**** BEGIN LOGGING AT Wed Sep 24 02:59:57 2008 Sep 24 03:01:04 greetings Sep 24 03:01:10 morning mwester-trip Sep 24 03:03:07 rwhitby: http://trac.freesmartphone.org/ticket/50 can be closed? Sep 24 03:03:07 Insufficient data points to be certain -- however, it seems that an incoming phone call is sufficient to cause the "bouncy" GSM (bug 1024) to behave normally, until next reset or power-on. Sep 24 03:03:33 hmm!? worth trying Sep 24 03:04:03 mickey|tw: actually, no. I haven't been able to get GPS to work on fso-testing in jffs2 since other changes to nmea.py and ogpsd have gone in after my changes. Sep 24 03:04:31 and havent' had time to debug it yet. Sep 24 03:05:02 I guess we could open a new bug instead. Sep 24 03:05:04 rwhitby: k, just checking Sep 24 03:16:26 ok done Sep 24 03:16:33 now lets hope rtc wakeups work... Sep 24 03:34:47 freesmartphone.org: 03mickey 07framework * rbd066c464603 10/tools/cli-framework: Sep 24 03:34:47 freesmartphone.org: tools: add usage and gps interfaces to cli-framework. closes fso #131 Sep 24 03:34:47 freesmartphone.org: patch by fso@mattb.net.nz, thanks! Sep 24 04:38:15 hmm Sep 24 04:38:16 still no luck Sep 24 04:38:23 even after restarting bluetooth init scripts Sep 24 04:38:34 dbus.exceptions.DBusException: org.bluez.Error.Failed: Operation not permitted Sep 24 04:38:42 while remoko is initializing Sep 24 04:44:39 hmm Sep 24 04:44:44 I'll come over Sep 24 05:40:25 * rwhitby imagines mickey|tw and alphaone in the same room, talking via IRC ;-) Sep 24 05:40:44 Nope, mickey|tw has an office. Sep 24 05:40:54 I'm sitting in a cubicle Sep 24 05:41:03 :-) Sep 24 07:03:03 freesmartphone.org: 03daniel 07framework * rb8b9889f4e71 10/ (5 files in 3 dirs): Sep 24 07:03:03 freesmartphone.org: ogsmd: Use UCS2 as default character set Sep 24 07:03:03 freesmartphone.org: This solves internationalization issues for contact names. Sep 24 07:11:58 ewww cubicles Sep 24 07:15:51 they put u in cubicles now? Sep 24 07:23:07 hi Sep 24 07:29:00 thats just so corporate Sep 24 07:29:35 a question: doses anyone have already touch RTAI on openmoko yet ? Sep 24 07:45:12 freesmartphone.org: 03mickey 07zhone * r39e48aa1bc1e 10/src/zhone: Sep 24 07:45:12 freesmartphone.org: Move GSM registration handling from Register() callback to network Status signal handler. Sep 24 07:45:12 freesmartphone.org: Patch courtesy hedora -- thanks! Closes FSO ticket #139 Sep 24 08:23:22 freesmartphone.org: 03charlie 07framework * rf6be7e427f5c 10/framework/controller.py: Sep 24 08:23:22 freesmartphone.org: Controller.object method raise a KeyError exception if the object is not present. Sep 24 08:23:22 freesmartphone.org: It makes more sens, since most of the time an object not present is an error. Sep 24 08:23:22 freesmartphone.org: 03charlie 07framework * r73068a192652 10/framework/subsystem.py: Print the traceback if an exception occures during a subsytem init. Sep 24 08:23:24 freesmartphone.org: 03charlie 07framework * r0ab8aae3b97b 10/tests/tasklet.py: Sep 24 08:23:26 freesmartphone.org: [tests] Updated tasklet module Sep 24 08:23:28 freesmartphone.org: In fact I should totaly remove it since we have a tasklet module in the framework now. Sep 24 08:23:31 freesmartphone.org: 03charlie 07framework * r6704245cdfad 10/ (5 files in 4 dirs): Sep 24 08:23:33 freesmartphone.org: Added the resource module. Sep 24 08:23:35 freesmartphone.org: Every subsystem that wants to register as a resource should subclass Resource, Sep 24 08:23:37 freesmartphone.org: and reimplement the _enable, _disable, _suspend and _resume methods. Sep 24 08:23:39 freesmartphone.org: it should also make sure to call Resource.