laptop: what is it about? :) Sep 01 14:54:41 simbian Sep 01 14:54:46 more specifically Sep 01 14:54:57 the simbian code release and the simbian foundation Sep 01 14:55:19 at some point he told things like changing the mind of the companies Sep 01 14:55:44 companies think that their assets are products,code and "intelectual property" Sep 01 14:56:10 if they release the code as free software they loose that buisness method Sep 01 14:56:20 but they gain another Sep 01 14:56:28 assets are then: expertise and people Sep 01 14:56:34 <[Rui]> while interesting in a way, how relevant is simbian still? last I checked it wasn't buildable with Free Software, and I seem to remember Welte complaining about a lot of interesting stuff being missing. Sep 01 14:56:35 <[Rui]> while interesting in a way, how relevant is simbian still? last I checked it wasn't buildable with Free Software, and I seem to remember Welte complaining about a lot of interesting stuff being missing. Sep 01 14:56:50 not relevant yet: Sep 01 14:56:55 my arguement will be, that they do not have to pay money for sw development Sep 01 14:56:58 *can't fully build with gcc//work on the way Sep 01 14:57:02 it will be done by the community Sep 01 14:57:05 and even better Sep 01 14:57:15 they will even earn money while development phase Sep 01 14:57:15 *no free simbian phone Sep 01 14:57:25 but there is that project that uses a beagleboard: Sep 01 14:57:28 because the people will pay money also for dev phones Sep 01 14:57:29 :-) Sep 01 14:57:33 http://developer.symbian.org/wiki/index.php/Wild_ducks_project Sep 01 14:57:54 ok Sep 01 14:58:06 leviathan, hmmm Sep 01 14:58:19 community management cost Sep 01 14:58:29 thats O2S part Sep 01 14:58:30 and at openmoko that's exactly what they thought Sep 01 14:58:38 and O2S in this project is basically.... Sep 01 14:58:38 me Sep 01 14:58:39 they ended up doing the work Sep 01 14:58:39 ok Sep 01 14:58:40 so. Sep 01 14:58:41 :-) Sep 01 14:59:05 o2s will adapt kernel drivers from the BSP etc... Sep 01 14:59:12 jup Sep 01 14:59:12 do oe integration Sep 01 14:59:15 etc...? Sep 01 14:59:24 hmmm I've an idea Sep 01 14:59:25 theres some people Sep 01 14:59:29 yes? Sep 01 14:59:31 which one? Sep 01 14:59:33 all that integration is valuable Sep 01 14:59:40 if someone is able to re-use it Sep 01 14:59:45 time to market is smaller Sep 01 14:59:50 jup Sep 01 14:59:51 so.... Sep 01 15:00:05 it will cost less Sep 01 15:00:11 why not tell samsung about that, see your phone as a dev phone or reference design Sep 01 15:00:11 to provide new devices Sep 01 15:00:24 and samsung could license new designs Sep 01 15:00:26 based on that Sep 01 15:00:42 thats the greatest idee! Sep 01 15:00:45 :-D Sep 01 15:00:52 I'll do that Sep 01 15:00:57 nice Sep 01 15:01:14 leviathan: 16:45 < leviathan> sometimes I dont even know what day it is in german Sep 01 15:01:27 or even use that work Sep 01 15:01:29 leviathan: semester break phenomenon. same here Sep 01 15:01:58 dcordes: lol, ok ^_^ Sep 01 15:02:09 GNUtoo|laptop: ok Sep 01 15:02:12 so it seems like we are not having the conf today Sep 01 15:02:24 I'm in chaostreff this evening Sep 01 15:02:30 ok Sep 01 15:02:32 better: I've gotta go in some minutes Sep 01 15:02:44 well if everyone is here Sep 01 15:02:53 hackrfunk :) Sep 01 15:02:54 I could make a break after the presentation Sep 01 15:03:05 yes ^^ Sep 01 15:03:08 mitm venty Sep 01 15:03:10 :-) Sep 01 15:03:13 we should get more ppl on board for the meeting anyway Sep 01 15:03:17 also it lacks some content Sep 01 15:03:20 jup Sep 01 15:03:28 I've got some good innovation Sep 01 15:03:32 we should talk about Sep 01 15:03:40 I'll write them down Sep 01 15:03:43 to me the most important thing is what I pointed out in the wiki already. the whole community thing, identity etc Sep 01 15:03:51 ok cool looking forward to check them in the wiki Sep 01 15:04:02 and applications writers or porters Sep 01 15:04:10 would make compatible software Sep 01 15:04:15 with their platform Sep 01 15:04:31 I thought about to write tickets for the features, to ennoy the developers a bit and motivate them to implement them xD Sep 01 15:05:33 GNUtoo|laptop: exactly Sep 01 15:05:51 GNUtoo|laptop: would make the development processes much more faster then slower Sep 01 15:06:01 s/then/than Sep 01 15:10:40 GNUtoo|laptop: leviathan: you guys need to check that: http://lists.linuxtogo.org/pipermail/mobile-linux-discuss/2010-August/000182.html Sep 01 15:10:55 http://lists.linuxtogo.org/pipermail/mobile-linux-discuss/2010-August/000183.html Sep 01 15:10:59 potential new SHR device Sep 01 15:11:02 Nexus One Sep 01 15:11:35 I wish I had the skill to port alsa to nexus (passion) and hd2 (leo= Sep 01 15:11:54 leviathan: I was all enthusiastic about the leo for you. sorry it didn't happen Sep 01 15:12:13 dcordes, alsa.... Sep 01 15:12:22 dcordes, I'll get a free one tough Sep 01 15:12:29 but I'll have to do some replicant work Sep 01 15:12:30 for it Sep 01 15:12:34 in exchange Sep 01 15:12:41 GNUtoo|laptop: cool. when will you get it ? Sep 01 15:12:58 dcordes: hint from my side Sep 01 15:13:01 when I get back to milan + the time it takes to come(shipping) Sep 01 15:13:03 dcordes: try kdevelop4 Sep 01 15:13:12 dcordes: even with not such amount of skill Sep 01 15:13:22 dcordes: its easy do analyse and write an alsa driver Sep 01 15:13:27 leviathan: thanks. I will look into that. I love kdiff3 :) Sep 01 15:13:30 you know it ? Sep 01 15:13:42 dcordes: jup Sep 01 15:13:52 Im only using kde4 on desktops Sep 01 15:14:00 leviathan, it's not so easy, there are caveats Sep 01 15:14:08 I'm loving the functionality variety Sep 01 15:14:23 leviathan, basically qsd is different Sep 01 15:14:29 hmm Sep 01 15:14:30 there are 2 modes for dsp Sep 01 15:14:36 android and codeaurora Sep 01 15:14:40 ok Sep 01 15:14:47 the codeaurora alsa driver uses the codeaurora mode Sep 01 15:14:55 so basically what is to be done is: Sep 01 15:15:02 dcordes: my offer: look into it, try it, show me your results, I'll help you Sep 01 15:15:05 or use codeaurora kenrel //not sure if it works Sep 01 15:15:06 now I've gotta go Sep 01 15:15:15 or adapt the msm7kv2 one Sep 01 15:15:21 and even worse.... Sep 01 15:15:29 I want good sound quality Sep 01 15:15:32 * leviathan needs to go to zurich now, for chaos computer club meeting Sep 01 15:15:39 * leviathan will come online there Sep 01 15:15:50 that is not solved even in dream Sep 01 15:15:50 some small buffer underrun persist Sep 01 15:15:55 k Sep 01 15:16:04 dcordes: try kdevelop4 Sep 01 15:16:06 :-) Sep 01 15:16:07 or Sep 01 15:16:09 uhhh Sep 01 15:16:11 ok Sep 01 15:16:12 brb Sep 01 15:16:36 leviathan, kdevelop4 does autocompletion ? Sep 01 15:16:46 like eclipse? Sep 01 15:17:00 and error checking with red thing? Sep 01 15:17:18 because it would be a nice thing for