**** BEGIN LOGGING AT Thu Jan 06 02:59:58 2011 Jan 06 04:01:02 Hello everyone Jan 06 04:01:18 Will SHR run on any openembedded supported platform? Jan 06 06:22:09 moin Jan 06 06:40:34 larsc: for me it also doesn't work.. after changing remote url from git:// to http:// I still get only .34 branches not .37 Jan 06 06:47:33 larsc: and can you please merge latest om-gta02-2.6.34 changes to om-2.6.34? Jan 06 07:49:47 mickeyl: ah, ok. int16, int32 ? Jan 06 07:52:09 heh, that will make another nice round of fixing... turning all gvalues into gvariants :P Jan 06 07:52:25 breakfast and daywork first though :/ Jan 06 09:28:33 hi, does openmoko git work for anyone? Jan 06 09:28:47 git.openmoko.org[0: 88.198.160.201]: errno=Connection refused Jan 06 09:29:14 this is what i am always getting Jan 06 09:30:01 http:// somehow does Jan 06 09:30:09 but still doesn't show new .37 branches Jan 06 09:35:04 probably missing the update-server-info hook Jan 06 09:35:24 (for http: that is... no idea why git: doesn't work) Jan 06 09:36:43 nmap doesn't show 9418 as open (maybe it's side-effect of moving that machine as larsc said) Jan 06 09:38:38 maybe someone with ssh access to git could push the new 2.6.37 branch somewhere? Jan 06 09:38:47 actually, the fact that both the git daemon and update-server-info failed simultaneously seems to indicate that there is something wrong with the git installation itself... Jan 06 09:39:12 radekp: I have diff against upstream .27 (created from downloaded snapshot) if it helps you Jan 06 09:39:50 .37 ofc, http://gitorious.org/~jama/angstrom/jama-shr-experimental/trees/master/recipes/linux/linux-2.6.37 Jan 06 09:39:58 JaMa: hmm thanks, but i'd rather prefer having also the history.. Jan 06 09:40:34 JaMa: btw have you already tried it? Jan 06 09:40:35 me too, this is just temporary for me to test it before git.om is fixed Jan 06 09:41:05 only built, fr is at home and I won't go home this week.. Jan 06 09:41:29 ahh oki Jan 06 10:15:32 morning Jan 06 10:20:54 moin Jan 06 10:26:51 freesmartphone.org: 03morphis 07msmcomm * rda0efbf8bcaf 10/libmsmcomm-next/msmcomm/messagetype.vala: libmsmcomm-next: refactor message type to own source file and add string conversion for it Jan 06 10:41:07 SHR: 03mok 07phonefsod * rab9715861f09 10/src/phonefsod-fso.c: phonefsod-fso: gdbus is using variants not values Jan 06 10:41:18 SHR: 03mok 07phonefsod * ra79b3bd7156c 10/src/phonefsod-fso.c: phonefsod-fso: actively request GSM device status after requesting the resource Jan 06 10:41:19 SHR: 03mok 07phonefsod * r6f9003d45ec5 10/src/ (phonefsod-dbus.c phonefsod-dbus.h phonefsod.c): make registering to the system bus work Jan 06 10:42:45 mrmoku: :) Jan 06 10:43:04 mickey|office: yeah, getting forward :-) Jan 06 10:43:04 freesmartphone.org: 03morphis 07msmcomm * rdc91a5afcf61 10/libmsmcomm-next/ (5 files in 2 dirs): libmsmcomm-next: refactor message result type to it's own source file too Jan 06 10:43:07 freesmartphone.org: 03morphis 07msmcomm * r00b0569752ee 10/libmsmcomm-next/msmcomm/ (4 files): libmsmcomm-next: add message class to expose type of a message Jan 06 10:43:11 SHR: 03mok 07phoneuid * rce5fc12bca87 10/src/phoneuid-dbus.c: phoneuid-dbus: fix owning the bus name Jan 06 10:43:45 still missing the switch to gvariants in libphon-ui Jan 06 10:46:02 mickey|office: a guess what CM_PH in HCI_SUBSYS_CM_PH_OPRT_MODE_CMD could mean? Jan 06 10:46:08 btw. good morning everybody! Jan 06 10:47:17 mickey|office: Palm implements it in a ServiceStateGsm class as SendChangeOperationModeReq Jan 06 10:47:49 I am asking as I am thinking about whats the correct category for such a command for a good naming convention in libmsmcomm-next