__init__ with a keyword 'name' parameter. Sep 24 08:23:42 freesmartphone.org: I updated ogsmd to behave this way, but we still need to implement the methods. Sep 24 09:02:20 freesmartphone.org: 03jluebbe 07zhone * r9478feff4823 10/data/themes/ (10 files): use fontconfig instead of the embedded font (fixes chinese text) Sep 24 09:02:20 freesmartphone.org: 03jluebbe 07zhone * r5224f9fdb20d 10/data/fonts/DejaVuSans-Bold.ttf: remove font file Sep 24 12:04:51 hi! Sep 24 12:05:15 does someone know if I can't have more output from qemu ? Sep 24 12:05:53 I flashed with success an asu image I build with MokoMakefile for om-gta01 Sep 24 12:06:10 ptitjes: and it just resets back to u-boot instead of booting the kernel? Sep 24 12:06:20 but the kernel is checked OK but qemu reboots at the very start of loading the kernel Sep 24 12:06:33 rwhitby: exactly!! Sep 24 12:06:34 yep, same here, both for ASU and FSO images Sep 24 12:06:46 arff Sep 24 12:07:17 seems that OM has let qemu-1973 die :-( Sep 24 12:08:25 grrr! I've got no neo yet!!! :-( Sep 24 12:11:30 rwhitby: can you confirm that that fresh u-boot + kernel + asu build runs well on a neo 1973 ? Sep 24 12:13:12 I haven't changed u-boot in a while. Sep 24 12:13:34 I built kernel and asu and ran it on gta01 and was severely underimpressed. Sep 24 12:13:42 So I went back to fso-testing. Sep 24 12:14:49 if you haven't got a neo yet, are you sure you're talking about a neo 1973 (gta01) and not a Neo FreeRunner (gta02) ? Sep 24 12:16:09 I'am waiting to receive my fr after the 20th of October Sep 24 12:16:24 in the mean time I try to have a working dev environnement Sep 24 12:16:50 (but you know that I started two weeks ago without any real success...) Sep 24 12:17:28 the official image starts well under qemu Sep 24 12:18:06 I am wondering whether I could flash a official u-boot/kernel but have my local rootfs Sep 24 12:18:42 rwhitby: what means "severely underimpressed" ? Sep 24 12:18:54 could you elaborate ? Sep 24 12:19:20 I wouldn't run ASU. To much bling, not enough usability. Sep 24 12:19:35 Look in the IRC logs about this time 24 hours ago to see. Sep 24 12:19:53 haaaa ok! Sep 24 12:20:03 !logs Sep 24 12:20:39 hmm - no logs, let me fix that. Sep 24 12:21:47 I just would like to have a working kernel + base rootfs with opkg to test cacao/swt/equinox/JFace/EMF/Databinding usability and have an idea if it could support that big stack in performance Sep 24 12:22:20 what could you advice me ? surely fso ? Sep 24 12:23:09 In fact I would like to make alternative front-end to fso Sep 24 12:24:02 amazing. that's exactly why we're doing this Sep 24 12:24:08 keep us posted Sep 24 12:24:21 dbus-java + EMF (data) + JFace forms + equinox OSGI plugin system Sep 24 12:24:21 if you have problems w/ the fso APIs, use trac and the lists Sep 24 12:24:28 * mickey|tw end of work today Sep 24 12:24:29 l8er Sep 24 12:24:59 bah! Sep 24 12:25:02 he ran Sep 24 12:25:05 :( Sep 24 12:25:14 he saw you coming Sep 24 12:25:19 hehehjeheheh Sep 24 12:25:28 lol Sep 24 12:25:38 * rwhitby runs too Sep 24 12:25:42 was going to pestern him on frameworkd's evil habit of listeing to touchscreen events Sep 24 12:26:23 :D Sep 24 12:26:44 like when i drag my finger around the screen - frameworkd is using 50+% cpu.... Sep 24 12:27:02 thats pretty nasty for just listeing to mouse events.. not actually even doing anything :) Sep 24 12:27:23 rwhitby: could u give me a tip so that I can use official u-boot/kernel and custom build rootfs ? Sep 