non-embedded companies Sep 01 15:17:40 like for people used to java or .net developments which write very bad C Sep 01 15:17:48 *who Sep 01 15:18:24 GNUtoo|laptop: yes, it does that all Sep 01 15:18:26 and much more Sep 01 15:18:29 try it! Sep 01 15:18:30 ^_^ Sep 01 15:18:31 so Sep 01 15:18:36 now really away Sep 01 15:18:37 bb Sep 01 15:18:38 ok Sep 01 16:04:45 strange....ogpsd works nearly out of the box, "gps" needs to be launched and an MSMDevice class needs to be written Sep 01 16:20:54 well, with several "possible" devices on the hacking horizon, it makes sense for "a manufacturer" to give away some devices to ensure "proper" direction. makes also sense for us not to fragment too much.... Sep 01 16:47:47 vanous: i think FR is underhacked :) Sep 01 16:48:19 gena2x: agree, but... you know... the evolution went on... Sep 01 16:53:59 evolution of what? Sep 01 16:54:08 cpu speeds? memory sizes? Sep 01 16:54:15 battery consumption? :) Sep 01 16:54:30 gena2x: should i commit 2-4-2 to Qi openmoko repo? Sep 01 17:00:30 PaulF: i am worring about sd card boot. u-boot has some strange troubles with it - 1/2 of attempts to boot need to retry because or loaded kernel 'crc error'. i can't say if this also influence qi. so, testing needed and may be more fixes. (i had to increase rootdelay after 2-4-2) Sep 01 17:01:12 gena2x: hm, why do not you use rootwait? Sep 01 17:02:15 PaulF: also android guys seem had some troubles with booting 2-4-2, but it may be everything - glamo look ok, but they can't provide any longs at all just 'see picture'. may be this is because someone rebuild it in some wrong way Sep 01 17:02:32 this happens during installation Sep 01 17:02:43 s/longs/logs/ Sep 01 17:03:19 PaulF: sure, just at the time of testing i didn't not know about rootwait, and now didn't switched to it Sep 01 17:03:50 gena2x: i see. I just wanted to let you know i'm not reluctant to commit any patches to the "official" qi repo. Sep 01 17:04:27 PaulF: so, may be need more testing and fixes. but so far nobody reported this problem about qi in which is in community-ml Sep 01 17:04:41 PaulF: thanks. Sep 01 17:07:54 gena2x: those folks on the community ml seem to not have a clue asking for a ready-patched qi etc. Sep 01 17:09:42 PaulF, is qi used in other devices than the fr? Sep 01 17:12:30 xrmx: i guess Andy is using it for some custom boards. Of course, he's keeping a fork somewhere, but i'm not aware of the details. Sep 01 17:45:34 <[Rui]> vanous: WRT fragmentation in the arm scene, there's an intention of making a standard in order to reduce costs associated with fragmentation Sep 01 17:45:49 gena2x: well, i am not necessarily happy about this... but it's like if a guitar player gets a new guitar Sep 01 17:46:02 it brings new sense of inspiration Sep 01 17:47:02 well, i think if people have a chance to "just do" the modem/sound etc integration and then install shr/oe/debian onto their phones, all is cool then Sep 01 17:47:12 integration into fso i mean Sep 01 17:47:34 so the system becomes standard and no integration work gets wasted Sep 01 17:49:41 vanous: i want phones to became something like PC - you just change hardware but software is same Sep 01 17:49:51 gena2x: 100% agree here Sep 01 17:50:04 vanous: i think this is way every linux distro is aiming Sep 01 17:50:20 so, shr on fr will be same to shr on htc Sep 01 17:50:24 in fact, i would hope that integration of