**** BEGIN LOGGING AT Sun Sep 13 02:59:56 2009 Sep 13 07:28:47 Ainulindale, could you approve the Mails from JaMa to shr-devel? Need approval due to 40kb limit... Sep 13 08:34:52 hi! any gta01 users around? did anybody successfully try out schwartz? Sep 13 08:36:10 balrog-kun wrote it but his build system remains totally obfuscated to me - even after he tried to explain it to me.. Sep 13 08:36:51 git clone http://folks.o-hand.com/andrew/schwartz.git/ Sep 13 08:38:41 if someone happens to have some freetime to look into it then let me know of the probably stupid catch that i missed to make it run Sep 13 09:24:46 cz_jc: man Sep 13 09:24:59 cz_jc: there're decent docs on the wiki "Manually using bluetooth" Sep 13 09:25:08 cz_jc: pand daemon is not required anymore. Sep 13 09:26:21 rtp: do you mean you're the author of the s3c usb device driver? Sep 13 09:30:23 rtp: if yes, sorry, i didn't realise it. Also i thought that nobody cares much about that driver since there was nobody willing to debug and fix windows bsod problems. Also i understand the difference between decent bugreport and a bad one, since i didn't spend enough time to produce a decent one, i decided to wait a little. Sep 13 09:35:11 larsc: i'm back Sep 13 09:52:54 PaulFertser: :-) Sep 13 09:54:20 DocScrutinizer-8: hey :D Sep 13 10:03:05 DocScrutinizer-8: btw, fyi atheros (at least Luis) does not forget about us, there're some internal "talks" ongoing (i don't know anything more about it, just the fact). Sep 13 10:03:22 SHR: 03seba.dos1 07geekogotchi * rc6b914aff7c9 10/geekogotchi_server: server: implement CleanPets method Sep 13 10:03:22 SHR: 03seba.dos1 07geekogotchi * r9a1382c70154 10/geekogotchi_client: client: add getPets function Sep 13 10:03:22 SHR: 03seba.dos1 07geekogotchi * r77dc6891e62f 10/geekogotchi_client: client: add cleanPets Sep 13 10:03:23 SHR: 03seba.dos1 07geekogotchi * rccddee349ebc 10/geekogotchi_server: server: rename Delete method to DeletePet Sep 13 10:03:26 SHR: 03seba.dos1 07geekogotchi * r95c503a817c0 10/geekogotchi_server: server: update pet state on resume from suspend Sep 13 10:03:31 SHR: 03seba.dos1 07geekogotchi * ra0231913e07d 10/geekogotchi_client: client: add killPet and deletePet functions Sep 13 10:04:07 PaulFertser: great news! Sep 13 10:04:46 DocScrutinizer-8: page 12 of AAT1275.pdf has a list of current limiting resistor values. Sep 13 10:05:16 PaulFertser: yes, I got an obsolete doc missing this Sep 13 10:05:32 DocScrutinizer-8: i just download from the internet the first i found. Sep 13 10:05:55 I had it on my PC ;-) Sep 13 10:06:14 rsync varaha ;-) Sep 13 10:06:34 test Sep 13 10:06:34 s/es/os/ Sep 13 10:06:34 dos1|neo meant: tost Sep 13 10:08:14 PaulFertser: maybe you also find a datasheet for the fsckng inductor choke Sep 13 10:08:58 DocScrutinizer-8: maybe, haven't tried yet. Sep 13 10:09:22 I've tried -- failed :-/ Sep 13 10:09:47 We should ask one of our chinese speaking friends probably. Sep 13 10:13:27 PaulFertser: the inductor is rated 800mA ~8MHz. This *might* mean you need 8*10^6 energy quantums of a fully saturated ferrite core to get an average current of 800mA. AAT is working at 2MHz iirc Sep 13 10:14:17 s/ums/ums per sec/ Sep 13 10:14:18 DocScrutinizer-8 meant: PaulFertser: the inductor is rated 800mA ~8MHz. This *might* mean you need 8*10^6 energy quantums per sec of a fully saturated ferrite core to get an average current of 800mA. AAT is working at 2MHz iirc Sep 13 10:15:16 Hi! I just flashed , but my WLAN/wifi is not working. I never get it to work. What shall I do, to connect to the internet using WLAN? Sep 13 10:15:25 ergo the choke might get saturated at >200mA Sep 13 10:15:57 Mathias__: what kernel? Sep 13 10:16:23 I think . Sep 13 10:17:28 Mathias__: 2d158aae9d8d36f575 won't do. you need at least 8851834038aec76c074 Sep 13 10:18:24 rumor has it you always should use matching kernel for your rootfs Sep 13 10:18:36 lindi-: There is no 8851834038aec76c074 on . Sep 13 10:18:56 Mathias__: there i can't help you much sorry, i don't use shr Sep 13 10:19:29 Mathias__: update your kernel to the latest in the repos! Sep 13 10:19:40 DocScrutinizer-8: right, it's working at 2MHz Sep 13 10:19:46 Mathias__: are you familiar with git? Sep 13 10:19:51 PaulFertser: via opkg update && opkg upgrade? Sep 13 10:19:58 lindi-: a bit Sep 13 10:20:00 Mathias__: and then probably you'll need to flash the kernel. Sep 13 10:20:22 Is it possible just to download the new kernel and flash it? Sep 13 10:20:32 Mathias__: then you can easily check if any of those kernels is later than 8851834038aec76c074? Sep 13 10:21:01 Mathias__: yes Sep 13 10:21:01 Mathias__: there IS the matching kernel in unstable Sep 13 10:21:24 lindi-: not recommended Sep 13 10:21:53 lindi-: rumor has it you always should use matching kernel for your rootfs Sep 13 10:22:04 Maybe it is . I flash it now. Sep 13 10:22:42 Mathias__: watch the file creation dates Sep 13 10:22:54 Mathias__: a3587e4ed77974a is older than 8851834038aec76c07 Sep 13 10:23:09 a3587e4ed77974a is from 2009-09-01. Sep 13 10:23:21 Mathias__: nope Sep 13 10:23:24 Mathias__: usually use most recent files (a lot are symlinks, so no worries which of those to use) Sep 13 10:23:24 Mathias__: 'commit a3587e4ed77974adfb057af261aaeea4022018e8 Author: Paul Fertser Date: Thu Aug 27 12:19:12 2009 +0400' Sep 13 10:23:54 Mathias__: a3587e4ed77974a is from 2009-08-27 Sep 13 10:24:04 lindi-: rumor has it you always should use matching kernel for your rootfs Sep 13 10:24:12 lindi-: not recommended Sep 13 10:24:22 DocScrutinizer-8: for booting, yes Sep 13 10:24:22 DocScrutinizer-8: but using a kernel that does not have the wlan fix won't work Sep 13 10:24:46 DocScrutinizer-8: but after one did opkg update and upgrade kernel, it'll fetch corresponding modules and you can flash the kernel it installed to /boot. Sep 13 10:24:54 s/can/must/ probably even Sep 13 10:25:05 If I use the latest files from I have a3587e4ed77974adfb057af261aaeea4022018e8. Sep 13 10:25:18 Mathias__: yes, it's ok, it has the fix Sep 13 10:25:23 PaulFertser: really? Sep 13 10:25:23 "fix" Sep 13 10:25:33 I just try it. Sep 13 10:25:38 lindi-: a35 has that fix, right, and it's the latest kernel SHR provides. Sep 13 10:25:46 PaulFertser: I see that the fix was added only three days later than that? Sep 13 10:26:05 Mathias__: but flashing the kernel doesn't make much sense since you need ar6000 module of the same revision. Sep 13 10:26:14 PaulFertser: ? Sep 13 10:26:16 lindi-: commit date 27 Aug, build date 1 Sep Sep 13 10:26:34 Mathias__: use the kernel and modules that match. All of them should be a35... Sep 13 10:26:34 PaulFertser: but 'Revert "AR6000: move low-level cleanup from ar6000_destroy to ar6000_close"' is from 30 Aug Sep 13 10:26:55 lindi-: and a35 build date is 1 Sep or so. 1Sep>30Aug Sep 13 10:27:13 waaah, we have an image of 20090906 at U Sep 13 10:27:14 PaulFertser: but why does build date matter? isn't the commit date only thing that matters here? Sep 13 10:27:45 PaulFertser: I thought, I can just download and flash the most recent files from . Am I wrong? Sep 13 10:27:49 PaulFertser: or have they backported the fix? Sep 13 10:27:57 lindi-: hm, shr-unstable "auto-revs" kernel from andy-tracking, so if the build date is > commit date it definetely includes the commit. Sep 13 10:28:18 PaulFertser: uh. so the git revision in the version number is just bogus? Sep 13 10:28:21 Mathias__: you need to opkg update && opkg upgrade to get the latest fixes and/or bugs. Sep 13 10:28:29 PaulFertser: ok Sep 13 10:28:35 yes Sep 13 10:29:13 lindi-: a3587e4ed77974adfb057af261aaeea4022018e8 is the commit where the wlan was "fixed" and it's the same string included in the latest kernel available from the shr repos, no? Sep 13 10:29:44 PaulFertser: no, a3587e4ed77974adfb057af261aaeea4022018e8 is about gta01 battery stuff Sep 13 10:30:13 lindi-: ah, yes. But it came after the wifi "fix". Sep 13 10:30:45 PaulFertser: nope? a3587e4ed77974adfb057af261aaeea4022018e8 is from 27 Aug, and wifi fix 8851834038aec76c07454f3620fa9fb1506e7217 is from 30 Aug Sep 13 10:30:57 or was there another wifi fix? Sep 13 10:31:09 * DocScrutinizer-8 mumbles swearwords git %#54\" Sep 13 10:31:33 lindi-: a35 was committed on 1 Sep 2009 20:50:41 Sep 13 10:31:53 lindi-: 885 was committed on 1 Sep 2009 20:22:23 Sep 13 10:32:06 lindi-: you confuse commit date and creation date Sep 13 10:32:25 PaulFertser: hmm, not according to my git: http://paste.debian.net/46420/ Sep 13 10:32:38 PaulFertser: ah, how do I see commit date? Sep 13 10:33:45 jme mumbles more swearwords even LOUDER Sep 13 10:34:07 lindi-: git show --pretty=fuller Sep 13 10:34:23 DocScrutinizer-8: git is great. But every powerful tool takes time to master. Sep 13 10:34:26 ;) Sep 13 10:34:33 With a3587e4ed77974adfb057af261aaeea4022018e8 my wifi is working. Sep 13 10:34:34 Thx! Sep 13 10:34:44 PaulFertser: thanks! Sep 13 10:34:46 lindi-: also you could have just took a look at "git log origin/andy-tracking" Sep 13 10:34:48 well, we have OE, bb, opkg. of course we need git to make the mess complete XD Sep 13 10:34:55 lindi-: to see what came after what. Sep 13 10:35:06 DocScrutinizer-8: did you prefer monotone back in times? Sep 13 10:35:17 waaaAAAAAHHH Sep 13 10:35:20 Mathias__: do you use some GUI/ Sep 13 10:35:31 PaulFertser: Mokonnect Sep 13 10:35:37 Mathias__: oh, it works? Interesting. Sep 13 10:35:40 PaulFertser: indeed. git log origin/andy-tracking | grep -E "(a3587e4ed77974adfb057af261aaeea4022018e8|8851834038aec76c07454f3620)" Sep 13 10:35:53 PaulFertser: At least it finds the WLAN AP. I try to connect. Sep 13 10:36:12 Mathias__: if it fails, go and read FSO_Resources wiki page at OM wiki. Sep 13 10:36:12 hehehe Sep 13 10:36:37 Mathias__: prepare for reboot ;-) Sep 13 10:37:58 PaulFertser: i hadn't even thought about the difference of commit/creation date before Sep 13 10:37:59 PaulFertser: isn't it funny the panic takes 30sec on my device, during which device is solidly frozen, after wlan wep connect Sep 13 10:41:31 Wifi is working using WPA! :-D Sep 13 10:42:58 as here ;) WPA and plain wpa_supplicant Sep 13 10:43:06 hmm who could depend on this? ERROR: Cannot satisfy the following dependencies for task-shr-apps: libwebkit-1.0-2 (>= 1.1.4) * Sep 13 10:43:22 Mathias__: never use WEP XD Sep 13 10:43:22 there is just Packaged contents of webkit-gtk into /tmp/gta/deploy/glibc/ipk/armv4t/libwebkit-1.0-1_0.1+svnr40814-r10_armv4t.ipk Sep 13 10:43:35 SHR: 03gmary 07shr-installer * rea8c0506d6e6 10/ (4 files in 3 dirs): Add french translation Sep 13 10:44:20 DocScrutinizer-8: Only if there is no WPA/WPA2. Sep 13 10:44:47 won't work nevertheless Sep 13 10:45:03 kernel panic here Sep 13 10:45:25 DocScrutinizer-8: hmm? WEP works in my school Sep 13 10:45:59 ooh, so it's the first report of WEP working (with mokonnect) Sep 13 10:46:02 i have really poor signal even close to AP (a bit better with maxperf), but it works Sep 13 10:46:18 maybe accusing WEP was wrong then Sep 13 10:46:24 The signal quality of the wifi chip is quite bad. :-| Sep 13 10:46:25 DocScrutinizer-8: not mokonnect, just wpa_supplicant Sep 13 10:47:01 Mathias__: are you sure? i think reported quality is just bogus Sep 13 10:47:22 it's about 30% on perfect signal strenght Sep 13 10:47:48 dos1|neo: I just walk around the flat. And my notebook's wifi is much better. Sep 13 10:47:53 still the chip sensitivity is poor Sep 13 10:48:39 my fr has chip quality similar to by old notebook Sep 13 10:48:51 but my new notebook beats them both ;p Sep 13 10:49:04 s/by/my/ Sep 13 10:50:38 I had a Orinoco Gold wlan-card, that had same poor sensitivity and same bogus S/N+Q readings ;-) Sep 13 10:51:14 3 years ago :-D Sep 13 10:54:59 prism-II chipset which can run hardmac as well as softmac :-) Sep 13 10:55:10 iirc Sep 13 10:56:00 ooh, no. sorry. Think that was siemens usb-wlan54 dongle Sep 13 11:00:29 PaulFertser: wpwrak: http://lekernel.net/prism54/mustumbler.html Sep 13 11:04:11 DocScrutinizer: interesting stuff Sep 13 11:04:30 PaulFertser: yes, isn't it :-D Sep 13 11:04:42 Too bad it doesn't work ;) Sep 13 11:05:13 PaulFertser: even worse the note about missing freemac sw is there for >2y now Sep 13 11:06:40 anyway I thought you might want to have a look into the main page (which moved somewhere else recently) as it contains a lot of details about softmac Sep 13 11:07:36 it's not exactly about ar6k, but nevertheless.... Sep 13 11:13:00 DocScrutinizer: neat :) Sep 13 11:15:16 wpwrak: most interesting is the approach of the manufacturers to ship devices like that Siemens wifi-dongle with a softmac though the hw is designed for hardmac and actually contains a hardmac/fullmac firmware Sep 13 11:16:03 DocScrutinizer: is the softmac on linux only or do they use a softmac also on windows ? Sep 13 11:16:13 I last read those pages some 2 years ago, so excuse me for being vague/incorrect in some points Sep 13 11:16:27 windows only afaik Sep 13 11:16:53 this project was about creating a freemac softmac for linux for those cards Sep 13 11:18:00 DocScrutinizer: most probably it's not what we're talking about: all modern softmac drivers should use mac80211 in-kernel implementation. Sep 13 11:18:58 wpwrak: looks like they don't provide linux drivers at all, and with those devices that can do hardmac, prism54 kernel driver works well. Sep 13 11:19:32 DocScrutinizer: (windows) interesting indeed. i guess softmacs are just easier to get right than these nasty device-specific hardmacs :) Sep 13 11:20:04 actually I had to use ndiswrapper to put that dongle to operation Sep 13 11:20:31 This has more recent info: http://wireless.kernel.org/en/users/Drivers/p54 Sep 13 11:21:11 yes, it moved recently, as I told Sep 13 11:21:23 * pabs3 finds http://wireless.kernel.org/en/users/Drivers/ar9170.fw and is amazed Sep 13 11:21:29 that's the site it moved to Sep 13 11:22:55 PaulFertser: of course that's not what we are dealing with for ar6k. It's just interesting background about how hardmac is implemented Sep 13 11:23:38 DocScrutinizer: yes, and it seems that recent drivers implement softmac in a proper way. Sep 13 11:23:53 PaulFertser: I'm also aware mac80211 should support hese chips now Sep 13 11:24:44 DocScrutinizer: and it does, judging by the page. Hey, if atheros gives us the ability to do softmac, who'll write the driver? ;) Sep 13 11:24:46 PaulFertser: 3 years ago it didn't Sep 13 11:27:07 Of course the best we can hope for is a way to update firmware to the latest release version :( Sep 13 11:27:43 first we would need a "latest release version" Sep 13 11:28:05 aiui we got a more recent version than officially available Sep 13 11:31:05 DocScrutinizer-8: there's nothing officially available because there's no way to update the firmware currently for end-users. Sep 13 11:32:09 yesyes. wpwrak told sth like that iirc. "official" fw is 1.x, we got 2.0 Sep 13 11:37:02 if atheros can GPL ar9170.fw, perhaps they could be convinced to do the same for the ar6k firmware? Sep 13 11:42:57 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * rb9a9cfe8556e 10framework/framework/subsystems/opimd/pimd_generic.py: opimd: GenericDomain: move new get_full_content function out of comment ;x Sep 13 11:46:25 pabs3: hard to tell. harald gave them a ping a little while ago, and there has been some indication that this won't get completely ignored, but we'll have to see what, if anything, crawls out of it Sep 13 11:51:41 wpwrak: how goes the tale about our fw version? Sep 13 11:58:47 DocScrutinizer: the 1.x version was broken beyond hope. atheros were making a 2.x version for some big customer. they let us have a development version of that one. dunno what happened with 2.x after that. Sep 13 12:00:47 * pabs3 wonders idly if a life-size gold statue of RMS would convince them to GPL it Sep 13 12:02:38 pabs3: if you can hit them on the head with it ... :) Sep 13 12:03:24 * DocScrutinizer-8 recalls pangalaktischer donnergurgler XD Sep 13 12:09:59 pabs3: so I suggest you wrap that statue with lemmon slices befor hitting their head with it ;-) Seems a proven good receipe Sep 13 12:12:44 morning folks Sep 13 12:12:52 hi mickeyl Sep 13 12:14:15 morning Sep 13 12:14:36 wpwrak: could we rip a 2.x fw version off some customer device then? Sep 13 12:15:04 wpwrak: if everything else fails Sep 13 12:16:01 wpwrak: assuming of course that version is more evolved than our development sample Sep 13 12:16:41 oh, that's why it fails on every assertion ;x (wifi fw) Sep 13 12:19:00 mickeyl: what about timestamp patch? it's quite long uncommited now ;/ Sep 13 12:19:21 dos1: yeah, sucks Sep 13 12:19:29 looks like the rest of FSO is in limbo :/ Sep 13 12:19:31 DocScrutinizer: no idea if the interface to the driver is the same Sep 13 12:20:12 dos1: (assertion) yep Sep 13 12:23:18 dos1: is the author here? Sep 13 12:31:03 mickeyl: dunno. i saw that patch on maillist, i don't know anything more that that ;x Sep 13 12:31:11 did anyone test the AT@POFF shutting down patch? Sep 13 12:31:58 PaulFertser: ^^ Sep 13 12:32:12 I bet he did Sep 13 12:33:39 freesmartphone.org: 03mickey 07framework * r24f52e8e00d0 10/framework/subsystems/ogsmd/gsm/convert.py: Sep 13 12:33:39 freesmartphone.org: ogsmd: fix SMS timezone being parsed incorrectly for places where GMT<0 Sep 13 12:33:39 freesmartphone.org: Fixes FSO #476. Patch courtesy "dkogan", thanks a lot! Sep 13 12:34:01 :) Sep 13 12:34:10 PaulFertser: ? Sep 13 12:34:44 is there a userland interface for getting an RTC irq? Sep 13 12:34:58 mickeyl: i'm testing now frameworkd a lot (fixing some issue in opimd), do you need also testing AT@POFF patch on gta02? Sep 13 12:35:04 mickeyl: rtcwake works so yes there is Sep 13 12:35:21 dos1: yeah, anywhere will do Sep 13 12:35:24 mickeyl: or what do you mean by "getting an RTC irq"? Sep 13 12:35:37 lindi-: being notified by the kernel that an RTC irq has occured Sep 13 12:35:49 some rtctest.c exists Sep 13 12:36:37 hmm Sep 13 12:36:45 mickeyl: yep in http://docs.openmoko.org/trac/ticket//2156 Sep 13 12:37:25 hmm Sep 13 12:37:29 i still don't see it Sep 13 12:37:39 mickeyl: attachment 1 Sep 13 12:37:42 how is the irq being told to userland? Sep 13 12:38:16 I'm not talking about suspend/resume Sep 13 12:38:16 mickeyl: select will return Sep 13 12:38:51 ah! Sep 13 12:38:59 so it comes via /dev/rtc Sep 13 12:39:27 * mickeyl trying Sep 13 12:39:33 mickeyl: http://pastebin.com/f4306f2e7. A teaser Sep 13 12:40:05 mickeyl: could you point me to that patch? Sep 13 12:41:03 DocScrutinizer: nice! more :) Sep 13 12:41:08 dos1: http://pastebin.ca/1564167 Sep 13 12:44:11 mickeyl: so what should i do to test? just request/release/request GSM resource? Sep 13 12:44:12 mickeyl: anything I should know about requirements that cause need for some ascenariolib.so? Sep 13 12:46:33 mickeyl: paul and me implemented a way to trigger a call to e.g. odeviced (or whatever deamon to play audio resource arbiter) that can be called exactly same hooks-way as above control-elems hook does for mixer controls Sep 13 12:47:21 dos1: yes, that should be enough Sep 13 12:47:51 dos1: verify that the modem really is off after release Sep 13 12:47:59 and check whether you can request it again afterwards Sep 13 12:48:11 mickeyl: but how to check it' Sep 13 12:48:14 +s really off? Sep 13 12:48:20 power consumption Sep 13 12:48:31 do you have measuring equipment attached to the neo? Sep 13 12:48:40 mickeyl: no Sep 13 12:48:40 dos1: call it? ;-) Sep 13 12:49:11 you don't need lab gear to probe gsm modem consumption Sep 13 12:49:20 CC will do perfectly Sep 13 12:49:43 on gta02 Sep 13 12:50:04 lindi-: yay! Sep 13 12:50:13 lindi-: 2009-09-13T12:49:44.939868Z Kernel26Rtc [INFO] : Setting RTC alarm to 2009-09-13 12:50:10 (dst=0) Sep 13 12:50:13 ** Message: plugin.vala:78: onInputEvent for RTC. source=8657072, condition=1 Sep 13 12:50:38 is it sending any meaningful info in the packet or just a 'ping' ? Sep 13 12:51:50 * mickeyl rips off mainloop notification Sep 13 12:52:45 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * r664e9533c5c9 10framework/framework/subsystems/opimd/ (docs/TODO pimb_sim_messages_fso.py pimd_generic.py): Merge commit 'origin/dos/opimd-tracking' into dos/opimd-tracking Sep 13 12:54:23 mickeyl, hi, the definitions for set_alarm seems to be out of sync in libfso-glib and the specs xml.. freesmartphone-time.vala's definition doesn't throw FreeSmartphone.Error... a regen of these files should fix it? Sep 13 12:55:44 Sup3rkiddo: yeah. actually you're supposed to rebuild libfso-glib directly from specs Sep 13 12:55:52 since i do not regenerate it every time Sep 13 12:55:55 you can do that with Sep 13 12:56:02 ./autogen.sh SPECS_PATH=/local/pkg/fso/specs/xml/ Sep 13 12:56:03 (e.g.) Sep 13 12:56:07 but i'll autogenerate it now Sep 13 12:56:32 mickeyl: note that the rtc clock drifts quite a lot Sep 13 12:57:29 lindi-: so you'd still recommend using system clock for appointments? Sep 13 12:58:19 mickeyl: my own solution is to compensate the rtc drift and then only use rtc time, never system time Sep 13 12:58:43 how do you compensate? Sep 13 12:58:48 mickeyl: since the rtc drift rate seems to be quite constant Sep 13 12:59:21 mickeyl: printf("%f\n", (16299/16298)*($ARGV[0] - 76283.33539)); converts from UTC to the RTC time :) Sep 13 12:59:34 aha Sep 13 12:59:40 mickeyl: then i just let rtc clock drift at its own rate where it wants and never _write_ to RTC Sep 13 12:59:44 mickeyl: standard meassure is to use hwclock --adjust Sep 13 12:59:58 but that standard measure is not very good imho Sep 13 13:00:19 lindi-: though you never made your point why Sep 13 13:00:31 DocScrutinizer-8: you can only set RTC clock at 1-second granularity Sep 13 13:00:44 so what? Sep 13 13:00:49 DocScrutinizer-8: if you set the RTC clock every time you suspend the device you can easily add 10 second error per day Sep 13 13:01:06 lindi-: that's incorrect Sep 13 13:01:24 DocScrutinizer-8: also, to schedule rtc alarms you need to anyway account for rtc drift Sep 13 13:01:53 error will average out, and hwclock doesn't adjust for too small skew amounts Sep 13 13:02:03 DocScrutinizer-8: but the system clock is not as accurate as the rtc clock Sep 13 13:02:11 exactly Sep 13 13:02:20 DocScrutinizer-8: so i think it's better to use rtc clock only and never set it Sep 13 13:03:09 and then just set_system_clock(compensate_for_drift(get_rtc_time())) as often as you like Sep 13 13:03:22 that's why you should do that RTC adjust only after correct timesource sync. However a hwclock --adjust is completely independent of sysclock Sep 13 13:03:49 DocScrutinizer-8: but I want accurate wakeups even when I'm not connected to internet for a week or use gps Sep 13 13:04:15 so what? Sep 13 13:04:18 freesmartphone.org: 03mickey 07specs * r6a6bdd50b19c 10/ (48 files in 2 dirs): regenerate Sep 13 13:04:18 freesmartphone.org: 03mickey 07specs * rcdb5e999fb15 10/org.freesmartphone.Device/org.freesmartphone.Device.RealtimeClock.xml.in: org.freesmartphone.Device.RealtimeClock: new signal org.freesmartphone.Device.RealtimeClock.Alarm(i) Sep 13 13:04:43 DocScrutinizer-8: how do you schedule your phone to wake up tomorrow at 2 pm? Sep 13 13:04:58 lindi-: hwclock --adjust is doing exactly what you suggest to do with own script Sep 13 13:05:19 DocScrutinizer-8: i think not Sep 13 13:05:39 DocScrutinizer-8: hwclock --adjust writes to RTC Sep 13 13:05:43 busybox... Sep 13 13:05:44 DocScrutinizer-8: I never write to RTC Sep 13 13:05:45 hwclock: unrecognized option `--adjust' Sep 13 13:05:46 :x Sep 13 13:07:22 lindi-: and you think that's a vlaid point in itself? I don't see any benefit from that Sep 13 13:07:29 freesmartphone.org: 03mickey 07libfso-glib * r7608c8d98bb6 10/src/ (6 files): regen Sep 13 13:07:44 DocScrutinizer-8: the benefit is that it is more accurate :) Sep 13 13:08:04 proof please Sep 13 13:08:18 I don't see that. Not at all Sep 13 13:08:21 DocScrutinizer-8: well, you can only use hwclock --adjust when you have external clock source, right? Sep 13 13:08:33 noooo Sep 13 13:08:38 that's why you should do that RTC adjust only after correct timesource sync. However a hwclock --adjust is completely independent of sysclock Sep 13 13:09:23 DocScrutinizer-8: ok, so imagine that for a month you have no external time source available Sep 13 13:09:27 s/adjust/systohc/ Sep 13 13:09:29 DocScrutinizer-8 meant: that's why you should do that RTC systohc only after correct timesource sync. However a hwclock --adjust is completely independent of sysclock Sep 13 13:10:40 DocScrutinizer-8: and your rtc will drift during that time. how do you set the rtc alarm then? Sep 13 13:11:47 lindi-: you call hwclock --adjust once a day, and that compensates for the evaluated drift. No matter if you have external timesource startum0 or not Sep 13 13:12:14 DocScrutinizer-8: but what if your phone is not awake every day? Sep 13 13:13:17 then shedule ONE even for 13:30 just to