Jan 06 10:52:07 morphis: hmm.. doesn't ring a bell offhand to me. Could have something to do with the CPHS (common pcn handset spec), but as it's an operation mode it makes no sense Jan 06 10:52:11 hm I should read more in the GSM specs; service state is what you have when you can any service from a PLMN Jan 06 10:52:31 s/can/get/ Jan 06 10:52:31 morphis meant: hm I should read more in the GSM specs; service state is what you have when you get any service from a PLMN Jan 06 10:52:44 as in TS 03.22 Jan 06 10:52:57 chapter 3.7 Jan 06 10:53:05 mickey|office: will treat it as a service state command Jan 06 10:53:08 as palm did Jan 06 10:54:04 yes, i had the urge to rename some of their commands, but i think until we're 100% sure what side-effects things have we should stick with their naming Jan 06 10:54:29 right Jan 06 10:54:33 PaulFertser: wrt. timeouts, did you have any chance to get those logs? i'd like to carry out the timeout changes soon Jan 06 10:55:02 ServiceStateChangeOperationModeRequstCommandMessage is a very long class name ... Jan 06 10:55:31 mayby I should use namespaces Jan 06 10:55:47 yes Jan 06 10:56:09 like class ServiceStateGsm.ChangeOperationModeRequestCommandMessage Jan 06 11:15:06 mickey|office: i can grep my logs for TIMEOUT but i can assure that subjectively it become much more stable with 5 second instead of 120s timeout. Jan 06 11:15:45 PaulFertser: ok, thanks. Jan 06 11:16:10 freesmartphone.org: 03morphis 07msmcomm * r9bff013e2c45 10/libmsmcomm-next/msmcomm/ (Makefile.am messagetype.vala statemessage.vala): libmsmcomm-next: first bits of change operation mode message Jan 06 11:16:25 so I am off for working Jan 06 11:16:26 bye Jan 06 11:32:51 mickey|office: ok, the log proved to be quite interesting: http://paste.debian.net/103867/ Jan 06 11:34:35 And even http://paste.debian.net/103868/ . Doesn't make any sense to me, damn it. Jan 06 11:37:58 And http://paste.debian.net/103869/ (omfg, it advanced to SUSPENDED before actually detecting the timeout?) Jan 06 11:38:44 I must note that '+COPS=0,0' and '+CMGL=4' are the top 2 for timeouts. Jan 06 11:42:25 I guess i'll pack my december log and will send to you via e-mail for your reference. Hold on, it's one big file, and FR has no fast way to transfer files. Jan 06 11:57:46 someone seen fsodeviced (gdbus) segfaulting like this? ** (process:5244): CRITICAL **: file dbusresource.c: line 287: uncaught error: An object is already exported for the interface org.freesmartphone.Resource at /org/freesmartphone/Resource/Bluetooth (g-io-error-quark, 2) Jan 06 11:58:57 JaMa, yes Jan 06 11:59:01 I know why Jan 06 11:59:01 mickey|office: sent Jan 06 11:59:09 JaMa, I'll explain Jan 06 11:59:41 on htcdream you need at least 2 plugins for handling wifi resource Jan 06 11:59:45 before gdbus Jan 06 12:00:03 adding the ifconfig plugin and the rfkill plugin in fsodeviced worked Jan 06 12:00:26 I did modprobe msm_wifi manually and launched iliwi and it was taking care of the rest Jan 06 12:00:38 but after gdbus it has the same error than you Jan 06 12:00:48 the fix was to remove one of the two plugins Jan 06 12:07:46 ok, you're right, if I remove ie rfkill it works again Jan 06 12:08:43 I think it should be fixed in fsodeviced, because we need the ability to run multiples plugins Jan 06 12:09:00 else we would need to : Jan 06 12:09:14 *develop a scripting language for it as mickeyl suggested Jan 06 12:09:15 or Jan 06 12:09:34 to make a plugin that takes cares of everything(code copying issue) Jan 06 12:10:18 how viable would be to transform rfkill and ifconfig plugins as libraries Jan 06 12:10:37 that one could use in the wifi/bluetooth plugin Jan 06 12:21:53 PaulFertser: thanks! Jan 06 12:22:07 mickey|office: thanks for caring Jan 06 12:25:59 PaulFertser: i will use the calypso as reference for all the timing values (different commands need much more than the 5 seconds timeout). I guess there is no slower modem still in use on the planet ;) Jan 06 12:40:32 mickey|office: there're issues much more serious than timeout values, at least that's how i percieve my log. Jan 06 12:41:53 103867 is what i'd accept Jan 06 12:42:01 let me check 103868 Jan 06 12:42:38 right Jan 06 12:42:44 no that looks quite normal Jan 06 12:42:51 the thing is Jan 06 12:43:11 if you enter a command while another one is operating, then the previous command is being cancelled and this will be reported with a plain OK Jan 06 12:45:42 mickey|office: a cancelled command is OK? Interesting. Jan 06 12:45:43 ;) Jan 06 12:45:50 yep Jan 06 12:45:54 welcome to the wonderful world of OK Jan 06 12:45:56 err Jan 06 12:45:56 AT Jan 06 12:46:00 :D Jan 06 12:46:11 it might be wise to add an additional command for every retried command in advance Jan 06 12:46:16 (without a timeout) Jan 06 12:47:00 It's OK, my phone hangs every now and then, that's just a wonderful world of OK. Be glad it doesn't bite :) Jan 06 12:47:23 The world of OK, starting to appreciate it. Jan 06 12:47:41 (in fact that's exactly my attitude to my cellphone usage patterns) Jan 06 12:48:05 to be fair, it's pretty much a calypsoism Jan 06 12:48:22 i never had any problems with deep sleep, not responding etc. with any other AT modem Jan 06 12:48:37 mickey|office: what still puzzles me a lot is that we didn't see anything like that with ogsmd. Jan 06 12:49:34 PaulFertser: well, there's only one theory to that... we're much faster. Jan 06 12:49:43 Timestamps in my log seem to not support the idea something is wrong with "awake you damn calypso" pinging. Jan 06 12:49:54 I implemented this minimum-delay-between-subsequent-packages in libgsm0710mux Jan 06 12:50:08 could be worth experimenting with Jan 06 12:50:13 it's not active by default Jan 06 12:50:19 but can be set to any number of milliseconds Jan 06 12:50:24 IIRC... been some time Jan 06 12:50:28 Hm, interesting, gotta give it a try. Jan 06 12:50:44 let me get you the config option name Jan 06 12:51:35 [libgsm0710mux] Jan 06 12:51:41 device_sendpause_threshold = 1000 Jan 06 12:51:55 IIRC that means don't send more than one package per second Jan 06 12:53:18 don't use 1000, that will be sllllllllllooooooooooowww ;) Jan 06 12:53:32 since it boils down to mux package level, not AT command Jan 06 12:53:50 (one AT command may be fragmented into multiple packages on MUX layer) Jan 06 12:54:18 ah well, I'll miss AT once we have all migrated to binary protocols ;) Jan 06 13:32:45 mrmoku, variants Jan 06 13:42:42 playya: yeah, digged into libfso-glib source Jan 06 13:42:57 lots of work to convert libphone-ui :/ Jan 06 13:43:02 good thing anyway Jan 06 15:05:12 moin Jan 06 15:06:31 morning Jan 06 16:05:17 SHR: 03mok 07libphone-ui * r282a6ca77e7c 10/src/ (3 files): big GValue --> GVariant move Jan 06 16:05:28 SHR: 03mok 07libphone-ui * r0b89f46d6584 10/src/ (dbus.c dbus.h): fix some dbus paths and add _fso_pim_notes Jan 06 16:06:19 SHR: 03mok 07libphone-ui-shr * ra8fe0a403a47 10/src/phoneui-contacts.c: phoneui_contacts: check if we actually got a contact path Jan 06 16:06:21 yo, looks much better now :-) Jan 06 16:06:23 SHR: 03mok 07libphone-ui-shr * r9cf077edcd57 10/src/view/ (4 files): big GValue --> GVariant move Jan 06 16:06:44 JaMa: if you have time... it's worth building Jan 06 16:08:02 ahh...maybe not yet... still some GVAlues left :/ Jan 06 16:22:10 heyho Jan 06 16:22:28 freesmartphone.org: 03morphis 