other phones goes fast so more work goes towards userspace apps Sep 01 17:51:03 and where is common subset of 'features' - so most of userspace stuff may be done and tested on fr Sep 01 17:51:08 hi PaulF Sep 01 17:51:15 yes Sep 01 17:51:30 so... field for hacking is huge :) Sep 01 17:51:30 PaulF, did you read that source file: Sep 01 17:51:34 hopefully hardware permitting :) Sep 01 17:51:46 GNUtoo|laptop: Hi :) Sep 01 17:51:49 drivers/hsi/clients/cmt_speech.c Sep 01 17:51:59 GNUtoo|laptop: no, why do you ask? Sep 01 17:52:00 i have to agree with [Rui] - have been spoiled by the freerunner's posibilities Sep 01 17:52:01 that defines: Sep 01 17:52:22 #define CS_QOS_LATENCY_FOR_DATA_USEC 20000 Sep 01 17:52:27 whith the following comment: Sep 01 17:52:30 <[Rui]> vanous: :) Sep 01 17:52:42 During data transfers, transactions must be handled within 20ms (fixed value in cmtspeech HSI protocol) Sep 01 17:53:03 * [Rui] feels to lazy to preserve db schema of elmdentica in upcoming changes... :| Sep 01 17:53:10 <[Rui]> s/ to / too / Sep 01 17:53:35 PaulF, was it you that pointed me to the pulse presentation and that told me the 20ms were overhead? Sep 01 17:53:41 gena2x: one thing did evolve a lot, gsm->3G Sep 01 17:53:43 <[Rui]> bot not working or regex confused it? :) Sep 01 17:53:54 GNUtoo|laptop: it was DocScrutinizer :) Sep 01 17:54:01 GNUtoo|laptop: and there was 5ms iirc Sep 01 17:54:08 <[Rui]> vanous: 4G will be awesome (standards based, IP, etc...) Sep 01 17:54:09 vanous: but how this influence development? Sep 01 17:54:15 vanous: or hacking? Sep 01 17:54:28 ok Sep 01 17:54:32 sorry then Sep 01 17:54:41 not much, but will draw more people towards other devices for daily usage Sep 01 17:54:43 vanous: almost no influence. you'll just get faster data rate on 3g, and that's all Sep 01 17:54:57 <[Rui]> I guess that with 4G the thin barrier between voice and voip will finally be totally diulluted (but for stupid operator greed). Sep 01 17:54:57 to whoever wrote rock paper scissors <--- Nice graphics. Sep 01 17:55:07 gena2x: you mean will cost more as it goes faster? Sep 01 17:55:20 GNUtoo|laptop: it's always nice to talk to you, so no need to sorry :) Sep 01 17:55:31 ok thanks a lot Sep 01 17:55:43 I'm evaluating my next targets Sep 01 17:55:49 or n900 or palm-pre-plus Sep 01 17:55:51 Moscow is already covered with WiMax almost entirely :P Sep 01 17:56:18 PaulF: will check soon again :)) Sep 01 17:56:38 But they do not provide a fixed (or even public) IP, so one can't get a phone number routed to it anyway. Sep 01 17:56:42 i mean, from developer POV nothing will change with 3g, so you just can't hack 5% of apps Sep 01 17:57:03 PaulF: you may tunnel via other provider Sep 01 17:57:41 yes vpn si an idea Sep 01 17:57:49 I use that Sep 01 17:58:01 gena2x: agree in principle. i just hope there will not be a general issue that would cause a big split or take devs to other directions... Sep 01 17:58:25 openmoko is made of splits already Sep 01 17:58:31 hehe Sep 01 17:58:44 tons if libs - qt, fso Sep 01 17:58:52 tons of bootloader - qt, u-boot Sep 01 17:58:56 tons of kernels Sep 01 17:59:04 tons of distributions Sep 01 18:00:18 hackers is hackers Sep 01 18:00:32 but fr is so amazing thing to hack Sep 01 18:00:52 it still has tons of things to fix and understand Sep 01 18:01:35 at any level - from kernel to userspace Sep 01 18:02:21 ah, bbs Sep 01 18:02:33 gena2x: i do