call hwclock --adjust, if you need a rtc wake at 14:00:00 Sep 13 13:13:21 freesmartphone.org: 03mickey 07cornucopia * r27c6d79ffc87 10/fsodeviced/src/plugins/kernel26_rtc/plugin.vala: Sep 13 13:13:21 freesmartphone.org: fsodeviced: kernel26_rtc: send new dbus signals Sep 13 13:13:21 freesmartphone.org: * org.freesmartphone.Device.RealtimeClock.WakeupTimeChanged(i) when the wakeup time has been changed Sep 13 13:13:22 freesmartphone.org: * org.freesmartphone.Device.RealtimeClock.Alarm(i) when the wakeup time has been reached Sep 13 13:13:29 s/even/event/ Sep 13 13:13:30 DocScrutinizer-8 meant: then shedule ONE event for 13:30 just to call hwclock --adjust, if you need a rtc wake at 14:00:00 Sep 13 13:14:01 DocScrutinizer-8: why not just compensate for the drift so that you can schedule the rtc wake up at the right time? Sep 13 13:14:38 duh?? that's exactly what above suggested scheme does Sep 13 13:14:47 DocScrutinizer-8: but you need to wake it up twice? Sep 13 13:16:09 lindi-: and *you* need to make sure *nobody* messes with RTC during these ten days. That's not what ever is going to happen Sep 13 13:16:30 DocScrutinizer-8: why not? I am sure nobody writes to rtc on my system Sep 13 13:16:55 haha. you never use FSO and GPS? Sep 13 13:17:00 DocScrutinizer-8: nope Sep 13 13:17:03 DocScrutinizer-8: i use gpsd Sep 13 13:17:09 DocScrutinizer-8: and only ogsmd Sep 13 13:17:16 they don't write to rtc Sep 13 13:17:26 so go ahead build your own proprietary system then Sep 13 13:17:36 done already :) Sep 13 13:18:01 ok, so please don't suggest FSO should do same way Sep 13 13:18:20 i think my idea offers more accuracy Sep 13 13:19:07 so maybe it could be integrated as an option to hwclock Sep 13 13:19:38 I think your system add 5 complexity points one end, to avoid 1 point at other end. result is identical Sep 13 13:20:19 except youur way is non-standard Sep 13 13:20:34 yes it's different Sep 13 13:20:46 freesmartphone.org: 03mickey 07cornucopia * rb47d0fede191 10/fsodeviced/src/plugins/kernel26_rtc/plugin.vala: fsodeviced: kernel26_rtc: throw dbus error, if requested wakeup time is not in the future Sep 13 13:21:02 and i started by trying to use hwclock --adjust but it wasn't accurate enough for me Sep 13 13:21:09 since i often don't have time source for days Sep 13 13:21:44 lindi-: !!!! hwclock --adjust *does NOT need* timesource !!! Sep 13 13:23:34 au contraire -- if you *had* timesource, you want to use --systohc Sep 13 13:24:26 we should really benchmark these approaches somehow Sep 13 13:24:32 ya, that'd rock Sep 13 13:24:34 scientific data Sep 13 13:24:41 lindi-: to me all this sounds like "too lame to RTFM, so invented it new" Sep 13 13:24:43 well i sent my mail about comparing clocks months ago Sep 13 13:24:48 DocScrutinizer-8: but i even read the source Sep 13 13:25:07 and discussed with others, i think this way is more accurate and actually simpler :) Sep 13 13:25:35 more accurate: I deal with floating point time and I'm not restricted to 1-second granularity of RTC in the drift correction Sep 13 13:26:06 lindi-: hwclock yields virtually infinite accuracy. So how can you be better? for "simpler" :X Sep 13 13:26:07 simpler: i never need to write to rtc Sep 13 13:26:30 coool Sep 13 13:26:43 I simply don't care Sep 13 13:27:07 as I also don't... hwclock does that for me Sep 13 13:28:32 freesmartphone.org: 03mickey 07cornucopia * r4ec6d36e4928 10/fsonetworkd/src/plugins/sharing/sharing.vapi: fsonetworkd: sharing.vapi no longer necessary, we do it all in vala now Sep 13 13:29:52 to benchmark this you can run http://iki.fi/lindi/openmoko/compare-clock-sources.pl on a phone for a few days and make sure you use no external time sources Sep 13 13:31:43 and what will we learn from such benchmark? Sep 13 13:31:58 freesmartphone.org: 03mickey 07cornucopia * r82b066330325 10/fsonetworkd/ (5 files in 2 dirs): fsonetworkd: ship dbus configuration and system activation files Sep 13 13:32:15 DocScrutinizer-8: if the hwclock approach can compensate for the rtc drift well enough Sep 13 13:32:27 that there's drift in RTC. we know that already Sep 13 13:32:50 lindi-: hwclock yields virtually infinite accuracy. So how can you be better? for "simpler" :X Sep 13 13:33:08 DocScrutinizer-8: but every time you write to rtc does it reset to the beginning of the second? Sep 13 13:33:30 if you write x to rtc, how long does it take for the rtc to read x+1? Sep 13 13:33:50 lindi-: that's a simple(!!!) plus-minus 0.5 issue. It will average out Sep 13 13:35:01 and even if you see a skew of 5 sec every time you write to RTC... hwclock will take that into account for calculation of adjust factor Sep 13 13:35:32 as long as you do the hwclock --adjust on a regular base Sep 13 13:35:41 e.g via at Sep 13 13:36:01 ok, one more advantage in my approach. I don't need to do anything regularly :) Sep 13 13:36:46 nah, I'll abort this useless sophisic bickering now Sep 13 13:37:18 wears my kbd for nothing Sep 13 13:37:21 freesmartphone.org: 03mickey 07cornucopia * r1c8a2a60317f 10/fsonetworkd/src/bin/main.vala: fsonetworkd: shut down can't load any plugins Sep 13 13:37:22 freesmartphone.org: 03mickey 07cornucopia * r069d9895894e 10/fsousaged/src/ (bin/main.vala plugins/controller/plugin.vala): fsousaged: shut down if can't load any plugins Sep 13 13:39:09 freesmartphone.org: 03mickey 07cornucopia * rda6f138b21d9 10/fsodeviced/src/bin/main.vala: fsodeviced: shut down if can't load any plugins Sep 13 13:41:42 mickeyl: sorry, i didn't been away for two days at the country side Sep 13 13:41:52 mickeyl: after request/release/request - everything works Sep 13 13:42:12 freesmartphone.org: 03mickey 07cornucopia * r9adf9a0e06d0 10/fsogsmd/src/bin/main.vala: fsogsmd: don't start if can't load any plugins Sep 13 13:42:16 PaulFertser: np Sep 13 13:42:26 dos1: ok, that looks good. i'll do the power consumption measurements on my own Sep 13 13:42:30 will commit meanwhile Sep 13 13:42:37 doesn't seem to break anything Sep 13 13:43:16 dos1: you got a GTA01? Sep 13 13:43:25 freesmartphone.org: 03mickey 07framework * r1e42fdcad51a 10/framework/subsystems/ogsmd/modems/ti_calypso/modem.py: ogsmd: [TI CALYPSO]: on closing, drop out of muxing mode and power off the modem (AT@POFF) Sep 13 13:43:44 DocScrutinizer-8: no, gta02 Sep 13 13:43:57 *phew* Sep 13 13:44:20 fsodeviced done, fsotimed done, fsonetworkd done Sep 13 13:44:22 now if we had async dbus, i could start on fsogsm Sep 13 13:44:27 *sigh* Sep 13 13:44:47 so, probably that's easy to check if modem is off, as CC in bat reports exactly current-now=p. But I guess Mickey is interested in GTA01 results Sep 13 13:45:17 s/p /0 / Sep 13 13:45:24 yea Sep 13 13:45:29 yup Sep 13 13:45:43 mrmoku: ping Sep 13 13:45:52 ~pester Vala team to work on stuff that matters Sep 13 13:46:09 mrmoku: i'll cry if we won't have testing image soon :'( Sep 13 13:46:13 mrmoku: do you want me to cry? :P Sep 13 13:47:53 Xorg/KMS stuff, newer EFL, more cornucopia inside... Sep 13 13:48:19 dos1: i pushed you patch to shr-today Sep 13 13:48:28 Slyon: thanks :) Sep 13 13:48:41 mickeyl: testing it on gta02 without commenting out output to power_on sysfs node wouldn't make much sense i guess. But if one does that (comment out sysfs access) it would be ~the same as gta01. Sep 13 13:48:54 PaulFertser: yeah Sep 13 13:49:01 pretty much Sep 13 13:49:06 Slyon: i registered to github (login: dos1), if you want you can give me commit access ;D Sep 13 13:50:05 PaulFertser: don't comment out too much. You still want to power up modem correctly ;-) Sep 13 13:51:20 dos1: yeah, added you as a collaborator Sep 13 13:51:30 could i force refreshing the accelerometer values without fiddling w/ the threshhold? Sep 13 13:51:39 i need a set of samples while they're still Sep 13 13:51:39 http://en.wikipedia.org/wiki/Programmers%27_Day Sep 13 13:51:41 dos1, do you have a webcam? Sep 13 13:51:45 happy programmers day everyone Sep 13 13:51:52 * mrmoku would really like to actually *see* dos1 crying ;) Sep 13 13:51:52 s/still/device not moving/ Sep 13 13:51:53 mickeyl meant: i need a set of samples while they're device not moving Sep 13 13:51:58 mickeyl: yes, there's ioctl to get current measurements. Sep 13 13:52:08 mrmoku: you bastard! Sep 13 13:52:09 PaulFertser: good! Sep 13 13:52:18 Standard for input devices i guess. Sep 13 13:52:21 PaulFertser: any sample code or should i check the source? Sep 13 13:52:21 dos1, :P Sep 13 13:52:33 ;) Sep 13 13:52:38 PaulFertser: well, i know about EIVIOCGKEY Sep 13 13:52:49 let me check whether there is more along those lines Sep 13 13:52:51 but i doubt it Sep 13 13:52:55 mrmoku: if you want you can start shr-experimental and put all new stuff to it, i don't care Sep 13 13:52:58 as we would need all the 3 axes Sep 13 13:53:17 mrmoku: but i think users won't be happy seing experimental branch without any good testing image :P Sep 13 13:53:32 dos1, what new stuff do we have to put in? Sep 13 13:53:36 iirc the input system only delivers the last known values Sep 13 13:53:49 it doesn't force the slave driver to actually reread 'em Sep 13 13:53:49 mickeyl: should<ät be there a sysfs node to get raw readings? Sep 13 13:53:59 mrmoku: KMS/Xorg, more cornucopia inside, newer EFL_SRCREV Sep 13 13:54:06 DocScrutinizer-8: is there? Sep 13 13:54:14 mickeyl: http://markmail.org/thread/7ydhpsazugygrd2n Sep 13 13:54:27 dunno. usually kernel guys have that for each dev Sep 13 13:54:29 mickeyl: EVIOCGABS Sep 13 13:54:30 dos1, so... would would you propose to get a decent testing? Sep 13 13:54:35 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * r7ff2214374fb 10framework/framework/subsystems/ogsmd/modems/ti_calypso/modem.py: Merge branch 'master' of git@git.freesmartphone.org:framework into dos/opimd-tracking Sep 13 13:54:59 mrmoku: dunno, you are buildhost expert ;x Sep 13 13:55:21 mrmoku: Ainulindale fixes some problems regarding generating images at all. what now is stopping us to get -testing? Sep 13 13:55:33 PaulFertser: this doesn't help Sep 13 13:55:34 s/fixes/fixed/ Sep 13 13:55:35 dos1 meant: mrmoku: Ainulindale fixed some problems regarding generating images at all. what now is stopping us to get -testing? Sep 13 13:55:35 "actually 'most recently reported') values." Sep 13 13:55:44 we need more than the most recent reported values Sep 13 13:55:45 mickeyl: let me put it this way: if there isn't, I consider that a mishap Sep 13 13:55:52 i need to force reading a new sample Sep 13 13:56:01 dos1, question is what refs to update for testing Sep 13 13:56:01 DocScrutinizer-8: hmm, there is some 'dump' sysfs node. could that be it? Sep 13 13:56:08 yes Sep 13 13:56:12 mrmoku: for testing... hmm Sep 13 13:56:14 usually Sep 13 13:56:33 mickeyl: frameworkd, shr-settings, opimd-utils for sure Sep 13 13:56:40 s/mickeyl/mrmoku/ Sep 13 13:56:40 dos1 meant: mrmoku: frameworkd, shr-settings, opimd-utils for sure Sep 13 13:56:44 hmm Sep 13 13:56:46 doesn't change at all Sep 13 13:56:53 :-( Sep 13 13:57:21 dos1, and the cornucopia stuff too I guess Sep 13 13:57:23 hmm Sep 13 13:57:26 mrmoku: yup Sep 13 13:57:33 could some leet kernel hacker implement a 'trigger' for us? Sep 13 13:57:51 mickeyl: but it's hardware, it just doesn't "sense" if the threshold is not reached. Sep 13 13:57:53 so that independent of threshhold, when you ping the trigger, it reads and sends a sample Sep 13 13:58:06 mmh Sep 13 13:58:15 bummer Sep 13 13:58:25 I'm not so 733t Sep 13 13:58:27 mickeyl: i can try to think about it but i must admit i've never read accel datashett. Sep 13 13:58:35 PaulFertser: :) Sep 13 13:59:10 usually those devices have an irq-based mode and a polling-mode Sep 13 13:59:23 if we can use both concurrently, we could just ask it to give us a sample Sep 13 13:59:36 if you're right though and the threshhold is global, then i'm out of luck and need to 'synthesize' this Sep 13 13:59:48 maybe that's push rather than poll Sep 13 14:00:23 and thresholds are implemented inside gmeter chips Sep 13 14:00:35 my FSO-slot if over for today. Please keep me posted on that accel thing, if you find something. need to spend time with wife now Sep 13 14:00:37 l8er Sep 13 14:01:02 DocScrutinizer-8: it has thresholds inside for sure, it can even generate and irq to wakeup the device. Sep 13 14:01:02 mickey_away: see ya l8er Sep 13 14:01:17 yes, I know Sep 13 14:01:42 I wonder what's exact data protocol, though - besides IRQ Sep 13 14:01:48 freesmartphone.org: 03seba.dos1 07dos/opimd-tracking * red29786dacec 10framework/framework/subsystems/opimd/pimb_sim_messages_fso.py: opimd: SIM-Messages-FSO: comment out not-really-well-thought (read as: why the hell is it there?) code Sep 13 14:02:55 last time I read these ds is maybe 15months ago Sep 13 14:06:02 mrmoku: btw, http://patchwork.dev.bearstech.com/project/SHR/list/ ;) Sep 13 14:06:31 mrmoku: http://patchwork.dev.bearstech.com/patch/119/ seems to be applicable now ;) Sep 13 14:06:42 PaulFertser: actually I guess you need to write some value *to* the chip anyway as that's the only poll-triger I can think of. So it's probably just as easy to set thresholds to 0 or sth like that, temporary Sep 13 14:08:54 for sure we wouldn't like those gmeters to spit out serial data all the time while suspend and no threshold tripped. so writing data to the device is only way to make it talk Sep 13 14:11:32 CSM in opimd fixed :) Sep 13 14:11:34 if this is a read-out command then, or a temporary reset of thresholds - not much difference Sep 13 14:11:52 whatever CSM might be Sep 13 14:12:08 DocScrutinizer-8: splitten sms :P Sep 13 14:12:18 aaah. :-))) Sep 13 14:12:25 concatenated short messages if i remember correctly Sep 13 14:12:41 sounds right Sep 13 14:12:43 migrating Messages domain to GenericDomain broken it, and i noticed it yesterday ;x Sep 13 14:13:24 dos1, so... which frameworkd version? Sep 13 14:13:33 mrmoku: newest, just tested it Sep 13 14:13:48 mrmoku: ed29786daceccefe918ce3911e3b6fb7f2efb08c Sep 13 14:14:22 mrmoku: it has also fix for timestamp handling :) Sep 13 14:15:00 mickey_away: taking into account my above pondering, it seems natural sysfs dump node is silent Sep 13 14:16:59 dos1, yeah, already set that... cornucopia? Sep 13 14:17:09 take the one from 6/9 image? Sep 13 14:17:42 mickey_away: but if you follow my suggestion to set thresholds to actual readout +/- a few percent, then why would you want to readout while no actual threshold trip anyway? You know the values didn't change beyond a marginal amount Sep 13 14:18:59 think so... all stuff after that is new stuff... no fixes as far as I can tell Sep 13 14:19:00 mrmoku: also take newest Sep 13 14:19:05 dos1, nah Sep 13 14:19:15 mrmoku: there are fixes about restartability Sep 13 14:19:31 and "fsousaged: shut down if can't load any plugins" :P Sep 13 14:19:36 Hi all! Sep 13 14:20:03 dos1, hmm... ok Sep 13 14:20:11 9adf9a0e06d07d970e9752ece3f4b9731b19ff24 then Sep 13 14:21:22 dos1: somehow opimd-messages are now not aware of incoming smses. 8( Sep 13 14:22:11 Q-Master: here it works(TM) Sep 13 14:22:32 rtp: ping Sep 13 14:22:48 dos1: hm Sep 13 14:23:02 Q-Master: i'll build newest frameworkd on buildhost Sep 13 14:23:11 ok Sep 13 14:23:14 mrmoku: are you doing something on buildhost? Sep 13 14:23:22 dos1, not yet Sep 13 14:23:25 ok Sep 13 14:23:35 dos1: BTW, will you fix sending smses from opimd-utils? Sep 13 14:23:36 dos1, right now I'm updating revs Sep 13 14:23:52 bbiab Sep 13 14:23:52 Q-Master: fix? here it works too (TM) Sep 13 14:24:25 dos1: strange Sep 13 14:26:57 mrmoku|away: you were right.. connman newer 0.39 starts ok, but than disables usb0 because no AutoStart enabled here and also mokonnect doesnt start it after connman restart :/ Sep 13 14:27:50 JaMa, and somehow it made X not start for me :( Sep 13 14:28:07 my guess was it kills the loopback but had no way to proof it Sep 13 14:29:45 JaMa: i hope we'll get your patches in -unstable soon ;) Sep 13 14:29:55 JaMa: (about KMS and Xorg) Sep 13 14:30:02 mrmoku: thats maybe because I use Xorg instead of Xglamo :) Sep 13 14:30:07 dos1: thanks :) Sep 13 14:30:21 dos1: and dont forget mplayer with glamo :P Sep 13 14:30:23 JaMa, hehe, one more reason to switch :) Sep 13 14:30:43 JaMa: http://patchwork.dev.bearstech.com/project/SHR/list/ - i can't see it here Sep 13 14:30:47 JaMa: could you resend? Sep 13 14:30:58 dos1, it's waiting masters approval because of size limit Sep 13 14:31:00 dos1: its waiting for moderator aproval too :/ Sep 13 14:31:05 oh Sep 13 14:31:07 Ainulindale, ping ^^^ Sep 13 14:31:07 ok then Sep 13 14:31:10 Ainulindale: ping Sep 13 14:31:11 :D Sep 13 14:31:21 mrmoku: also that KMS patch is quite old.. I sent better yesterday.. Sep 13 14:31:40 JaMa, yeah, will apply the new one Sep 13 14:31:41 or i can send it bziped directly to you.. if it helps :) Sep 13 14:31:46 first fix revs for testing though Sep 13 14:32:06 JaMa, I hope Ainulindale pops up sooner or later Sep 13 14:32:11 I know he's watching us ;) Sep 13 14:32:14 * dos1 is excited with breaking -unstable ;> Sep 13 14:32:41 at last we'll provide Xorg and KMS by default! Sep 13 14:32:41 yes it would be great.. although I tried to make Xorg or even Xorg+KMS optional.. just choose which config to include to shr distro config :) Sep 13 14:32:53 could any of the linux system level cracks share a pointer to a site explaining foundation rationale and best practices of sysfs, please? I seen /dev drivers that simply interface to a /sys node - maybe do some sanity checking or muxing. And aiui sysfs is all owner:root and userland not supposed to have any access whatsoever Sep 13 14:32:55 but tested just Xorg+KMS :) Sep 13 14:33:47 now bisecting connman to see which function/setting is missing in mokonnect or how to change default autostart for "unknown" services like usb0 Sep 13 14:33:50 JaMa: did your mplayer patches also enable tremora (or anything else which i don't remember exactly and it makes ogg playing less cpu-sucking?) Sep 13 14:33:59 dos1: sure Sep 13 14:34:14 JaMa: even nicer :))) Sep 13 14:34:36 dos1: as I sent comparision between all interesting mplayer builds only shr one lacked tremor :) Sep 13 14:35:16 dos1: strange but pressing on shr-notifier's incoming message box doesn't followed by loading of shr-messages 8/ Sep 13 14:36:34 mrmoku: I'm there Sep 13 14:37:16 Ainulindale: mail on ml is waiting for moderator approval due to attachment size Sep 13 14:37:31 Ainulindale: accept it (it contains patches about mplayer and glamo support) Sep 13 14:37:50 Done Sep 13 14:38:01 mrmoku, JaMa: ^^^ :) Sep 13 14:38:03 Ainulindale: thanks :) Sep 13 14:38:21 dos1: http://pastebin.com/m583fd2d8 Sep 13 14:39:26 Q-Master: shr-today? Sep 13 14:39:40 seems so Sep 13 14:39:45 we will break unstable? bump EFL? Sep 13 14:39:48 Ainulindale: thanks, I'll try to split to smaller parts next time :) Sep 13 14:39:57 No problem :-) Sep 13 14:40:02 I'll raise the filter Sep 13 14:40:06 mrmoku: by the way Sep 13 14:40:13 Ainulindale: even better :) Sep 13 14:40:13 I think you do have my GSM number Sep 13 14:40:19 You should phone me for that kind of thing or text me Sep 13 14:40:27 Do not hesitate Sep 13 14:40:41 dos1: and this seems yours: Sep 13 14:40:41 Heinervdm: i want to have EFL bumped, Xorg instead of kdrive, KMS support and more cornucopia than only fsousaged used in unstable Sep 13 14:40:42 org.freesmartphone.PIM.NoMoreEntries: All results have been submitted Sep 13 14:40:42 dbus error! org.freesmartphone.PIM.InvalidBackend: Backend does not feature add_entry Sep 13 14:40:42 kabum Sep 13 14:40:46 Although I could give you the accesses as well Sep 13 14:41:01 Q-Master: add_entry? are you using SIM-Messages-FSO? Sep 13 14:41:03 dos1: yes that would be nice Sep 13 14:41:15 dos1: yep. should I switch to sqlite? Sep 13 14:41:30 Q-Master: if you want to use opimd-utils, then of course Sep 13 14:41:32 Ainulindale, ok :) Sep 13 14:41:46 dos1: will shr-messages work then? Sep 13 14:42:03 it will work, but new messages won't be displayed in shr-messages Sep 13 14:42:37 dos1: ok. I'll test it right now Sep 13 14:43:07 There's a lot of spam Sep 13 14:48:42 SHR: 03david 07shr * r767cc2cd22f5 10/libframeworkd-phonegui-efl/po/ (es.po gl.po): libframeworkd-phonegui-efl: Updated spanish/galician translations Sep 13 14:49:24 heh Sep 13 14:50:48 dos1: funny. there were no sound of incoming message. 8( Sep 13 14:51:20 Q-Master: http://openmoko.opendevice.org/~dos/opimd/oeventsd/fso_triggers.diff Sep 13 14:51:20 Q-Master: you have to apply it Sep 13 14:51:41 dos1: again??? Sep 13 14:54:08 again? Sep 13 14:54:18 SHR: 03david 07shr-settings * r5681b441ca4e 10/data/po/ (4 files in 2 dirs): [po] Updated spanish/galician translations Sep 13 14:54:24 well, every frameworkd upgrade from repo :P Sep 13 14:55:36 dos1: and there's no possibility to integrate this patch to framework? Sep 13 14:55:47 Q-Master: not yet Sep 13 14:56:47 pity Sep 13 15:00:13 ok... dinner time... bbl Sep 13 15:03:50 dos1: what about starting shr-experimental in org.openembedded.dev branch as mickey now suggested on ML? Sep 13 15:04:52 I have KMS patch somehow ready for org.oe.dev, but cannot apply the same, because shr stuff is missing there and also some bumped xorg versions were there already.. Sep 13 15:05:44 ok, so IHMO we should do that: Sep 13 15:05:49 maybe you could let shr/import be branch for shr-testing and continue with -unstable in org.oe.dev "soon" Sep 13 15:05:56 a) build testing image with stuff we have already Sep 13 15:06:09 b) apply JaMa stuff, bump EFL in unstable Sep 13 15:06:15 current image should be stable enough for testing Sep 13 15:06:32 c) bring shr-oemerge back to life, now with org.openembedded.dev branch Sep 13 15:07:08 dos1: good idea Sep 13 15:07:09 when shr-oemerge will be the same or better than unstable, then do testing<-unstable<-oemerge migration, and remove oemerge Sep 13 15:07:33 dos1: the only problem I have with my patches for org.oe.dev.. is that I can test them only for arm4(neo) and arm5(spitz) in shr/import it was safe and easy :) Sep 13 15:08:14 JaMa: that will be problem for people which will do migration ;) Sep 13 15:08:26 to have neo specific stuff applicable only to neo Sep 13 15:08:30 and s/neo/shr/ too Sep 13 15:09:02 then let's start with testing ;)( Sep 13 15:09:22 Heinervdm: mrmoku already started, i hope nothing will stop him now :) Sep 13 15:09:33 good :) Sep 13 15:09:38 I dont understand "testing<-unstable<-oemerge migration" testing and unstable will live still in shr/import and oemerge will track org.oe.dev? Sep 13 15:10:09 i think unstable will die when we do that Sep 13 15:10:22 and then we will rename oemerge to unstable Sep 13 15:10:29 JaMa: it will be as you said, but before that t<-u<-o migration Sep 13 15:10:50 JaMa: after it, testing will be shr/import based, unstable org.oe.dev, and oemerge - deleted Sep 13 15:11:31 dos1: then what is oemerge for? couldn't we just commit it directly to org.oe.dev? Sep 13 15:11:32 (by org.oe.dev i mean our new, shiny SHR feature branch which will track org.oe.dev) Sep 13 15:11:51 JaMa: shr-oemerge, kind of distro (as testing and unstable), not branch in oe Sep 13 15:12:06 dos1: ah now I see Sep 13 15:12:15 we already had shr-oemerge when migrating from oe overlay to shr/import Sep 13 15:13:44 content of images is now upgraded in repo, you can opkg update; opkg upgrade Sep 13 15:13:54 now feed is building (talking about unstable) Sep 13 15:18:04 mrmoku|away: subversion success? Sep 13 15:19:11 rhkfin_mobi: you want subversion in feed? Sep 13 15:19:49 dos1: yes please Sep 13 15:20:18 mrmoku|away tried but it didnt build Sep 13 15:20:36 rhkfin_mobi: hmm... i built it on my buildhost some time ago and it built Sep 13 15:27:06 dos1: ok, nice Sep 13 15:27:08 freesmartphone.org: 03seba.dos1 07phonesim * r20c231f304a4 10/lib/ (63 files in 6 dirs): remove .svn files Sep 13 15:31:45 dos1: after this patch nothing happens on incoming sms at all Sep 13 15:33:19 that sounds good.. Sep 13 15:35:20 dos1, you're stopping me ;) Sep 13 15:35:34 why? :x Sep 13 15:35:41 dos1, buildhost activity? Sep 13 15:35:50 subversion didn't build for me too :/ tried more versions.. but then used svn from debian on second partition :) Sep 13 15:36:08 rhkfin_mobi, apr-utils does not build due to some libtool problem... Sep 13 15:36:18 mrmoku: buildhost is yours Sep 13 15:36:20 ;) Sep 13 15:36:35 you updated OE for testing too? Sep 13 15:36:59 yep you did :) Sep 13 15:38:30 ok.. Sep 13 15:38:45 mrmoku: i'm watching you! do something, "cd" won't do everything for you lazy ass! :P Sep 13 15:38:47 :D Sep 13 15:39:13 lol Sep 13 15:40:23 SHR: 03gmary 07opimd-utils * r11536ff92e44 10/data/po/fr/ (opimd-notes.mo opimd-notes.po): Update french translation Sep 13 15:41:05 dos1, you're watching the wrong window my friend ;) Sep 13 15:41:14 testing is already building Sep 13 15:41:47 god dammit! Sep 13 15:41:49 :D Sep 13 15:42:27 time to go to gym... happy building and see you later :) Sep 