07msmcomm * r605ee4b66d88 10/libmsmcomm-next/ (15 files in 6 dirs): libmsmcomm-next: cleanup structure specification Jan 06 16:27:27 SHR: 03mok 07libphone-ui-shr * rd16cb6184548 10/src/view/ (7 files): kill the remaining GValues Jan 06 16:28:28 SHR: 03mok 07libphone-ui-shr * r15a069bc8a4a 10/src/view/sim-manager-view.c: sim-manager: don't try to free gvariants with common_utils_gvalue_free Jan 06 16:37:38 freesmartphone.org: 03morphis 07msmcomm * rb934cf10a844 10/libmsmcomm-next/ (7 files in 3 dirs): libmsmcomm: complete change operation mode message implementation and add it's response Jan 06 16:57:52 SHR: 03mok 07libphone-ui-shr * r544eb4195c50 10/src/ (8 files in 2 dirs): fix up last commit and finally eliminate ALL gvalues Jan 06 17:29:13 can anyone clone git://git.openmoko.org/git/kernel.git ? does it work for you? My result is remote: Counting objects: 1841605, done. Jan 06 17:29:14 remote: fatal: Out of memory? mmap failed: Cannot allocate memory Jan 06 17:29:14 remote: aborting due to possible repository corruption on the remote side. Jan 06 17:29:39 but 4g of memory free, strace shows no syscalls failing Jan 06 17:30:14 top shows git doesn't use significant amount of memory Jan 06 18:04:51 gena2x, you've access to the remote server? Jan 06 18:13:15 GNUtoo|laptop: this is just usual clone procedure Jan 06 18:13:26 I know Jan 06 18:13:34 but the message is about issue on the remote server Jan 06 18:13:37 not on your computer Jan 06 18:13:59 hm... Jan 06 18:14:51 i got idea. so, my quesion if anyone can try to clone it remains Jan 06 18:19:40 i mean _question_ remains :) Jan 06 18:20:39 it'll do the same for anyone trying to clone Jan 06 18:20:51 better contacting the people who know who is in charge of the server Jan 06 18:21:09 larsc: hi. still no luck with git :( Jan 06 18:21:23 it worked when i tried it Jan 06 18:21:30 maybe the server crashed again Jan 06 18:22:25 may be it is possible to checkout without cloning? Jan 06 18:22:57 works fine here Jan 06 18:23:04 oook. Jan 06 18:23:38 git clone git://git.openmoko.org/git/kernel.git linux-2.6 Jan 06 18:23:42 yeah Jan 06 18:23:50 i tried this Jan 06 18:24:05 i even get new git from git Jan 06 18:24:14 "Cloning into linux-2.6..." Jan 06 18:24:56 yeah, it 'counts remote objects' ok then fails immediately Jan 06 18:25:47 maybe to many remote objects ;) Jan 06 18:26:40 gena2x: you hacked u-boot, would enabling watchdog on resume be easy? ;) Jan 06 18:27:52 lindi-: still fighting your misterious resume bug, given up, annoyed, crying, and now thinking about workaround? :) Jan 06 18:29:03 lindi-: i already have too long todo list. cpu-freq still not working. Jan 06 18:29:29 lindi-: want also to fight with sd speed and fix sd in u-boot. Jan 06 18:31:02 larsc: too many for me, not for you? ;) Jan 06 18:32:51 gena2x: nah, i never tried to the end, only tested whether the server is reachable. Jan 06 18:32:58 gena2x: i get this http://pastebin.com/5M0R8uYg Jan 06 18:33:37 yeah, yeah. same here Jan 06 18:33:53 so object count too much for both of us Jan 06 18:34:30 and look like GNUtoo is right and lack of memory is on remote side Jan 06 18:35:02 may be someone read interesting article about vm.overcommit_memory and set it to 2 :) Jan 06 18:35:35 or no swap... Jan 06 18:37:32 gena2x, maybe time to get a nokia900? Jan 06 18:37:44 and hack on easier stuff? Jan 06 18:38:17 freerunner fully satisfies my hacking demand :) Jan 06 18:38:29 ah I thought you were fedup Jan 06 18:38:33 lindi-: still fighting your misterious resume bug, given up, annoyed, crying, and now thinking about workaround? :) Jan 06 18:38:41 ah no sorry Jan 06 18:38:46 you were talking to lindi- Jan 06 