**** BEGIN LOGGING AT Mon May 31 02:59:56 2010 May 31 07:16:40 Heinervdm: hi, just noticed that koen bumped webkit-gtk and webkit-efl seems quite close http://www.cmake.org/pipermail/cmake/2010-April/036494.html May 31 07:22:02 JaMa|Wrk: cc was keeping an eye on webkit-efl i think May 31 07:22:15 we can ask him if he tried it with ventura May 31 07:22:28 and if it works we can switch May 31 07:22:46 but we should wait for his ack, as he is the only user of webkit-efl May 31 07:27:27 ok, I was remembering you keeping eye, so I wanted to ask you first :) May 31 07:47:22 wget@drugs! .... 129% 259K -48s May 31 07:51:42 mh... still getting more data despiting passing the Content-Length:-Limit? May 31 07:51:49 despite. May 31 07:53:46 yes.. but gcc mirrors had it usually right.. or at least this is first time I noticed May 31 10:29:33 heyho May 31 10:30:15 hi May 31 10:34:17 moin May 31 10:59:20 morphis: I seen your mail on openmoko devel ML, I am very interested and I would like to know if it's working also with Palm Pre Plus devices May 31 11:00:09 valos: it should be, cause the modem is as far as I know nearly the same May 31 11:00:42 we only don't support cdma Pre's atm May 31 11:00:59 so Pre and Pre Plus should be work May 31 11:01:16 ok cool May 31 11:01:58 valos: you are interested in using FSO or developing? May 31 11:02:24 morphis: both May 31 11:02:44 great May 31 11:03:08 tell me if you want to know more about where to start etc. May 31 11:04:22 we need every help we can get at the moment as there is very much to do until we can do a first release May 31 11:04:35 and we need something visible to get more people involved May 31 11:09:48 valos: do you have a Pre Plus? May 31 11:10:06 morphis: no, not yet May 31 11:10:15 so only a Pre? May 31 11:11:03 no, neither of the two May 31 11:12:54 but, I intend to buy a Palm Pre Plus this week May 31 11:14:28 ah great so you can start to work on FSO really soon :) May 31 11:14:41 YEAH! May 31 11:14:49 * mickey|office desperately needs help May 31 11:14:58 (and time, but that's another story...) May 31 11:15:25 mickey|office: :) May 31 11:15:42 valos: do you have any preference with which thing you want to start? May 31 11:16:13 morphis: absolutely no preference May 31 11:16:35 mickey|office: I am currently implementing a simple state machine for the audio routing on the Pre May 31 11:16:51 mickey|office: playya already thought about to redesign the audio API May 31 11:18:19 valos: ok, then we will find something suitable for you :) May 31 11:30:47 morphis: OK ;-) I contact you as soon as I receive my pre May 31 11:31:05 great May 31 11:31:25 hi, I've got an issue rebuilding a single package with bitbake / SHR-makefile. Building the whole image works ok though. Is this a good place to ask? May 31 11:33:17 fsfs: go ahead May 31 11:33:46 namely, configure fails with "NOTE: nothing to configure", and the directory this is executed in is empty (not the source directory) May 31 11:34:48 ie, sources in shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/gnutls-2.8.5-r0/gnutls-2.8.5, devshell in shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/gnutls-2.8.5-r0/ZEDAT May 31 11:35:34 I've set (in local-builds.inc): S_pn-gnutls = "${WORKDIR}/ZEDAT/${PN}" May 31 11:36:07 (ah, devshell might drop me in .../ZEDAT/gnutls/, empty just the same) May 31 11:36:24 morphis: ok, if you have a proposal, please forward it to smartphones-standards May 31 11:36:51 fsfs: sounds more OE-ish than SHRish May 31 11:37:01 either way, that's too hardcore OE for me :) May 31 11:37:08 all the run-scripts in temp/ cd into .../ZEDAT/gnutls/ as well May 31 11:37:28 zub: yea, that's what I suspected... May 31 11:38:05 fsfs: ${WORKDIR}/ZEDAT/${PN} May 31 11:38:21 ${PN} = gnutls May 31 11:38:33 so it does exactly what you told it May 31 11:39:00 mickey|office: currently no real proposal as I don't have enough time, all code is currently very Pre-specific May 31 11:39:08 Heinervdm: I suppose, but how come the sources end up elsewhere? how should configure get there? May 31 11:39:12 mickey|office: pushed it May 31 11:39:22 freesmartphone.org: 03morphis 07cornucopia * r2dc480605051 10/fsodeviced/src/plugins/router_palmpre/plugin.vala: fsodeviced: rework route_palmpre plugin