13 15:42:28 aw, shucks mickey is away.. Sep 13 15:43:07 can anyone tell me how a Null Logger is useful? shouldn't we default to stderr if no logger setting was specified? Sep 13 15:43:13 JaMa, have fun :) Sep 13 15:45:48 dos1: strange things with python. now everything works. BTW, how to switch from shr-contacts to opimd-contacts in lower 3 buttons in your theme? Sep 13 15:46:30 Q-Master: remove "Bar" category in shr-contacts, and add it to opimd-contacts desktop file Sep 13 15:46:53 dos1: same with messages? Sep 13 15:47:03 Q-Master: yup Sep 13 15:47:13 dos1: hm. doesn't help Sep 13 15:47:21 do I need to restart x? Sep 13 15:47:39 just restarting enlightenment should be enough Sep 13 15:47:57 how to restart enlightment? Sep 13 15:48:03 killall -HUP enlightenment Sep 13 15:48:09 ok Sep 13 15:52:12 webkit-efl :x Sep 13 15:52:18 it'll take ages ;/ Sep 13 15:53:17 is there any web browser which uses elementary and webkit? Sep 13 15:53:24 Q-Master: eve Sep 13 15:53:25 ? Sep 13 15:54:38 is it in repo? Sep 13 15:55:14 Q-Master: yup Sep 13 15:55:20 Q-Master: but it lacks finger scrolling ATM Sep 13 15:55:58 and what about rss reader? Sep 13 16:05:32 eve doesn't use emelentary Sep 13 16:05:44 it's just evas Sep 13 16:07:42 well Sep 13 16:08:04 browser using plain elementary... i don't think it would be possible :P Sep 13 16:09:48 there is a webkit widget on the elementary todo list ;) Sep 13 16:12:55 Heinervdm: yup, i know ;) Sep 13 16:19:41 hehe Sep 13 16:19:49 -3 unread messages Sep 13 16:21:22 heh. seems eve lacks a desktop file Sep 13 16:21:43 Q-Master: you read messages that will be sent in the future? ;) Sep 13 16:22:09 Q-Master: yes, eve has no desktop file, but it's not really useable Sep 13 16:22:22 so it's just for testing and that needs no icon Sep 13 16:27:01 ok Sep 13 16:27:03 thnx Sep 13 16:30:37 * DocScrutinizer grunts about stupid arrogant sw-devels thinking "what, hw? the hw has to work, nothing else". And forget about the fact it's the sw that has to control the hw properly :-/ Sep 13 16:31:32 Luckily not on OM Sep 13 16:31:45 DocScrutinizer: i think high-level devels can think something like that Sep 13 16:31:58 DocScrutinizer: of course low-level sw devels can't, as it's plain stupid ;x Sep 13 16:32:27 dos1: http://docs.openmoko.org/trac/ticket/2310#comment:8 Sep 13 16:35:02 DocScrutinizer: who is timeless_mbp? Sep 13 16:36:00 PaulFertser: the guy with the N900, in #maemo and in #2310 Sep 13 16:36:35 rather unfriendly guy Sep 13 16:36:38 DocScrutinizer: ah. :( Sep 13 16:36:59 DocScrutinizer: is he sure that the device doesn't conform to the specs or is it rather an unfounded speculation? Sep 13 16:37:00 so kickit Sep 13 16:37:20 no idea. I don't care anymore Sep 13 16:38:13 DocScrutinizer: the evidence from the lindi- is enough to try to simulate this behaviour even if we lack n900, imho. Sep 13 16:40:19 eeeek, kicked for excessive flooding :o) Sep 13 16:41:06 * mwester fetches sandbags and a large pump, just in case doc does it again. Sep 13 16:41:22 PaulFertser: n800 and n810 might fit Sep 13 16:42:13 NEWSFLASH: VALA HAZ ASYNC DBUS! Sep 13 16:42:23 wooooow Sep 13 16:42:36 apparantly pestering enough helps Sep 13 16:42:41 mickey_away: so finally everything will be fine Sep 13 16:42:43 ? Sep 13 16:42:43 although i feel kind of bad for that ;) Sep 13 16:43:05 async dbus on server side was the showstopper for ogsmd in Vala Sep 13 16:43:12 i will now start working on that Sep 13 16:43:18 and we'll get a really fast GSM server Sep 13 16:43:19 btw, is ffalarms maintained? Sep 13 16:43:28 mickey_away: great news!!! Sep 13 16:43:41 Q-Master: i think it is, but not that version from SHR Sep 13 16:43:44 it's going to be lots of hard work though, so don't expect it to be around the corner :) Sep 13 16:43:53 Q-Master: it was rewritten to vala. shr uses python version Sep 13 16:43:56 mickey_away: I just learned my lesson why Nokia/N900/maemo isn't such a great thing wrt FOSS Sep 13 16:44:03 DocScrutinizer: really,why? Sep 13 16:44:09 mickey_away: but still great news ;) Sep 13 16:44:15 dos1: *nod* Sep 13 16:44:36 mickey_away: http://docs.openmoko.org/trac/ticket/2310#comment:8 Sep 13 16:45:03 :/ Sep 13 16:46:06 I'm not surprised. "opensource" is just like it. Sep 13 16:47:10 dos1: vala version also lacks for continuing alarms for several days... Sep 13 16:47:16 lacked Sep 13 16:47:19 when I've seen it Sep 13 16:49:52 mrmoku: build finished! Sep 13 16:51:10 * DocScrutinizer away for brunch Sep 13 16:59:42 mrmoku: i started build of full testing image Sep 13 17:02:05 dos1, ok Sep 13 17:03:02 mrmoku: you can test lite in meantime :D Sep 13 17:05:28 duhh... forgot my FR downstairs :( Sep 13 17:05:35 * mrmoku makes stairs again.... Sep 13 17:12:20 * mrmoku is flashing Sep 13 17:12:54 dos1, hmm... why is it rebuilding ophonekitd? Sep 13 17:13:27 mrmoku: hmm Sep 13 17:13:38 * mrmoku checks testlab of lite build Sep 13 17:14:43 dos1, correct version included... don't understand why it's rebuilding it Sep 13 17:14:45 mrmoku: maybe that's due to commited translation? Sep 13 17:14:48 hmm Sep 13 17:14:55 ah, no, it has fixed rev Sep 13 17:14:57 dunno ;x Sep 13 17:15:10 the same with e-wm-theme-illume-shr Sep 13 17:15:19 well, i have once the same Sep 13 17:15:31 it was building the same versions of packages everytime Sep 13 17:15:50 but after few builds it stopped, and i didn't notice any pattern Sep 13 17:16:17 heh Sep 13 17:17:15 as long as it builds correctly :P Sep 13 17:22:22 http://scap.linuxtogo.org/files/bc079a18802e2a100bd59f991c68b4ac.png Sep 13 17:22:26 noooooooooooo Sep 13 17:22:27 :( Sep 13 17:22:28 dos1, ^ Sep 13 17:22:52 you doomed us all. Sep 13 17:22:55 TAsn: :DDDDD Sep 13 17:23:03 tamagochi is the devil Sep 13 17:23:12 you are helping the devil! Sep 13 17:23:31 dos1, make it happy and plug in the wallcharger ;) Sep 13 17:24:14 TAsn: hmm Sep 13 17:24:16 dos1, you should totally incorporate the accelerometer and gps!!! Sep 13 17:24:22 "Geekogotchi - it comes from hell!" Sep 13 17:24:23 ;) Sep 13 17:24:27 (/me changed his mind) Sep 13 17:24:36 i.e add a Sep 13 17:24:43 "take out for a walk mode" Sep 13 17:24:55 that'll add happiness and fitness by the milage you pass. ;) Sep 13 17:25:09 TAsn: nice idea :DDD Sep 13 17:25:27 dos1, use the accels for that... the more you shake it... the happier it will get :P Sep 13 17:25:44 well, geekogotchi have rather low priority in developing, but i'll try to add some cool features after reaching basics ;) Sep 13 17:26:19 dos1, and add touch screen stroking Sep 13 17:26:20 ): Sep 13 17:26:22 :) Sep 13 17:26:28 I've seen somewhere fix for the GPS+GSM trashing the uSD Sep 13 17:26:35 anyone remember? Sep 13 17:26:48 mrmoku: how's image? Sep 13 17:27:05 TAsn: shr-lite-testing image is built, shr-fat-testing is building ATM :) Sep 13 17:27:20 TAsn: we want to finally break unstable! :D Sep 13 17:27:45 let mrmoku at it, then! :p Sep 13 17:27:58 dos1, booting Sep 13 17:28:15 mwester, nah... I don't break images... I break builds ;) Sep 13 17:28:18 dos1, Sep 13 17:28:21 hehe! Sep 13 17:28:29 I won't be testing it at least until tomorrow evening Sep 13 17:28:33 ciao I'm off. Sep 13 17:28:50 mwester, how's life by the way? Sep 13 17:30:50 dos1, it booted... waiting for it to register :) Sep 13 17:38:07 mrmoku: and? :) Sep 13 17:39:10 dos1, have you already uploaded geegochi code somewhere? Sep 13 17:39:17 Sharwin_F: git.shr-project.org Sep 13 17:39:41 dos1, bad Sep 13 17:39:51 fso-abyss does not start (or segfault) Sep 13 17:40:07 last line from ogsmd is Sep 13 17:40:07 2009.09.13 18:32:03.40 ogsmd.modems.ti_calypso INFO Requesting new channel from 'fso-abyss' Sep 13 17:40:11 mrmoku: hmm Sep 13 17:40:13 and no fso-abyss running Sep 13 17:40:25 mrmoku: is fso-monitord running? Sep 13 17:40:37 mrmoku: we should remove it from image, or bug someone to fix it Sep 13 17:40:48 root@om-gta02 ~ $ ps ax | grep fso 2193 ? Ss 0:00 /usr/sbin/fso-gpsd -S localhost gpsd -P /var/run/fso-gpsd.pid 2468 ? R 0:00 /usr/bin/fsousaged Sep 13 17:40:53 not running Sep 13 17:40:59 hmm, strange Sep 13 17:41:04 what about gsm0710muxd? Sep 13 17:41:34 dos1, do you use master for fso-monitord? Sep 13 17:41:41 mrmoku, life is ok... I have some contract work, but it is far distant (I travel to Canada weekly), and once there I have limited network access to speak with the world (due to security requirments). And I can't really bring my GTA0x devices there, at all. Sep 13 17:41:43 playya: dunno Sep 13 17:41:51 i should merge dbus-hlid to master Sep 13 17:41:52 So I feel sort of isolated from this project anymore. Sep 13 17:42:16 playya: but problem is still in master Sep 13 17:42:41 playya: http://git.freesmartphone.org/?p=fso-monitord.git;a=blob;f=src/gsm.vala;hb=HEAD Sep 13 17:43:01 playya: it requests gsm0710muxd,without looking which muxer is selected in config Sep 13 17:43:22 master is quite old Sep 13 17:43:31 playya: it results in starting gsm0710muxd before fso-abyss, then fso-abyss tries to use gsm, and it fails as it's already used by gsm0710muxd Sep 13 17:43:37 dbus-hlid scans all busses for supporte interfaces Sep 13 17:43:57 mwester, I see... sorry for that Sep 13 17:44:27 mrmoku: just start fso-abyss manually Sep 13 17:44:31 It has a good side, though -- I am not around to torment you! Sep 13 17:44:35 :p Sep 13 17:44:43 mrmoku: I've seen sometimes a dependency problem Sep 13 17:44:57 ans listens on NameOwnerChanged to get new bus names Sep 13 17:44:57 of not installing the lib needed by fso-abyss Sep 13 17:45:40 dos1, hmm... I think ogsmd part of frameworkd hangs somehow Sep 13 17:45:50 stopping frameworkd does not work... have to kill it Sep 13 17:46:02 mwester, :P Sep 13 17:46:04 mrmoku: i have it often even without any problem :P Sep 13 17:46:12 (having to kill frameworkd) Sep 13 17:46:26 mwester, I would prefer you to torment me and also fix our busybox config ;) Sep 13 17:46:38 * mrmoku reboots Sep 13 17:47:19 as sting would sing... put up the red light ;) Sep 13 17:49:37 mrmoku: so now fix issues ;P Sep 13 17:52:23 mrmoku: (ogsmd hangs) you used gprs? :-P Sep 13 17:52:24 btw. there's no http plugin for gstreamer in shr. mms, too Sep 13 17:53:48 DocScrutinizer-8, I never use GPRS ;) Sep 13 17:57:42 mrmoku: (as long as it builds correctly) alas it upgrades for ages, due to that (plus each and every kernel module inherits a new version string of course) :-/ Sep 13 17:58:07 DocScrutinizer, nah, not if it is the *same* version Sep 13 17:58:17 dos1, any way to get debug from fso-abyss? Sep 13 17:58:29 mrmoku: my last upgrade told me different Sep 13 17:58:36 mrmoku: start it from console and then request gsm resource Sep 13 17:58:45 dos1, ok Sep 13 18:01:03 yup fso-abyss :( Sep 13 18:01:03 2009-09-13T18:00:49.576544Z: server.vala:127: AllocChannel requested for name ogsmd.misc, requested channel 3 Sep 13 18:01:06 Segmentation fault Sep 13 18:04:26 mrmoku: why midori was built after shr-image???? Sep 13 18:04:55 dos1, ok libfsotransport needs rebuilding Sep 13 18:05:31 mrmoku: something is still broken in build process Sep 13 18:07:57 dos1, it is doing the tasks like rm_remove afterwards Sep 13 18:08:07 rm_work I mean Sep 13 18:08:17 package_stage_all too? Sep 13 18:08:35 well... no idea :P Sep 13 18:09:04 I interrupt the build... as we need newer fsotransport first Sep 13 18:09:45 hmm, i learned how to scroll in screen :D Sep 13 18:09:51 dos1, or better... you rebuild libfso-transport and I will be back in a bit ;) Sep 13 18:09:54 dos1, yeah screen is great Sep 13 18:13:55 hmm, messages stopped unexpectedly, dialer doesn't dial. Seems my tests yesterday once more borked parts of gsm system. Though I didn't start gprs this time Sep 13 18:14:28 DocScrutinizer-8: timeless_mbp is josh? Sep 13 18:14:34 or it borked out of no apparent reason Sep 13 18:14:49 lindi-: timeless_mbp is some unfriendly jerk it seems :( Sep 13 18:15:08 lindi-: dunno, the guy with N900 you allowed to heat your FR Sep 13 18:15:22 DocScrutinizer-8: then that's josh probably. bear in mind that he's a web browser developer Sep 13 18:16:16 lindi-: in no way that's a good reason to talk in a way he did, imho. Sep 13 18:16:26 I give a shit on that. He made me feel really unhappy about guys being in a position to state "WE gave out devices to some" Sep 13 18:16:27 PaulFertser: what did he say? Sep 13 18:16:59 DocScrutinizer-8: can i reference your latest pastebin? Sep 13 18:17:06 yep Sep 13 18:17:09 lindi-: http://pastebin.com/f3ccf5688 Sep 