18:38:55 and this were not really serious :) Jan 06 18:39:02 * GNUtoo|laptop is doing too much thing at the same time Jan 06 18:39:12 ah ok Jan 06 18:41:17 n900 is works good i guess, and fr has much things to fix Jan 06 18:42:50 with fr i forgot that linux may just work very well, so then installed it to forgotten by god MIPS on wifi router, i were completely surprised that it works flawlessly :) Jan 06 18:44:32 lol ok Jan 06 18:45:38 gena2x: watchdog would make it easier to trace the execution since i could get trace buffer contents without having to remove battery every time Jan 06 18:49:07 gena2x: on the positive side, I have not found any new 2.6.34 specific bugs Jan 06 18:50:31 lindi-: watchdog interface is simple just two registers. have you just tried adding it? Jan 06 18:50:56 lindi-: + clock selection i think Jan 06 18:51:43 lindi-: may be it will be need to fix wd init in kernel somehow Jan 06 18:51:57 lindi-: just few guesses Jan 06 18:54:12 lindi-: set control and dat to maximum timeout, (42 sec i guess?), enable clock Jan 06 18:54:21 3 lines of code i guess Jan 06 18:56:24 ah, seems just 2 registers Jan 06 18:56:45 you may even not care about resetting it as you should resume in 42 second Jan 06 18:57:19 not care about PCLK timing also as it will never be >50*10^6 Jan 06 18:57:42 so, what's the problems with enabling it? Jan 06 18:57:54 if you _really_ want Jan 06 18:58:34 lindi-: we not have .37... can't clone, but still... Jan 06 19:00:45 gena2x: could you try again Jan 06 19:00:47 ? Jan 06 19:03:23 larsc: same thing Jan 06 19:05:08 larsc: but ok, i'll just track it as remote branch Jan 06 19:07:37 gena2x: yes I was just asking how difficult you see it :) Jan 06 19:07:52 lindi-: 2-liner :) Jan 06 19:08:04 lindi-: moderate difficulty Jan 06 19:10:41 SHR: 03mok 07libphone-ui-shr * r02f37ca3ebde 10/src/util/common-utils.c: common_utils: fix debug_dump_hashtable to work with gvariants Jan 06 19:10:52 SHR: 03mok 07libphone-ui-shr * r19788a9ea35a 10/src/util/ui-utils-contacts.c: ui-utils-contacts: fix number selection for gvariants Jan 06 19:10:53 SHR: 03mok 07libphone-ui-shr * r8b0b6a6719c7 10/src/view/ (message-new-view.c message-show-view.c): cosmetics Jan 06 19:11:58 freesmartphone.org: 03morphis 07msmcomm * r7a0566c5e675 10/libmsmcomm-next/ (7 files in 3 dirs): libmsmcomm-next: implement all currently known misc subsystem messages Jan 06 19:12:47 SHR: 03mok 07libphone-ui-shr * r490fd851cbe9 10/src/view/message-list-view.c: message-list-view: use int64 variants for genlist item pointers Jan 06 19:15:16 freesmartphone.org: 03morphis 07msmcomm * rbc96b1028296 10/libmsmcomm-next/msmcomm/messagedisassembler.vala: libmsmcomm-next: register all available message groups to disassembler Jan 06 19:15:20 freesmartphone.org: 03morphis 07msmcomm * r0b0c922900c1 10/libmsmcomm-next/.gitignore: libmsmcomm-next: update .gitignore Jan 06 19:19:11 GarthPS: good news, palmpre2 support for OE is only two steps away Jan 06 19:20:19 larsc: so, here it works if i add remote tracking branch from context of already-clonned repo. in case of empty repo it doesn't work Jan 06 19:21:25 morphis: hi! kweeel! fso-install gains : switch-boot (tested and working) , restore-media(on its way), setup-init (on its way), install-all(to test), and wget on its-way Jan 06 19:21:39 larsc: i suggest checking vm.overcommit_memory sysctl and add swap file on git server Jan 06 19:21:39 yeah! Jan 06 19:21:46 GarthPS: that are even great news! Jan 06 19:23:06 morphis: so if I succed to work great this evening I hope it should be near total automaticaly installer like : wget Makefile, put your device in dev mode and make install-all and tyhat 's