as simple state machine reacting on several events May 31 11:39:27 Heinervdm: I just followed the local-builds.inc setup from the wiki... May 31 11:40:39 fsfs: $WORKDIR is always shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/${PN}-${PV}-${PR} May 31 11:41:06 and the source is extracted in that directory May 31 11:41:08 morphis: ok. in general i don't mind revamping the audio API. it was based on the 'scenario' thought which -- what we now know -- is too limited for a smartphone May 31 11:41:22 jepp May 31 11:41:36 Heinervdm: ok. Sources were extracted / copied to shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/${PN}-${PV}-${PR}/${PN}-${PV} it seems May 31 11:41:56 I even take a look at the scenarios used on the om-phones and think we can work on a gernal state-machine-like approach May 31 11:42:06 good May 31 11:42:24 but first we have do some more important things :) May 31 11:42:26 fsfs: i think you pointed SRC_URI to a directory, then bitbake will copy this directory to $WORKDIR May 31 11:42:26 one thing i learned from the iPhone OS is that the only sane thing to follow is a 'last-in-counts' approach May 31 11:42:46 Heinervdm: yes May 31 11:43:25 Heinervdm: now should S_pn-gnutls in local-builds.inc point to that directory as well? May 31 11:43:26 fsfs: so your sources end up in shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/${PN}-${PV}-${PR}/$DIRECTORYNAME May 31 11:43:35 Heinervdm: good you are here, do you have some time to work on tslib for the Pre? May 31 11:43:59 morphis: i still have no FSO image on my pre... May 31 11:44:00 Heinervdm: it need some finetunning May 31 11:44:15 Heinervdm: hm, install one? :) May 31 11:44:25 fsfs: depends on what you are trying to do May 31 11:44:48 morphis: wich sort of finetuning? May 31 11:44:55 Heinervdm: I want to change a constant in the source, and see if that fixes a bug I experience May 31 11:45:29 fsfs: and why do you change $S for that? May 31 11:45:32 Heinervdm: moving the cursor works, but clicking not really May 31 11:45:45 and I don't have the time to do that May 31 11:45:55 Heinervdm: got me. what's $S? May 31 11:46:12 morphis: hmm, i thougth that tslib will do that... May 31 11:46:27 fsfs: S_pn-gnutls defines $S May 31 11:46:27 Heinervdm: http://shr-project.org/trac/wiki/Getting%20started%20developing%20SHR says I need to change that line in conf/local-builds.inc... May 31 11:46:44 Heinervdm: and we have to work on the problem with the margins May 31 11:46:54 Heinervdm: $S is "location of source"? May 31 11:47:17 fsfs: location of the WORKINGSOURCE May 31 11:47:44 morphis: yes, but i have no idea how to get rid of them... May 31 11:47:47 Heinervdm: I can upload the image if you need one May 31 11:47:49 hm May 31 11:48:09 morphis: i will wirte an email to tslib ml and ask them how to do a click May 31 11:48:28 or grab kergoth in #e May 31 11:48:31 #oe, that is May 31 11:48:45 he's looking for a new maintainer btw. May 31 11:48:46 clicking is working sometimes, but not everytime May 31 11:49:18 mickey|office: maintaining tslib .. ohoh ... we need multitouch support there :) May 31 11:50:30 mickey|office: i asked him some weeks ago and he said that he hasn't looked at tslib code for a very long time :) May 31 11:52:38 oh great, mainline tslib has cy8mrln support now :) May 31 11:52:49 how? May 31 11:53:24 kergoth merged playyers branch May 31 11:53:33 but not my last changes May 31 11:53:43 fsfs: where do you have your sources with that changed variable? May 31 11:55:20 JaMa|Wrk: the sources get correctly copied to shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/${PN}-${PV}-${PR}/gnutls-2.8.5 (from shr-unstable/ZEDAT/gnutls-2.8.5) May 31 11:55:30 fsfs: if it's the same directory name as unpackad source archive then just update SRC_URI May 31 11:55:59 mickey|office: btw. when you write an entry with AT commands on the SIM card, do you get the result as direct response or event? May 31 11:56:17 JaMa|Wrk: Heinervdm: I just don't set S_pn-gnutls at all now, and it seems configure is doing something :-) May 31 11:56:29 fsfs: now i see what's your problem, just remove the S line May 31 11:56:40 fsfs: updated SRC_URI will just cp -ra shr-unstable/ZEDAT/gnutls-2.8.5 to shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/${PN}-${PV}-${PR}/ May 31 11:56:49 :) May 31 11:56:50 fsfs: and S _should_ stay the same May 31 11:58:34 works ok now, great! May 31 11:58:39 Heinervdm: JaMa|Wrk: THANKS May 31 11:59:16 I don't understand why http://shr-project.org/trac/wiki/Getting%20started%20developing%20SHR talks about overriding S then - why would that be necessary? May 31 12:01:37 fsfs: yes it is in that case May 31 12:01:52 fsfs: if you have ie git checkout of gnutls in shr-unstable/ZEDAT/gnutls May 31 12:02:01 fsfs: without -2.8.5 May 31 12:02:34 JaMa|Wrk: because default S is $PN-$PV? May 31 12:02:41 fsfs: then you have to update S which is by default set in recipe to S=${WORKDIR}/${P} to S=${WORKDIR}/gnutls May 31 12:02:52 i see May 31 12:04:04 fsfs: you missunderstood the meaning of S that was the problem :) May 31 12:04:40 and for libphone-ui-shr it's overwritten in recipe to S = "${WORKDIR}/git", but if you have it checkouted in libphone-ui-shr dir (as SRC_URI shows) then you have to update it too May 31 12:06:17 makes a lot more sense now. I see that the "Building SHR" example is actually functional, whereas the "Getting started.." example leaves room for mistakes such as mine May 31 12:16:55 freesmartphone.org: 03morphis 07msmcomm * rd426a182d648 10/ (10 files in 5 dirs): libmsmcomm: rename simbook to phonebook May 31 12:42:53 freesmartphone.org: 03morphis 07msmcomm * r77f46c484a7c 10/ (10 files in 5 dirs): libmsmcomm: rename simbook to phonebook May 31 12:53:30 morphis: direct response. In AT, everything but USSD is request->response. There are only few URCs. May 31 12:53:47 except USSD an URC will never be a response May 31 12:54:07 hm ok May 31 12:54:16 short example: May 31 12:55:01 CMD: Write Phonebook (test, 1337) -> Response (ModifyId = 4) .... Event: Phonebook (ModifyId = 4, EntryId = 6) May 31 12:55:19 so you get the real result of the write operation via the URC May 31 12:56:21 I hope that is not to much work for you in fsogsmd May 31 12:56:48 interesting. As long as we get a response, I'm sure we can deal with the additional URC May 31 12:57:26 but all three messages have the same refId set May 31 12:57:31 ok May 31 12:57:41 so it should be easy to trac May 31 12:58:06 but thats not a default manner May 31 12:58:30 seems to be different in some cases May 31 13:19:58 mickey|office: what do you think about this patch? http://paste.pocoo.org/show/220337/ maybe it's needed to update in some other recipe too (build still running) May 31 13:30:36 JaMa|Wrk: looks good May 31 13:30:50 fso-specs are the only case IIRC May 31 13:34:59 ok, thanks, I'll push it if my build finish fine May 31 13:36:39 mrmoku|away: is phonefsod missing dependency on libfso-glib? imho yes, but iirc you said you had it in your tree already May 31 13:52:38 mickey|office, after upgrade it seem to come back after suspend May 31 13:52:44 hi btw May 31 13:52:53 fso says unknow reason May 31 13:53:31 FSO_USAGE_RESUME_REASON_Unknown May 31 14:05:06 yes May 31 14:05:12 Android/Linux sucks May 31 14:05:18 soorry wifi issues May 31 14:05:29 i told you i wanted to know the resume reason from BP May 31 14:05:35 now you know why May 31 14:05:50 ok May 31 14:05:52 we need to fix the kernel to get the resume reason May 31 14:05:58 ah ok May 31 14:06:06 the android model is a distributed one May 31 14:06:17 every subsystem has the chance to check whether it was responsible for wakeup May 31 14:06:23 we have a centralized model for suspend/resume May 31 14:06:24 do you have an idea on how to fix the knerel for that May 31 14:06:27 which is incompatible May 31 14:06:36 ok May 31 14:06:38 i propose adding a sysfs node May 31 14:06:45 that is populated during the resume path May 31 14:06:52 from the data that's gather by the BP May 31 14:07:58 do you have an idea on how to do that? May 31 14:08:06 is it complicated/time consuming? May 31 14:09:02 mmm May 31 14:09:13 what would contain the sys node? May 31 14:09:31 do you want something like all thoses printk during resume with debugging May 31 14:09:44 in the sys node May 31 14:11:06 there's a header file that contains the wakeup reasons somewhere May 31 14:11:10 i'm not at home, so can't look May 31 14:11:27 it's more or less like an enum of IRQ sources May 31 