13 18:17:35 mrmoku: libfsotransport rebuilt Sep 13 18:18:07 mrmoku: do you want me to build new shr-lite? Sep 13 18:20:03 mrmoku: building :P Sep 13 18:24:02 dos1, yep building is fine Sep 13 18:25:22 heh, libphone-utils rebuilding Sep 13 18:26:10 and ophonekitd again :P Sep 13 18:26:25 dos1: will it work ?:) Sep 13 18:26:38 soltys: no :P Sep 13 18:26:40 PaulFertser: ok, I guess we need a way to reproduce this bug somehow Sep 13 18:26:47 i knew that ;) Sep 13 18:49:51 lindi-: have you tried with another FR? Sep 13 18:51:15 what i need to have gcc working on neo with SHR? Sep 13 18:51:45 root@(none) ~ $ arm-angstrom-linux-gnueabi-gcc test.c Sep 13 18:51:47 arm-angstrom-linux-gnueabi-gcc: error trying to exec 'as': execvp: No such file or directory Sep 13 18:51:51 dos1: binutils Sep 13 18:52:12 PaulFertser: nope Sep 13 18:52:12 yep Sep 13 18:52:18 binutils Sep 13 18:52:27 lindi-: can you? Sep 13 18:52:33 PaulFertser: thanks, installing Sep 13 18:52:38 i thought i have it installed ;x Sep 13 18:53:11 Is there any possibility to make a keyboard make sounds while pressed or just vibrate 8) Sep 13 18:53:18 PaulFertser: if I find somebody who's willing to risk his device :) Sep 13 18:53:29 Q-Master: yes. you can patch it ;> Sep 13 18:53:37 8) Sep 13 18:53:42 that suxx 8) Sep 13 18:56:17 Q-Master: literki has vibration feedback. I don't think it's that hard to add russian to it. Sep 13 18:57:12 PaulFertser: need to contact to author. I like 1 button switch to ru/en not selecting a keyboard from a list Sep 13 19:03:30 mrmoku: lite image to test! Sep 13 19:03:31 ;) Sep 13 19:03:47 yo Sep 13 19:06:14 and i have c++ compiler on neo, as we have now c++ in school :P Sep 13 19:06:56 :) Sep 13 19:07:21 dos1, mmmm... interested to write a c++ frontend? :P Sep 13 19:07:31 :DD Sep 13 19:07:45 i'll think about it ;) Sep 13 19:07:51 good :) Sep 13 19:10:00 larsc, PaulFertser, lindi- (and others): so, about this 2.6.31 kernel... (apologies if this has been covered..) Sep 13 19:10:37 Weiss: i'm waiting for larsc's directions basically Sep 13 19:11:15 Weiss: i'm mostly only testing things anyway. just tell me when I can test something :-) Sep 13 19:12:01 PaulFertser: same here, but i'm willing to "pull my weight" a bit, at least for the next week or so before my big move to germany Sep 13 19:13:21 Weiss: move? Are you going to work there? Sep 13 19:13:37 Weiss: i'm not able to coordinate efforts so i'm just waiting for larcs Sep 13 19:14:00 freesmartphone.org: 03mickey 07cornucopia * r315389465450 10/libfsobasics/fsobasics/ (fsobasics-2.0.vapi logger.vala): fsobasics: don't assert_not_reached() on logger.critical; exit(FAILURE) instead and dump a message Sep 13 19:14:14 PaulFertser: yep, starting a post-doc at DESY Sep 13 19:14:19 (in Hamburg Sep 13 19:14:21 +) Sep 13 19:14:32 Weiss: Neighbour :) Sep 13 19:14:43 Weiss: lucky you Sep 13 19:15:01 Defiant: you're from HH as well? there seems to be a bit of a group there (larsc as well) Sep 13 19:16:19 (subject, of course, to not flunking my PhD viva in a week's time) Sep 13 19:16:20 Weiss: its not a small town anyway Sep 13 19:18:09 * mrmoku booting Sep 13 19:23:23 dos1, still bad :( Sep 13 19:24:13 Defiant: indeed. but there will be OM Beers within a few weeks of my arrival :) Sep 13 19:24:34 OpenGL renderer string: Mesa DRI glamo 20090402 Sep 13 19:25:02 dos1: ping Sep 13 19:25:14 Weiss: Gröninger if you doesn't know it yet Sep 13 19:26:15 khiraly1: pong Sep 13 19:26:49 dos1: do you know when you create a .edc file with a swallow element, and you inject an another group into. Why the other group's size is 0,0 ? Sep 13 19:27:09 uhh... fso-abyss is wrong version :P Sep 13 19:27:20 ;xxx Sep 13 19:28:20 mrmoku: hey, is kernel package for gta01 supposed to update nand flash automatically? Sep 13 19:28:49 PaulFertser, no idea about gta01... it updates it for gta02 :P Sep 13 19:29:17 PaulFertser, update on gta01 is... impossible anyway (when running from flash) Sep 13 19:29:36 mrmoku: why do you think so? Sep 13 19:29:55 PaulFertser, to less space left :( Sep 13 19:30:15 Defiant: that's where you are, or where OM Beers happen? Sep 13 19:30:31 Weiss: thats where you can find good beer Sep 13 19:30:45 mrmoku: can you clarify a bit? Sep 13 19:31:15 PaulFertser, opkg needs some space on disk to do the update Sep 13 19:31:27 and I remember somebody stating it does not work on gta01 Sep 13 19:31:51 o shit Sep 13 19:31:51 mrmoku: for the update of only nand kernel? Sep 13 19:31:54 s/o/oh/ Sep 13 19:31:54 dos1 meant: oh shit Sep 13 19:32:03 PaulFertser, update in general Sep 13 19:32:11 mrmoku: :((( Sep 13 19:32:26 starting time of simple C app is incredible when you got used to starting time of simple python app ;x Sep 13 19:32:38 PaulFertser, it might depend on how much it has to update... and maybe updating one package at a time might work Sep 13 19:32:42 no idea though Sep 13 19:35:03 dos1, that version is better :) Sep 13 19:35:27 dos1, please build the index before rebuilding the image... so that I can try meanwhile Sep 13 19:37:08 mrmoku: NOTE: package libgsm0710mux-0.3.5+gitr42+df01e9e81775c1d7b515b9caff7c420fae59f140-r1: task do_configure: failed Sep 13 19:37:29 :( Sep 13 19:37:58 mrmoku: http://tinderbox.openembedded.net/public/logs/task/2861670.txt Sep 13 19:38:04 Requested 'vala-1.0 >= 0.7.5' but version of Vala is 0.7.3-fso1 Sep 13 19:38:10 mickey_away: !!!!! Sep 13 19:39:45 Defiant: Groniger's Braukeller? i'll keep that in mind.. Sep 13 19:40:13 configure: WARNING: In the future, Autoconf will not detect cross-tools Sep 13 19:40:15 whose name does not start with the host triplet. If you think this Sep 13 19:40:16 configuration is useful to you, please write to autoconf@gnu.org. Sep 13 19:40:18 hmm? Sep 13 19:40:59 freesmartphone.org: 03mickey 07cornucopia * radffd6f1fe59 10/fsogsmd/src/ (lib/modem.vala plugins/gsm_device/plugin.vala): fsogsmd: remove assert; use logger.critical instead Sep 13 19:41:43 dos1, nothing important Sep 13 19:41:55 it just says, that it will detect only Sep 13 19:42:15 armv4tl-softfloat-linux-gnueabi-[gcc,ld,...] Sep 13 19:42:40 mrmoku: so what about vala? Sep 13 19:42:42 e.g. first part in canonical way before [ will become mandatory Sep 13 19:43:56 dos1, hmm... let me check Sep 13 19:45:17 mrmoku: and what about setting BB_NUMBER_THREADS to "1" ? Sep 13 19:45:53 dos1, there is vala-0.7.5-fso2 Sep 13 19:46:03 last time I bumped vala was horrible :P Sep 13 19:48:03 dos1, ok I did it Sep 13 19:48:18 no idea what happens now ;) Sep 13 19:48:39 mrmoku: ok, building fso-abyss again Sep 13 19:48:55 mrmoku: do i need to update git trees? Sep 13 19:49:18 assuming yes Sep 13 19:49:19 :D Sep 13 19:49:38 dos1, yep update Sep 13 19:50:48 ERROR: Could not inherit file classes/shr.bbclass while parsing /home/shr/shr Sep 13 19:50:50 -testing/openembedded/recipes/libhito/libhito_git.bb Sep 13 19:50:51 mrmoku: ^^^ Sep 13 19:51:06 uhh... forgot one Sep 13 19:51:20 why does it want to build that anyway? or is it just parsing it? Sep 13 19:51:28 * mrmoku fixes Sep 13 19:52:00 dos1, well just remove the shr inherit Sep 13 19:52:02 mrmoku: just parsing Sep 13 19:52:04 shr.bbclass is no more Sep 13 19:52:12 mrmoku: but it continues Sep 13 19:52:23 mrmoku: just commit fix ;) Sep 13 19:54:36 mrmoku: E Sep 13 19:54:37 RROR: Could not inherit file classes/shr.bbclass while parsing /home/shr/shr Sep 13 19:54:39 -testing/openembedded/recipes/neod/neod_git.bb Sep 13 19:55:14 uhhh Sep 13 19:55:18 :P Sep 13 19:55:34 mrmoku: ;) Sep 13 19:55:39 those two are obsolete anyway (as far as I can tell) Sep 13 19:55:48 we should just dump them into the trashbin Sep 13 19:56:12 ;] Sep 13 19:59:13 mrmoku: patch to oe sent ;) Sep 13 19:59:47 * mrmoku checks patchwork Sep 13 20:00:34 freesmartphone.org: 03mickey 07cornucopia * r039e54425350 10/fsogsmd/ (3 files in 2 dirs): fsogsmd: depend on libfso-glib, add skeleton for implementing org.freesmartphone.GSM.Device.* Sep 13 20:04:20 xorg master broken again.. but we dont need dmx on neo :).. just disabled it in EXTRA_OECONF Sep 13 20:06:24 JaMa: r.e. your WSOD.. it's just temporary until you restart X? Sep 13 20:08:38 mrmoku: package-index built, building shr-lite-image Sep 13 20:10:30 JaMa: (that happens when KMS's scanout buffer gets "lost" somehow. currently that happens straight away if you try to use xrandr - the KMS driver is totally unprepared for that at the moment) Sep 13 20:11:51 i'll be back soon from neo Sep 13 20:12:28 Weiss: suspending now to test Sep 13 20:13:14 there isn't, for some reason, any xrandr stuff hiding in your suspend/resume scripts? Sep 13 20:14:59 neo is too nice today:) it refuses to suspend at all :) good boy Sep 13 20:15:29 It feels something bad will happen when resuming Sep 13 20:16:03 hmm... Sep 13 20:16:45 does anybody know where is the keyboard in QTMoko Sep 13 20:24:54 dos1|neo, works :) Sep 13 20:25:04 (at least it registers) Sep 13 20:25:19 dos1|neo, I will flash the new image tomorrow... off to bed now Sep 13 20:25:24 yay! Sep 13 20:25:30 dos1|neo, thanks and good night :) Sep 13 20:25:35 image is building ;) Sep 13 20:25:41 mrmoku|away: night! Sep 13 20:25:42 dos1|neo: image for gta01 is ready too ? ;) Sep 13 20:27:02 dos1|neo: ? Sep 13 20:27:33 soltys: no, will build after lite for gta02 Sep 13 20:28:02 dos1|neo: ok so I'll have sth to do tomorrow morning ;) Sep 13 20:28:21 dos1|neo: just send me a link to imaga via notice or sth ;) Sep 13 20:30:19 soltys: it will be in tests/mrmoku/ Sep 13 20:30:30 k Sep 13 20:30:40 I'll try to remember that ;P Sep 13 20:32:10 is it possible to instruct bitbake where I want to unpack source archive? Sep 13 20:33:20 freesmartphone.org: 03mickey 07cornucopia * r96b3570957a1 10/fsogsmd/src/plugins/gsm_device/plugin.vala: fsogsmd: add preliminary interface Sep 13 20:39:08 JaMa: there's "-c unpack" probably Sep 13 21:01:21 PaulFertser: you mean bitbake -c unpack? Sep 13 21:01:38 dos1|neo: what do you use on the neo as irc client? Sep 13 21:02:11 JaMa: yes Sep 13 21:02:33 PaulFertser: I need something else.. in SRC_URI, when is http://mirror/something.tar.gz and that tar.gz doesnt contain any directory like something-version then its extracted directly to $WORKDIR.. which can be partialy solved by S = ${WORKDIR}.. Sep 13 21:03:31 JaMa: oh... I see. Probably it's better to ask upstream maintainer to do the right thing in the first place. Sep 13 21:03:46 JaMa: and sorry for useless advice -- my OE knowledge is quite limited. Sep 13 21:04:03 PaulFertser: but ie pisi has own setup.py now.. which tries to pack everything in ${S}, which works for git when there is S = ${WORKIR}/git, but ends in neverending loop when installing ${WORKDIR}/image to ${WORKDIR}/image :) Sep 13 21:04:35 khiraly1: pidgin Sep 13 21:04:58 JaMa: i think PISI author is quite responsive, probably it's more productive to tell him about proper usage of python distutils or something... Sep 13 21:05:02 khiraly1: but i think i'll look at something lighter Sep 13 21:05:21 PaulFertser: Yeah I asked Michael to pack it with directory as its common for source tarballs, but from my curiosity I tried to found how to do that with bitbake :) Sep 13 21:06:09 PaulFertser: I even prepared setup.py for python setuptools for him.. but he doesn't want pisi in site-packages of python directory and prefers /opt/pisi :/ Sep 13 21:07:00 JaMa: i guess it goes against established practices... Sep 13 21:07:24 probably he needs a link to somewhere authoritative enough that describe that. Sep 13 21:07:25 PaulFertser: /opt or site-packages? Sep 13 21:07:35 JaMa: /opt Sep 13 21:07:38 (23.05.02) dos1|neo: khiraly1: but i think i'll look at something lighter <-- writing a simple client using elementary and libtelepathy Sep 13 21:08:02 khiraly1: i think someone does that Sep 13 21:08:08 its pity there is just one response tohttp://lists.openmoko.org/pipermail/community/2009-August/053413.html Sep 13 21:08:09 apathy IIRC Sep 13 21:08:21 /opt is reserved for the installation of add-on application software packages. Sep 13 21:08:35 ^^ quote from FHS Sep 13 21:08:55 dos1|neo: with some sip on top of it?;) Sep 13 21:09:17 from my understanding of FHS is /opt reserved for packages which cannot be integrated to FHS properly.. which I guess python applications can be :) Sep 13 21:09:18 khiraly1: dunno Sep 13 21:09:57 PaulFertser: still there's quite different POVs about what is a add-on package :-P Sep 13 21:10:02 http://www.amacleod.me.uk/blog/ Sep 13 21:10:44 JaMa: especially when they're installed and removed by package manager so yes, i agree. Sep 13 21:12:05 khiraly1: gimmiegimmegimmie! Sep 13 21:12:26 dos1|neo: there is a video too Sep 13 21:12:41 http://www.amacleod.me.uk/blog/wp-content/uploads/2009/08/apathy1.ogv Sep 13 21:12:42 JaMa: (only one response) no wonder, he asked in a wrong place. Sep 13 21:16:45 PaulFertser: he said that he will think about it during weekend trip, so we'll see if he changes his mind :), you can sent that quote from FHS as reply :) Sep 13 21:16:57 ~opimd-utils Sep 13 21:16:57 somebody said opimd-utils was not phone suite, it's only set of test scripts for developers. You can use it as phone apps, but without any warranty. Many things are not implemented, *AND THEY PROBABLY WON'T BE*. That's what dos1 said. Sep 13 21:18:46 dos1|neo: http://www.openapathy.org/ Sep 13 21:20:50 dos1|neo: everything is python;-) Sep 13 21:22:16 khiraly1: looks interesting, especially when they get the SIP part workin Sep 13 21:22:20 freesmartphone.org: 03Frederik.Sdun 07dbus-hlid * rbbb08f7e462f 10fso-monitord/src/monitor.vala: Listen on NameLost instead of NameOwnerChanged Sep 13 21:22:21 freesmartphone.org: 03Frederik.Sdun 07dbus-hlid * r6c1ada012a95 10fso-monitord/src/monitor.vala: Listen on NameAqcuired instead of NameOwnerChanged Sep 13 21:22:22 freesmartphone.org: 03Frederik.Sdun 07dbus-hlid * r470097b67b65 10fso-monitord/src/monitor.vala: Add workaround for Vala Bug with construct properties Sep 13 21:22:22 working* Sep 13 21:22:49 sybren: yeah, and as everything is in python and elementary, there should be no big surprise in the code Sep 13 21:23:27 khiraly1: I'm not too fond of elementary, mostly because you can only focus edit fields when you specifically click on the text in the field instead of the field itself or its label. Sep 13 21:24:07 sybren: hmm? Sep 13 21:24:33 sybren: look at pim changing in shr-settings Sep 13 21:24:37 dos1|neo: at least many apps on SHR work that way, and I've been told that it's caused by the way elementary works Sep 13 21:24:39 or shr-installer Sep 13 21:24:48 or opimd-contacts Sep 13 21:24:50 sybren: so how set focus on an empty textfield then? Sep 13 21:25:04 clicking on field just works Sep 13 21:25:29 dos1|neo: good, so it's not elementary but lazy, usability-oblivious application developers then? Sep 13 21:25:44 sybren: those are SHR apps problems. but they are obsolete and going to be rewritten Sep 13 21:25:46 DocScrutinizer-8: that's exactly what I'm having trouble with on so many apps Sep 13 21:26:03 dos1|neo: that's good to hear :) Sep 13 21:27:42 and I'm very fond of Python, so I'm eagerly looking forward to SIPping with Apathy Sep 13 21:28:02 sybren: if your diagnosis were correct, it's not difficult but impossible Sep 13 21:28:34 DocScrutinizer-8: it was near-impossible, there were a few pixels left where I could click Sep 13 21:28:58 sybren: what I know is you need to click several times to set focus to name field in contacts Sep 13 21:29:35 DocScrutinizer-8: yeah, I had trouble adding a new contact with shr-contacts Sep 13 21:29:40 or you simply don't notice focus is set, as cursor isn't removed from prev field Sep 13 21:29:56 sth like that Sep 13 21:30:04 DocScrutinizer-8: omg Sep 13 21:30:17 anyway that's a known problem I think Sep 13 21:31:30 DocScrutinizer-8: as long as "a known problem" means "will be solved in the not-so-far future" I'm a happy man :) Sep 13 21:32:40 well, seems I got used to it somehow, as I never noticed that issue last few months Sep 13 21:34:09 DocScrutinizer-8: to replay, start shr-contacts, click on 'New'. You'll see two huge edit boxes. If you click in the middle of the lower one, the top one still has focus Sep 13 21:34:37 you can't even press the tab button to move to the "Number" field, as it'll just insert a tab in the name Sep 13 21:36:36 yes. I got used to click to the place where cursor would appear. And it usually does, one pixel wide ;-P. TAB is a real flaw as well Sep 13 21:37:39 DocScrutinizer-8: having to click on a single pixel while there is a HUGE box outlined is horrible. Sep 13 21:37:57 ac Sep 13 21:38:03 k Sep 13 21:38:27 I don't know who could even think of it as "working properly" Sep 13 21:39:06 I mean, I've seen bad designs coming from a graphics designer "because it looks cool", but aren't we programmers not supposed to enjoy things that work well? Sep 13 21:39:38 Weiss: restarting Xorg doesn't change WSOD.. its there until reboot... but Xorg starts ok (my guess based on xorg logs, but nothing shown). Sep 13 21:40:10 DocScrutinizer-8: anyway, I'm sure things will get better with the opimd based utils ;-) Sep 13 21:40:50 Weiss: some glamo reset voodoo on your mind? Sep 13 21:40:57 sybren: try opimd-utils' contacts, it works :-D Sep 13 21:41:24 sybren: at least wrt focus :-P Sep 13 21:41:51 DocScrutinizer-8: great :) Sep 13 21:42:06 DocScrutinizer-8: now I'll join the missus in bed, ttyl! Sep 13 21:42:26 cya Sep 13 21:42:35 soltys: gta01 image is built Sep 13 21:42:50 now eten-m800, i wonder if it'll build ;D Sep 13 21:43:05 dos1|neo: ok will see it tomorrow ;) now I'm going to bed ;) Sep 13 21:43:41 i'm already in bed ;D Sep 13 21:43:52 dos1|neo: if only it would really work (modem) I'd give M800 image a try Sep 13 21:44:15 PaulFertser: thanks for wifi once again :) Sep 13 21:44:39 DocScrutinizer-8: dunno how it works Sep 13 21:44:46 maybe ask morphis Sep 13 21:44:56 dos1|neo: i wonder why it didn't work before the patch that was reverted... Is it possible your school changed AP without you noticing? Sep 13 21:45:00 it was built once Sep 13 21:45:28 PaulFertser: hehe, I suggested same Sep 13 21:45:42 DocScrutinizer-8: i'm just repeating your suggestion Sep 13 21:45:44 :) Sep 13 21:45:45 PaulFertser: a) no, AP is the same. b) it didn't worked in home, and now it works Sep 13 21:46:24 freesmartphone.org: 03mickey 07cornucopia * r1407a59e2250 10/fsogsmd/src/plugins/gsm_device/plugin.vala: fsogsmd: gsm_device: register dbus object Sep 13 21:46:35 PaulFertser: with 2.6.24 it worked at home and in school Sep 13 21:46:50 dos1|neo: the offending commit was applied on Apr 2. Sep 13 21:46:50 PaulFertser: with 2.6.28 only in school Sep 13 21:46:52 PaulFertser: listing APs works great now here. Just it borks on associating to my AP. Before it didn't list any scan result Sep 13 21:46:56 dos1|neo: so it should have worked before that. Sep 13 21:46:56 freesmartphone.org: 03mickey 07cornucopia * rc9724c50cedd 10/libfsoframework/fsoframework/interfaces.vala: fsoframework: interfaces: add well-known paths for GSM object Sep 13 21:47:08 PaulFertser: then i didn't tested it Sep 13 21:47:15 ~seen ptitjes Sep 13 21:47:16 ptitjes was last seen on IRC in channel #openmoko-cdevel, 2d 11h 18m 37s ago, saying: 'Ainulindale: ping'. Sep 13 21:47:34 PaulFertser: when i tested it again, it wasn't working at all on 2.6.29 Sep 13 21:47:37 DocScrutinizer-8: the backtrace would be interesting... Sep 13 21:47:42 and now it works Sep 13 21:47:56 PaulFertser: :-S Sep 13 21:48:41 PaulFertser: I rather attach debug board than fiddling with enabling ramconsole by hand (push push hint hint) Sep 13 21:48:41 PaulFertser: that's whole story, probably i just missed time when it worked on 2.6.29 Sep 13 21:48:46 DocScrutinizer-8: probably you can just hook up any of your devices to a dboard and save the output Sep 13 21:50:00 mickeyl: i see fsogsmd is going well ;) Sep 13 21:50:32 dos1|neo: hehe, nah, just some cleanup, didn't touch it for some months Sep 13 21:50:44 need to think now how to fit the logic in async Sep 13 21:50:46 BTW, funny thing: it's now an official holiday in russia, the day of the programmer. (not that it's a day off for anybody but still) Sep 13 21:50:58 mickeyl: tell me if there's anything to test wrt gprs Sep 13 21:51:15 Will be celebrated on 13 Sep on regular years and 12 Sep on leap years. Sep 13 21:51:32 PaulFertser: ouch XD Sep 13 21:51:33 DocScrutinizer-8: will do. didn't find a chance to reproduce the hang yet Sep 13 21:51:36 mickeyl: any points against using fsodeviced, fsonetworkd and fsotimed in shr? Sep 13 21:51:47 dos1|neo: no points against fsonetworkd and fsotimed Sep 13 21:51:47 DocScrutinizer-8: yep, 255th day counting from 0. Sep 13 21:51:58 dos1|neo: fsodeviced is big, needs testing Sep 13 21:52:18 dos1|neo: if you don't mind a regression, start using it Sep 13 21:52:18 PaulFertser: lol Sep 13 21:52:47 DocScrutinizer-8: and on regular years it will match my fucking "birthday", so probably everybody who knows me won't forget about it :( Sep 13 21:53:07 mickeyl: are recipes in shr/import ok? Sep 13 21:53:11 hehehehe Sep 13 21:53:31 DocScrutinizer-8: i have no idea. i fixed everything in org.oe.dev, didn't look at shr/imporrrt Sep 13 21:53:37 PaulFertser: you have birthday on 255th day? ;D Sep 13 21:53:39 err Sep 13 21:53:43 s/doc/dos1/ Sep 13 21:53:52 mickeyl: heh Sep 13 21:54:08 dos1|neo: only on non-leap years. Sep 13 21:54:15 nah, never!! Sep 13 21:54:16 PaulFertser: happy birthday, I'm not too late :-) Sep 13 21:54:23 Ainulindale: screw you Sep 13 21:54:32 Well did you think I'd miss that? :-p Sep 13 21:54:35 Ainulindale: i get very upset every year Sep 13 21:54:38 about it Sep 13 21:54:41 Heh Sep 13 21:54:44 PaulFertser: but still nice ;D Sep 13 21:54:45 substitute me by dos... that won't fly :-P Sep 13 21:54:46 Don't be mad Sep 13 21:54:51 At least you're not losing your hair Sep 13 21:54:52 I do :-p Sep 13 21:54:57 Ainulindale: yet Sep 13 21:55:20 PaulFertser: *n*yet would be a better thing considering what the hair loss will do to your morale :-) Sep 13 21:55:22 Ainulindale: but after i loose all my hair i can make a nice GNU tatoo on scalp, would be fun. Sep 13 21:55:43 PaulFertser: By the way I'm nuts with locative/accusative/genetive right now :-p Sep 13 21:56:46 mickeyl: i'll try to build them and test in this week Sep 13 21:56:56 dos1|neo: excellent Sep 13 21:57:22 and i hope we'll get shr based on oe.dev sokn ;/ Sep 13 21:57:30 s/sokn/soon/ Sep 13 21:57:31 dos1|neo meant: and i hope we'll get shr based on oe.dev soon ;/ Sep 13 21:57:32 yeah, my hope as well Sep 13 21:57:44 Ainulindale: prepositions in english are not much better Sep 13 21:57:53 PaulFertser: well we have the same in french Sep 13 21:57:55 mickeyl: is it worth to give M800 a try (will it work as a phone? with AHR/FSO?) Sep 13 21:57:56 PaulFertser: So I'm used to it Sep 13 21:58:25 PaulFertser: Though last week I bought a phonetic book I improved a lot :-) Sep 13 21:58:32 DocScrutinizer-8: eten? forget about it. no way to communicate with the modem and no more momentum. Sep 13 21:58:47 ok :-S Sep 13 21:58:48 PaulFertser: as well as a very good book about grammar difficulties for french with russian language Sep 13 21:58:56 DocScrutinizer-8: the only hope is to wait for the palm pre Sep 13 21:59:04 next month according to my info Sep 13 21:59:20 Did you get a preview mickeyl? Sep 13 21:59:47 Ainulindale: unfortunately not, but hanging around @ #webos-internals Sep 13 21:59:50 so who need a fsckng winmoob phone with an excellent 640*480 screen? Sep 13 22:00:09 I have a "brick" here to get rid off Sep 13 22:00:17 of even Sep 13 22:01:05 DocScrutinizer-8: send it to me if you want ;D Sep 13 22:01:59 DocScrutinizer-8: SHR, even without phone functionality, will be nice on it to just code something ;D Sep 13 22:02:12 hmmm Sep 13 22:02:39 probably the kbd is way too crappy Sep 13 22:03:14 DocScrutinizer-8: now i'm using illume terminal keyboard Sep 13 22:03:29 what's better? ;D Sep 13 22:03:42 illume Sep 13 22:03:48 hmm Sep 13 22:03:59 is it really as crappy? ;x Sep 13 22:05:20 yep. completely flat surface black in black with keycode printed upper left corner. so if you press on the symbol, you probably hit the wrong key :-(( Sep 13 22:05:48 dos1|neo: have you tried literki? Sep 13 22:06:10 PaulFertser: yup, i have it even installed Sep 13 22:06:29 mickeyl: what if some guys found out how to communicate with the modem on m800? Would it immediately turn it into something useful? Sep 13 22:06:45 PaulFertser: but for some reason i don't like it... maybe i just got used to illume kbd Sep 13 22:07:20 but literki saved me few times when theming enlightenment Sep 13 22:07:35 fwew times i broke illume ;) Sep 13 22:07:48 and i was fixing it by literki ;D Sep 13 22:08:35 PaulFertser: good question. Yes and No. Yes: it would be useful since the rest seems pretty much the same as a GTA02 hence it would