it! Jan 06 19:23:26 great Jan 06 19:23:34 morphis: oh I forget "plug it"! :p Jan 06 19:23:42 what about a config.mk file which can the user use to set it's own settings? Jan 06 19:23:52 include it in Makefile when available Jan 06 19:24:08 and then set every variable only with =? Jan 06 19:24:22 so it gets not overwritten when already set by config.mk Jan 06 19:24:22 morphis: oh for the moment i just define strings at the top of the makefile Jan 06 19:24:28 ok Jan 06 19:25:25 mrmoku, gdbus is there? :P Jan 06 19:25:26 morphis: like all needed fields already feededs and just need to switch TARGET to pre or pre 2? does it ok for you ? Jan 06 19:26:26 jepp Jan 06 19:26:42 morphis: anyway i wil show you for your opinion and then we will see Jan 06 19:33:33 TAsn: almost :P Jan 06 19:33:38 :P Jan 06 19:33:49 still working on gvariants Jan 06 19:34:02 those bastards. Jan 06 19:34:11 they are much nicer than the gvalue bastards ;) Jan 06 19:34:25 but we used gvalues everywhere... and I replaced them all Jan 06 19:35:16 i.c Jan 06 19:35:21 we sucked :P Jan 06 19:35:47 so much nice stuff Jan 06 19:35:55 are going to get in for e17 1.1 :P Jan 06 19:35:56 morphis: oh I have a question for you! Jan 06 19:35:58 (i.e after the release) Jan 06 19:36:15 aboout the makefile there is line that I don't understand Jan 06 19:36:37 TAsn: no thanks :P Jan 06 19:36:47 huh? Jan 06 19:36:51 some day we should try to stabilize something Jan 06 19:36:58 oh, lol. Jan 06 19:37:03 I'm the worst man for that. Jan 06 19:37:09 all of us Jan 06 19:37:12 I hate stabilizing stuff. Jan 06 19:37:29 morphis: http://pastebin.com/mzCsG10i what does this portion "downloads/$(WEBOSJAR)" Jan 06 19:37:32 I prefer using unstable everything :) Jan 06 19:37:41 * mrmoku is using kdevelop from git ;) Jan 06 19:38:13 TAsn: we can talk about e17 1.1 after we have a SHR release Jan 06 19:38:20 when is that due anyway? Jan 06 19:38:24 (both :P) Jan 06 19:39:20 1.1 ~ 6 months after 1.0 Jan 06 19:39:26 which is in Jan 06 19:39:35 1 to 89 months from now Jan 06 19:40:35 ok, gives us some time :P Jan 06 19:40:36 or not Jan 06 19:40:48 shr release, a couple of years after that :P Jan 06 19:40:56 taking history into account we probably have lots of time ;) Jan 06 19:41:03 yeah Jan 06 19:41:19 first world domination... then release Jan 06 19:41:35 everything else would be easy ;) Jan 06 19:43:40 :P Jan 06 19:47:26 * GarthPS eating Jan 06 19:49:44 i guess we have about 3 months to release ;) Jan 06 19:50:12 1 april seems to be perfect date ;D Jan 06 19:58:37 which year dos1, which year? :P Jan 06 19:58:57 mrmoku: hmm, good question ;) Jan 06 19:59:09 and then... 11.11.11 would be a nice date too Jan 06 19:59:18 gives us some more time to introduce even more bugs Jan 06 20:03:21 GarthPS: it is a dependency on the file downloads/$(WEBOSJAR) Jan 06 20:03:37 if the file is not there make runs the rule downloads/$(WEBOSJAR): Jan 06 20:10:50 evening :) Jan 06 20:12:05 long time no see :) Jan 06 20:12:20 i a experiencing a strange issue in python-elementary Jan 06 20:13:24 mrmoku: the thing is: on 1.04 we can announce whatever we want :D Jan 06 20:13:35 not sure shat widget is affecting this, but generally, pager page longer then screen, therefore scrolled, when tapped at it's bottom section, instead of registering double tap suddenly scrolls up Jan 06 20:13:59 is this known to any of the E mags? Jan 06 20:30:40 morphis: yeah I understaood that but there is no downloads/$(WEBOSJAR) rule/target Jan 06 20:30:59 that is why i did not understood Jan 06 20:31:49 morphis: so it is juste a way to check that the file is there right ? Jan 06 20:32:07 without a real downloads/$(WEBOSJAR) rule