appreciate your view, i think it also brings up levels of interest Sep 01 18:03:27 vanous: i think it is possible to even attract new devs. need just to show that fr is in good state now, it is fast and hacking-attractive platform. Sep 01 18:04:37 ok, really gone Sep 01 18:06:40 gena2x: i have had success with my fr plea to the ml, i am hoping my hacking colleague will pick up the towel :) Sep 01 18:09:56 there is a guy from the uk shiping soon as this was faster then me being there next week Sep 01 18:10:13 there is another guy in moscow, perhaps i will consider it for another colleague and arrange handing over at the end of september Sep 01 18:10:53 * vanous hopes for some success in this direction... Sep 01 18:15:02 * dos1 is building nokia900 SHR image on buildhost Sep 01 18:15:26 i hope nothing will break, as JaMa is gone :] Sep 01 18:16:18 dos1: fingers X :) Sep 01 18:16:21 <[Rui]> I can't understand why I still can't build shr-lite-image, now it's eglibc with "hiccups" again (and bb -c clean all eglibc and gcc recipes didn't solve it) :( Sep 01 18:16:39 <[Rui]> dos1: you'll do just fine :) Sep 01 18:16:46 <[Rui]> dos1: did you bump omnewrotate? Sep 01 18:17:26 [Rui]: no, i don't have OE access and I don't want to mess with local checkout of oe, sorry :x Sep 01 18:17:30 <[Rui]> dos1: you'll need to bump and remove the patch (as it's already applied in trunk). It also contains fixes so it works better with current accelerometer data... I missed the relative to transition Sep 01 18:17:37 <[Rui]> dos1: ah... ok, no problem then... Sep 01 18:17:46 Does someone know how i can make an empty char**? Sep 01 18:17:51 [Rui]: but if it'll be commited to OE, then i'll of course build it Sep 01 18:17:52 <[Rui]> dos1: you mean that then I could try and ask someone at #oe ? Sep 01 18:18:12 [Rui]: yup, and maybe just send patch to oe-devel maillist Sep 01 18:18:16 [Rui]: i haven't had time, but will test locally :) Sep 01 18:19:18 [Rui]: i can commit it Sep 01 18:20:51 <[Rui]> Heinervdm: pasted the pastebin'ized diff in #oe Sep 01 18:21:42 [Rui]: ok, but i have to pull first, and i will commit it with maxrev instead of deleting that patch Sep 01 18:21:43 <[Rui]> vanous: it's still not perfectly detecting the change, I still need to shuffle it a bit in order to get there, I don't understand what made detection harder. Sep 01 18:22:00 <[Rui]> Heinervdm: the patch isn't needed anymore, as it's included Sep 01 18:22:08 [Rui]: are you sure you're using correct accelerometer? Sep 01 18:22:16 <[Rui]> dos1: yes, /dev/input/event3 Sep 01 18:22:28 [Rui]: FR has two of them, and order of input devices has changed with 2.6.32 Sep 01 18:22:37 <[Rui]> the order changed as well?= Sep 01 18:22:42 i think yes Sep 01 18:22:44 <[Rui]> I didn't know that... Sep 01 18:23:04 as without changing configs X used accels as mouse for me ;) Sep 01 18:23:06 <[Rui]> Heinervdm: if you can wait a minute or two, I'll just try the other accelerometer and check if it ... Sep 01 18:23:11 <[Rui]> damn, I have to go out. Sep 01 18:23:12 <[Rui]> bbl Sep 01 18:23:23 [Rui]: ok, i will wait Sep 01 19:14:49 vanous: heh, i wanted to make some movie about current state of freerunner Sep 01 19:15:37 prepared environment to capture both iteraction from host and view to device Sep 01 19:15:48 and started capturing qtmoko Sep 01 19:17:17 i spent much time on preparations (my web camera is not really in excellent