24 12:28:10 should I copy my jffs file to the official images location and rename it ? is that easy ? (not sure whether i will crack it all!!) Sep 24 12:37:00 are u all still using u-boot 1.3.1 ??? Sep 24 12:41:44 Hey Sep 24 12:41:48 Ainulindale, here? Sep 24 12:42:37 all: could u please tell me what u-boot versions you are using ???? Sep 24 12:44:53 ... ??? Sep 24 12:45:05 ptitjes, what for? Sep 24 12:45:14 for comparison Sep 24 12:45:37 official images downloaded by MokoMakefile uses u-boot 1.3.1 Sep 24 12:46:13 qemu does not work anymore with 1.3.2, I was wondering whether the 1973 and fr were running with 1.3.2 Sep 24 12:46:32 I'd like to narrow the problem (qemu, u-boot or kernel...) Sep 24 12:48:30 quickdev: there Sep 24 12:48:33 Saw your question Sep 24 12:48:42 I don't see what could give you that situation Sep 24 12:48:51 Ainulindale, I don't know how to debug that Sep 24 12:48:56 rwhitby told he did not update his u-boot, so... did someone updated to u-boot 1.3.2 ? Sep 24 12:49:03 Ainulindale, did do a little test? Sep 24 12:49:04 CallStatus isn't triggered in lfg, in any way ? Sep 24 12:49:12 quickdev: I can't I'm in a work rush Sep 24 12:49:16 Ainulindale, yes, it isn't Sep 24 12:49:44 hmmm Sep 24 12:49:47 Ainulindale, ok...when would it be possible for you? :) Sep 24 12:49:57 saturday or sunday Sep 24 12:50:00 but wait Sep 24 12:50:00 it's quite strange... Sep 24 12:50:04 do you use ophonekitd ? Sep 24 12:50:16 are you sure you defined the callback in frameworkhandler ? Sep 24 12:50:51 Ainulindale, it even says "Added call call status handler" Sep 24 12:52:28 and yes, I use ophonekitd, but a highly modified version Sep 24 12:54:11 Ainulindale, could the refactoring of the dbus/*.h have destroyed something? Sep 24 12:54:35 Ainulindale, http://rafb.net/p/LeVo2722.html Sep 24 12:59:35 okay: qemu works with u-boot 1.3.1 +kernel from official image + custom build asu Sep 24 13:00:21 I will now try with official u-boot 1.3.1 + custom build kernel and asu Sep 24 13:02:55 quickdev: did you refactor dbus/*.h by hand ? Sep 24 13:03:17 Ainulindale, no, I deleted my lfg directory and cloned the git one with your renamings Sep 24 13:04:38 Hmmm Sep 24 13:04:49 Could you debug step by step and see if the handler is actually attached ? Sep 24 13:05:02 It may be a dbus error due to an invalid handler for the signal Sep 24 13:05:07 I'll have a look Sep 24 13:05:21 Ainulindale, how do I debug step by step? If I know, I'll do it. Sep 24 13:06:19 And it is Sep 24 13:06:26 void (*callCallStatus)(const int, const int, GHashTable *); Sep 24 13:06:45 Oh no, wrong call there for me =) Sep 24 13:06:51 quickdev: using gdb ? :-) Sep 24 13:07:46 quickdev: did you add some printf in ogsmd_call_status_handler ? Sep 24 13:07:54 Ainulindale, have you a good introduction to gdb? Used it some time ago but haven't found a good introduction Sep 24 13:08:01 yes, I did Sep 24 13:08:06 g_error() or g_debug() Sep 24 13:08:24 quickdev: No I haven't Sep 24 13:08:26 quickdev: and ? Sep 24 13:08:29 Nothing ? Sep 24 13:08:35 nothing...that's the problem.. Sep 24 13:09:45 That must mean that the signal handler couldn't register properly Sep 24 13:10:42 It must be a type problem again Sep 24 13:11:13 and with gdb I'll see if there's an error, right? Sep 24 13:11:15 qemu boots with u-boot 1.3.1 + custom build kernel and asu Sep 24 13:11:31 should I add that info to the wiki ? Sep 24 13:12:28 quickdev: with gdb you'll be able to step by step the execution of lfg Sep 24 13:12:35 though stopping on the interesting line Sep 24 13:12:50 ... nothing