defined Jan 06 20:34:37 SHR: 03mok 07libphone-ui-shr * rd43481932605 10/src/ (9 files in 2 dirs): unref variants when destroying hashtables with'em in it Jan 06 20:38:10 happy new year everyone! :) Jan 06 20:41:19 GarthPS: right Jan 06 20:42:57 morphis: so I can rename it by a wget-webosdoctor: and defined it . is that ok for you ? Jan 06 20:45:31 right Jan 06 20:45:58 but you don't do the wget-webosdoctor Jan 06 20:46:14 the user should have to put it there Jan 06 20:46:36 no automatically process for at it as thats something not expected as legal as when you do it manual on your own Jan 06 20:46:59 and if you do something like this Jan 06 20:47:23 let it be the target downloads/$(WEBOSJAR) don't make it phony Jan 06 20:47:33 as if the file is there the target don't get executed Jan 06 20:47:39 as it depends on the file Jan 06 20:53:09 onen|openBmap: for you too Jan 06 20:56:23 morphis: ok anderstood :) Jan 06 21:00:31 SHR: 03mok 07libphone-ui * r4ad47c4db9ad 10/src/ (5 files): unref variants when destroying hashtables Jan 06 21:01:44 playya_: so... we finally got rid of libframeworkd-glib :-D Jan 06 21:13:04 nice. Jan 06 21:13:08 mrmoku, ^ Jan 06 21:14:11 ok building :) Jan 06 21:14:14 thanks Jan 06 21:15:06 hmm new 2.6.37 kernel boots with suppiled kernel config, but if i try mine used for qtmoko it's just WS after boot Jan 06 21:15:11 ah some build is already running :) Jan 06 21:22:28 yup :) Jan 06 21:26:25 heh, shit Jan 06 21:29:43 anybody knows the correct way to store pointers in gvariants? Jan 06 21:41:36 mrmoku, yay. now we should start writing fsopimd Jan 06 21:42:03 GarthPS: could you please test the kernel from here (http://amethyst.openembedded.net/~morphis/oe/palmpre/deploy/images/palmpre2/) Jan 06 21:42:24 GarthPS: in some minutes, still compiling Jan 06 21:43:30 its ready in two minutes Jan 06 21:43:33 so I am off now Jan 06 21:43:35 byebye Jan 06 21:44:00 playya_: oh yes :) Jan 06 21:44:37 larsc: what config should I use with 2.6.37? with gta02_defconfig I get 'FIC Neo1973 GSM Phone (GTA01 Hardware) (MACH_NEO1973_GTA01) [N/y/?] (NEW)' Jan 06 21:45:04 larsc: is it the gta02_drm_defconfig that you update? Jan 06 21:54:08 lindi-: hi. i think i've digged the sources enough so i'd like you to share how you managed make gpsd use ubx with fr's receiver. Jan 06 21:58:09 SHR: 03mok 07libphone-ui-shr * r8723ff875bbc 10/src/ (3 files in 2 dirs): add common_utils to stuff a pointer arch friendly into a gvariant Jan 06 21:59:23 PaulFertser: I just switched to ubx mode with a python program Jan 06 21:59:42 PaulFertser: the ubx mode switching code in gpsd is quite broken at the moment Jan 06 22:00:07 PaulFertser: iki.fi/lindi/git/ubx.git i think is the tools I used Jan 06 22:01:09 lindi-: cheater :) Jan 06 22:01:17 lindi-: ok, i got your point, thanks :) Jan 06 22:01:23 PaulFertser: I can help with ubx stuff Jan 06 22:01:40 PaulFertser: I was just first trying to get gpsd include the hooks that would let me do all this Jan 06 22:01:46 lindi-: i know. Jan 06 22:02:10 lindi-: i think i'll try to understand how exactly to use control channel for bidirectional communication as esr said it's possible. Jan 06 22:02:36 lindi-: btw, i'm such a looser, i made gpsd segfault somehow but core dumps were disabled on my laptop :/ Jan 06 22:02:42 :) Jan 06 22:02:53 using a laptop is clearly cheating too Jan 06 22:04:56 lindi-: it asks because lars added gta01 support today Jan 06 22:05:24 lindi-: you can say no (and it should be default) for the questions Jan 06 22:06:12 radekp: my build process isn't interactive :P Jan 06 22:06:31 lindi-: so interactively regenerate the config and commit it :) Jan 06 22:06:56 there is