**** BEGIN LOGGING AT Wed Jul 27 02:59:57 2011 Jul 27 06:56:47 heh Jul 27 06:56:49 moin Jul 27 07:00:21 moin Jul 27 07:12:52 JaMa|Off: time to give oe-core another try :-) Jul 27 07:13:06 good Jul 27 07:15:14 * mrmoku will start with a gta02 lite image and see if it works Jul 27 07:15:48 wait a sec for my next push Jul 27 07:15:54 ok :) Jul 27 07:20:51 pushed Jul 27 08:59:41 SHR: 03Martin.Jansa 07meta-smartphone * rcd880e03ea51 10/meta-shr/conf/distro/shr.conf: shr: drop DISTRO_SSH_DAEMON as here we're using IMAGE_FEATURES for that Jul 27 08:59:41 SHR: 03Martin.Jansa 07meta-smartphone * rab466c6e9576 10/meta-shr/recipes-shr/images/shr-image.inc: shr-image: move core-image import to keep POKY_BASE_INSTALL in IMAGE_INSTALL Jul 27 08:59:41 SHR: 03Martin.Jansa 07meta-smartphone * r3b07fac475d4 10/meta-shr/conf/distro/shr.conf: SHR: use initscripts-shr as initscripts provider, maybe convert it later to initscripts_1.0.bbappend Jul 27 08:59:41 SHR: 03Martin.Jansa 07meta-smartphone * r78428b1343aa 10/meta-shr/recipes-shr/tasks/task-shr.bb: task-shr: reenable eve Jul 27 08:59:52 SHR: 03andrea.adami 07meta-smartphone * r6e38c776a17f 10/meta-zaurus/ (102 files in 40 dirs): meta-zaurus: apply changes from Andrea Adami Jul 27 12:40:54 hah, another vacation next week :) Jul 27 12:46:20 JaMa: yo :) Jul 27 12:52:33 JaMa: NOTE: Running task 3710 of 6597 Jul 27 12:52:37 all fine so far :) Jul 27 12:53:02 good Jul 27 12:53:02 SHR: 03Martin.Jansa 07meta-smartphone * r920468633f93 10/meta-zaurus/recipes-bsp/klibc/klcc-cross_1.5.23.bb: klcc-cross: use TUNE_PKGARCH instead of BASE_PACKAGE_ARCH Jul 27 13:01:40 Hi everyone Jul 27 13:02:25 SHR: 03Martin.Jansa 07meta-smartphone * rf682c8a85c00 10/meta-zaurus/recipes-bsp/zaurus-updater/zaurus-updater.bb: zaurus-updater: update LIC_FILES_CHKSUM so calling sed in do_configure won't break checksum Jul 27 13:05:54 SHR: 03Martin.Jansa 07meta-smartphone * r39207e20c6ce 10/meta-zaurus/recipes-bsp/zaurus-legacy-tar/zaurus-legacy-tar.bb: zaurus-legacy-tar: fix SRC_URI and add checksums Jul 27 14:52:34 JaMa: is that a problem on meego git or with the fetcher? Jul 27 14:52:36 ERROR: Function 'Fetcher failure for URL: 'git://git.gitorious.org/meego-cellular/libcmtspeechdata.git;protocol=http'. Unable to fetch URL git://git.gitorious.org/meego-cellular/libcmtspeechdata.git;protocol=http from any source.' failed Jul 27 17:00:51 mrmoku: I'll try to redownload Jul 27 17:04:08 mrmoku: redownloaded just fine now Jul 27 17:04:46 JaMa: hmm... maybe it was just a bad moment... retrying Jul 27 17:06:02 NOTE: Running task 7192 of 7497 on buildhost :) Jul 27 17:11:39 JaMa: yup, looks like I just hit a bad moment in time Jul 27 17:11:41 works now Jul 27 17:25:41 GNUtoo|laptop: hey, I'm building a shr-core image :) Jul 27 17:27:56 me too Jul 27 17:28:04 | svn: REPORT of '/svn/e/!svn/vcc/default': Could not read response body: connection was closed by server (http://svn.enlightenment.org) Jul 27 17:28:10 I've that however Jul 27 17:28:15 I guess it's my connection Jul 27 17:28:28 btw harald was right about samsung Jul 27 17:28:32 they seem great Jul 27 17:28:38 I've no samsung yet Jul 27 17:28:46 but someone who is a dev has one Jul 27 17:28:53 paulk on #replicant Jul 27 17:29:16 we were even told about a modem lib Jul 27 17:29:27 (a free modem lib) Jul 27 17:29:30 for samsung h1 Jul 27 17:29:38 yo Jul 27 17:29:46 where's [RUI] Jul 27 17:29:48 he has one Jul 27 17:30:01 the GPS seem connected to the main CPU Jul 27 17:30:10 the modem seem to be a network protocol Jul 27 17:30:27 altough the nexus S is a bit different from the H1 on that point Jul 27 17:30:29 there is alsa Jul 27 17:30:31 there is v4l2 Jul 27 17:30:37 both seem standard Jul 27 17:30:43 standard sound/soc Jul 27 17:31:03 but we need to finish the n900 first Jul 27 17:31:43 bbiab Jul 27 17:35:46 GNUtoo|laptop: N900? who cares about legacy hardware?... :-P Jul 27 17:36:45 I even care about gta02 Jul 27 17:36:54 gta02 is the freest phone Jul 27 17:36:54 *GPS attached to the CPU Jul 27 