relatively easily be on the same level but with a much better modem and a hardware keyboard. No: since a) it has bad manufacturing quality. b) the hardware is already outdated hence it's unlike a community would build around this device. Sep 13 22:09:15 i'd rather have us focusing on selling hardware Sep 13 22:09:25 err Sep 13 22:09:29 hardware that's still in production Sep 13 22:09:29 i mean Sep 13 22:09:32 mickeyl: i see... asking because it would probably take a day or two for some really bright hackers to finally reverse the modem interface, i guess. Sep 13 22:09:45 PaulFertser: unlikely. very bright hackers already looked for months Sep 13 22:09:48 including Harald Welte... Sep 13 22:10:15 mickeyl: i wonder how much time he spent on actual looking at this; he has so many other important tasks... Sep 13 22:10:24 quite a substantial amount of time Sep 13 22:10:27 but anyways Sep 13 22:10:32 there's some short timeframe of 'momentum' Sep 13 22:10:37 for this device, it's passed IMO Sep 13 22:10:38 I guess there's no way to use an ndiswrapper approach here? Sep 13 22:10:40 what did they do on it? o_O Sep 13 22:11:27 no, that's probably too hard to recreate all the API the dll expects Sep 13 22:11:58 someone designing m800 said "let's hide modem for every linux hacker, let's make them looking stupid!", or what? ;x Sep 13 22:12:44 dos1|neo: it's basically what you would call a softmac if that were wifi Sep 13 22:12:57 aiui Sep 13 22:12:57 mickeyl: from my uneducated look at http://gnufiish.org/trac/wiki/Ericsson_3.5G_Modem i see that there's a lot known about the modem already, that's why it seems that just a little bit left. Sep 13 22:13:11 PaulFertser: yes, they just can't talk to it Sep 13 22:13:17 "just" Sep 13 22:13:26 DocScrutinizer-8: oh. that's different case then Sep 13 22:13:55 i'd see more value in helping the HTClinux folks Sep 13 22:14:00 _if_ someone wants to do RE Sep 13 22:14:14 otherwise, waiting for the Palm Pre, freeing that Sep 13 22:14:34 although you have a hard time fighting against the androidisms :D Sep 13 22:14:48 so i can only hope for someone funding me palm pre with ability to install shr ;x Sep 13 22:14:48 (on htclinux) Sep 13 22:14:50 eew Sep 13 22:15:16 mickeyl: yup, how's FSO on HTC Dream? Sep 13 22:15:37 mickeyl: what's left? Sep 13 22:15:39 phone works fine Sep 13 22:15:43 problems with X Sep 13 22:15:45 touchscreen Sep 13 22:15:47 audio Sep 13 22:15:53 camrea Sep 13 22:16:01 BT / wifi unknown Sep 13 22:16:10 ask Gnutoo when you have a chance Sep 13 22:16:17 what kind of problems? Sep 13 22:16:20 yayayay Sep 13 22:16:35 i don't know details. varies between not working, no refresh, slow, etc. Sep 13 22:16:47 eh, dunno if i'm lowlevel enough Sep 13 22:17:32 hi Sep 13 22:17:33 Ainulindale proposed me (well, proposed an idea) funding htc dream by bearstech to get fso working on it Sep 13 22:17:55 but i still doubt igjf i can really help Sep 13 22:17:56 yeah, not that the dream would be good quality hardware, but better than many alternatives Sep 13 22:18:10 s/igjf/if/ Sep 13 22:18:11 dos1|neo, imprecisions on the touchscreen on kdrive and xorg that doesn't work...I'll retry as soon as I can Sep 13 22:18:26 dos1|neo meant: but i still doubt if i can really help Sep 13 22:18:33 xorg not working means not even with xfbdev Sep 13 22:18:44 audio should work now as qualcomm has alsa in its repo Sep 13 22:19:13 Gnutoo: nice! Sep 13 22:19:17 I'll re-start working on SHR on the htc dream soon Sep 13 22:19:28 I was working on android to get it working fast Sep 13 22:19:50 and freely(in freedom,100% free software running on the main CPU) Sep 13 22:20:09 Gnutoo: eeh? fast and android in same sentence? Sep 13 22:20:32 I'll go to sleep soon Sep 13 22:20:36 you mean free? Sep 13 22:20:38 hehe Sep 13 22:20:40 yes it's possible Sep 13 22:20:49 basically you use base android Sep 13 22:20:58 patch the sound not to use libhtc_acoustic Sep 13 22:21:18 use the reference ril... Sep 13 22:21:22 patch it for the pin Sep 13 22:21:24 etc... Sep 13 22:22:33 i would like to help... but as i have some amount of time for hacking in python, i'm not sure i'll have enough time to learn lowlevel details and to sitting near PC, as i won't be able to hack kernel in bed using just that device ;P Sep 13 22:23:46 but having SHR running on Palm Pre or HTC Dream would be really, really nice! Sep 13 22:23:54 there is an initial FSO implementation for the modem Sep 13 22:24:09 so Gnutoo, keep the good work going! :) Sep 13 22:24:24 palm pre would be great too...unfortunately not available in europe Sep 13 22:25:57 Gnutoo: won't it be available at all in europe? Sep 13 22:27:40 dos1|neo: it's announced for christmastime by o2 in germany Sep 13 22:27:48 ok, going to sleep now Sep 13 22:28:07 Slyon: oh. nice :) Sep 13 22:28:29 I don't know Sep 13 22:28:37 I didn't follow palm pre closely Sep 13 22:28:39 probably it'll be too expensive for me Sep 13 22:28:55 but if someone buys it...do not use it as a phone as is Sep 13 22:29:01 as it spys the user a lot Sep 13 22:29:11 send a lot of personal details on the internet Sep 13 22:29:18 but it'll be still nice to have SHR working on it Sep 13 22:29:28 Gnutoo: i heard about it Sep 13 22:29:33 ok Sep 13 22:29:44 Gnutoo: so having SHR on it will be even nicer ;) Sep 13 22:29:50 it would liberate it and make an usable phone of it Sep 13 22:30:15 Gnutoo: apparently it sends your df -h and partition layout and maybe a few other things Sep 13 22:30:22 Gnutoo: there are a few articles on it Sep 13 22:30:28 some fsf guys are very interrested in palm pre...I wanted to help there too but I was too busy with the htcdream Sep 13 22:31:18 what about FSO on iPhone? ;D Sep 13 22:32:04 for that i would maybe consider buying iPhone ;) Sep 13 22:32:10 lol the GNU/Linux port is not very advanced Sep 13 22:32:25 hehe, i know Sep 13 22:32:37 but still, it could be nice ;D Sep 13 22:33:10 as now i'm not interested in iPhone at all ;) Sep 13 22:33:46 moreover it seems abandonned Sep 13 22:33:54 I'll go bye Sep 13 22:34:04 bye Sep 13 22:34:07 * dos1|neo too Sep 13 22:34:09 cu Gnutoo Sep 13 22:34:12 good night ;) Sep 13 22:34:17 thanks bye Sep 13 22:46:45 freesmartphone.org: 03mickey 07cornucopia * ra84200925708 10/fsogsmd/ (12 files in 7 dirs): fsogsmd: convert to using Gee, add typefactory for mediators in Modem class Sep 13 23:03:20 mickeyl: what's the process here? how do you go from python to vala(or c) Sep 13 23:04:18 in general, this depends on the amount of lessons learned during the python implementation Sep 13 23:04:33 for many subsystems, it's a pretty straightforward process Sep 13 23:04:46 i could use most algorithms Sep 13 23:04:54 and just needed to translate them into Vala Sep 13 23:05:02 (by hand, of course, there's no automated process here) Sep 13 23:05:12 in some cases, the abstractions don't match Sep 13 23:05:22 e.g. when you're using lot of the python dynamicity Sep 13 23:05:27 fsogsmd will be the hardest part Sep 13 23:05:50 since i used lots of magic here, sticking callbacks into queues, processing stuff, getting results, mapping them to callbacks, calling back with results Sep 13 23:06:54 so with the new Vala dbus async serverside stuff, i'm checking how to best fit the ogsmd architecture in Vala Sep 13 23:08:33 at the end of the day though, it's a complete rewrite from scratch Sep 13 23:10:16 same API? Sep 13 23:11:02 mickeyl: I wonder if all this gsmd stuff isn't just a lexem analyzer Sep 13 23:11:40 on the dbus level, yes. apart from minor improvements, cornucopia is going to be a drop-in replacement. API changes will come in the phase after the rewrite again. Sep 13 23:12:33 DocScrutinizer-8: hehe, yeah. i thought long about all that stuff, seriously considering to formulate all that with high level state machines and transitions. the result was that it's not worth it Sep 13 23:12:44 there's too many exceptions to rules Sep 13 23:12:47 too many special cases Sep 13 23:12:50 too many vendor specifics Sep 13 23:13:06 :-( Sep 13 23:13:06 especially when you move from At to integrating non-At modems Sep 13 23:13:19 ouch Sep 13 23:13:23 yeah, it's a lot of boring tedious handwriting code Sep 13 23:13:53 i spent a month to come up with a generic framework Sep 13 23:14:00 threw it all away after implementing 10 commands Sep 13 23:14:08 since it didn't fit Sep 13 23:14:09 :/ Sep 13 23:14:17 so, back to good handwritten code :) Sep 13 23:14:25 hrhrr Sep 13 23:14:34 same thing that happened to the ofono people Sep 13 23:14:43 hehehe Sep 13 23:14:51 *nod* Sep 13 23:14:54 they couldn't handle the msm's because of the strange line endings Sep 13 23:14:58 hehe Sep 13 23:15:02 sounds so familiar to me Sep 13 23:15:07 the architecture i have in ogsmd is very good Sep 13 23:15:15 it already shows that it handles lots of different types Sep 13 23:15:28 yes, it's a bit more layers than i'd love it to be, but it's necessary Sep 13 23:15:36 and this fine architecture will still be present in fsogsmd Sep 13 23:15:42 just fast as lightening ;) Sep 13 23:15:49 [compared to ogsmd] Sep 13 23:15:57 they also went the same way on mux it looks like, with a userspace library Sep 13 23:16:19 hmm, i guess they didn't even look at abyss Sep 13 23:16:20 if we had a lightning fast UMTS on the other end.... *sigh* Sep 13 23:16:22 and my libgsm0710mux Sep 13 23:16:51 DocScrutinizer-8: we'll have. sooner or later FSO will run on competitive hardware Sep 13 23:16:55 i'm not willing to give up that dream Sep 13 23:17:08 and then I'm starting with UI Sep 13 23:17:22 UI? Sep 13 23:17:23 UI for which? that didn't follow Sep 13 23:17:48 the day FSO runs fine on competetive current hardware, I'm dropping out of middleware Sep 13 23:17:52 starting to develop applications Sep 13 23:17:59 aah Sep 13 23:17:59 since that's what i enjoy more Sep 13 23:18:06 yep Sep 13 23:18:35 i work on middleware to enable us to concentrate on User Interfaces Sep 13 23:18:37 I don't know, with dzo's progress it might not be time to give up on msm Sep 13 23:18:52 well Sep 13 23:18:56 i haven't given up Sep 13 23:18:59 i'm just very frustrated Sep 13 23:19:03 that it all takes so long Sep 13 23:19:05 it's been very frustrating getting the kernel from google and finding that it would take another year to get where we are Sep 13 23:19:11 exactly Sep 13 23:19:16 maybe the investment to send one to dzo would have been worth it Sep 13 23:19:24 when dcordes told me they had modem access on the Touch Pro Sep 13 23:19:26 i bought one Sep 13 23:19:28 but then non-A people would have potentially lost out Sep 13 23:19:36 we did have modem access, not audio Sep 13 23:19:38 thought it would take the kernel guys some weeks to finish the remaining bits Sep 13 23:19:40 and we said that Sep 13 23:19:44 i know i know Sep 13 23:19:45 no pun here Sep 13 23:19:47 my own fault Sep 13 23:20:02 having this great hardware not being able to use it Sep 13 23:20:13 that was just a huge downer Sep 13 23:20:26 actually, I don't know if dzo is interested in qwerty devices but I'm sure he doesn't have a touch pro Sep 13 23:20:38 he's working on gsm diamond Sep 13 23:21:15 well, you heard druidu Sep 13 23:21:29 not sure what you are referring to Sep 13 23:21:32 I seriously doubt he will do anything after android works Sep 13 23:21:35 maybe I missed something Sep 13 23:21:52 well, druidu is moving to aurora base which is much more linux-like Sep 13 23:21:56 i don't buy his idea of "lets get it to work with android, then rewrite it for open interfaces" Sep 13 23:21:57 alsa, v4l, etc. Sep 13 23:22:06 if that happens, I'll be glad Sep 13 23:22:10 i just have serious doubts Sep 13 23:22:29 but anyways, the hardware is still sitting here and waiting for the day it can be used. Sep 13 23:22:30 I get it, I'm too much of an optimist until hitting the same walls Sep 13 23:22:45 :) Sep 13 23:24:05 if this was the first time, i'd not be so frustrated Sep 13 23:24:08 but you see i'm coming from openezx Sep 13 23:24:19 4 years being 'nearly there' Sep 13 23:24:24 but yet so far away Sep 13 23:24:30 so much promise Sep 13 23:24:55 yep Sep 13 23:26:24 * DocScrutinizer-8 wants a food replicator ala startrek in hos next phone :-/ Sep 13 23:27:17 no 24/7 gasstation in vicinity Sep 13 23:30:03 hmm, don't talk about food Sep 13 23:30:07 getting hungry as well Sep 13 23:51:58 Weiss: where is glamo_gem_name_buffer defined? cannot build latest xf86-video-glamo-kms **** ENDING LOGGING AT Mon Sep 14 02:59:57 2009