Sep 24 13:13:16 quickdev: your g_error are there in lfg or ophonekitd ? Sep 24 13:13:21 s/there/they/ Sep 24 13:13:30 ophonekitd call status handler Sep 24 13:14:44 put it in lfg Sep 24 13:14:48 see if the handler is called Sep 24 13:15:07 ok Sep 24 13:15:08 If it isn't, it must be because the signal handler didn't register properly Sep 24 13:15:17 frameworkd-glib-ogsmd-call.c Sep 24 13:15:29 line 50 Sep 24 13:16:26 I just saw that I already did that Sep 24 13:16:31 void ogsmd_call_status_handler (DBusGProxy *proxy, const int id, const char *status, GHashTable * properties, gpointer user_data) Sep 24 13:16:31 { Sep 24 13:16:31 g_error("call status handler!"); Sep 24 13:18:53 And nothing then ? Sep 24 13:19:29 And sim auth status works , Sep 24 13:21:43 yes, right Sep 24 13:22:08 that's mightily weird Sep 24 13:22:35 Can you use call methods ? Sep 24 13:28:24 Ainulindale, yes, I can Sep 24 13:28:35 I have to go for some time Sep 24 13:28:39 I'll be back don't worry Sep 24 13:29:10 ok, I'll run gdb later Sep 24 13:31:04 we should test things precisely Sep 24 13:31:55 I mean, I'll hopefully see what's wrong then Sep 24 16:28:12 wurp2|working: so, complains about MS3 ? Sep 24 16:28:32 Even if everything is not working right now, I think it's pretty good, because I like the way it's going Sep 24 16:28:37 But that's only my point of view Sep 24 16:31:47 Ainulindale: I am not using it Sep 24 16:31:52 I know Sep 24 16:31:55 Ainulindale: I heard too many complaints :-( Sep 24 16:31:59 So what ? Sep 24 16:32:07 You hear complaints all the time Sep 24 16:32:08 So, I use my neo as my primary phone Sep 24 16:32:15 No one is going to tell "yay, this is great ! I have no problem !" Sep 24 16:32:20 So do I Sep 24 16:32:22 And I use MS3 Sep 24 16:32:23 So I'm running qtopia Sep 24 16:32:28 And it's working well for you? Sep 24 16:32:29 And right now I can't even receive SMSes Sep 24 16:32:31 And I don't care Sep 24 16:32:39 First because I know the reason Sep 24 16:32:44 sms is not terribly important for me Sep 24 16:32:47 Secondly because I know this is going to settle soon enough Sep 24 16:32:55 And third because I don't care about SMSes either Sep 24 16:33:12 I think the OM initiative is great, and the FSO initiative is even better Sep 24 16:33:16 but... I don't have time to burn the image, then burn it again when sms is fixed, just so I can say I'm running FSO Sep 24 16:33:25 because its goal isn't to build something based on open things Sep 24 16:33:30 even though I much prefer FSO in concept to anything out there Sep 24 16:33:32 But to identify and build standards Sep 24 16:33:39 although I think I prefer FSO on debian :-) Sep 24 16:33:56 Which is far better to my industrial-son-of-a-bitch-professionnal point of view Sep 24 16:34:07 And in fact, that's going to be my next burn, once I get an 8gb card Sep 24 16:34:08 and time Sep 24 16:34:16 I have to go to Arkansas yet again this weekend Sep 24 16:34:19 Debian is bad Sep 24 16:34:21 For my stepdad's memorial Sep 24 16:34:23 Debian kills kitten Sep 24 16:34:30 I hate kittens anyway Sep 24 16:34:34 I don't Sep 24 16:34:40 They taste good Sep 24 16:34:44 little yowling hairballs Sep 24 16:34:50 ah, kitten tartar Sep 24 16:34:53 Anyway Sep 24 16:34:54 straight off the flank Sep 24 16:34:59 If you have a look at 2007.2 Sep 24 16:35:02 Then 2008.9 Sep 24 16:35:06 Everything is, well Sep 24 16:35:09 Everything Sep 24 16:35:18 There's not one central nervous system Sep 24 16:35:26 And that's what I'm interested in Sep 24 16:35:32 Yeah, me too Sep 24 16:35:38 Because I think that