17:36:54 *audio attached to the CPU(the CODEC) Jul 27 17:37:07 *the modem can't access the CPU's RAM Jul 27 17:37:24 *free bootloader Jul 27 17:37:24 *nearly everything works: Jul 27 17:37:24 **xrandr -o 1 works Jul 27 17:37:49 **xrandr can even change resolutions Jul 27 17:37:50 ** the programs are obtimized for it Jul 27 17:37:50 etc... Jul 27 17:38:16 :) Jul 27 17:43:28 sorry connection issues Jul 27 17:43:34 (vpn etc...) Jul 27 17:43:40 the wifi AP is far away Jul 27 18:17:01 PaulFertser: ping Jul 27 18:17:19 PaulFertser: do you know a terminal that supports selecting text with the keyboard Jul 27 18:17:22 ? Jul 27 18:17:39 PaulFertser: I know I can do that with screen... but in case I would like to do it without - is there one? Jul 27 18:26:42 JaMa: dfu-util-native failed with the following linking error: http://pastebin.com/2D3tqsxL Jul 27 18:26:43 moo Jul 27 18:26:54 macht die Kuh :-) Jul 27 18:27:03 ~moo Jul 27 18:27:05 * apt mooooooooo! I am cow, hear me moo, I weigh twice as much as you. I am cow, eating grass, methane gas comes out my ass Jul 27 18:27:10 :-D Jul 27 18:27:18 DocScrutinizer: do you know such an terminal? with keyboard text selection? Jul 27 18:27:48 err, like "hold shift, move cursor, to highlight"? no Jul 27 18:28:39 yup Jul 27 18:29:00 can't work with proper terminal. See line discipline iirc Jul 27 18:29:36 all key codes are passed to the shell rsp the process shell has spawned Jul 27 18:29:47 hmm Jul 27 18:30:00 that's why ctrl+c doesn't copy in xterm Jul 27 18:30:20 and that's also why shift+cursor won't highlight text Jul 27 18:30:33 unless your shell supports it Jul 27 18:31:17 ok, have to accept that then :) Jul 27 18:31:50 on xterm you have the benefit of shell not aware of any concept of a pointer device, so xterm can use mouse to highlight text Jul 27 18:32:17 which usually "just works" Jul 27 18:33:03 to do same via kbd you need to implement it to shell, editor, whatever-your-program Jul 27 18:33:33 e.g Qt text edit boxes have it implemented Jul 27 18:33:56 same for GTK I bet Jul 27 18:34:51 DocScrutinizer: xfce terminal has Ctrl+Shift+A to select the complete buffer Jul 27 18:35:14 mrmoku: I have work around for that and GNUtoo was working on dfu-util-native_0.4.bb to fix it properly Jul 27 18:35:17 mhm, that's implemented somewhere and intercepts the keycode Jul 27 18:35:30 JaMa: ok, I'll wait for the real fix :) Jul 27 18:35:57 DocScrutinizer: yeah, but the same _could_ be done to switch to a 'select from the buffer with keys' mode Jul 27 18:36:00 kbd events are passed thru from one handler to next handler until one handler decides its "end of chain" and doesn't pass the event to next one Jul 27 18:36:19 just like screen does it Jul 27 18:36:35 mrmoku: yes, it can be done, no problem Jul 27 18:36:44 just it usually isn't Jul 27 18:36:46 mrmoku: sed -i 's/-lusb/-lusb -lusb-1.0 -lpthread -lusbpath -lrt/g' ../tmp/sysroots/x86_64-linux/usr/lib/pkgconfig/libusb.pc Jul 27 18:36:58 mrmoku: bitbake -c cleansstate dfu-util-native; bitbake dfu-util-native Jul 27 18:37:42 JaMa: ok Jul 27 18:37:45 thx Jul 27 18:38:11 mrmoku: and of course when you teach xterm to intercept and act on e.g ctrl+C for "copy" then this key combo isn't available for shell anymore Jul 27 18:38:13 DocScrutinizer: yeah, and I'm looking for the unusual guy who did that unusual terminal which does it :) Jul 27 18:38:35 DocScrutinizer: the problem is not copying... the problem is selecting text with the keyboard Jul 27 18:38:46 shortcuts for copying they do have all Jul 27 18:39:18 quite useless though if you have to select the text with the mouse prior to copying... as then it's already in one of x's clipboards Jul 27 18:39:20 holding shift+(usual key combo) *often* works to escape the lower level meaning and make one level up intercept the code as if it was without shift Jul 27 18:39:58 check mc for that Jul 27 18:40:03 shift+up scrolls linewise back in the buffer for the terminals i tried Jul 27 18:40:29 brb Jul 27 18:40:35 mc is a xterm program, yet