14:11:30 this we would need May 31 14:11:32 so we can do May 31 14:11:38 cat /sys/wakeup_reason May 31 14:11:38 5 May 31 14:12:00 ah nice May 31 14:12:00 ok May 31 14:12:04 should be relatively simple once you find out where in the resume path you have access to this info May 31 14:13:48 in the meantime is there a trick to do to fsousaged.conf? May 31 14:16:21 what is the exact problem? May 31 14:16:31 the same than before May 31 14:16:43 you syspend and it resumes not so long after May 31 14:17:03 more fun with the broken async. suspend May 31 14:17:13 no idea offhand, hand me a log when you get it May 31 14:17:16 (from fsouaged) May 31 14:17:21 [with log_level=DEBUG] May 31 14:17:25 I get it all the time May 31 14:17:55 trying stayalive instaed of android...let's see May 31 14:20:04 stayalive doesn't even suspend May 31 14:20:16 rebooting with android low_level_type May 31 14:22:45 mickey|office, http://pastebin.com/tF72K4Jt May 31 14:24:34 mickey|office, strange: I suspend + I press a key like volume and it keep suspended,as soon as I stop pressing a key...it wakes up May 31 14:24:55 freesmartphone.org: 03morphis 07msmcomm * r8d17816d9697 10/ (12 files in 5 dirs): libmsmcomm: implement write/delete phonebook messages May 31 14:25:18 mickey|office, the strange thing is : DBus Error while talking to IdleNotifier: Method "SetState" with signature "s" on interface "org.freesmartphone.Device.IdleNotifier" doesn't exist May 31 14:25:31 according to someone here it was a missing /0 in deviced May 31 14:25:36 dbus interface May 31 14:27:05 uncommenting [fsodevice.kernel_idle] May 31 14:27:09 and rebooting May 31 14:29:22 now it's even worse with [fsodevice.kernel_idle] uncommented May 31 14:29:59 http://pastebin.com/xpF1GGEM with apm -s May 31 14:38:21 grrr against my version of compat-wireless on my dekstop computer May 31 14:38:29 sorry again May 31 14:38:39 I think I'll download a new compat wireless version May 31 14:38:43 it's unbearable May 31 14:38:59 mickey|office, with [fsodevice.kernel_idle] uncommented and apm -s http://pastebin.com/xpF1GGEM May 31 14:39:34 the strange thing is that I've no bluetooth or gps in my configs May 31 14:39:48 I really must go May 31 14:40:32 I'll go with the half-broken suspend....hmmm May 31 16:23:21 <[Rui]> hi all May 31 16:24:01 hi [Rui] May 31 16:24:50 <[Rui]> JaMa: hi, any new kernel? :) May 31 16:26:15 yes, but only for me :P May 31 16:26:50 [Rui]: no new patch for your and my other issues.. just fixed build with gcc-4.5.0, so you don't need it.. May 31 16:27:16 * [Rui] grins... "I was about to claim bad form! ;)" May 31 16:43:17 <[Rui]> JaMa: did it generate a stable kernel? May 31 16:43:51 <[Rui]> I remember, int he past, Linux was a bit picky about the GCC versions compiling it... May 31 16:47:31 there are issues all over image with gcc-4.5.0 May 31 16:47:41 one in kernel was already fixed by khem May 31 16:48:20 one in busybox was fixed in newer gcc snapshot I'm using instead released 4.5.0 tarbal May 31 16:48:49 but on desktop it's ok and kernel is not picky for gcc different thant 2.95 anymore :) May 31 16:49:47 <[Rui]> gonna get me some cherries :) May 31 17:14:48 freesmartphone.org: 03morphis 07msmcomm * ra5143e45122a 10/ (9 files in 5 dirs): libmsmcomm: add change_pin message May 31 17:18:06 I got shr on qemuarm ... see no apps and can do nothing special! May 31 17:19:09 Maybe I should try ssh into the emulator ... May 31 17:23:35 guys can someone help to interpret this numbers: May 31 17:23:43 test May 31 17:24:22 cat /sys/class/power_supply/battery/uevent May 31 17:24:37 ... POWER_SUPPLY_CURRENT_NOW=-88500 May 31 17:24:44 or POWER_SUPPLY_CURRENT_NOW=219937 May 31 17:25:14 what does it mean? May 31 17:25:33 i mean units May 31 17:25:40 thanks in advance May 31 17:26:13 X is running ... and an impressive thing to look at but can do nothing! May 31 17:26:28 illume ... brb May 31 17:27:34 gena2x: IIRC 1000000=1A, and if negative, then it means charging May 31 17:28:08 dos1: thanks May 31 17:32:18 dos1: i get opposite :( May 31 17:32:45 dos1: oh, no just some delay... May 31 17:34:17 dos1: yes, just huge 10second delay. May 31 17:34:25 gena2x: yup, that's typical May 31 17:34:56 GNUtoo|oeee: hmm May 31 17:35:04 mickeyl, hi May 31 17:35:20 