on mobile with a specific set of functionnalities Sep 24 16:35:47 So I'll run FSO Sep 24 16:35:49 on debian Sep 24 16:35:50 What is interesting and necessary is a meta operating system Sep 24 16:36:08 But once again Sep 24 16:36:20 That's because I'm a fervent zealot of optmization and factorization Sep 24 16:36:32 Only thing I don't like in frameworkd is that it's python Sep 24 16:36:35 Not that I don't like python Sep 24 16:36:44 factorization? Sep 24 16:36:45 Just that I prefer C for this kind of software Sep 24 16:36:52 Or even better : OCaml Sep 24 16:37:04 wurp2|working: Yeah, because it leads to code factorization Sep 24 16:37:10 I prefer C once it's been proven by 5 years or so in the market w/ active development :-) Sep 24 16:37:12 you don't have to manage gsm here and there Sep 24 16:37:19 Ainulindale: and I suppose you actually have an example of a main UI written in ocaml that should be persuasive to us? ;-) Sep 24 16:37:28 cjb: who talked about UI yes ? Sep 24 16:37:32 UI are for sissies Sep 24 16:37:34 And old people like wurp2|working Sep 24 16:37:38 ah, yeah. Sep 24 16:37:45 s/yes/here/ Sep 24 16:37:46 by the way Sep 24 16:37:53 cjb: but more seriously Sep 24 16:37:56 I'm interested in primitives Sep 24 16:38:01 Of course, I use UIs Sep 24 16:38:05 But I don't like to build them Sep 24 16:38:11 And I don't care how they're built Sep 24 16:38:17 Ainulindale: hah, I was writing shell scripts when you were a twinkle in your mom's eye Sep 24 16:38:21 As long as they can use a set of defined primitives which I can understand Sep 24 16:38:32 wurp2|working: Possible Sep 24 16:38:34 I like UIs when they get me customers Sep 24 16:39:01 Nah, not possible since I've only been doing unix for maybe 17 years Sep 24 16:39:07 But I might have been writing 6502 assembler Sep 24 16:39:24 Anyway as I said, I don't like systems where you have to look at several separate bits of code to do a thing you're supposed to do and will do on almost every piece of software Sep 24 16:40:15 Yeah, but I'm not writing serious code for the neo yet :-( Sep 24 16:40:57 Do you seriously think I am ? Sep 24 16:41:07 I'm struggling with quickdev's patches because I don't have time Sep 24 16:41:18 Stop whinning granpa Sep 24 16:41:45 When I start writing serious code, I'll be more concerned that I'm on a good platform Sep 24 16:42:01 As opposed to now, when I'm more concerned that I can make & answer calls & sms Sep 24 16:42:12 And survive more than 6 hours away from a charger Sep 24 16:42:31 That I can understand Sep 24 16:42:38 And relate to :-) Sep 24 16:43:20 cjb: Though to troll a little more Sep 24 16:43:26 OCaml is a great and marvelous language Sep 24 16:43:49 And match x with is the prettiest piece of language keyword I ever saw Sep 24 16:44:22 Are OCaml pointers all 'raw'? Sep 24 16:44:28 Pointers ? Sep 24 16:44:29 What for ? Sep 24 16:44:39 e.g. can you fuck up memory by writing to an uninitialized variable Sep 24 16:44:51 Again : Pointers ? What for ? :-) Sep 24 16:44:56 I dunno nuttin about OCaml. Except what's on the front page for the inria site Sep 24 16:45:07 wurp2|working: it's mainly a functional language Sep 24 16:45:10 For to write to set memory locations to communicat w/ hw Sep 24 16:45:13 Works great for algorithmic problems Sep 24 16:45:20 s/cat w/cate w/ Sep 24 16:45:39 And I like strong types Sep 24 16:46:02 And lambda calculus is fun, isn't it ? Sep 24 16:46:12 OK, I'm a big fan of functional... as much as I can be given that my only real experience is limping by with what C++ and python support Sep 24 