state) Sep 01 19:18:25 then i finally managed to capture i had some plan what to focus. and i started recording every interesting moment about fr Sep 01 19:18:36 i mean qtmoko Sep 01 19:19:04 i realized that i do not know nothing about it :) i do not know how to use arora. how to watch youtube Sep 01 19:19:16 and what to do with it at all Sep 01 19:19:28 except calls/contacts/messages :) Sep 01 19:20:24 then i finished shooting, i realized that movie source is >1hr Sep 01 19:20:50 opened cinelerra and realized that task to shoot all major distro is huge Sep 01 19:21:31 because i need to reshoot even qtmoko version and learn how to use it's 'cool' features first :) Sep 01 19:28:18 gena2x: question: are the arora changes specific to qtmoko? Sep 01 19:29:33 lindi-: need to ask radek or check git Sep 01 19:45:56 gena2x: lightspark? Sep 01 19:46:23 tmzt: youtube-dl + mplayer or similar afaik Sep 01 19:46:29 tmzt: lightspark uses opengl afaik Sep 01 19:47:19 ah true Sep 01 19:47:20 tmzt: what? wow, i didn't even know it exists! Sep 01 19:47:30 but I think it only uses gstreamer, opengl is optional Sep 01 19:48:05 whooh, seem opensource now is more than one man can insert to his head Sep 01 19:48:54 no, in qtmokov26 radek did download+recoding+play sheme Sep 01 19:49:06 lindi-: recoding to theora btw Sep 01 19:49:18 and i still didn't tried it. Sep 01 19:49:32 where were instruction on his announcement Sep 01 19:49:53 tmzt: from #gnash I heard opengl Sep 01 19:50:13 gena2x: recording? you mean recoding? Sep 01 19:50:52 gena2x: hah, I'm still trying, I'm still young :) Sep 01 19:51:00 tmzt: "< rsavoye> it's also mostly a youtube player, other sites don't work < rsavoye> also OpenGL *only*" Sep 01 19:51:04 lindi-: sure, damn ENTER key :) Sep 01 19:51:17 they don't work yet Sep 01 19:51:22 it' s a new vm using llvm Sep 01 19:51:39 and that depends on the selected gst sink, it should work without gl Sep 01 19:51:50 tmzt: hehe. it is funny, but 1000 photo editor appear last few years. Sep 01 19:52:16 I only care about interesting products Sep 01 19:52:20 projects Sep 01 19:52:25 heh Sep 01 19:52:35 may be my interests too wide... Sep 01 19:53:22 i opened cinelerra and realized that this will be very very very uneasy task to make video :) Sep 01 19:57:34 tmzt, lindi- it seems you are speaking about youtube on openmok? you might like this :) http://downloads.tuxfamily.org/3v1deb/git/etube.git Sep 01 19:57:53 elementary youtube client (stub, but it quite works) Sep 01 19:58:01 or used too :P Sep 01 19:58:35 efl? Sep 01 20:03:44 pespin: E is too bleeding edge for me Sep 01 20:04:48 tmzt, yep Sep 01 20:05:29 guess clutter will never have 2d gpu/2.5d backend Sep 01 20:05:29 lindi-, then kick their asses to fix bugs and rest a bit! :P Sep 01 20:05:40 pespin: fix E? Sep 01 20:05:54 pespin: I think I'll rather stick to known and working solutions :) Sep 01 20:05:56 lol Sep 01 20:07:48 pespin: does etube have a bug tracker? Sep 01 20:08:49 lindi-, it doesn't have anything, it's just a stub from TreviƱo :P Sep 01 20:09:15 anything I know of, of course Sep 01 20:09:19 pespin: 1) there are no build instructions 2) some files lack licensing information 3) parsing HTML, HTTP _and_ RSS in C? that is scary and buggy Sep 01 20:09:37 pespin: 4) hardcoded #define TMP_PATH "/tmp/.etube" is a security problem Sep 01 20:10:09 pespin: 5) hardcoded "mplayer -fs -dr -vo xv -vf framestep=2 " "\"http://3v1n0.net/youtube-full-dl/index.php?download=flv&ytv=%s\"", vpage->video->id); sounds like a privacy issue and should at least be documented Sep 01 20:10:13 as I said, it's just a stub (which mostly works) Sep 01 20:10:46 pespin: I assumed you wanted feedback so here it is :) Sep 01 20:11:10 lindi-: why do that ? youtube-dl just extracts from the youtube page Sep 01 20:11:25 tmzt: do what? Sep 01 20:12:47 lindi-: using a third party youtube url resolver Sep 01 20:13:20 tmzt: exactly, that's why I reported it as a bug using this irc channel :) Sep 01 20:14:13 pespin: can you forward the bug reports 1-5? Sep 01 20:15:08 parsing in C look ok btw. why not? fast and reliable. Sep 01 20:15:24 gena2x: but it only implements a small corner of the http spec Sep 01 20:17:21 lindi-, uhm I don't know the author. I may send him a mail? Though as i said, to me this is only a kind of stuff and it's not intented for user use right now Sep 01 20:17:28 pespin: 6) request->target = (char *)malloc(2); request->target = "/\0"; leaks memory Sep 01 20:17:39 pespin: calling free() on that also probably crashes Sep 01 20:17:52 I'll send him a mail with a log of this :) Sep 01 20:17:57 pespin: I can mail him Sep 01 20:18:11 pespin: but if you can do it, please cc timo.lindfors@iki.fi Sep 01 20:31:40 lindi-, sent :) Sep 01 20:54:12 pespin: replied Sep 01 20:54:22 hi, Sep 01 20:54:34 DocScrutinizer, told me that there was a kickoff project about SHR on n900 Sep 01 21:04:54 mickey|zzZZzz, hi Sep 01 21:04:57 <[Rui]> well, I think it works as well with /dev/input/event3 as with /dev/input/event4 Sep 01 21:06:48 GNUtoo|laptop: n900 doesn't use audio routing exactly, it's a binary protocol either over serial or usb to the modem, but the audio packets pass through the cpu and echo cancellation is done on the cpu Sep 01 21:08:37 tmzt, that's waht I thought after reading the driver so my questions are: Sep 01 21:09:19 ARM<->MODEM is SSI Sep 01 21:09:39 so now...usually SSI are connected to CODECS Sep 01 21:09:53 what do we send to the SSI then? Sep 01 21:11:24 you'll have to ask nokia people for specifics, but I know the audio is handled by patches to pulseaudio Sep 01 21:11:35 which might be enough to set up a call Sep 01 21:11:48 I think it's a netdev but I'm not sure Sep 01 21:11:58 basically cmt_speech.c is a pass-trough Sep 01 21:12:37 and I bet their pulseaudio plugin is not free software Sep 01 21:16:21 I don't know, talk to Stskeeps in #mer about that Sep 01 21:16:29 it's just aec Sep 01 21:18:22 aec? Sep 01 21:19:59 aes means audio eccho cancellation? Sep 01 21:20:30 acoustic? Sep 01 21:21:20 ok Sep 01 21:21:32 acoustic on msm is the library to program adsp for audio Sep 01 22:36:59 http://meego.gitorious.org/maemo-multimedia/pulseaudio-modules-meego/trees/master/src/common Sep 01 22:37:35 might help Sep 01 22:38:46 http://meego.gitorious.org/maemo-multimedia/pulseaudio-modules-meego/trees/master/src/voice actually Sep 01 22:39:09 I think he left, but I'll tell him Sep 01 22:40:07 http://meego.gitorious.org/maemo-multimedia/pulseaudio-modules-meego/trees/master/src/cmtspeech Sep 01 22:48:09 668 * Sends an UL frame using SSI audio interface 'sal'. Sep 01 22:48:21 http://meego.gitorious.org/maemo-multimedia/pulseaudio-modules-meego/blobs/master/src/cmtspeech/cmtspeech-connection.c Sep 01 22:49:43 he's not in the channel **** ENDING LOGGING AT Thu Sep 02 02:59:58 2010