GNUtoo|oeee: the code i wrote needs to wait until suspend actually has been performed May 31 17:35:25 i have no idea how to do that atm. May 31 17:35:30 ah ok May 31 17:35:45 that's the reason why it immediately triggers the wakeup May 31 17:35:45 you need hooks maybe May 31 17:35:48 ok May 31 17:36:11 we need to find out how to find out when suspend is actually being performed May 31 17:36:23 i can try looking at hardware clock May 31 17:36:28 but that feels not right ;) May 31 17:36:32 indeed May 31 17:36:41 by suspend you mean? May 31 17:36:46 real or fake? May 31 17:36:51 CPU halt May 31 17:36:54 ah ok May 31 17:38:49 if we had a suspend reason sysfs node we could do it by looking at it May 31 17:38:59 since this would be cleaned upon suspend request May 31 17:39:06 and would only have a value upon real resume May 31 17:40:00 mmm May 31 17:40:07 ok I'll do that in prioriry May 31 17:42:23 maybe I should look at your code because I don't understand why we have to wait for beeing really suspended? May 31 17:43:26 mmm May 31 17:43:31 maybe: May 31 17:43:38 /do nothing while not suspended yet May 31 17:43:50 //resume after beeing suspended May 31 17:43:59 or something like that? May 31 17:45:23 upon resume we need to inspect the input nodes May 31 17:45:29 and check whether a button woke us up May 31 17:45:31 and if so, which button May 31 17:45:40 and only resume if power button May 31 17:46:11 if we are alive, but have no input keycode May 31 17:46:15 apparantly someone else woke us up May 31 17:46:16 (GSM) May 31 17:46:19 so we wake up fully May 31 17:46:49 that means we should only look at the input nodes after resume May 31 17:46:57 which would all be fine if suspend was sync May 31 17:47:05 since it is async though, we need to delay looking at the input nodes May 31 17:47:08 but for how long? May 31 17:47:11 until we are realyl asleep May 31 17:47:19 that's why we need to know _when_ we are really suspended May 31 17:47:28 or when we are really resumed May 31 17:47:30 either one will do May 31 17:48:10 reading May 31 17:48:53 when we are really resumed seam easier May 31 17:49:00 yep May 31 17:49:06 because how can you tell when you're suspended if the cpu sleeps May 31 17:49:38 I bet the code doesn't even run when the cpu is suspended May 31 17:49:49 only listen for irq or something like this May 31 17:50:09 I've another idea May 31 17:50:14 get the resume reason May 31 17:50:35 it would need a kernel part altough May 31 17:51:25 so you would know when gsm woke it up no? May 31 17:51:44 basically the plan would be: May 31 17:51:59 *enable suspend/wakelocks etc...debug May 31 17:52:04 look what it prints May 31 17:52:13 and grep the source May 31 17:52:22 and do something instead of printing May 31 17:52:27 yep May 31 17:52:58 it would be the easiest thing to do I think May 31 17:54:33 for now I've to replace my old pentium III 500 router in order to host a plone website May 31 17:55:34 have fun May 31 17:55:48 thanks May 31 17:56:14 I would have prefered working on the dream or on the bug device altough May 31 17:56:29 but I really need to do that May 31 17:56:33 hehe, i see May 31 17:56:45 the plone website is not for me May 31 17:57:17 it's for sommeone + some people who don't know html May 31 17:57:32 the pentium III 500 is too slow for that May 31 18:09:03 PaulFertser: are you here? have few minutes to answer few qustions about bluetooth? May 31 18:24:58 freesmartphone.org: 03mickey 07cornucopia * r2d680cdf574f 10/fsousaged/src/plugins/lowlevel_android/plugin.vala: fsousaged: lowlevel_android: give grace period to fall asleep May 31 18:48:43 gena2x: hopefully May 31 18:49:04 gena2x: though i'm not fully here, partially busy having dinner, but go ahead :) May 31 18:49:12 PaulFertser: first, i want to tell. i finally got my bt to work. May 31 18:49:26 gena2x: good news May 31 18:49:35 Surprisingly, i got several today. May 31 18:49:46 PaulFertser: with help of Gabrys who did daemon May 31 18:49:51 gena2x: i'm glad you succeeded :) May 31 18:50:06 PaulFertser: but i were very close to solution. May 31 18:50:06 Ah, i remember talking with him about the matters :) May 31 18:51:07 yeah, he told that he told with you :) May 31 18:52:17 PaulFertser: i understood my mistake - he talk