16:46:27 :-) Sep 24 16:46:36 wurp2|working: In fact I wrote a compiler in OCaml Sep 24 16:46:44 Greatest experience I ever had in development Sep 24 16:47:02 And probably the more straightforward (/simple) too Sep 24 16:48:02 OCaml looks great just looking at the manual Sep 24 16:48:17 Anyway I have to go Sep 24 16:48:21 But, why OCaml instead of haskell, erlang, etc? Sep 24 16:48:24 k, later Sep 24 16:48:26 I'm seeing Rigoletto tonight at the Opéra Bastille Sep 24 16:48:39 wurp2|working: because haskell is the tool of the devil Sep 24 16:48:52 (And because I never used haskell and don't know at all how it looks) Sep 24 16:48:59 Ainulindale: fucking cultured bastard Sep 24 16:49:03 (But I heared it was great) Sep 24 16:49:14 I'm not cultured Sep 24 16:49:20 I want to be but it's hard work Sep 24 16:49:28 Like reading books without pictures Sep 24 16:49:30 I'm not sure I want to be Sep 24 16:49:39 But I would love to go see a good opera, at least once Sep 24 16:49:43 (I'm reading Churchill's biography these times) Sep 24 16:49:46 (Truly amazing) Sep 24 16:49:52 I'm not sure any opera showing in TX would qualify as good Sep 24 16:50:03 wurp2|working: I saw pretty bad operas here Sep 24 16:50:11 I'm sure Sep 24 16:50:12 In fact, pretty bad interpretations Sep 24 16:50:35 Ainulindale: autobiography, or written by someone else? Sep 24 16:50:52 But I saw a Don Giovanni version at the Teatro Real in Madrid which was tremendous, and I hold this feeling tight Sep 24 16:51:01 biography by François Kersaudy Sep 24 16:51:06 I prefer Ben Franklin's biography, mostly 'cause it's online for free :-) Sep 24 16:51:08 And next I'll move to his memoirs Sep 24 16:51:14 (autobiography) Sep 24 16:51:15 (twelve books) Sep 24 16:51:27 hah, that's too much Churchill for me Sep 24 16:51:51 Heh Sep 24 16:52:17 On my desk there's also a book about atheism by Dawkins, the biography of Hitler, a book about Yalta and some novels Sep 24 16:52:20 Anyway Sep 24 16:52:24 I'm going to be late Sep 24 16:52:36 Don't forget your medication granpa, this arthritis won't heal by itself Sep 24 16:52:46 And see you later ;-) Sep 24 16:54:47 later Sep 24 16:55:00 I've read several Dawkins Sep 24 17:45:37 freesmartphone.org: 03daniel 07framework * r3d55eed21bfa 10/framework/subsystems/ogsmd/modems/abstract/mediator.py: Sep 24 17:45:37 freesmartphone.org: ogsmd: Fix SimRetrieveMessage and remove some unneeded code. Sep 24 17:45:37 freesmartphone.org: Patch by dv_mlist@verizon.net, thanks! Sep 24 17:45:37 freesmartphone.org: This closes FSO Ticket #153 Sep 24 17:45:37 freesmartphone.org: 03daniel 07framework * rf60b9e85d0ac 10/framework/subsystems/ogsmd/gsm/convert.py: ogsmd: Fix calculating the timezone for the SCTS and VP field Sep 24 17:45:40 freesmartphone.org: 03hadara 07framework * r5c1cfd6b58f7 10/framework/subsystems/ogsmd/gsm/convert.py: Sep 24 17:45:43 freesmartphone.org: ogsmd: Fix bug in bcd_decode when an address is empty Sep 24 17:45:45 freesmartphone.org: When you have SMS messages with empty sender address the bcd_decode() Sep 24 17:46:35 freesmartphone.org: function will fail since it expects string to be at least 1 char long. Sep 24 17:46:35 freesmartphone.org: This patch fixes it. Sep 24 17:46:35 freesmartphone.org: Thanks to hadara@bsd.ee! This fixes FSO Ticket #145 Sep 24 19:19:44 <|FiLiP|> whois GuruMeditation Sep 24 21:37:10 wurp2|working: well, it was good Sep 24 21:40:31 Ainulindale: Great! Sep 25 00:10:15 morning Sep 25 01:15:18 morning? Sep 25 01:15:20 morning! Sep 25 01:42:56 mooning **** ENDING LOGGING AT Thu Sep 25 02:59:57 2008