it cares about mouse events. So if you want to highlight text in mc, you have to hold shift Jul 27 18:41:59 so even in shell shift+ctrl+c might copy the highlighted text Jul 27 18:42:18 shift+ctrl+a selects all text Jul 27 18:42:20 etc Jul 27 18:43:08 I dunno if unbinding keycode for e.g. shift-left in shell will free the code for usage in xterm Jul 27 18:44:16 or if you explicitly need to define that code as a relevant one to intercept, in xterm config/sourcecode, so xterm intercepts and stops the event before it even reaches shell Jul 27 18:44:56 for sure it's feasible, I just don't know if it actually is already implemented, and if so then how to enable it Jul 27 18:45:12 bbl Jul 27 18:47:11 xterm is so 90s :) Jul 27 18:50:55 says someone using irc ;) Jul 27 18:52:09 I've been away for a couple weeks. Any news on the issue where using the accelerometers in SHR-U on gta02 freezes? Jul 27 18:59:12 mrmoku: last tune overhaul will cause another rebuild unless we try hard enough :/ Jul 27 18:59:27 JaMa: hehe, ok Jul 27 18:59:45 SHR: 03Martin.Jansa 07meta-smartphone * r6c6b099cceff 10/meta-shr/conf/distro/shr.conf: SHR: use PKGARCHCOMPAT_ARMV7A as Angstrom does Jul 27 18:59:51 SHR: 03Martin.Jansa 07meta-smartphone * rbadb57b7c4e7 10/meta-shr/recipes-shr/tasks/task-shr-minimal.bb: task-shr-minimal: add hires-icon-theme to make postinst in gdk-pixbuf-loaders happy Jul 27 19:00:02 mrmoku: http://git.shr-project.org/git/?p=meta-smartphone.git;a=commit;h=6c6b099cceff98796ac5a8e84b4d5fbcce9be3c6 Jul 27 19:00:18 lindi-: I'm a 69er... so xterm is rocket science for me :-P Jul 27 19:00:57 JaMa: ah, ok Jul 27 19:01:37 mrmoku: heh Jul 27 19:01:47 mrmoku: not sure if it's better to maintain backward compatibility or just switch it now before more people are using oe-core based shr Jul 27 19:03:00 JaMa: for me a rebuild would be fine too Jul 27 19:03:13 good Jul 27 19:03:56 so I'll push updated oe-core-contrib and start rebuild Jul 27 19:04:02 before I leave for vacation Jul 27 19:04:49 ok Jul 27 19:11:34 SHR: 03Martin.Jansa 07meta-smartphone * r935e74f49d1e 10/meta-shr/recipes-shr/tasks/task-shr-minimal.bb: task-shr-minimal: add hires-icon-theme to make postinst in gdk-pixbuf-loaders happy Jul 27 19:41:34 jeepingben: using accelerometers I'd say "freeze" aka block is the right behaviour on reading in new values, depending on the thresholds you've set for the lis302 IRQ engine Jul 27 19:42:35 jeepingben: you don't usually want to read in raw values at a random sampling rate, just to find nothing happened Jul 27 19:45:25 hte correct way to handle lis302 is: set up thresholds and highpass-filter so any movement of device that is large/fast enough to matter for your usecase is triggering an IRQ, then block on read of accelerometer values, and when the read returns from blocking you know a movement happened Jul 27 19:47:07 jeepingben: odds are that when you failed on setting up thresholds, filter, and other config inside lis302 chip correctly, you'll never see an IRQ getting fired by the chip and your read will block forever Jul 27 19:49:52 there are very silly concepts of reading accelerometer with IRQ engine configured to shot an IRQ on data-ready which means it will sample on a constant rate of either 100 or 400Hz iirc. Quite usually that is NOT what you want to use for a embedded mobile device, as it keeps CPU busy all the time Jul 27 19:53:04 Alas iirc "our" gta02 driver (as well as some upstream drivers) does not support configuration of highpass filter, which makes handling quite awkward as you need to read out current absolute values and then set your thresholds accordingly, and to make matters worse you need to redo that every single time the accelerometer fired Jul 27 19:58:32 DocScrutinizer: he is talking about system freezes I think Jul 27 21:10:49 that'd be a rather nasty kernel bug then Jul 27 21:12:12 could only be a Ooops or Panic though Jul 27 21:12:42 or some extremely nasty collision on I2C bus **** ENDING LOGGING AT Thu Jul 28 02:59:57 2011