problem that with my 2.6.34 config i am getting WS Jan 06 22:07:17 but the kernel is working probably Jan 06 22:07:22 i can ssh to it Jan 06 22:07:36 hmm but i am too tired for today Jan 06 22:08:23 SHR: 03mok 07libphone-ui-shr * r76eb8102f9fd 10/src/util/common-utils.c: next try to get pointers to gvariants Jan 06 22:17:41 lindi-: gta02_defconfig should work Jan 06 22:19:11 larsc: it does ask ''FIC Neo1973 GSM Phone (GTA01 Hardware) (MACH_NEO1973_GTA01) [N/y/?] (NEW)'' Jan 06 22:21:56 hm Jan 06 22:24:19 larsc: $ grep NEO1973_GTA01 arch/arm/configs/gta02_defconfig Jan 06 22:24:21 $ Jan 06 22:24:28 larsc: so it's not in the config Jan 06 22:27:25 hm Jan 06 22:33:07 lindi-: ah. you pull the om-2.6.37 tree Jan 06 22:33:22 the config is only for om-gta02-2.6.37 Jan 06 22:34:01 larsc: aha, wrong tree Jan 06 22:35:07 maybe i should add an update to the om-2.6.37 tree adding the newly undefined symbols to the config Jan 06 22:37:54 larsc: which tree should I test? ;) Jan 06 22:38:24 lindi-: does fr's gps accept UBX by default via serial? afaict, yes. And gpsd has a trick (sending a special ubx message to request version) that if answered would switch the driver. Jan 06 22:38:38 Strange it's not working. Jan 06 22:39:18 PaulFertser: UBX is one of the input protocols that are enabled by default Jan 06 22:39:33 PaulFertser: not strange at all :) Jan 06 22:39:48 PaulFertser: as I told you, gpsd's ubx stuff is pretty broken at the moment Jan 06 22:40:00 lindi-: yes, so because gpsd requests a version via ubx, the chip should answer in UBX too, and gpsd will recognise the packet and switch the driver. Jan 06 22:40:06 lindi-: at least that's the intent of the code :) Jan 06 22:40:12 afaict Jan 06 22:40:22 PaulFertser: at least there were quite many off-by-two's in the code Jan 06 22:40:42 PaulFertser: I didn't report those since there was this other bug that was blocking me from ever being able to execute that part of the code Jan 06 22:40:50 lindi-: ah, ok, i see. Thanks for confirming ubx is enabled. Jan 06 22:40:51 now there's a suggested patch I haven't tested yet Jan 06 22:40:55 lindi-: om-gta02-2.6.37 Jan 06 22:41:05 PaulFertser: you probably want to play around with my python tools Jan 06 22:41:08 larsc: ok Jan 06 22:41:12 PaulFertser: just to learn ubx Jan 06 22:41:19 (they are forked from ogpsd) Jan 06 22:41:37 larsc: hi :) And i want to express i'm rather glad to see you being still very active in OM land. Jan 06 22:42:05 PaulFertser: hi :) Jan 06 22:42:18 hopefully most of the stuff is now ready to go upstream Jan 06 22:42:23 lindi-: i guess i already have cloned them, months ago. TBH, i was mostly impressed by that one-liner to enable RAW reporting. Jan 06 22:42:42 my RTK stuff is still below snow :( Jan 06 22:43:09 maybe in a few months I can retry Jan 06 22:43:48 lindi-: you mean snow as in "it'd require helluva work to dig them out"? Jan 06 22:44:08 PaulFertser: yes and it's in a place with restricted access Jan 06 22:44:27 Hey, that sounds cryptic now :) Jan 06 22:44:29 on a roof top of a building Jan 06 22:44:52 lifted there from a balcony Jan 06 23:11:42 * lindi- testboots 2.6.37 Jan 06 23:12:07 larsc: no such file or directory while trying to open /dev/mmcblk0p1 Jan 06 23:14:47 hm, the driver should be built-in Jan 06 23:15:00 does /dev/mmcblk0 exist? Jan 06 23:15:09 and it did mount root Jan 06 23:15:14 it's stuck at fsck failing Jan 06 23:17:27 I guess I need to retry later with serial console Jan 07 01:10:17 larsc: http://paste.debian.net/103947/ Jan 07 01:10:30 larsc: so udev is not creating the mmcblk* stuff Jan 07 01:10:39 larsc: also ext3 support is missing **** ENDING LOGGING AT Fri Jan 07 02:59:58 2011