for 2 hours, i am only for 1 :) May 31 18:52:36 PaulFertser: that's way i need 2 weeks to be close to solution :) May 31 18:52:49 gena2x: LOL May 31 18:52:51 :D May 31 18:53:15 So, what still bothers you now? May 31 18:53:21 PaulFertser: ok, now my question :) May 31 18:53:42 PaulFertser: now about digital variant May 31 18:53:53 my favourite one May 31 18:54:00 gena2x: SCO over HCI? May 31 18:54:03 one that routes audio through USB ;-) May 31 18:54:06 PaulFertser: i've read you mail about pskey May 31 18:54:21 in mailing list May 31 18:54:33 so, you changed hciconfig a bit. May 31 18:54:46 anything else or this is all to get it work? May 31 18:55:14 or need to reflash something? May 31 18:55:50 / i think now i have much better understanding about this topic/ May 31 18:57:06 PaulFertser: (I remember some troubles about routing bi-directional audio with HCI and a fix being reflashing bluetooth firmware) May 31 18:57:37 i'm talking about main in jan 2009 May 31 18:57:41 *mail May 31 18:57:45 gena2x: i have never researched a proper solution because it was a quick hack i did to get it working. May 31 18:58:03 so it finally worked without any problems? May 31 18:58:10 gena2x: i just patched hciconfig a little, that was enough to get SCO over HCI. May 31 18:58:19 very nice! May 31 18:58:23 It worked the same as SCO over PCM, no problems with known-good headsets. May 31 18:58:40 And i've never tried that special command that temporarily switches it back, though it should work. May 31 18:59:11 ok, nice to hear. May 31 18:59:55 thanks. May 31 19:00:20 and, btw how not-known-good headsets do not work? just no sco channel? May 31 19:01:17 gena2x: those listed on the wiki May 31 19:01:34 I've spent quite some time trying to figure out what makes those incompatible. May 31 19:01:47 Still no slightest idea. May 31 19:02:00 gena2x: bbiab May 31 19:09:06 freesmartphone.org: 03mickey 07cornucopia * r462733e9e810 10/fsousaged/src/plugins/lowlevel_android/plugin.vala: fsousaged: lowlevel_android: use select and wait for ERESTARTHAND to block until we have resumed May 31 19:14:57 freesmartphone.org: 03mickey 07cornucopia * r71ade1245539 10/fsousaged/src/plugins/lowlevel_android/plugin.vala: fsousaged: lowlevel_android: debug++ May 31 19:16:01 gena2x: if you want to try to make SCO over HCI work, i think i can provide you with a patched binary. Are you pondering voip idea? May 31 19:16:59 PaulFertser: thanks, i've build bluez several times, so no problem with building. May 31 19:17:32 PaulFertser: still not thinking about voip May 31 19:18:18 PaulFertser: more thinking about recording conversation. just wanted to see how much system will be loaded. May 31 19:19:34 hmm, garrrr May 31 19:19:58 the glibc select implementation automagically relaunches the select call if it received a signal May 31 19:20:22 SA_RESTART? May 31 19:20:37 gena2x: you can record with a modified gsmhandset, why can't you do that with gsmbluetooth after appropriate modifications? May 31 19:21:02 -ERESTARTNOHAND May 31 19:21:10 ah, _p_select might be it May 31 19:22:32 PaulFertser: wait, last time we talk that it's impossible to ADC signal from Voice back to Hi-Fi May 31 19:23:13 gena2x: indeed, you're right. It was a wrong suggestion of mine apparently. May 31 19:23:27 gena2x: what middleware do you use? May 31 19:23:49 PaulFertser: no middleware May 31 19:24:01 PaulFertser: just understanding concept May 31 19:24:36 gena2x: i mean you can indeed spare effort to get recording conversations going. But if your code will be as popular as my emacs interface, i'd not suggest to even start. May 31 19:24:59 PaulFertser: :) May 31 19:25:34 PaulFertser: Paul, i am doing that for me. if other as interested - very good, nice and funny. if not - i see no problem. May 31 19:25:46 *others are May 31 19:26:17 PaulFertser: i am curretly using qtmoko. May 31 19:26:40 PaulFertser: and happy with it. As always :) May 31 19:28:41 gena2x: so, do you need any more assistance from me? ;) May 31 19:29:18 PaulFertser: thanks for your assistance. and have fun :) May 31 19:30:42 gena2x: lol, i'll get some sleep instead May 31 19:37:44 ~logs May 31 19:37:45 logs are on http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged. logs are updated daily --- livelogs are http://logs.nslu2-linux.org/livelogs/openmoko-cdevel.txt May 31 19:39:07 !logs May 31 19:39:08 Channel logs for #openmoko-cdevel are archived at: May 31 19:39:09 http://hentges.net/tmp/logs/irc/%23openmoko-cdevel May 31 19:39:10 Live-logs are available at May 31 19:39:11 http://hentges.net/tmp/logs/irc/livelogs/%23openmoko-cdevel.livelog May 31 19:39:13 See ?? help-logs for usage instructions May 31 19:43:00 mickeyl: seems its another LOGLEVEL bug, like in fso-abyss. I've enabled DEBUG mode and fsogsmd now not loosing the net. May 31 19:43:13 mickeyl: deep sleep enabled May 31 19:47:40 is SA_RESTART the default? May 31 19:47:43 for all system calls? May 31 19:54:47 GNUtoo|oeee: hi May 31 19:55:00 I did not get an answer from this guy yet May 31 19:55:06 perhaps he is ignoring me?? May 31 19:55:15 leviathan, who? May 31 19:55:23 the nand guy? May 31 19:55:23 freesmartphone.org: 03mickey 07cornucopia * re5487dc58363 10/fsousaged/src/plugins/lowlevel_android/plugin.vala: fsousaged: lowlevel_android: -ERESTARTNOHAND is not being delivered to the application, but rather intercepted by glibc :/ May 31 19:55:24 the one who wrote the nand driver... May 31 19:55:27 yes May 31 19:55:32 ok I didn't try May 31 19:55:39 I bet it would be the same for me May 31 19:56:08 I "#if 0"-ed out May 31 19:56:15 GNUtoo|oeee: sleeping a couple of seconds to give it a chance to fall into suspend improves it May 31 19:56:17 the if condition which produces the error May 31 19:56:33 but... the functionality is missing May 31 19:56:48 so it hangs while try'n to write subpage-block May 31 19:56:51 in ubiformat May 31 19:59:05 leviathan: when did you write him the mail? May 31 20:01:00 Kensan: lets see May 31 20:01:04 yesterday? May 31 20:01:05 ok May 31 20:01:10 perhaps I should wait May 31 20:01:14 some additional days May 31 20:01:32 leviathan: I would suspect he is busy doing paid work... May 31 20:02:37 Kensan: hmm, right May 31 20:02:41 its most likely May 31 20:02:44 ^^ May 31 20:02:49 just waitn' May 31 20:03:21 leviathan: did you look at the msm fb driver by any chance? May 31 20:04:22 Kensan: which part exactly? ;-) May 31 20:05:57 leviathan: I got it to work on nexus but I had to disable request_vsync in mdp_lcdc.c because otherwise I would get May 31 20:06:07 leviathan: "scheduling while atomic" BUGs May 31 20:06:23 leviathan: did you ever run the with lockdep enabled? May 31 20:07:07 I guess you should ask GNUtoo|oeee for that, he has done much more work, concerning framebuffer driver for msm May 31 20:07:45 when I remember correctly we had such a similiar problem once May 31 20:08:10 Kensan: hmm, does msm-gnuconform from leviathan-incoming run for you? May 31 20:08:17 I'm not sure about the compatibility May 31 20:08:21 is it trout? May 31 20:08:24 would be a nice try May 31 20:08:36 leviathan: no I took dcordes' htcleo tree May 31 20:08:44 http://gitorious.org/htc-msm-2-6-32/leviathan-incoming/commits/msm-gnuconform May 31 20:08:48 leviathan: it's based on htc-msm-2.6.32 May 31 20:08:49 :-) May 31 20:09:27 leviathan: nexus one and htcleo are very similar. I need to talk to dcordes how we are going to organize the kernel-tree May 31 20:09:48 I don't want to add another tree to the party and have more fragmentation... May 31 20:13:33 leviathan: btw, what's the reason for the refresh thread? May 31 20:13:59 uhm May 31 20:14:01 X May 31 20:14:04 :-) May 31 20:14:29 wow, the shortest answer I ever gave, since I live May 31 20:14:38 one single letter, say'n everything May 31 20:14:41 =D May 31 20:24:21 Kensan, refresh thread is necessary for xorg,else the screen is not refreshed and xorg doesn't work May 31 20:25:30 GNUtoo|oeee: would x not have been sufficient? ;-) May 31 20:25:46 but ok, now we have also the technical reason May 31 20:26:01 in fact, we would have an x-server running May 31 20:26:14 but the write process would hold May 31 20:26:24 so we would not see any reaction on the screen May 31 20:26:34 what is so much worth like not running at all May 31 20:26:53 so we deploy a worker thread, which makes frame refreshment, regulary May 31 20:27:05 no May 31 20:27:11 yes? May 31 20:27:24 its in kernel space May 31 20:27:32 x would have not been sufficent May 31 20:27:43 no was not for the later thing you said May 31 20:27:48 hehe May 31 20:27:54 ok May 31 20:28:01 lets be it good and go on May 31 20:28:05 :-) May 31 20:28:13 now, we need another thing May 31 20:28:19 funny enough android booted fine with my kernel May 31 20:28:34 android is a ugly tinkering May 31 20:28:43 and the kernel is optimized for this tinkering May 31 20:28:45 android is non-standard and userspace refresh the screen May 31 20:28:46 sure it boots May 31 20:28:47 somehow the nexus boots from flash if nothing else is specified May 31 20:28:53 yes May 31 20:29:08 you need to build a custom command line directly in the boot.img May 31 20:29:20 Kensan, if it boots from flash without root= that's normal May 31 20:29:26 uhm May 31 20:29:34 no, also with root i guess May 31 20:29:35 I think it's because of the bootloader May 31 20:29:47 I also observed such strange behaviour May 31 20:29:48 I want to fix msm_fb first before moving on. May 31 20:29:49 so do like leviathan said May 31 20:29:50 from time to time May 31 20:29:51 if you want to boot from sd May 31 20:30:01 every time you press power May 31 20:30:12 http://wiki.openmoko.org/wiki/OpenMoko_on_HTC-Dream#For_SD-card_rootfs May 31 20:30:16 GNUtoo|oeee: yes, I was just a bit surprised that android came up with the first kernel that I cooked up. May 31 20:30:25 lol May 31 20:30:36 GNUtoo|oeee: I was more expecting to see a panic and a reboot. May 31 20:31:18 ok May 31 20:32:46 Kensan: surprise, surprise ;-) May 31 20:33:27 leviathan: indeed. I've rarely got something running on one of the first tries ;) May 31 20:35:42 man this guy is slick May 31 20:35:46 Steve Jobs May 31 20:36:17 He talks about ordinary things and make you think it's *the* revolution the tech world has been waiting for. May 31 20:36:18 Kensan, what again did he do? May 31 20:36:35 GNUtoo|oeee: Just saw a recording of a presentation on tv. May 31 20:36:38 lol I'm more the kind of people makiing the revolution against apple May 31 20:38:19 I'm the one who is just not prepared to use software, where he can't not proof the code May 31 20:38:22 everybody is going crazy over the iPad May 31 20:38:27 uhm May 31 20:38:33 I saw it some days ago May 31 20:38:39 not very impressive May 31 20:39:05 I prefer to repair my hp tablet pc May 31 20:39:07 :-) May 31 20:41:17 ah the i.bad? May 31 20:41:18 wow May 31 20:41:24 Apple takes 30% May 31 20:41:31 brilliant May 31 20:41:34 of? May 31 20:41:44 whatever you charge for your app May 31 20:41:49 ah ok May 31 20:43:26 no wonder they make so much money May 31 20:44:18 MyPod, Jeve Mobs, Lisa Simpson May 31 20:44:19 xD May 31 20:44:25 ohh, yeah May 31 20:44:28 MyTunes May 31 20:44:29 ;-) May 31 20:44:50 Mapple May 31 20:44:52 ^^ May 31 20:55:38 GNUtoo|oeee: I'll sleep May 31 20:55:49 tomorrow I have physics lecture May 31 20:55:57 where we talk about surface holograms May 31 20:56:02 it was my idea May 31 20:56:14 because the subject perfectly fits in May 31 20:56:23 in this part of quantum physics May 31 20:56:33 its a practicle usage of the theory May 31 20:56:47 and I guess, it will be now a subject of the basis excersice May 31 20:57:27 lucky me, that I've already have used the formulas, to recalc the equations for this construction May 31 20:57:30 ok May 31 20:57:33 so May 31 20:57:50 after friday, the holydays begins May 31 20:57:55 so next weeks and later May 31 20:58:01 I've got a lot of free time May 31 20:58:21 beneat the preparations for the examina May 31 20:58:45 I hope the NAND guy will answer soon May 31 20:59:17 I'm very excited to soon use the whole sd card as MP3/OGG space May 31 20:59:20 ^^ May 31 21:00:03 ok... well, rotation and suspend are more important issues,... i know May 31 21:00:07 :-) May 31 21:00:20 ok, enough spam from my side, I'll go sleep May 31 21:00:21 gn8 May 31 21:00:23 ;-D May 31 21:02:29 leviathan: good night May 31 21:02:48 leviathan: semesterferien rock btw May 31 21:02:49 heh May 31 21:02:55 or in my case: used to rock. May 31 21:05:43 hello good evening May 31 21:06:01 is anybody here using navit ? May 31 21:14:06 <[Rui]> hi May 31 21:21:53 [Rui]: hi May 31 21:29:58 JaMa: hello i have problems (freezing and drawing issues)with navit, can you help me ? **** ENDING LOGGING AT Tue Jun 01 02:59:57 2010